INTERNATIONAL JOURNAL OF EARTH SCIENCES AND ENGINEERING CAFET-INNOVA Technical Society Hyderabad, INDIA

June 2014 Volume 07 No 03 ISSN 0974-5904

INTERNATIONAL JOURNAL OF EARTH SCIENCES AND ENGINEERING

Indexed in: Scopus Compendex and Geobase (products hosted on Engineering Village)

Elsevier, Amsterdam, Netherlands, Chemical Abstract Services-USA, Geo-Ref Information

Services-USA, List B of Scientific Journals in Poland, Directory of Research Journals

Scopus Journal Rating (SJR) 0.15 (2012); H-index: 2 (2012);

CSIR-NISCAIR, INDIA Impact Factor 0.042 (2011)

EARTH SCIENCE FOR EVERYONE

Published by

CAFET-INNOVA Technical Society

Hyderabad, INDIA

http://cafetinnova.org/


1-2-18/103, Mohini Mansion, Gagan Mahal Road, Domalguda

Hyderabad – 500 029, Andhra Pradesh, INDIA

Website: http://www.cafetinnova.org

Mobile: +91-7411311091

Registered by Government of Andhra Pradesh

Under the AP Societies Act., 2001 Regd. No.: 1575

The papers published in this journal have been peer reviewed by experts. The authors are solely

responsible for the content of the papers published in the journal.

Each volume, published in six bi-monthly issues, begins with February and ends with December

issue. Annual subscription is on the calendar year basis and begins with the February issue every

year.

Note: Limited copies of back issues are available.

Copyright © 2014 CAFET-INNOVA Technical Society

All rights reserved with CAFET-INNOVA Technical Society. No part of this journal should be

translated or reproduced in any form, Electronic, Mechanical, Photocopy, Recording or any

information storage and retrieval system without prior permission in writing, from CAFET-

INNOVA Technical Society.

INTERNATIONAL JOURNAL OF EARTH SCIENCES AND ENGINEERING

The International Journal of Earth Sciences and Engineering (IJEE) focus on Earth

sciences and Engineering with emphasis on earth sciences and engineering.

Applications of interdisciplinary topics such as engineering geology, geo-

instrumentation, geotechnical and geo-environmental engineering, mining engineering,

rock engineering, blasting engineering, petroleum engineering, off shore and marine

geo-technology, geothermal energy, resource engineering, water resources and

engineering, groundwater, geochemical engineering, environmental engineering,

atmospheric Sciences, Climate Change, and oceanography. Specific topics covered

include earth sciences and engineering applications, RS, GIS, GPS applications in earth

sciences and engineering, geo-hazards such as earthquakes, landslides, tsunami, debris

flows and subsidence, rock/soil improvements and development of models validations

using field, laboratory measurements.

Professors / Academicians / Engineers / Researchers / Students can send their papers

directly to: [email protected]

CONTACT:

For all editorial queries:

D. Venkat Reddy (Editor-in-Chief)

Professor, Department of Civil Engg.

NIT-Karnataka, Surathkal, INDIA

+91-9739536078

[email protected]

All other enquiries:

Hafeez Basha. R (Managing Editor)

+91-9866587053

[email protected]

Raju Aedla (Editor)

+91-7411311091

[email protected]

EDITORIAL COMMITTEE

D. Venkat Reddy

NITK, Surathkal, Karnataka, INDIA

EDITOR-IN-CHIEF

Trilok N. Singh

IIT-Bombay, Powai, INDIA

EXECUTIVE EDITOR

P. Ramachandra Reddy

Scientist G (Retd.), NGRI, INDIA

EXECUTIVE EDITOR

R. Pavanaguru

Professor (Retd.), OU, INDIA

EXECUTIVE EDITOR

Joanna Maria Dulinska Cracow University of Tech., Poland

EXECUTIVE EDITOR

Hafeez Basha R


MANAGING EDITOR

Raju Aedla


EDITOR

INTERNATIONAL EDITORIAL ADVISORY BOARD

Zhuping Sheng Texas A&M University System

USA

Choonam Sunwoo

Korea Inst. of Geo-Sci & Mineral

SOUTH KOREA

Hsin-Yu Shan National Chio Tung University

TAIWAN

Hyun Sik Yang Chonnam National Univ Gwangu

SOUTH KOREA

Krishna R. Reddy University of Illinois, Chicago

USA

L G Gwalani NiPlats Australia Limited

AUSTRALIA

Abdullah MS Al-Amri King Saud University, Riyadh

SAUDI ARABIA

Suzana Gueiros Dra Engenharia de Produção

BRAZIL

Shuichi TORII Kumamoto University, Kumamoto

JAPAN

Luigia Binda DIS, Politecnico di Milano, Milan

ITALY

Gonzalo M. Aiassa Cordoba Universidad Nacional

ARGENTINA

Nguyen Tan Phong Ho Chi Minh City University of

Technology, VIETNAM

Ganesh R. Joshi University of the Rykyus, Okinawa

JAPAN

Kyriakos G. Stathopoulos DOMI S.A. Consulting Engineers Athens,

GREECE

U Johnson Alengaram University of Malaya, Kuala Lumpur,

MALAYSIA

Robert Jankowski Gdansk University of Technology

POLAND

Paloma Pineda University of de Sevilla, Seville

SPAIN

Vahid Nourani Tabriz University

IRAN

Anil Cherian United Arab Emirates

DUBAI

P Hollis Watts WASM School of Mines

Curtin University, AUSTRALIA

Nicola Tarque Department of Engineering

Catholic University of Peru

S Neelamani Kuwait Institute for Scientific

Research, SAFAT, KUWAIT

Jaya naithani Université catholique de Louvain

Louvain-la-Neuve, BELGIUM

Mani Ram Saharan National Geotechnical Facility

DST, Dehradun, INDIA

Abdullah Saand Quaid-e-Awam University of Eng.

Sc. & Tech., Sindh, PAKISTAN

Subhasish Das IIT- Kharagpur, Kharagpur

West Bengal, INDIA

S Viswanathan IIT- Bombay, Powai, Mumbai

Maharashtra, INDIA

Katta Venkataramana NITK- Surathkal

Karnataka, INDIA

Ramana G V IIT– Delhi, Hauz Khas

New Delhi, INDIA

Usha Natesan Centre for Water Resources

Anna University, Chennai, INDIA

K U Maheshwar Rao IIT- Kharagpur, Kharagpur

West Bengal, INDIA

Kalachand Sain National Geophysical Research Institute,

Hyderabad, INDIA

G S Dwarakish NITK- Surathkal

Karnataka, INDIA

M K Nagaraj NITK- Surathkal

Karnataka, INDIA

R Sundaravadivelu IIT- Madras

Tamil Nadu, INDIA

S M Ramasamy Gandhigram Rural University

Tamil Nadu, INDIA

M R Madhav JNTU- Kukatpally, Hyderabad

Andhra Pradesh, INDIA

Chachadi A G Goa University, Taleigao Plateau

Goa, INDIA

R Bhima Rao IMMT, Bhubaneswar

Odissa, INDIA

Gholamreza Ghodrati Amiri Iran University of Sci. & Tech.

Narmak, Tehran, IRAN

C Natarajan NIT- Tiruchirapalli,

Tamil Nadu, INDIA

N Ganesan NIT- Calicut, Kerala

Kerala, INDIA

Shamsher B. Singh BITS- Pilani, Rajasthan

Rajasthan, INDIA

Pradeep Kumar R IIIT- Gachibowli, Hyderabad


Vladimir e Vigdergauz ICEMR RAS, Moscow

RUSSIA

D P Tripathy National Institute of Technology

Rourkela, INDIA

E Saibaba Reddy JNTU- Kukatpally, Hyderabad


Chowdhury Quamruzzaman Dhaka University

Dhaka, BANGLADESH

Parekh Anant kumar B Indian Institute of Tropical

Meteorology, Pune, INDIA

Datta Shivane Central Ground Water Board

Hyderabad, INDIA

Gopal Krishan National Institute of Hydrology

Roorkee, INDIA

Karra Ram Chandar NITK- Surathkal

Karnataka, INDIA

Prasoon Kumar Singh Indian School of Mines, Dhanbad

Jharkhand, INDIA

A G S Reddy Central Ground Water Board,

Pune, Maharashtra, INDIA

Rajendra Kumar Dubey Indian School of Mines, Dhanbad

Jharkhand, INDIA

Subhasis Sen Retired Scientist

CSIR-Nagpur, INDIA

M V Ramanamurthy Geological Survey of India

Bangalore, INDIA

A Nallapa Reddy

Chief Geologist (Retd.)

ONGC Ltd., INDIA

Bijay Singh Ranchi University, Ranchi

Jharkhand, INDIA

B R Raghavan Mangalore University, Mangalore

Karnataka, INDIA

C Sivapragasam Kalasalingam University,

Tamil Nadu, INDIA

Xiang Lian Zhou ShangHai JiaoTong University

ShangHai, CHINA

K. Bheemalingeswara Mekelle University

Mekelle, ETHIOPIA

Kripamoy Sarkar Assam University

Silchar, INDIA

Anand V. Shivapur SDM College of Engg. and Tech.

Karnataka, INDIA

S Suresh Babu Adhiyamaan college of Engineering

Tamil Nadu, INDIA

Nandipati Subba Rao Andhra University, Visakhapatnam


M Suresh Gandhi University of Madras,

Tamil Nadu, INDIA

Debadatta Swain National Remote Sensing Centre

Hyderabad, INDIA

H K Sahoo Utkal University, Bhubaneswar

Odissa, INDIA

R N Tiwari Govt. P G Science College, Rewa

Madhya Pradesh, INDIA

B M Ravindra Dept. of Mines & Geology, Govt. of

Karnataka, Mangalore, INDIA

M V Ramana CSIR NIO

Goa, INDIA

N Rajeshwara Rao University of Madras

Tamil Nadu, INDIA

R Baskaran Tamil University, Thanjavur

Tamil Nadu, INDIA

Salih Muhammad Awadh College of Science

University of Baghdad, IRAQ

Sonali Pati Eastern Academy of Science and

Technology, Bhubaneswar, INDIA

Nuh Bilgin

Istanbul Technical University

Maslak, ISTANBUL

Naveed Ahmad University of Engg. & Technology,

Peshawar, PAKISTAN

Raj Reddy Kallu University of Nevada

1665 N Virginia St, RENO

Manish Kumar Tezpur University

Sonitpur, Assam, INDIA

Raju Sarkar

Delhi Technological University

Delhi, INDIA

Jaya Kumar Seelam National Institute of Oceanography Dona

Paula, Goa, INDIA

Safdar Ali Shirazi University of the Punjab,

Quaid-i-Azam Campus, PAKISTAN

C N V Satyanarayana Reddy Andhra University

Visakhapatnam, INDIA

S M Hussain University of Madras

Tamil Nadu, INDIA

Glenn T Thong Nagaland University

Meriema, Kohima, INDIA

T J Renuka Prasad Bangalore University

Karnataka, INDIA

Deva Pratap National Institute of Technology

Warangal, INDIA

Samir Kumar Bera Birbal sahni institute of palaeobotany,

Lucknow, INDIA

Mohammed Sharif Jamia University

New Delhi, INDIA

A M Vasumathi K.L.N. College of Inf. Tech.

Pottapalayam, Tamil Nadu, INDIA

Vladimir Vigdergauz ICEMR, Russian Academy of Sciences

Moscow, RUSSIA

C J Kumanan Bharathidasan University

Tamil Nadu, INDIA

B R Manjunatha Mangalore University

Karnataka, INDIA

Ranjith Pathegama Gamage Monash University, Clayton

AUSTRALIA

Ch. S. N. Murthy NITK- Surathkal

Karnataka, INDIA

K. Subramanian Coimbatore Institute of Technology

Tamil Nadu, INDIA

INDEX

Volume 07 June 2014 No.03

RESEARCH PAPERS

Study on Fresh and Hardened Properties of Concrete Incorporating Steel Slag as

Coarse Aggregate

By P S KOTHAI and R MALATHY

1024-1030

Geological Hazards in Deep Tunneling (A Case Study: Beheshtabad Water

Conveyance Tunnel)

By R. BAGHERPOUR and M.J. RAHIMDEL

1031-1040

Watershed Based Drainage Morphometric Analysis in A Part of Landslide Incidence

Areas of Coorg District, Karnataka State

By D.N. VINUTHA, S. RAMU, M.M. MUZAMIL and M.R. JANARDHANA

1041-1048

Impact of CO2 fertilization on growth and biomass in marine diatom Nitzschia sp

By RAJANANDHINI, K., P. SANTHANAM, S. JEYANTHI, A. SHENBAGA DEVI, S.

DINESH KUMAR AND B. BALAJI PRASATH

1049-1054

A Study on the Early UCC Strength of Stabilized Soil Admixed with Industrial

Waste Materials

By JIJO JAMES and P. KASINATHA PANDIAN

1055-1063

An analysis of earthquake information extraction based on GIS and RS

By ZUOWEI HUANG, WEI HUANG and ZOU YU

1064-1071

Flexural Behaviour of Stiffened Cold-Formed Steel Rectangular Hollow Sections

By PRABOWO SETIYAWAN, MOHD HANIM OSMAN, A. AZIZ SAIM, AHMAD

BAHARUDDIN ABD.RAHMAN and IMAN FARIDMEHR

1072-1081

Response and Reliability Analysis of pile foundation under Strong excitation

By YONGFENG XU, HAILONG WANG, LIQUN ZHANG and JIANLIN HU

1082-1088

Settlement algorithm research of pile foundation under load impact

By YONGFENG XU, LIQUN ZHANG, HAILONG WANG and MINFENG LI

1089-1095

Empirical Study on the Co-integration Relationship between Urban construction

land, Economic Growth and Urbanization Development of Jiangxi Province

By WEI LIU, YANG AN BAO and CHANG XIN XU

1096-1103

Discussion on Empirical Formula of Vertical Bearing Capacity of Single Press-in

Pipe Pile

By JIAFU YANG, YUZHENG LV, KUN LIU and ZHAOJUN CHU

1104-1109

Research and Experiment of Sticking Dust Isolation Curtain for Dust Reduction

By YANG XIUDONG and JIN LONGZHE

1110-1117

Predicting frost penetration of high-speed railway subgrade in seasonally frozen

regions based on empirical method

By ZHANG YU-ZHI, DU YAN-LIANG and SUN BAO-CHEN

1118-1126

Comprehensive Evaluation of the Development Quality of New-Type of

Urbanization of Jiangxi Province Based on Ecological Views

By WANG YONG XIANG

1127-1134

Simulation study of dense coherent tower sintering flue gas desulfurization system

By QIAN JIA, DONGHUI ZHANG, CUNYI SONG, ZHENSONG TONG and BAORUI

LIANG

1135-1140

Vehicle Information Compression and Transmission Methods Basis on Mixed Data

Types

By YANG JINGFENG, LI YONG, ZHANG NANFENG,HE JIARONG and XUE YUEJU

1141-1150

Numerical investigation on effect of pile tip shape on soil crushing behavior

By YANG WU

1151-1157

Estimation of Reservoir Capacity Using Remote Sensing Data – A Soft Classification

Approach

By JEYAKANTHAN V S

1158-1163

Experimental Investigation on RC and Retrofitted RC Column under Cyclic

Loading

By A.MURUGESAN AND G.S.THIRUGNANAM

1164-1170

An Experimental Study on Hybrid Fibre Reinforced Concrete

By NAZEER. M and GOURI MOHAN. L

1171-1177

Roof Failure Mechanism of Salt-solution Goaf in bedded salt deposits

By LIU JX, LIU YT, CHEN J, SHI XL, HOU JJ and CHEN XL

1178-1185

Effective Utilisation of Pond Ash as Fine Aggregate in Cement Concrete under

Flexure

By K.ARUMUGAM, R.ILANGOVAN and A.V.DEEPAN CHAKRAVARTHI

1186-1191

Variable Tap Parameter (α) Techniques for Sato based Blind Equalizer

By K SUTHENDRAN and T ARIVOLI

1192-1198

Inverse deduction of unsaturated soil parameters based on RBF neural network

model

By LIU J. X, W. LIU, Y. T. LIU and NI JUNJIE

1199-1204

Integral Action of Slab and Beam in Grid Floor Systems

By S. VIMALA and P. GOPALSAMY

1205-1209

Strength, Swelling and Durability Characteristics of Fly-Lime Stabilized Expansive

Soil-Ceramic Dust Mixes

By AKSHAYA KUMAR SABAT and BIDULA BOSE

1210-1215

Workability and Mechanical Performance of Concrete with copper slag

By M.VELUMANI and K.NIRMALKUMAR

1216-1222

Economic Comparison of Semi Continuous Mining System with Intermittent System

in Special Cycle Time

By ZHAODONG WANG

1223-1230

www.cafetinnova.org

Indexed in

Scopus Compendex and Geobase Elsevier, Chemical

Abstract Services-USA, Geo-Ref Information Services-USA,

List B of Scientific Journals, Poland,

Directory of Research Journals

ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P. 1024-1030

#02070331 Copyright ©2014 CAFET-INNOVA TECHNICAL SOCIETY. All rights reserved.

Study on Fresh and Hardened Properties of Concrete Incorporating

Steel Slag as Coarse Aggregate

P S KOTHAI1 AND R MALATHY

2

1Department of Civil Engineering, Kongu Engineering College, Perundurai, Erode-638 052, Tamilnadu, India

2Department of Civil Engineering, Sona College of Technology, Salem-636 005, Tamilnadu, India

Email: [email protected], [email protected]

Abstract: Steel slag, a by-product of steel making, is produced during the separation of the molten steel from

impurities in steel-making furnaces. The slag occurs as a molten liquid melt and is a complex solution of silicates

and oxides that solidifies upon cooling. It is estimated that 115-180Mt of steel slag is poured out annually worldwide

and in addition to this previous accumulation of the material has created mountains of steel slag. In India, Steel slag

output is approximately 20% by mass, of the crude steel output. The slag in India is used mainly in the cement

manufacture and in other unorganized works, such as, landfills and railway ballast. In order to reduce the pollution

load on landfill for the disposal of steel slag at steel industries, it can be effectively utilized in construction as

aggregates in concrete. In this research work an attempt is made to utilize the steel slag as partial replacement

material for natural aggregates in concrete. Coarse aggregate in concrete is replaced by coarse slag in M20 grade

concrete. 10% to 100% replacement in 10% increment is made and fresh concrete properties such as slump

value,density and air content were determined. Hardened concrete properties such as compressive strength, split

tensile strength, flexural strength and modulus of elasticity of the concrete with steel slag were determined.

Optimum strength of the concrete is achieved at 30% replacement proportion of natural coarse aggregates by steel

slag.

Keywords: Concrete, Aggregates, Steel slag, Replacement, Strength, Durability.

1. Introduction

Steel slag, a by-product of steel making, is produced

during the separation of the molten steel from impurities

in steel-making furnaces. The slag occurs as a molten

liquid melt and is a complex solution of silicates and

oxides that solidifies upon cooling. Slags are named

based on the furnaces from which they are generated.

Shetty (1982) have reported that aggregates are the

important constituents in concrete. They give body to

the concrete, reduce shrinkage and effect economy. The

mere fact that the aggregates occupy 70-80 percent of

the volume of concrete, their impact on various

characteristics and properties of concrete is undoubtedly

considerable. NSA (1998) reported the results of the

risk assessments which demonstrate that BF (Blast

furnace), BOF (Basic Oxygen Furnace), and EAF

(Electric Arc Furnace) slags are safe for use in a broad

variety of applications and pose no significant risks to

human health or the environment. Maslehuddin et al

(2003) made a comparison study of steel slag and

crushed limestone aggregate. Their results showed that

the durability and physical properties of concrete with

steel slag aggregates was better than limestone

aggregates. Mindness et al (2003) states that aggregate

provide dimensional stability and wear resistance for

concrete. Not only do they provide strength and

durability to concrete, but they also influence the

mechanical and physical properties of concrete.

Aggregates act as a filler material and lower the cost of

concrete. Aggregates should be hard, strong, free from

undesirable impurities and chemically stable. They

should not interfere with the cement or any of the

materials incorporated into concrete. They should be

free from impurities and organic matters which may

affect the hydration process of cement. Mindness et al

(2003) identified a wide range of materials can be used

as an alternative to natural aggregates. When any new

material is used as a concrete aggregate, three major

considerations are relevant:(1) economy, (2)

compatibility with other materials and (3) concrete

properties. Shekarchi et al (2004) carried out tests

regarding the utilization of steel slag in concrete. The

results indicated that utilization of steel slag as

aggregate is advantageous when compared with normal

aggregate mixes. Zeghichi (2006) explained that when

the slag is allowed to cool slowly, it solidifies into a

grey, crystalline, stony material, know as air cooled, or

dense slag. This forms the material used as a concrete

aggregates, it is a real silico calcareous rock, similar to

the basalt, of angular aspect, rugous and of micro

1025 P S KOTHAI AND R MALATHY

International Journal of Earth Sciences and Engineering

ISSN 0974-5904, Vol. 07, No. 03, June, 2014, pp. 1024-1030

alveolar structure. Pajgade et al (2013) reported that the

steel slag must be allowed to undergo the weathering

process before using as an aggregate in construction

because of its expansive nature. This is done in order to

reduce the quantity of free lime to acceptable limits. The

steel slag is allowed to stand in stockpiles for a period

of at least 4 months and exposed to weather. Chinnaraju

et al (2013) discussed the effect of steel slag, a by-

product from steel industry as replacement for coarse

aggregate in concrete and eco sand, which is a

commercial by-product of cement manufacturing

process. Tests on compressive strength, flexural

strength, split tensile strength at 7days and 28days, and

water absorption at 28days were conducted on

specimens. It was concluded that replacing some

percentage of coarse aggregate with steel slag enhances

the strength.

2. Materials and Methods

Steel slag: Steel slag is obtained from the Basic Oxygen

Furnace of Agni Steels Private Limited, Ingur,

TamilNadu, India and its specific gravity in coarse form

is 3.1. The slag was collected from the open stocking

yard of the industry where the slag was exposed to

atmosphere over a period of more than 1.5years. The

chemical composition of steel slag is expressed in terms

of simple oxides calculated from elemental analysis

determined by Le-Chatlier Method (IS: 228, 1987).

Table 1. lists the chemical compounds present in steel

slag from a typical basic oxygen furnace and the

chemical composition satisfies ACI 233 R-03, 2003.

Table 2. Lists the physical and mechanical properties.

Sieve analysis of coarse aggregate and coarse slag was

conducted and found that there is no significant

variation in the particle size distribution of the coarse

aggregate and coarse slag. Particle size distribution

curve is given in Figure 1. Fraction of steel slag which

passes through 20mm size sieve and retained on

4.75mm size sieve is only used. Other materials used

are given below:

Cement: Ordinary Portland cement of 43 grade

conforming to IS: 8112 – 1989 and similar to ASTM

type III (C150 – 95) is used to find the effect of steel

slag in concrete without admixtures.

Fine aggregate: Natural river sand with fraction

passing through the 4.75mm sieve and retained on

600µm sieve is used. The specific gravity of fine

aggregate is 2.62, fineness modulus is 2.6 and density is

1654kg/m3.

Coarse aggregate: Crushed granite stone aggregates of

20mm maximum size (passes through 20mm size sieve

and retained on 4.75mm size sieve) having specific

gravity of 2.70, fineness modulus of 2.73 and density of

1590kg/m3 is used.

Water: Potable tap water available in the laboratory

with pH value of 7.0±1 and confirming to the

requirements of IS: 456 - 2000 was used for mixing

concrete and also for curing the specimens.

3. Experimental Methodology

The quantity of ingredients used in the M20 concrete

mix is given in Table 3. The reference mixture (CC)

was completely prepared with natural aggregates like

granite jelly and river sand, while the other mixtures

were prepared with the slag from steel plant. For the

entire test, the designed mix was taken with the w/c

ratio 0.5.

Table 4. gives the mix designation for various mixes of

concrete for coarse aggregate replacement with coarse

form of slag. Fresh concrete properties such as slump

value, density and air content were found out as per

IS: 1199 – 1959, ASTM C 231 specifications and

ASTM C 138 guidelines respectively. Hardened

concrete properties such as compressive strength, tensile

strength, flexural strength and modulus of elasticity

were determined as per IS: 516 –1959, IS: 5816-1970,

IS: 516-1959 and California test 522 (2000) procedure

respectively. Tests were conducted for all the

replacement proportions (10% to 100% in 10%

increment) of coarse aggregate by coarse slag for M20

grade concrete.

Figure 1. Particle size distribution curve

1026 Study on Fresh and Hardened Properties of Concrete Incorporating Steel Slag

as Coarse Aggregate



Table 1.Chemical composition of steel slag

Constituent Composition

(%)

Composition (%)

as per ACI 233

R-03

CaO 32.5 32 to 45

SiO2 34 32 to 42

Fe2O3 0.3 0.1 to 0.5

MgO 9 5 to 15

Al2O3 22 7 to 16

P2O5 0.56 --

SO3 0.7 --

Table 2. Properties of Steel slag

Test particlars Results

Specific gravity 3.10

Bulk

density(kg/m3)

1650

Aggregate impact

value

13.20

Aggregate crushing

value

26.70

Water absorption

(%)

0.79

Table 3. Mix Proportion

Cement

(kg/m3)

Fine

Aggregate

(kg/m3)

Coarse

aggregate

(kg/m3)

Water-

cement

ratio

372 592.29 1146.16 0.5

Table 4. Mix designation for various replacement

proportions

S.No

Mix Coarse

aggregate

%

Coarse

slag

%

Fine

aggregate

%

1. CC 100% - 100%

2. CS1 90% 10% 100%

3. CS2 80% 20% 100%

4. CS3 70% 30% 100%

5. CS4 60% 40% 100%

6. CS5 50% 50% 100%

7. CS6 40% 60% 100%

8. CS7 30% 70% 100%

9. CS8 20% 80% 100%

10. CS9 10% 90% 100%

11. CS10 - 100% 100%

4. Results and Discussion

Fresh concrete properties such as slump value, density

and air content of the various replacement proportions

are graphically represented in the following figures.

Figure 3. Slump value

Figure 4. Density

Figure 5.Air content

Slump value of conventional concrete is 100mm and

increase in the proportion of steel slag decreases the

workability of concrete. The density values of all the

replacement proportions were higher than conventional




concrete density of 22.5kN/ m3. All the mixes show a

density value lies between 22.6kN/m3

to 22.85kN/m3.

Regarding air content value, mix CS1 shows an equal

percentage air content with conventional concrete and

all the other mixes show a decreased percentage of air

content when compared with the value of conventional

concrete which is 5.3%.

Results of the mechanical properties such as

Compressive strength, Split tensile strength, Flexural

strength and Modulus of Elasticity were graphically

represented in Figures 6, 7, 8 and 9.

Figure 6.Compressive strength

Figure 7.Split tensile strength

Figure 8. Flexural strength

Figure 9.Modulus of Elasticity

In case of M20 grade concrete, the compressive strength

of conventional concrete is 24.5MPa. For coarse

aggregate replacement by coarse slag, maximum

compressive strength is achieved in the mix CS3 and its

strength is 26.7MPa. At 28 days there is an increase in

compressive strength of 8.98% is achieved when

compared with conventional concrete. Up to 50%

replacement level, increase in compressive strength is

observed. Beyond 50% replacement proportion,

decrease in strength is observed when compared with

conventional concrete.

The split tensile strength of conventional concrete is

3.02MPa. For coarse aggregate replacement by coarse

slag, maximum split tensile strength is achieved in the

mix CS3 and its split tensile strength is 3.05MPa at

28days. At lower replacement levels of 10% and 20%, a

slight increase (0.02MPa) in split tensile strength is

observed when compared with conventional concrete.

40% replacement level shows the same split tensile

strength as that of conventional concrete but beyond that


as Coarse Aggregate



proportion split tensile strength decreases when

compared with that of conventional concrete.

For coarse aggregate replacement by coarse slag in M20

grade concrete, replacement proportions from 10% to

50% shows flexural strength value higher than that of

conventional concrete and an optimum strength is

observed in the mix CS3 at 28days. An increase of

0.08MPa is observed at 28days in the mix CS3 when

compared with conventional concrete. CS6 mix shows

strength similar to that of conventional concrete at 28

days. For replacement proportions beyond 60% the

flexural strength decreases when compared with

conventional concrete.

Up to 40% replacement level the value of modulus of

elasticity is higher than that of conventional concrete

and a maximum of 22.28GPa is observed in the mix

CS3 at 28 days curing. An increase of 0.45% is attained

in the same mix compared with conventional concrete.

Mix CS5 shows an equal value of modulus of elasticity

with conventional concrete. For replacement

proportions beyond 50% decrease in the value of

modulus of elasticity is observed when compared with

conventional concrete value.

Relationship between Compressive strength and

Flexural strength:

Figure10. Compressive strength Vs Flexural strength

The flexural strength and compressive strength test results

are plotted as a graph, and given in Figure 10. An analytical

equation for the relationship between these two was

derived from the test results as given below:

fr = 0.995√fck

Whereas for normal concrete as per IS 456 –2000, fr =

0.7√fck. It shows that the flexural strength increases in

concrete with steel slag aggregates.

Relationship between Compressive strength and

Modulus of Elasticity:




Figure11. Compressive strength Vs Modulus of Elasticity

The modulus of elasticity and compressive strength test

results are plotted as a graph, and given in Figure 11. An

analytical equation for the relationship between these two

was derived from the test results as given below:

E = 5036.33√fck

whereas for normal concrete as per IS 456 – 2000, E =

5000√fck. It shows that the modulus of elasticity slightly

increases in concrete with steel slag aggregates.

5. Conclusions

From the experimental investigations on fresh concrete

properties, following conclusions were derived:

Concrete with steel slag aggregates in M20 grade

concrete, significantly reduces the workability of

concrete. More angularity and rough texture of the steel

slag aggregates affects the workability of the concrete.

It is suggested that the workability of steel slag

aggregate can be increased by use of plasticizers or air

entraining admixtures.

The density of steel slag aggregate concrete is higher

when compared with conventional concrete. The rough

texture and angular particles of steel slag aggregates

create better interlocking between the particles and the

cement paste which helps in improving the density of

concrete.

The air content in steel slag aggregate is comparatively

lower and hence workability was reduced. But increase

in air content affects the strength of the concrete.

Strength is the basic requirement in concrete and for

improving workability one can use plasticizers or air

entraining admixtures.

From the experimental investigations on hardened

concrete properties, following conclusions were

derived:

Replacing of 30% coarse aggregate by coarse slag in

M20 grade concrete has higher compressive strength,

split tensile strength, flexural strength and modulus of

elasticity values.

Up to 50% replacement the compressive strength is

observed to be more than that of conventional concrete.

40% replacement shows equal split tensile strength

value with conventional concrete but beyond 40%

replacement the split tensile strength decreases when

compared with conventional concrete.

Up to 50% replacement, flexural strength is higher than

conventional concrete. 60% replacement shows equal

value with conventional concrete and beyond that

proportion flexural strength decreases when compared

with conventional concrete.

Similar to that of flexural strength, when compared with

conventional concrete modulus of elasticity values also

increases up to 50%, equals at 60% and decreases

beyond that proportion.

For further improvement in strength water cement ratio

may be reduced by adding chemical admixtures like

plasticizers. Strength can also be improved by adding

mineral admixtures.

The relationship between the mechanical properties of

the concrete indicates that the strength of the steel slag


as Coarse Aggregate



aggregates concrete is comparatively higher than the

normal concrete strength specified by IS 456 – 2000.

Hence it is recommended that, steel slag aggregates can

be used as a replacement material for conventional

coarse aggregates in concrete in order to reduce the

exploitation of natural aggregates and to reduce the cost

of construction. Also industrial waste material

utilization in construction helps in reducing pollution

and achieving sustainability.

References

[1] Chinnaraju, K.., Ramkumar, VR., Lineesh, K.,

Nithya, S. and Sathish, V. 2013.Study on concrete

using steel slag as coarse aggregate replacement

and eco sand as fine aggregate replacement.

International Journal of Research in Engineering

and Advanced Technology. vol. 1, no.3, 1-6.

[2] Maslehuddin, M., Alfarabi Sharif, M., Shameem,

M., Ibrahim, M. and Barry, MS. 2003.Comparison

of Properties of Steel Slag and Crushed Limestone

Aggregates Concrete. Construction and Building

materials.Vol,17,105-112.

[3] Mindness, S., Young, J.F. and Darwin,D. 2003.

Concrete second edition. Pearson Education Inc.

[4] National Slag Association. 2003. Iron and Steel

making Slag Environmentally Responsible

Construction Aggregates. NSA Technical Bulletin.

[5] Pajgade,P.S. and Thakur.N.B. 2013. Utilisation of

Waste Product of Steel Industry. International

Journal of Engineering Research and Applications

(IJERA). Vol. 3, Issue 1, 2033-2041.

[6] Rustu S.Kalyoncu. 2001. Slag iron and Steel.US

Geological Survey Minerals Yearbook.

[7] Shekarchi, M., Soltani, M., Alizadeh, R., Chini ,

M., Ghods, P., Hoseini, M. and Montazer, Sh.

2004. Study of the mechanical properties of heavy

weight preplaced aggregate concrete using electric

arc furnace slag as aggregate. International

Conference on Concrete Engineering and

Technology, Malaysia.

[8] Shetty, M. S.1982. Concrete Technology Theory

and Practice. S.Chand and Company Ltd.

[9] Zeghichi, L. 2006. The effect of replacement of

naturals aggregates by Slag products on the

strength of concrete. Asian Journal of Civil

Engineering (Building and Housing). Vol. 7, 27-

35.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1031-1040


Geological Hazards in Deep Tunneling (A Case Study: Beheshtabad

Water Conveyance Tunnel)

R. BAGHERPOUR1 AND M. J. RAHIMDEL

2

1Department of Mining Engineering, Isfahan University of Technology, Isfahan 8415683111, Iran 2Department of Mining Engineering, Sahand University of Technology, Tabriz 513351996, Iran


Abstract: Knowledge of geology conditions and its hazards can play an important role in the selection of support

and suitable excavation method in underground structures. Water transport tunnel is one of the most important

structures with regard to the goal of excavation, special conditions and limitations considered in the design and

execution of them. Beheshtabad water Conveyance tunnel with 64930 meters length, 6 meters final diameter is the

largest water Conveyance tunnel in IRAN. Because of high over burden and weak rock in the most of tunnel path,

the probable geological hazards such as squeezing and rock burst must be studied. Squeezing stands for large time-

dependent convergence during tunnel excavation. This phenomenon occurs in weak rocks and deep conditions.

Besides, the height of overburden in some zone of the tunnel is about 1200 meters. The occurrence of this

phenomenon is always together with the instantaneous release of strain energy stored in the rock materials, causing

the harm to personal equipment and the collapse of underground structures. The existence of high thickness

overburden in some zones of this project indicates the high potential of rock burst hazard. In this research, the length

of the tunnel has been partitioned into sections using the interpreted geological, geophysical studies and borehole

data. After evaluating rock burst and squeezing potential with alternative analytical and experimental methods for

each section, the results of different methods are compared with each other. Results predict low to moderate

squeezing potential and moderate to high rock burst potential for some panels of the tunnel.

Keywords: Central plateau of Iran, geological hazards, rock burst, squeezing, Beheshtabad Water Conveyance

Tunnel.

1. Introduction:

Tunnels are one of the vital arteries that, because of a lot

of expenses spent for introduction of them and also

derangement of passing traffic as a result of perfect

demolition or serious damages, need the observation of

technical geotechnical considerations in design and

performance. Zayandehrud River is the only permanent

river in the Central Plateau of Iran. Water demand in

this area is constantly growing due to population

growth, key industries, withdrawal of ground water

tables and reduction of its quality. So, Beheshtabad

tunnel, by transporting 1070 millions of cube meters of

water per year to Iran central plateau, is considered in

order to eliminate the shortages in parts of drinking

water, industry and agriculture. This plan, consisting of

a dam with 184 meters height and water transport tunnel

with the length of about 65 km and 6 meters diameter, is

expected to be the longest water transport tunnel in

IRAN.

In this research, at the First, the tunnel was paneled

using the interpretation of geological, geophysical

studies and boreholes. Then, the squeezing and rock

burst potential was studied using empirical and

analytical methods for each panel. Finally, the results

were compared with each other.

1.1. Literature Review:

The Rock burst and squeezing are two main modes of

underground instability caused by overstressing of the

ground. Both modes are generally related to continuous

ground. Squeezing can occur both in massive (weak and

deformable) rocks and in highly jointed rock masses as

a result of overstressing. It is characterized by yielding

under the redistributed state of stress during and after

excavation [1]. The squeezing can be very large;

deformations as much as l7% of the tunnel diameter

have been reported in India [2]. According to the

unexpected geotechnical hazards during tunneling,

Singh et al., Goel et al., Jethwa et al., Hoek and Marinos

have studied the squeezing phenomenon for deep

tunnels in weak rocks and derived some criteria to

recognize it [2, 3, 4, 5, 6].

In most criteria, the overburden load plays an important

role in developing the squeezing conditions.

Furthermore, when an excavation for a deep

mailto:[email protected]


1032 Geological Hazards in Deep Tunneling (A Case Study: Beheshtabad Water

Conveyance Tunnel)



underground tunnel or chamber is undertaken in a

strong and brittle rock, the change in stress results in

dynamic damage to the adjacent rock. This is referred to

as rockburst or break ways. Such rock bursts are a major

hazard for the safety of engineers and engineering

equipment, as well as affecting the shape/size of the

structure [7]. Hoek and Brown, Myrvang and Grimstad,

Hatcher, Haramy, Qiao and Tian, Wang and Park and

Amberg have been working to identify rock burst in

deep tunnels with brittle rocks [8, 9, 10, 11, 12, 13, 14].

2. Beheshtabad Water Conveyance Tunnel:

Beheshtabad Water Conveyance Tunnel, with about 65

kilometer length and 6 meter width, is one of the biggest

water supplying projects for transporting water to the

central plateau of Iran. This tunnel is located near Ardal

city with east north-west south direction. From the

entrance to 17 km of the tunnel, it is located in Zagros

zone and the output of it is in Sanandaj-Sirjan zone.

This tunnel is expected to transfer water to resolve

water deficiencies and shortcomings in industrial and

agriculture use in the central plateau of Iran, 1070 cubic

million meters annually [15].

Most important problems in path of this tunnel refers to

its cross to numerous fractures, resulting in many

problems and troubles during drilling and in the stages

of maintenance coverage of tunnel.

With regard to 19 boreholes in the tunnel path, tunnel

has been paneled to 16 sections. Engineering geological

properties for each panel are summarized in Table 1.

The rock engineering classification is shown in Table 2

[16].

Table1: Rock engineering geological characteristics for each tunnel section [16]

Secti

on Kilometer (m) Rock mass

Overburden

(m)

Density

(gr/cm3)

UCS

(MPa) RQD

I 5941-7800 Limestone with dolomite 600 2.530 65-75 95-100

II 7800-8116 Marl stone 781.58 2.968 20-40 95-100

III 8116-10790 Lime stone and Marl stone 1205.5 2.509 65-75 95-100

IV 10790-12129 Marl stone and conglomerate 340 2.488 70-90 95-100

V 12129-15492 Mud stone and conglomerate 294 2.450 30-45 95-100

VI 15492-17574 Weathered and altered andesitic 285 2.491 20-30 50-60

VII 17574-18013 Crushed limestone and Marly limestone 327 2.651 20-40 40-50

VIII 18013-20862 Marly and shale limestone 349 2.464 20-30 50-85

IX 20862-21730 Marl and Shale 477 2.733 25-35 85-90

X 21730-24174 Marl and Shale 621 2.646 20-40 85-90

XI 24174-29030 Alteration of massive limestone 654.45 2.646 40-50 75-85

XII 29030-31604 Shaly limestone 381 2.651 25-60 25-60

XIII 31604-34912 Melonitic limy sand stone with quarts

lenses 335.6 2.667 10-30 25-45

XIV 34912-37490 Melonitic limy sand stone with quarts

lenses 481 2.690 25-50 25-50

XV 37490-37892 Limestone and dolomite 571 2.690 50-80 90-100

Table2: Rock engineering classification of the studied tunnel [16]

Tunnel Section RMR Q

Value Rating Value Rating

I 54-55 Fair 1.65-2.67 Poor

II 60-64 Good 1.35-4 Poor

III 53-60 Fair 1.1-2 Poor

III 57-60 Fair 1.35-3 Poor

IV 50-71 Fair 2.4-13.3 Poor-Fair

V 56-61 Fair 2.3-9 Poor-Fair

VI 58-69 Good 3.92-9 Fair

1033 R. BAGHERPOUR AND M. J. RAHIMDEL



VII 55-60 Fair 3.4-9 Poor-Fair

VIII 57-59 Fair 4.3-9 Fair

IX 19-21 Poor 0.006-0.015 Exceptionally Poor-

Extremely poor

X 23-28 Poor 0.006-0.02 Exceptionally Poor-

Extremely poor

XI 18-20 Poor 0.37-6 Fair

XII 50-64 Fair 2.1-6 Poor-Fair

XIII 50-57 Fair 0.95-2 Poor

XIV 49-59 Fair 1.1-3 Poor

XV 30-35 Poor 0.2-0.4 Poor

From Table 1, it can be seen that the classification

grading by Q system is lower than that by the RMR for

the same type rock. That is because Q system takes the

high stress field into consideration, and to some extent,

it causes rock mass instability.

Regarding researches in the studied area, stability

analysis and leakage quantity investigation have been

conducted. Rahimdel and et al. proposed the primary

support for tunnel section based on geology section and

rock masses of tunnel using RMR, Q and VNIMI

methods. The results based on VNIMI method are given

in Table 3 [17].

Table3: Primary support estimation for tunnel rock

masses

Rock Mass Primary Support

Limestone with dolomite,

marl stone, mud stone and

conglomerate

Using rock bolt or

shotcrete lining by 5 cm in

Thickness.

Crushed limestone and

marly limestone, Marly and

shale limestone and Shaly

limestone

Application of rock bolt

2.5 m in length with 1×1

distance together and

shotcrete lining by 5 cm or

more in Thickness with

mesh and rock bolt

Rafiee and et al. [15] used the Fuzzy Analytical

Hierarchy Process (FAHP) to support the estimation of

tunnel. In this study, regarding the numerical analysis

(finite difference program FLAC2D), six support

systems were considered as the decision alternative are

shown in Table 4 and support cost, factor of safety,

applicability, time, displacement and mechanization

were considered as the criteria. Calculations showed

that the alternative "E" should be selected as the

optimum support system to satisfy the goals and

objectives of Behashtabad Tunnel.

Table4: Explanation of Model Notations [15]

Support

system

(Alternative)

Explanation

A Supporting by shotcrete lining by 25 cm

in thickness together with IPE18

B Supporting by shotcrete lining by 30 cm

in thickness together with IPE16

C Supporting by shotcrete lining by 20 cm

in thickness together with wire mesh

D

This system is the combination of

shotcrete with steel fiber by 20 cm in

thickness

E

Application of rock bolt 3 m in length

with 1×1 distance together with shotcrete

lining by 10 cm in thickness

F

Application of rock bolt 3 m in length

with 2×2 distance together with shotcrete

lining by 20 cm in thickness

3. Squeezing:

The magnitude of tunnel convergence, the rate of

deformation and the extent of the yielding zone around

the tunnel depend on the geological and geotechnical

conditions, the in-situ state of stress relative to rock

mass strength, the groundwater flow and pore pressure,

and the rock mass properties [18]. Squeezing is,

therefore, synonymous with yielding and time-

dependence; its cost depends on the excavation and

support techniques adopted. If the support installation is

delayed, the rock mass moves into the tunnel and stress

redistribution take place around it. On the contrary, if

deformation is restrained, squeezing will lead to long-

term load build-up of rock support.

For the evaluation of the potential of squeezing,

empirical and semi-empirical methods have been

introduced via deferent researchers. These methods are

explained below.

3.1. Prediction of Squeezing:

3.1.1. Empirical approaches:

The empirical approaches are essentially based on

classification schemes. Two of these approaches are


Conveyance Tunnel)



mentioned below in order to illustrate the uncertainty

still surrounding the subject, notwithstanding its

importance in tunneling practice.

Singh et al. (1992) approach:

This method, which is based on the results of 39 case

histories, by collecting data on rock mass quality Q,

overburden and height, proposes that Squeezing

potential is predictable by using Equation 1 and Table

5 [2].

H=350Q1/3 (1)

Table5: Classification of squeezing behavior

according to Singh et al. (1992)

H Type of behavior

>350Q1/3

Squeezing conditions

<350Q1/3

Non squeezing conditions

Goel et al., (1995) approach:

A simple empirical approach developed by Goel et al.

(1995) is based on the rock mass number N, which is

defined as stress-free Q as follows [3].

N = (Q)SRF = 1 (2)

This is used to avoid the problems and uncertainties

in obtaining the correct rating of parameter SRF in

Barton et al. (1974) Q. By considering the tunnel

depth H, the tunnel span or diameter B, and the rock

mass number N from 99 tunnel sections, Goel et al.

(1995) plotted the available data on a log-log diagram

(Figure 1) between N and H×B0.1

.

Figure1: Goel et al’s approach for predicting squeezing

conditions [3]

3.1.2. Semi-Empirical Approaches:

The common starting point of all these methods for

quantifying the squeezing potential of rock is the use

of the “competency factor”, which is defined as the

ratio of uniaxial compressive strength σc/σcm of

rock/rock mass to overburden stress γH. Two of such

methods are briefly discussed below.

Jethwa et al. approach (1984):

As mentioned above, the degree of squeezing is

defined by Jethwa et al. [4] on the basis of Equation

(3) and Table6:

Nc= σcm/P0= σcm/H (3)

where σcm is rock mass uniaxial compressive strength,

P0 is in situ stress and H is tunnel depth below

surface.

Table6: Classification of squeezing behavior according

to Jethwa et al. (1984)

NC Type of behavior

0.4> Highly squeezing

0.4-0.8 Moderately squeezing

0.8-2 Mildly squeezing

>2 Non squeezing

Aydan et al. approach (1993):

Aydan et al. [19], based on the experience with

tunnels in Japan, proposed to relate the strength of the

intact rock σci to the overburden pressure γH by the

same relation as (3), implying that the uniaxial

compressive strength of the intact rock σci and that of

the rock mass σcm are the same. The fundamental

concept of the method is based on the analogy

between the stress-strain response of rock in

laboratory testing and tangential stress-strain response

around tunnels. As illustrated in Figure 2, five distinct

states of the specimen during loading are experienced,

at low confining stress σ3 (i.e., σ3 ≤ 0.1σci) The

following relations, as defined, give the normalized

strain levels ηP, ηs and ηf.

ηP=εP/εe=2σci-0.17, ηs=εs/εe=3σci-0.25,

ηf=εf/εe=5σci-0.32 (4)

where εP, εs and εf are the strain values shown in

Figure 2, as εe is the elastic strain limit.

Figure2: Idealized stress-strain curve and the

associated states for squeezing rocks

Based on a closed form analytical solution, which has

been developed for computing the strain level εϴa




around a circular tunnel in a hydrostatic stress field,

the five different degrees of squeezing are defined as

shown in Table 7. In this Table, εϴa is the tangential

strain around a circular tunnel in a hydrostatic stress

field [19], whereas εϴe is the elastic strain limit for the

rock mass.


to Aydan et al. (1993)

Theoretical expression Squeezing degree

εϴa / εϴ

e ≤1 Non-squeezing

1≤ εϴa / εϴ

e ≤ ηP Light-squeezing

ηP ≤ εϴa / εϴ

e ≤ ηs Fair-squeezing

ηs ≤ εϴa / εϴ

e ≤ ηf Heavy-squeezing

εϴa / εϴ

e ≥ ηf Very heavy squeezing

3.1.3. Analytical-Theoretical Approaches:

Barla and International Society of Rock Mechanics

(ISRM) approaches:

The squeezing potential in these methods can be

expected in accordance with Table 8 by considering

the values of tangential stress (σϴ), uniaxial

compressive strength (σcm) and the maximum stress

(σ1).


to Barla and ISRM approaches

Evaluation Method

Squeezing degree ISRM

(σθ/σcm)

Barla

(σcm/σ1)

<1 >1 Non-squeezing

1-2 1-0.4 Light-squeezing

2-4 0.4-0.2 Fair-squeezing

>4 0.2> Heavy-squeezing

3.2. Evaluation of squeezing potential in Beheshtabab

Water Conveyance Tunnel:

The results of assessing squeezing potential for the zone

of the tunnel in which there was the occurrence of this

phenomenon using different criteria have been shown in

Figure 3. To study the different criteria results, the

percentage of each category of the squeeze zones

studied was calculated as shown in Table 9. In average,

71, 20, 5 and 4 percent of total panels were in none,

light, moderate and heavy squeezing conditions,

respectively. So, most section of the tunnel was in none

squeezing potential.


Conveyance Tunnel)



Figure3: The results of the squeezing potential using Singh (A), Goel (B), Jethwa (C), Aydan (D), Barla (E) and

ISRM (F) criteria.

Table9: The results of the squeezing potential in Beheshtabad Water Conveyance Tunnel

Percentage of tunnel sections in each squeezing condition Evaluation

criteria Non-squeezing Light-squeezing Fair-squeezing Heavy-

squeezing Very heavy-

squeezing

61 39 0 0 0 Singh

66 0 17 17 0 Goel

72 28 0 0 0 Jethwa

72 0 11 17 0 Aydan

78 22 0 0 0 Barla

72 28 0 0 0 ISRM

4. Rock Burst:

A rock burst is one of the most complicated dynamic

geological phenomena, with intricate mechanisms and

numerous affecting factors, which account for the

difficulty of predicting its characteristics. In the past

few years, many methods of forecasting rock bursts

have been proposed, including rock mechanics

assessment, stress detection and modern mathematical

theories.

The prevention of rock bursts is one of the key problems

in the construction of deep tunnels in which rock burst

prediction is a basic problem. In the construction of

underground engineering, it is of great importance for

the safety and the optimization of support measures to

make correct and timely predictions of the possibility,

as well as the scope and intensity of rock bursts in the

rock mass surrounding the ground to be excavated.

4.1. Rock Burst Prediction:

Regarding available and valid references,

comprehensive researches have carried out done in

classification and evaluation of rock burst phenomenon.

In most of them, Linear elastic criterion, Method of

Tensile Stress, Method of Brittleness coefficient and

Method of Stresses have been used for rock burst

prediction [7, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30,

31, 32, 33, 34].

Linear Elastic Criterion:

Linear elastic energy (LE) stored in rock before

reaching the peak strength can be defined by the

following equation.

2E

2cσEL (5)

Where E is unloading tangent elastic modulus of rock,

and σc is uniaxial compressive strength. Rock burst

potential is predictable by using Table 10.

Table10: Classification of Rock burst behavior

according to linear elastic criterion

50> 50-

100 100-150

150-

200 200< LE

(MPa)

Very

Low Low Moderate High

Very

High

Rock

bust

potential

Method of Tensile Stress:

Rock burst predictions using this method can be defined

by Equation (6). Rock burst potential is predictable by

using Table 11.

cσθσ

sT (6)

where σɵ: Tensile stress, σc: Uniaxial compressive

strength.





according to the Method of Tensile Stress

0.3> 0.3-

0.5 0.5-0.7

0.7-

0.9 0.9< TS

Non-

Rock

Burst

Low Moderate High Very

High

Rock

bust

potential

Method of Brittleness Coefficient:

This method evalutes the tendency of rock burst through

the brittleness coefficient R of Rocks. This coefficient is

defined as the ratio of σc over σt (σc and σt are the

uniaxial compressive strength and the tensile strength of

the rock), i.e., β=σc/σt. In general, the grater β, the

higher the rock burst tendency (see Table 12).


according to the method of brittleness coefficient

40< 40-26.7 26.7-14.5 14.5> Β

Non-

rock

burst

Low Moderate High

Rock

bust

potential

Method of Stresses:

Method of stresses combines the lithological character

of a rock mass (including tensile and compressive

strength) to judge the possibility that rock burst can take

place. This method introduces two factors of and β to

serve as criteria. and β are defined, respectively, as

the ratio of the rocks uniaxial compressive strength, σc,

over the major principle geostress, σ1, i.e., = σc/σ1 and

as the ratio of the rocks uniaxial tensile strength, σt, over

σ1, i.e., = σt/σ1. Because the index of the uniaxial

compressive can be determined easily, the value of is

generally used for a criterion having the following

Table.


according to the Method of Stresses

10< 10-5 5-2.5 2.5>

Non- rock

burst Low Moderate High

Rock

bust

potential

4.2. Evaluation of rock burst potential in Beheshtabad

Water Conveyance Tunnel:

The results of the assessing rock burst potential for the

zone of the tunnel in which the occurrence of this

phenomenon was achieved using different criteria, as

shown in Figure 4. To study the Different criteria

results, the percentage of each category of rock burst

zones studied was calculated as shown in Table 14.

Figure4: The results of the rock burst potential using the method of stresses (A), linear elastic criterion (B),

brittleness coefficient (C) and tensile stress (D) criteria.


Conveyance Tunnel)



Table14: The results of the rock burst potential

Percentage of tunnel sections in each of rock burst conditions

Evaluation criteria Non-

rock burst Light- rock

burst Fair-

rock burst Heavy- rock

burst Very heavy-

rock burst

40 24 20 12 4 Linear elastic criterion

16 12 40 16 16 Tensile Stress

8 8 60 20 0 Stresses

0 0 0 100 0 Brittleness coefficient

Regarding Table 14, Linear elastic criterion predicts no

rock burst potential for more sections of the tunnel,

while Tensile Stress and Stresses methods assume the

major sections of tunnel to be in the fair rock burst

potential. According to brittleness coefficient, all tunnel

sections are unfortunately in heavy rock burst condition.

In average, 30, 12, 34, 17 and 7 percent of total panels

are in none, light, moderate, heavy and very heavy rock

burst conditions, respectively. So, most section of tunnel

is in none to fair rock burst condition. To have a better

comparison, the results have been shown in Figure 5.

Regarding Figure 5, more of sections are in high

squeezing potential condition. So, in this tunnel, the

squeezing potential is more important than the rock

burst. These results are in agreement with high

overburden and weak sedimentary rock masses in these

sections.

Figure5: Comparison of the squeezing and rock burst

potential results

5. Discussion and Conclusion:

Squeezing and rock burst potential was addressed in this

article using different empirical, semi-empirical and

analytical approaches. The results showed that empirical

and analytical methods were almost accommodated with

each other. In squeezing potential research, according to

Singh, Jethwa, Barla and ISRM approaches, a great

number of tunnel sections fell into non-squeezing

potential category. Aydan and Goel criteria, similar to

the recently mentioned approaches, have predicted

moderate to heavy squeezing potential for a small

percentage of sections. Based on our researches, the

results showed that 71, 20, 5 and 4 percent of total

panels were in none, light, moderate and heavy

squeezing conditions, respectively. Thus, the rock

masses in this tunnel path were in none to light

squeezing potential. In rock burst potential research,

according to forbear linear elastic criterion that

predicted moderate rock burst potential for all sections,

according to other Methods of Tensile Stress, Tensile

Stress and Method of Stresses, 30, 12, 34, 17 and 7

percent of total panels were in none, light, moderate,

heavy and very heavy rock burst conditions

noticeability. So, the rock masses in this tunnel path

were in none to moderate rock burst potential.

According to the precise prediction of these phenomena,

it is not possible to have a safe environment during the

deep exploration and mining. So some necessary

prevention measures are proposed.

1) The construction methods can be improved. The

impact of blasting vibration should be minimized as

far as possible to avoid bringing about various

factors inducing rock burst.

2) Rock can be strengthened by grouting to change the

mechanical properties of wall rock. Grouting bolt

nets and plastic bolts can also be applied to the

underground chamber or wall rock.

3) In very poor squeezing conditions, using heavy

support and monitoring the displacements of roof

and bottom of tunnel and in moderate to high

squeezing conditions, using flexible support are

essential.

6. Reference:

[1] A. Palmström, Characterizing rock burst and

squeezing by the rock mass index, Design and

Construction of underground Structures, New

Delhi, 1995, 23-25 February.

[2] B, Singh, J.L. Jethwa, A.K. Dube, B. Singh,

Correlation between observed support pressure and

rock mass quality, Tunnelling and Underground

Space Technology, 7, 1992, p. 59- 74.

[3] R.K. Goel, J.L. Jethwa, A.G. Paithakan, Tunnelling

through the young Himalayas – a case history of the

Maneri-Uttarkashi power tunnel. Engrg. Geol., 39,

1995, p. 31-44.




[4] J.L. Jethwa, B. Singh, Estimation of ultimate rock

pressure for tunnel linings under squeezing rock

conditions – a new approach. Design and

Performance of Underground Excavations, ISRM

Symposium, Cambridge, E.T. Brown and

J.A.Hudson eds., 1984, p. 231-238.

[5] E. Hoek, P. Marinos, Predicting Tunnel Squeezing

Problems in Weak Heterogeneous Rock Masses,

Tunnels and Tunneling International, 45 -51: part

one, 2000a, p. 33-36: Part 2.

[6] E. Hoek, P. Marinos, Predicting tunnel squeezing

problems in weak heterogeneous rock masses,

Tunnels and Tunnelling International Part two

(December), 2000b, p. 33–36.

[7] Q. Jiang, X.T. Feng, T.B. Xiang, G.S. Su, Rock

burst characteristics and numerical simulation

based on a new energy index: a case study of a

tunnel at 2500 m depth, Bulletin of Engineering

Geology and the Environment, 69, 2010, p. 381–

388.

[8] E. Hoek, E.T. Brown, Underground excavations in

rock, Institution of mining and metallurgy, London

1980, p. 527.

[9] A. Myrvang, E. Grimstad, Rock burst Problems in

Norwegian Highway Tunnels- Recent Case

Histories, Rock burst Prediction and Control. The

Institution of Mining and Metallurgy Oct., 1983, p.

133-139.

[10] R.D. Hatcher, Structural geology– principles,

concepts, and problems, Prentice Hall, Englewood

Cliffs, second edition, New Jersey, 1985.

[11] K.Y. Haramy, Underground strucures in rock bursts

zones – In: Underground structures, design and

instrumentation, Elsevier Science publishers B.V.,

Amsterdom, 1998.

[12] C.S. Qiao, Z.Y. Tian, Study of the possibility of

rockburst in Dong-gua-shan Copper Mine, Chinese

J. Rock Mech. Eng. Žexp. 17, 1998, p. 917-921.

[13] J.K. Wang, H.D. Park, Comprehensive Prediction

of Rock Burst Based on Analysis of Strain Energy

in Rock, Tunneling and Underground Space

Technology, 2001, p. 7-49.

[14] F. Amberg, Tunneling in high overburden with

reference to deep tunnels in Switzerland, Tunneling

and Underground Space Technology (19), 2004.

[15] R. Rafiee, M. Ataei, S.M.E. Jalali, The Optimum

Support Selection by Using Fuzzy Analytical

Hierarchy Prosess Method for Beheshtabad Water

Transporting Tunnel in Naien, Iranian Journal of

Fuzzy Systems, 10(6), 2013, p. 39-51.

[16] M. Hashemi, Rock mechanics reports, Water

Supply Project of the Central Plateau, Zayan-dehab

Consulting, 2007.

[17] M.J. Rahimdel, S. Mahdevari, R. Bagherpour,

Stability Assessments and Support Estimation of

Beheshtabad Water Transport Tunnel by VNIMI

Method, The 1st Asian and 9th Iranian Tunneling

Symposium, 2011.

[18] G. Barla, Tunnelling under Squeezing Rock

Conditions, 2002, p. 1-75.

[19] Ö. Aydan, T. Akagi, T. Kawamoto, The squeezing

potential of rock around tunnels: theory and

prediction, Journal of Rock Mechanics and Rock

Engineering, 26(4), 1993, p. l37-163.

[20] N.G.W. Cook, E. Hoek, J.P.G. Pretorius, W.D.

Ortlepp, M.D.G. Alamon, Rock mechanics applied

to the study of rock bursts, Journal of the South

African Institute of Mining and Metallurgy, 66,

1996, p. 435–528.

[21] U. Casten, Z. Fajklewicz, Induced gravity

anomalies and rockburst risk in coal mines – a case

history, Geophysical Prospecting, 41(1), 1993, p.

1–13.

[22] K. Shivakumar, M.V.M.S. Rao, C. Srinivasan, K.

Kusunose, Multifractal analysis of the spatial

distribution of area rock bursts at Kolar Gold

Mines, International Journal of Rock Mechanics &

Mining Sciences, 33(2), 1996, p.167–172.

[23] A.M. Linkov, Rockburst and the instability of rock

masses, International Journal of Rock Mechanics

and Mining Science, 33, 1996, p. 727–732.

[24] S.K. Sharan, A finite element perturbation method

for the prediction of Rock burst, Computers and

Structures, 85, 2004, p. 1304–1309.

[25] H.B. Zhao, Classification of rock burst using

support vector machine, Rock and Soil Mechanics

26(4), 2005, p. 642–644 (in Chinese).

[26] F.Q. Gong, X.B. Li, A distance discriminant

analysis method for prediction of possibility and

classification of rockburst and its application,

Chinese Journal of Rock Mechanics and

Engineering, 26(5), 2007, p. 1012–1018 (in

Chinese).

[27] W.C. Zhu, Z.H. Li, L. Zhu, C.A. Tang, Numerical

simulation on rock burst of underground opening

triggered by dynamic disturbance, Tunneling and

Underground Space Technology, 25, 2010, p. 587–

599.

[28] Y.C. Wang, Y.Q. Shang, H.Y. Sun, X.S. Yan,

Study of prediction of rock burst intensity based on

efficacy coefficient method, Rock and Soil

Mechanics, 31(2), 2010, p. 529–534 (in Chinese).

[29] J.L. Yang, X.B. Li, Z.L. Zhou, Y. Lin, A Fuzzy

assessment method of rock-burst prediction based

on rough set theory, Metal Mine, 6, 2010, p. 26–29

(in Chinese).

[30] J. Zhou, X.Z. Shi, L. Dong, H.Y. Hu, H.Y. Wang,

Fisher discriminant analysis model and its

application for prediction of classification of rock

burst in Deepburied Long Tunnel, Journal of Coal

Science and Engineering (China), 16 (2), 2010, p.

144–149.


Conveyance Tunnel)



[31] H. Hu, D. Linming, L. Xuwei, Q. Qiuqiu, C.

Tongjun, G. Siyuan, Active velocity tomography

for assessing rock burst hazards in a kilometer deep

mine, Mining Science and Technology (China), 21,

2011, p. 673–676.

[32] C. Xuehua, L. Weiqing, Y. Xianyang, Analysis on

rock burst danger when fully mechanized caving

coal face passed fault with deep mining, Safety

Science, 50, 2012, p. 645–648.

[33] F. Jun, D. Linming, H. Hua, D. Taotao, Z. Shibin,

G. Bing, S. Xinglin, Directional hydraulic

fracturing to control hard-roof rock burst in coal

mines, International Journal of Mining Science and

Technology, 22, 2012, p. 177–181.

[34] N. Baisheng, L. Xiangchun, Mechanism research

on coal and gas outburst during vibration blasting,

Safety Science, 50, 2012, p. 741–744.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1041-1048


Watershed Based Drainage Morphometric Analysis in a Part of

Landslide Incidence Areas of Coorg District, Karnataka State

D N VINUTHA1, S RAMU

2, MOHAMMAD MUZAMIL AHMAD

3 AND M R JANARDHANA

1

1Department of Earth Science & Resource Management, Yuvaraja’s College, University of Mysore,

Mysore – 570005, Karnataka, India 2Department of Civil Engineering, KVG College of Engineering, Sullaya, Dakshina Kannada, India

3Department of Geology, University BDT College of Engineering, Davangere, Karnataka, India

Email: [email protected]

Abstract: Quantitative morphometric analysis of a watershed provides a description of the drainage system and

involves the quantification of the channel network and related parameters such as drainage area, gradient and relief.

These parameters in addition to the geology and the degree of weathering of rocks have a bearing on the occurrence

of landslides in the hilly terrains. The present study deals with the morphometric characteristics of the sub-

watersheds of Harangi watershed which is a part of Cauvery river mega basin in Coorg district of Karnataka State.

Morphometric characteristics comprising linear, areal and relief aspects of Harangi watershed which has been

subdivided into 10 sub-watersheds, were evaluated based on the Survey of India toposheets on 1:50000 scale,

orthorectified Landsat MSS and ETM imageries. Digital Elevation Model (DEM) was prepared by using ASTER

data and Geographical Information System (GIS) tool was used in the evaluation of linear, areal and relief aspects

and in the preparation of thematic maps. Landslide incidences are restricted to sub-watersheds 7 and 9 located in the

southwestern part of the Harangi basin. These sub-watersheds are located in highly weathered granitic gneiss region

with thick soil cover and are characterized by relatively low relief ratio and drainage density, stream frequency,

texture ratio, form factor, circulatory ratio and correlates well with the amount of vegetation and water absorption

capacity of the soil of the region.

Key words: Morphometry, Landslide, Harangi watershed, GIS, Western Ghat.

1. Introduction:

Morphometry is the measurement and mathematical

analysis of the configuration of the earth’s surface,

shape and dimension of its landforms (Clarke, 1966).

Morphometric analysis of a watershed provides a

quantitative description of the drainage system, which is

an important aspect of the characterization of

watersheds (Strahler, 1964).The close relationship

between hydrology and geomorphology play an

important role in the drainage morphometric analysis

(Horton, 1932). The influence of drainage morphometry

is very significant in understanding the landform

processes, soil physical properties and erosional

characteristics (Iqbal et al. 2012), which are all

fundamental features in the assessment of the possible

occurrence of landslides in any given area. Watershed is

taken as a basic unit in morphometric analysis as all the

hydrologic and geomorphic processes can be evaluated

within a finite area. Morphometric analysis can be

achieved through measurement of linear, areal and relief

aspects of the basin and slope contribution (Nag and

Chakraborty, 2003).

Harangi watershed region is witnessing a number of

landslides during and soon after heavy rain falls

disrupting normal life. The area is also known for

intense human activities related to plantation. The

region is hilly with varied geomorphologic features and

occurrence of landslides during monsoon season

warrants study of drainage morphometry vis-à-vis

landslide occurrences. However, the present authors

have restricted their studies to evaluate basin

morphometric characteristics encompassing relevant

linear, area and relief aspects of the Harangi watershed,

which is a part of Cauvery River in the SW part of

Karnataka state. The study serves as a platform to carry

out the natural hazard investigations in the region.

1.1. Study Area:

The study area lies in the Western Ghats encompassing

Coorg district and NW part is in South Kanara district.

The study area is bounded by latitudes 1406590.32 to

1372348.92 N and longitudes 610716.58 to 569484.17

East (UTM). It forms a part of Survey of India

toposheets 48P/10, P/11, P/14 and P/15 and covers an

area of about 664.76 Km2. The area under consideration


1042 Watershed Based Drainage Morphometric Analysis in a Part of Landslide Incidence




is known for picturesque hills and has a general

northwest-southeast trend. The location map of the

study area is depicted in Fig1. The geological

formations of the study area belong to the Archaean

metamorphic complex. The main rock types are

charnockites, granitic gneisses and dolerite dykes.

Major part of the study area is covered by granitic

gneisses (Fig. 2). The rocks are highly weathered and at

places, the soil cover extends to a depth of more than 12

m. The Coorg district is one of the major places that get

heavy rainfall in India. The river Harangi, a tributary of

the river Cauvery, originates in the Pushpagiri hills of

Western Ghats in Coorg district, Karnataka State. The

total length of the Harangi River is 50 km and the

southwest monsoon brings a lot of rains and is the

source of water in the catchment area. The present

landscape of the Coorg district under consideration is

viewed as the product of a series of interactions between

fluvial and denuded processes operating on underlying

geology that has been subjected to past vertical tectonics

(Vinutha et al., in press).

Fig1: Location map of the Harangi watershed

Fig2: Geological map of the Harangi watershed area

2. Methods of Investigation:

In the present study, an attempt has been made to study

the morphometric characteristics of Harangi basin. GIS

technique is employed for assessing various terrain and

morphometric parameters of the drainage basins and

watersheds, as they provide a flexible environment and

a powerful tool for the manipulation and analysis of

spatial information (Pankaj and Kumar, 2009). Drainage

map of the Coorg district as represented over the

topographical maps (scale 1:50,000), was acquired from

the Karnataka State Remote Sensing Centre (KSRSC),

Bangalore, in jpg format. Harangi river basin was

delineated in terms of surface – watershed area based on

topographic divides. The streams were georegistered to

the GIS using Universal Trans Mercator (UTM)

coordinates. Sub-watersheds numbering ten were

identified (Fig. 3). Determined watersheds were

imported into ArcView 3.2 and shape files were created

through manual digitizing. With all stream inventoried

and the Harangi watershed clearly defined and sub-

1043 D N VINUTHA, S RAMU, MOHAMMAD MUZAMIL AHMAD AND M R JANARDHANA



watersheds demarcated, different morphometric

parameters were calculated for each of the sub-

watershed. Linear, areal and relief aspects of the basin

were computed in GIS environment by using ArcGIS

9.3 software. Ground truthing was carried out randomly

and the maps created in GIS environment were modified

as necessary. Linear, areal and relief parameters of the

watershed of the present study area were evaluated as

per the procedure detailed in Vittala et al, (2004). The

results of the analysis are shown in tables from 1 to 5.

Fig3: Drainage map of the Harangi watershed showing

sub-watershed boundary

3. Results and Discussion:

The morphometric analysis is carried out through the

evaluation of linear, areal and relief aspects of the basin.

The results of each of the parameters are presented and

discussed below.

3.1. Linear Aspects:

The linear aspects include the linear scale measurements

of the drainage networks and the dimensionless

numbers such as bifurcation ratio and stream length

ratio. The linear scale parameters analyzed in the

present study are stream order, stream length, mean

steam length, stream length ratio and bifurcation ratio

which have a direct relationship with the erodibility of a

region.

3.1.1. Stream Order (U):

Ordering of stream is the first stage of basin analysis

and stream ordering involves applying numerical

value to a streams position and size in the basin. In

the present study, ranking of streams has been carried

out according to Strahler (1964) stream ordering

system. The order wise stream numbers, area and

stream length of the 10 sub-watersheds evaluated in

GIS environment are presented in Table 2. Out of

these, sub-watershed 1is under III order, 4, 7, 8 and

10 are IV order whereas the remaining 2, 3, 5, 6 and 9

are V order. In Table 2 it is noticed that the maximum

frequency is in case of first order streams. It is also

observed that there is a decrease in stream frequency

as the stream order increases. Highest and lowest I

order streams are seen in 9th

and 4th

sub-watersheds

respectively. I and II order streams dominate in 9th

sub-watershed which incidentally has witnessed

landslides, moderate in 6th

and 3rd

sub-watersheds.

These sub-watersheds are occupied by hillocks.

Table2: Stream orders and Stream lengths of Harangi watersheds

Sl.

No.

Sub-water

shed name

Stream

Order

Stream Order (U) Stream Length (Lu) in km

I II III IV V I II III IV V

1 1 III 53.00 12.00 3.00 0.00 0.00 25.67 12.97 0.37 0.00 0.00

2 2 V 45.00 13.00 4.00 2.00 1.00 24.77 14.86 4.50 3.60 0.90

3 3 V 179.00 41.00 7.00 2.00 1.00 90.00 36.10 25.67 6.75 5.58

4 4 IV 41.00 9.00 2.00 1.00 0.00 20.36 5.48 4.86 1.35 0.00

5 5 V 124.00 29.00 6.00 2.00 1.00 51.98 18.00 8.10 5.40 4.50

6 6 V 248.00 47.00 13.00 2.00 1.00 124.40 38.70 18.90 6.39 6.30

7 7 IV 96.00 21.00 3.00 1.00 0.00 41.30 15.31 4.68 2.88 0.00

8 8 IV 126.00 25.00 7.00 1.00 0.00 61.71 18.00 4.95 4.50 0.00

9 9 V 406.00 89.00 25.00 6.00 1.00 145.94 38.01 24.32 8.20 7.10

10 10 IV 133.00 30.00 6.00 1.00 0.00 48.80 29.90 9.00 7.10 0.00

Total 1451.00 316.00 76.00 18.00 5.00 634.93 227.33 105.35 46.17 24.38

3.1.2. Stream Length (Lu):

The total length of individual stream segments of each

order is the stream length of that order. This has been

computed based on the law proposed by Horton

(1945) for all the sub-watersheds of the study area.

The stream length in each order increases

exponentially with increasing stream order. From

Table 2, it is clear that the overall drainage





development in the study area is more in sub-

watershed No. 9 which is covered with granitic

gneisses and is less in the watershed no 4. It reflects

the frequency of the drainage development is high in

areas which have witnessed landslides. Further, these

areas are highly dissected and over saturation in the

interfluves between the drainage lines may possibly

be the triggering factor for the inducement of

landslides. Generally, the total length of stream

segments is maximum in first order streams and

decreases as the stream order increases. In conformity

with the stream order maximum Lu is noticed in sub-

watershed No. 9 and minimum is seen in sub-

watershed No. 4.

3.1.3. Mean Stream Length (Lsm):

Mean stream length (Lsm) is characteristic property

related to the drainage network components and its

associated basin surface (Strahler, 1964); this has

been calculated by dividing the total stream length of

order (U) by the number of streams of segments in the

order. The mean stream length is presented in Table

3. It is seen that Lsm values exhibit variation from

0.12 - 9.00. The mean stream lengths of stream

increase with the increase of the stream order. But

some sub-watersheds shows opposite relation, higher

order stream has a small mean length. These sub-

watersheds show variable relative relief and

asymmetry in the degree of weathering.

3.1.4. Stream Length Ratio (RL):

The Stream Length Ratio (RL) is the ratio of the mean

length of the one order (Lu2) to the next lower order

(Lu1) of the stream segments. The RL values are

presented in Table 3. The stream length ratio between

the streams of different orders of the study area shows

a change in each sub-watershed. This change might

be attributed to variations in slope and topography

(Singh & Singh, 1997), indicating the late youth stage

of geomorphic development in the streams of the

study area.

Table3: Mean stream lengths and Stream length ratios of Harangi watershed

Sub -

Watershed

Name

Mean Stream Length (Lsm) in km Stream Length Ratio (RL)

Mean

Bifurcation

Ratio

I II III IV V II/I III/II IV/III V/IV VI/

V 2.81

1 0.48 1.08 0.12 0.00 0.00 0.51 0.03 0.00 0.00 0.00 2.68

2 0.55 1.14 1.13 1.80 0.90 0.60 0.30 0.80 0.25 0.00 3.93

3 0.50 0.88 3.67 3.38 5.58 0.40 0.71 0.26 0.83 0.00 3.69

4 0.50 0.61 2.43 1.35 0.00 0.27 0.89 0.28 0.00 0.00 3.53

5 0.42 0.62 1.35 2.70 4.50 0.35 0.45 0.67 0.83 0.00 4.35

6 0.50 0.82 1.45 3.20 6.30 0.31 0.49 0.34 0.99 0.00 4.86

7 0.43 0.73 1.56 2.88 0.00 0.37 0.31 0.62 0.00 0.00 5.20

8 0.49 0.72 0.71 4.50 0.00 0.29 0.28 0.91 0.00 0.00 4.57

9 0.36 0.43 0.97 1.37 7.10 0.26 0.64 0.34 0.87 0.00 5.14

10 0.37 1.00 1.50 7.10 9.00 0.61 0.30 0.79 0.00 0.00 2.81

3.1.5. Bifurcation Ratio (Rb):

Bifurcation ratio is the index of relief and dissection

(Horton, 1945) and is defined as the ratio of the

number of the stream segments of given order to the

number of segments of the next higher orders

(Schumn, 1956). Bifurcation ratio shows a small

range of variation for different regions or for different

environment except where the powerful geological

control dominates (Strahler, 1957). The higher values

of bifurcation ratio suggest that the area is tectonically

active and indicates structural control on the drainage

pattern. As per Horton (1945) bifurcation ratio having

a less value of about 2 to 3 is of flat region. When

geology is reasonably homogeneous throughout basin,

Rb values usually range from 3.0 to 5.0. High

variation in the bifurcation ratio of a watershed

indicates ruggedness topography. The mean Rb

values in the sub-watersheds of the study area (Table

4) range from 2.68 to 5.20. The study area is largely

covered with granitic gneisses (Fig. 2) and the mean

bifurcation value is by and large well within the

suggested limits of homogeneity of lithology. Further,

the Rb values of study area indicate that there is a

uniform decrease in Rb values of micro watershed 1,

2 and 4 from one order to the next order indicating

that the lower part of the sub-watershed of the basin is

relatively flat. Whereas in other sub-watersheds, the

Rb values are not same from one order to next order.

This irregularity is significant in watershed numbers 7




and 9 and the ratio of the Rb of the lower orders is

variable suggesting that the streams are of highly

dissected drainage basins. In the study area, the

moderate values of Rb indicate a structural control in

the drainage pattern whereas the lower values indicate

that the sub-watersheds are less affected by structural

disturbances (Nag, 1998 & Chopra et al, 2005). This

is evident as the study area is characterized by

dendritic and sub-parallel drainage patterns (Fig. 3).

3.2. Relief Aspects:

Relief is the maximum elevation difference between the

highest and lowest points of a basin. Basin relief is an

important factor in understanding the denudational

characteristics of the basin (Sreedevi et al, 2009). The

relief measurements such as relief ratio (Rh) and total

relief have been carried out and the computed data are

presented in Table 5. The elevations of the maximum

and minimum points in the study area are 1620 m and

120 m above msl (Fig. 3) and hence the total relief of

the Harangi watershed is 1400 m. Most of the landslide

incidences have occurred at an elevation of 1200 m

above msl.

The maximum relief to horizontal distance along the

longest dimension of the basin parallel to the principal

drainage line is termed as relief ratio (Schumn, 1963).

Rh value has direct relationship between the relief and

channel gradient. The Rh value normally increases with

decreasing drainage area and size of the watersheds of a

given drainage basin (Gottaschalk, 1964). In the study

area, the values of relief ratio range from 0.03 – 0.14.

High to very high relief and slope are characterized by

high values of relief ratios. The Rh value of the sub-

watershed No. 9 is low suggesting low relative relief in

the area and as the soil cover is thick, facilitates high

infiltration of rain water.

Fig4: DEM map of Harangi Watershed

Table5: Relief aspects of Harangi watershed

Sl.

No.

Sub-

watershed

name

Basin

length (Lb)

(km)

Total

relief

(Mtrs)

Relief

ratio

(Rh)

1 1 8.00 260.00 0.03

2 2 6.70 380.00 0.06

3 3 20.30 480.00 0.02

4 4 11.05 200.00 0.02

5 5 20.85 400.00 0.02

6 6 16.18 680.00 0.04

7 7 12.33 400.00 0.03

8 8 14.42 680.00 0.05

9 9 22.18 420.00 0.02

10 10 41.70 360.00 0.01

3.3. Areal Aspects:

The morphometric parameters considered for the

present study to understand the areal aspects are sub-

watershed areas, perimeter, drainage density, stream

frequency, texture ratio, form factor, Circularity ratio,

elongation ratio and the length of the overland flow. The

computed values of these parameters are presented in

table 6 and discussed.

Table6: Areal aspects of sub-watersheds of Harangi watershed

Sl.

No

.

Su

b-w

ate

rsh

ed

na

me

Su

b-w

ate

rsh

ed

are

a

(km

2)

Per

imet

er(P

) (k

m)

Dra

ina

ge

den

sity

(D)

(km

-1)

Str

eam

Fre

qu

ency

(Fs)

Tex

ture

Ra

tio

(Rt)

Fo

rm F

act

or

(Ff)

Cir

cu

lari

ty R

ati

o

(Rc)

Elo

ng

ati

on

Ra

tio

(Re)

Len

gth

of

Ov

erla

nd

flo

w(L

g)

1 1 30.88 29.47 1.26 2.20 2.31 0.48 0.45 0.78 0.40

2 2 33.08 28.25 1.47 1.96 2.30 0.74 0.52 0.97 0.34

3 3 91.10 47.55 1.80 2.52 4.84 0.22 0.51 0.53 0.28





4 4 24.42 26.13 1.31 2.17 2.03 0.20 0.45 0.50 0.38

5 5 56.53 53.57 1.56 2.87 3.02 0.13 0.25 0.41 0.32

6 6 107.75 61.23 1.81 2.89 5.08 0.41 0.36 0.72 0.28

7 7 32.77 26.32 1.96 3.69 4.60 0.22 0.59 0.52 0.26

8 8 52.87 47.92 1.69 3.01 3.32 0.25 0.29 0.57 0.30

9 9 108.40 49.20 2.06 4.86 10.71 0.22 0.56 0.53 0.24

10 10 126.96 104.96 0.75 1.34 1.62 0.07 0.14 0.30 0.67

664.76

3.3.1. Watershed Area:

Harangi watershed covers an area of 664.76 sq. kms.

There are ten sub-watersheds in the study area

varying in their areal extent from 24.42 sq. kms. to

126.96 sq. kms. Sub-watershed No. 4 is the smallest

and sub-watershed No. 10 is the largest.

3.3.2. Drainage Density (Dd):

Drainage lines are numerous over impermeable areas

than permeable areas. Drainage density indicates the

closeness of spacing of channels and provides a

numerical measurement of the landscape dissection

and surface runoff potential. It is defined as the total

stream length per unit area in a region of the

watershed (Strahler et al, 1954). Density factor is

related to the interaction between climate and type of

rocks, relief, infiltration capacity, vegetation cover

and run-off intensity index. High Dd values are the

result of weak or impermeable sub-surface material,

sparse vegetation and mountainous relief. Low

density leads to coarse drainage texture signifying the

area of having permeable sub-soil material while high

drainage density leads to fine drainage texture. The

computed Dd values (Table 6) in the study area range

from 1.06 to 2.06 Km-1

reflecting high infiltration

capacity of the land/underlying rock and increased

transmissivity of the soil with thick vegetation. The

obtained data corroborates well with the field

condition of the study area.

3.3.3. Stream Frequency (Fs):

The total number of stream segments of all orders per

unit area is known as stream frequency (Horton,

1932). The Fs values of the sub-watersheds of the

study area are presented in Table 6. It is noted that

values of Fs vary from.1.34 - 4.86. It is also seen that

the drainage density values of the sub-watersheds

exhibits positive correlation with the stream

population with respect to increasing drainage

density. The Fs values of sub-watershed numbers 9

and 7 are relatively high (Table 6) in conformity with

the number of stream order and stream lengths.

3.3.4. Drainage Texture (Dt):

Drainage texture is the total number of stream

segment of all orders per perimeter of that area and

infiltration capacity is the single important factor

which influences drainage texture (Horton, 1945).

The amount and type of vegetation, precipitation,

infiltration capacity viz., absorption capacity of soil,

underlying lithology, and relief aspect of the terrain

influences the rate of surface run-off and affects the

drainage texture of an area. Based on the drainage

density five drainage textures have been classified

(Smith, 1950). The drainage density less than 2

indicates very coarse, between 2 and 4 is related to

coarse, between 4 and 6 is moderate, between 6 and 8

is fine and greater than 8 is very fine drainage texture.

The soft or weak rocks unprotected by vegetation

produce a fine texture, whereas massive and resistant

rocks cause coarse texture. The maximum value of

drainage texture ratio of the study area is 10.71 (Table

6) and is seen in the sub-watershed No. 9.

3.3.5. Form factor (Ff):

Form factor may be defined as the ratio of basin area

to square of basin length (Horton, 1932). The values

of form factor would always be greater than 0.78 for a

perfect circular basin. Smaller the value of form

factor, more elongated will be the basin. Rf values of

the study area are presented in Table 6. It is noted that

Ff values vary from 0.07 - 0.74. The value of Ff in

sub-watershed 2 suggest it is almost circular in shape

whereas Ff of remaining sub-watersheds lies well

below 0.78 suggesting elongated in shape and flow

for longer duration. Thus, there is more scope for

infiltration.

3.3.6. Circulatory Ratio (Rc):

The circulatory ratio is concerned with the length and

frequency of streams, geological structures, land

use/land cover, climate, relief and slope of the basin.

It is the ratio of the watershed area to the area of

circle having the same perimeter of the watershed. In

the study area, the Rc values range from 0.14 - 0.59

(Table 6). The Rc values of all the sub-watershed

suggest that the sub-watersheds are more or less

strongly elongated and comprise highly permeable

homogeneous geologic materials.

3.3.7. Elongation Ratio (Re):




Elongation ratio is the ratio between the diameter of

the circle of the same area as the drainage basin and

the maximum length of the basin (Schumm, 1965).

The varying shapes of watershed can be classified

with the help of the index of elongation ratio i.e.,

circular (0.9 – 1.0), oval (0.8 – 0.9), less elongated

(0.7 – 0.8), elongated (0.5 – 0.7), and more elongated

(<0.5) (Pareta & Pareta, 2012). A circular basin is

more efficient in the discharge of run-off than an

elongated basin (Singh and Singh, 1997).

The elongation ratio values of the sub-watersheds of

the study area vary from 0.41 - 0.97 (Table 6). The

sub-watersheds 7 and 9 witnessing landslides have

elongation ratios 0.52 and 0.53 respectively

suggesting them to be elongated. Analysis of

elongation ratio (Re) indicates that these areas have

high infiltration capacity and low runoff.

3.3.8. Length of overland flow (Lg):

It is the length of water over the ground before it gets

concentrated into definite stream channels. This factor

basically relates inversely to the average slope of the

channel and is quit synonymous with the length of the

sheet flow to the large degree. The Lg value of the

study area shows variation from 0.24 to 0.67 kms

(Table 6). The values of Lg are relatively higher in1,

2, 4, 5 and 8 indicating low relief and lower values in

remaining sub-watersheds indicate relatively high

relief.

3.4. Conclusions:

The quantitative morphometric analysis at the sub basin

level enables to understand the nature of watersheds and

sub-watersheds, relationships among the different

aspects of the drainage patterns and their influence on

the vulnerability of the area for landslides. The study

was undertaken as a preliminary step to assess the role

of drainage development and its impact on the mass

movement in the Harangi watershed area. The

morphometric analysis of the drainage network of the

Harangi basin exhibits largely dendritic pattern and at

places sub-parallel pattern. The study area was divided

into 10 sub-basins on the basis of water divide and the

sub-basins were evaluated, both by conventional

methods and modern tools for their linear, relief and

areal parameters. The mean stream length of the river

was found ranging from 1.68 to 18.97 km among

different stream orders. The variation in stream length

ratio might be due to changes in slope and topography. I

and II order streams dominate in 9th

sub-watershed

which incidentally has witnessed more number of

landslides, moderate in 6th

and 3rd

sub-watersheds.

These sub-watersheds are occupied by hillocks. The

bifurcation ratio (Rb) of the lower orders is variable in

sub-watershed numbers 7 and 9 suggesting that the

streams are of variably dissected drainage basins. The

mean bifurcation ratio (Rbm) of the sub-watersheds No.

7 and 9 are 4.86 and 4.57 (Table 4) which indicates that

the geological structures are less disturbing the drainage

pattern.

The channel network morphometry indicates that the

basin and sub basins conforms Horton’s law of stream

number and stream lengths. The relief ratio Rh value of

the sub-watershed No. 9 known for incidences of

landslides is low suggesting low relative relief in the

area and as the soil cover is thick facilitates high

infiltration of rain water. All the sub-watersheds are

associated with low drainage density (Dd) reflecting

high infiltration capacity of the /underlying rock and

increased transmissivity of the soil with thick vegetation

while stream frequency is high in sub-watersheds 7 and

9 thus pointing at the streams of the first order are short

compared to other sub-watersheds. It is noted that Ff,

Rc and Re values show the elongative nature of the

basin suggesting flow for longer duration, low discharge

of runoff and highly permeable sub soil condition.

These factors in addition to human activity might have

induced the mass movement at places in the study

region.

4. Acknowledgement:

The first author wishes to thank the Karnataka State

Remote Sensing Centre, Bangalore for the soft copy of

the drainage map of the study area. Ms. Annapoorna H

and Mr. Abdul - Aleam Ahmed Ahmed Dihwa Al-

Qadhi are thanked for logistic support.

References:

[1] Chopra R, Raman Deep Dhiman and Sharma PK

(2005) Morphometric analysis of sub-watersheds in

Gurudaspur district, Punjab using Remote sensing

and GIS techniques, J Indian Soc of Remote

Sensing, 33(4): 531-539.

[2] Clarke JI (1966) Morphometry from maps, Essays

in geomorphology, 235-274.

[3] Gottschalk LC (1964) Reservoir Sedimentation in:

V.T. Chow (Ed.) Handbook of Applied Hydrology.

McGraw Hill Book Company, New York, Section

7-I.

[4] Kuldeep Pareta and Upasana Pareta (2012)

Quantitative Geomorphological Analysis of a

Watershed of Ravi River Basin, H.P. India, J

International Journal of Remote Sensing and GIS,

1: Issue 1,

[5] Horton RE (1932) Drainage basin characteristics,

Trans. Am. Geophys. Union 13.350-361.

[6] Horton RE (1945) Erosional development of

streams and their drainage basins Hydro physical

approach to quantitative morphology, Geol Soc

Am. Bull, 56: 275-370.





[7] Miller VC (1953) A Quantitative Geomorphic

Study of Drainage Basin Characteristics on the

Clinch Mountain Area, Virgina and Tennessee,

Proj. NR Tech Rep 3, Columbia University, 389-

402.

[8] Mohd Iqbal , Haroon Sajjad and Bhat FA (2012)

Morphometric analysis of Dudhganga catchment,

Kashmir valley, India using geographical

information system, J International Journal of

Current Research, 4: Issue 12, 410-416.

[9] Nag SK (1998) Morphometric analysis using

remote sensing techniques in the Chaka milli

watershed, Purulia district, West Bengal, J Indian

Soc Remote Sensing 26(1&2): 69-76.

[10] Nag SK and Chakraborty S (2003) Influences of

rock types and structures in the development of

drainage network in hard rock area, J Indian Soc

remote Sensing 31(1): 25-35.

[11] Pankaj A and Kumar P (2009) GIS-based

morphometric analysis of five major sub-

watersheds of Song River, Dehradun District,

Uttarkhand with special reference to landslide

incidences, J Indian Soc. Remote Sens, 37: 157 –

166.

[12] Schumn SA (1956) Evolution of drainage systems

and slopes in Badlands at Perth Amboy .New

Jersey, Geol Soc Am Bull 67: 597-646.

[13] Schumm SA (1963) Sinuosity of alluvial rivers on

the Great Plains, Bull. Geol. Soc. Amer. 74: 1089-

1100.

[14] Schumm SA and Lichty RW (1965) Time, Space,

and Casualty in geomorphology, J American

Journal of Science, 263: 110-119.

[15] Singh S and Singh MC (1997) Morphomertric

Analysis of Kanhar River Basin, National

Geographical Jour. of India, 43 (1): 31-43.

[16] Smith KG (1950) Standards for grading textures of

eritonal topography, Am Jour Sci 248: 655-668.

[17] Sreedevi PD, Owais S, Khan S.S and Ahmed S

(2009) Morphometric Analysis of a Watershed of

South India Using SRTM Data and GIS, J of the

Geological Society of India 73 (4): 543-552 .

[18] Strahler AN (1957) Quantitative analysis of

watershed geomorphology, Trans. Am Geophys

Union 38: 913-920.

[19] Strahler AN (1954) Quantitative slope, analysis,

Bull. Geol. Societ. Am. 67: 571-596.

[20] Strahler AN (1964) Quantitative geomorphology of

drainage basins and channel networks. In V. T.

Chow (Ed.), Handbook of Applied Hydrology. New

York: McGraw Hill. (pp. 4, 39-4, 76).

[21] Vinutha DN, Sridhara Raje Ars K & Janardhana

MR (2014), Landform studies & Geomorphological

mapping of a part of Coorg district, Karnataks state,

Communicated to International Journal of Geology,

Earth and Environmental science, (ISSN-2277-

2081) (In press).

[22] Vittala SS, Govindaiah S and Honne Gowda H

(2004), Morphometric analysis of sub-watersheds

in the Pawagada area of Tumkur district, South

India, using remote sensing and GIS techniques. J

Indian Soc of Remote Sensing 32(4):351–362.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1049-1054


Impact of CO2 Fertilization on Growth and Biomass in Marine

Diatom Nitzschia sp.

RAJANANDHINI K, P SANTHANAM, S JEYANTHI, A SHENBAGA DEVI, S DINESH KUMAR AND

B BALAJI PRASATH Department of Marine Science, Bharathidasan University, Tiruchirappalli, INDIA


Abstract: Despite their microscopic size, marine phytoplankton is responsible for about half of the global primary

production and represents the basis of the marine food web. Microalgae are of particular interest because of their

rapid growth rates, tolerance to varying environmental conditions and can also fix greater amounts of CO2 per land

area than higher plants. Diatoms are the major photosynthesizing organisms contributing over 80% and serve as a

vital first link in the food chain. Nitzschia sp. is a unicellular diatom in the class Bacillariophyceae. Marine diatoms

are key players in the ocean carbon cycle they have been found to be capable of concentrating CO2 within the cell.

This experiment illustrates the importance of CO2 for the growth of the diatoms in sea water. Photosynthesis of

diatom Nitzschia sp. is likely to be simulated by increased availability of CO2 which has resulted in high growth and

biomass content in CO2 fertilized culture.

Keywords: Diatoms, Nitzschia sp., CO2 and Photosynthesis.

1. Introduction:

Carbon is the element of life. The oceans store much

more carbon than the atmosphere and the terrestrial

biosphere (plants and animals). The growth and

photosynthetic CO2 fixation of marine phytoplankton

were found to be enhanced by elevated levels of CO2

(Riebesell et al., 1993; Hein and Sand-Jensen, 1997).

Photosynthesis serves as a source of energy for

metabolism and growth for basically all forms of life on

Earth either in a direct or indirect way (Masojidek et al.,

2004). Photosynthesis is by far the largest component of

primary production (Littler et al., 1979; Heine, 1983;

Levitt and Bolton, 1990). Phytoplankton has been

recognized as an extremely important source of food for

larval forms of aquatic organisms, since they constitute

the microscopic plant life of the sea. They play an

important role in the biosynthesis of organic matter

(primary production) in the aquatic ecosystem which

directly or indirectly serves all the aquatic life as the

basic food.

Algae are photosynthetic, non-vascular plants

containing chlorophyll as a pigment (Vonshak and

Maske, 1982). Nearly half of the total photosynthesis

taking place on earth is associated with marine

phytoplankton (Camacho et al., 2003). Microalgae

promises to be a suitable energy resource because the

photoautotrophic mechanism that can convert

atmospheric carbon dioxide into biomass, fatty acid and

lipids (Spolaore et al., 2006, Chisti, 2007). Microalgae

are of particular interest because of their rapid growth

rates, tolerance to varying environmental conditions and

can also fix greater amounts of CO2 per land area than

higher plants (Brown, 1996).

Diatoms are unicellular, silica-forming autotrophic

organisms which can fix energy from the sun and thus

form the base of the food chain (Round et al., 1990).

Diatoms along with other micro algal species have been

found to be capable of concentrating CO2 (CO2

Concentrating Mechanisms, CCMs) within the cell

(Tortell et al., 1997; Badger et al., 1998; Raven and

Falkowski, 1999; Reinfelder et al., 2000; Giordano et

al., 2005), which usually involves an extra cellular

carbonic anhydrase (CA) that facilitates the conversion

of bicarbonate (HCO3ˉ) to CO2 (Nimer et al., 1999).

Marine diatoms are key players in the ocean carbon

cycle, accounting for atleast 40% of the marine primary

production (Nelson et al., 1995). Diatoms are important

marine photoautotrophic protists that account for up to

25% of the primary production on Earth (Falkowski and

Raven, 1997). Theoretically, photosynthesis and growth

of diatom species can be limited by the availability of

CO2 (Riebesell et al., 1993), and oceanic primary

production might be enhanced by increasing

atmospheric CO2 concentration (Hein and Sand-Jenson,

1997; Riebesell and Tortell, 2011; Schippers et al.,

2004).

Marine micro algae need carbon dioxide (CO2) in order

to grow. Like plants on land, they help to remove CO2

1050 Impact of CO2 Fertilization on Growth and Biomass in Marine Diatom Nitzschia sp.



from the air and thus contribute to the function of the

ocean as a carbon “sink”. This experiment illustrates the

importance of CO2 for the growth of the diatoms in sea

water. The aim of the study was to examine the effect of

CO2 fertilization on growth and biomass of the marine

diatom, Nitzschia sp.

2. Methods:

Laboratory Experiments:

Diatom Culture:

The algal samples were collected from the Muthukuda

backwaters (Lat 9.8˚ N and Long. 79.1˚ E) using

phytoplankton net of 25μm. After the initial cultivation

of mixed cultures with the f/2 medium (Guillard and

Ryther, 1962) pure cultures were isolated by performing

serial dilutions and kept at 25ºC in a thermostatically

controlled room, illuminated with white florescence

lamps with 12:12 h of light/ dark regime. The

temperature and salinity was maintained in the range of

23 to 25oC and 28 to 30 ppt. Among the diatoms

Nitzchia sp. was isolated and identified based on its

morphological characters (Subrahmanyan, 1946;

Desikachary and Prema, 1987; Santhanam and Perumal,

2008) and the indoor stock culture was maintained.

(Plate.1)

Plate1: The fine structure of unicellular Nitzschia sp.

observed under magnifications of 400×

Experimental Design (Plate.2):

The following treatment series were maintained for the

CO2 fertilization studies on diatom

Treatment 1 – Control (No CO2 Fertilization) – flask

covered with a cotton wad.

Treatment 2 - CO2 Fertilized – flask covered with a

cotton wad made airtight by wrapping it in a transparent

film with a straw inserted in the middle for CO2

introduction by blowing

Treatment 3- CO2 Limited – flask covered tightly using

parrafilm

Plate2: Experimental set up of Nitzschia sp. (Control,

CO2 Fertilized and CO2 limited

CO2 Bubbling:

Bubbling was performed in the treatment 2, with the air

blown into the culture for 30 – 60 seconds twice a day.

The bubbling of the cultures was performed on day 1

and continued until the end of the experiment

Analytical Procedures:

Data was collected on daily basis for: pH,

Photosynthetic measurements and cell density. The pH

was measured using an ELICO grip pH meter. A Motic

microscope equipped with a digital camera was used for

microscopic analysis of sample purity and Cell counting

was conducted using a Sedgewick counting chamber

followed by method of Venugopalan and Paulpandian

(1989). Chlorophyll ‘a’ and ‘b’ was estimated by the

standard method of Strickland and Parsons (1972).

Statistical Analysis:

Statistical significance of the data was tested for

Pearson correlation with the significant level at 0.01

using SPSS 14.0 software.

3. Results:

Changes in the Medium:

In the present study, the pH was found to be low (7.00

to 8.5) in CO2 fertilized when compared with Control

(8.00 to 9.00) and CO2 limited (8.00 to 9.03). At the

control and CO2 fertilized treatment groups, the pH was

found to be in large correlation [r=0.89, N=11; P < 0.01]

suggesting quite a strong relationship between the two.

The pH of Control and CO2 limited showed large

correlation [r=0.97, N=11; P < 0.01] suggesting a

positive relationship between the two treatments. The

correlation is highly significant at 0.01 (Fig.1, 2 &3)

1051 RAJANANDHINI K, P SANTHANAM, S JEYANTHI, A SHENBAGA DEVI, S DINESH KUMAR

AND B BALAJI PRASATH



Fig.1 Measurement of pH in marine diatom Nitzschia sp. culture

0

1

2

3

4

5

6

7

8

9

10

1 2 3 4 5 6 7 8 9 10 11

Days

pH

Control

Fertilised

Limited

Fig1: Measurement of pH in marine diatom Nitzchia sp.

culture

Fig2: Correlation between Control and CO2 fertilized

pH

Fig3: Correlation between Control and CO2 limited pH

Growth Study:

The overall range of cell growth of Nitzchia sp. in CO2

fertilized was found highest (86488 cells/ml) followed

by the CO2 limited (56927 cells/ml) and the control

(53704 cells/ml). The correlation between Control and

CO2 fertilized growth showed positive value [r=0.98,

N=11; P < 0.01] whereas the statistical values among

the cell density of Control and CO2 limited showed large

correlation [r=0.94, N=11; P < 0.01]. The correlation is

highly significant at 0.01 (Fig.4, 5&6)

Fig.4 Growth of marine diatom Nitzschia sp.

0

20000

40000

60000

80000

100000

120000

140000

160000

180000

1 2 3 4 5 6 7 8 9 10 11

Days

Cell

Den

sit

y (

cell

s/m

l)

Control

Fertilised

Limited

Fig5: Correlation between the growth of Nitzschia sp.

in Control and CO2 fertilized

Fig6: Correlation between the Nitzschia sp. growth in

Control and CO2 limited

Analysis of Chlorophyll Pigments:

The Nitzchia sp. biomass in terms of Chlorophyll ‘a’

concentration was observed in the average ranges of




0.05 mg/ml in Control, 0.07 mg/ml in CO2 fertilized and

0.06 mg/ml in CO2 limited. Similarly the Chlorophyll

‘b’ concentration was found to be in the average range

of 0.06 mg/ml in comparison with CO2 fertilized and

CO2 limited with the ranges of 0.07 mg/ml (Fig.7&8)

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

Co

ncen

trati

on

(m

g/m

l)

1 2 3 4 5 6 7 8 9 10 11

Days

Fig.7 Measurement of Chlorophyll ' a' in marine diatom

Nitzschia sp.

Control

Fertilised

Limited

0

0.05

0.1

0.15

0.2

0.25

Co

ncen

trati

on

(

mg

/ml)

1 2 3 4 5 6 7 8 9 10 11

Days

Fig. 8 Measurement of Chlorophyll 'b' in marine diatom

Nitzschia sp.

Control

Fertilised

Limited

4. Discussion:

This study showed that the primary producer Nitzschia

sp. used in this study increased their growth and

biomass when cultured in CO2 fertilized flask. It has

been reported that pH increases when algae grow

(Anderson, 2005). The pH of Nitzschia sp. cultivated

with the initial pH of 7.32 continued to increase and

reached up to 8.8. Meanwhile, the pH of control reached

a maximum of 9.27 and then started to decrease. This

shows that bicarbonate ions are the dominant inorganic

carbon species and have been reported to be utilized for

their growth. In the past, pH has not been considered to

be an important chemical parameter influencing biotic

processes in marine environments. However, a number

of studies have shown that pH and, in some cases,

inorganic carbon may be important in regulating the

growth rate and distribution of marine algae (Yoo 1991,

Hinga 1992, Riebesell et el., 1993). Therefore pH

mediated through [CO2] may be an important abiotic

factor affecting the ecology of marine phytoplankton.

Thus, atmospheric CO2 increase could potentially

promote phytoplankton productivity (Schippers et al.,

2004). He suggested that increased phytoplankton

productivity because of atmospheric CO2 elevation was

proportional to the increased atmospheric CO2, though it

was reported that unlike terrestrial plants, phytoplankton

would not show a significant response to the

atmospheric CO2 increase (Raven and Falkowski, 1999;

Tortell et al., 2000). Our results showed that biomass of

marine diatom increased in CO2 fertilized cultures in

response to the CO2 rise. Schippers et al., (2004)

predicted doubling of atmospheric CO2 could result in

an increase of the productivity of 10%-40% and Hein

and Sand-Jenson (1997) found 15-19% stimulation of

primary production in response to elevated CO2

concentration. The growth in CO2 limited cultures was

that the residual CO2 was still in the water and in the air

inside the flask. This was sufficient to trigger initial

growth of the algae but this could not be sustained

because of absence of CO2.

The growth curve of microalgae in Control is

characterized by an initial log phase and at about the 7th

day the growth enters the plateau stage. With an excess

of CO2 in CO2 fertilized cultures the plateau stage was

reached faster than with the control where CO2 was

available but not in excessive concentrations. In the

experiment limited enrichment of CO2 has enhanced the

photosynthetic activity showing a bright coloration in

the culture flask of CO2 fertilized with respect to the

other cultures. Morphological characteristics did not

affect the diatoms as they are highly tolerant to the wide

fluctuations of pH levels. Phytoplankters have different

affinities to CO2 and will use it in different ways.

Diatoms are highly tolerant to wide range of pH

fluctuations as their cell walls are made up of silica

frustules. As a result the diatoms grown in the acidified

cultures were capable of compensating the pH change

and they survived showing high productive results.

Some diatoms showed enhanced growth rate with

enriched CO2 and acidity (Riebesell et al., 1993;

Riebesell, 2004; Sobrino et al., 2008; Wu et al., 2010)

like the results have been observed with the presently

studied diatom Nitzschia sp.

Photosynthesis of diatoms is likely to be stimulated by

increased availability of CO2, lower pH might enhance

their respiration too (Wu et al., 2010), which would

down-regulate their contribution to the marine

biological CO2 pump. Different algal species or growth

conditions show differential preferences to CO2 or

HCO3− (Burkhardt et al., 2001; Nimer et al., 1997).

However, other studies show insignificant effects

(Boelen et al., 2011; Chen and Gao, 2003; Nielsen et

al., 2010, 2012; Tortell et al., 2002; Tortell and Morel,

2002) and even negative effects (Feng et al., 2009;

Low-DÉCarie et al., 2011; Torstensson et al., 2012) on

photosynthesis, growth or primary productivity in

diatoms or phytoplankton community.

1053 RAJANANDHINI K, P SANTHANAM, S JEYANTHI, A SHENBAGA DEVI, S DINESH KUMAR

AND B BALAJI PRASATH



5. Conclusion:

It is understood that the CO2 is essential for the

photosynthesis of algae and terrestrial plants. This study

shows that photosynthetic activity has been triggered by

enrichment of CO2 .The potential for marine organisms

to adaptation for increasing CO2 and broader

implications for ocean ecosystems are not well known;

both more high priorities for future research.

6. Acknowledgements:

Authors are grateful to Head, Department of Marine

Science and authorities of Bharathidasan University,

Tiruchirappalli, for facilities provided. The Authors also

acknowledge to Dr. M. Anand, Assistant Professor,

Department of Marine Science and Coastal Studies,

Madurai Kamaraj University, Tamil Nadu, for his

valuable suggestions to carry out future studies with

different combination of atmospheric air viz., 100%

pure CO2 or CO2 + air mixture and the authors are also

thankful to Dr. D. Venkat Reddy, Editor-in-chief,

Professor of Geology, National Institute of Technology,

Karnataka for bringing out this paper into publishing.

References:

[1] Anderson, R.A (2005) Algal culturing techniques,

Elsevier Inc., San Diego, USA, pp.35-64.

[2] Badger, M.R., T.J. Andrews, S.M. Whitney, M.

Ludwig, D.C. Yellowlees, W. Leggat and G.D

Price (1998). The diversity and coevolution of

Rubisco, Plastids, Pyrenoids and Choloroplast-

based CO2-concentrating mechanisms in algae.

J.Bot.76: 1052-1071.

[3] Boelen, (2011). Neither elevated nor reduced CO2

affects the photo physiological performance of the

marine Antarctic diatom Chaetoceros brevis.

Journal of Experimental Marine Biology and

Ecology 406 (2011): 38–45.

[4] Brown, L.M. (1996). Uptake of Carbon Dioxide

from Flue Gas by Microalgae. Energy Conversion

Management, 37, 1363-1367.

[5] Burkhardt, S., Amoroso, G., Riebesell, U.,

Sultemeyer, D., (2001). CO2 and HCO3- uptake in

marine diatoms acclimated to different CO2,

concentrations. Limnology and Oceanography 46,

1378-1391.

[6] Camacho, R. F., Garcia, C. F., Fernandez, S. J. M.,

Chisti, Y., Molina, G. E. (2003). A mechanistic

model of photosynthesis in microalgae. Biotechnol

Bioeng. 81(4): 459-473.

[7] Chen X, Gao K (2003): Effect of CO2

concentrations on the activity of photosynthetic

CO2 fixation and extracellular carbonic anhydrase

in the marine diatom Skeletonema costatum. Chin

Sci Bull 48: 2616−2620.

[8] Chisti, Y, Biodiesel from microalgae, Biotech.

Adv., 2007, 25, 294-306.

[9] Desikachary, T.V., Prema, P., 1987. Diatoms from

the Bay of Bengal., In: Desikachary, T.V. (Ed.)

Atlas of Diatoms. Fascicles III and lV. TT Maps

and Publications Limited, Madras.

[10] Falkowski PG, Raven JA (1997) Aquatic

photosynthesis. Blackwell Science, Malden, MA.

[11] Feng, Y.Y., Hare, C.E., Leblanc, K., Rose, J.M.,

Zhang, Y.H., DiTuillio, G.R., Lee, P.A., Wilhelm,

S.W., Rowe, J.M., Sun, J., Nemeck, N., Gueguen,

C., Passow, U., Benner, I., Brown, C., Hutchins,

D.A., (2009). Effects of increased pCO2 and

temperature on the North Atlantic spring bloom. I.

The phytoplankton community and biogeochemical

response. Marine Ecology progress Series 388, 13-

25.

[12] Giordano, M., J. Beardall & J.A. Raven. (2005).

CO2 concentrating mechanisms in algae:

mechanisms, environmental modulation, and

evolution. Annu. Rev. Plant Biol. 56: 99-131.

[13] Guillard RRL, Ryther JH (1962). Studies of marine

planktonic diatoms. Can J Microbiol 8: 229-239.

[14] Hein, M., Sand-Jensen, K., (1997). CO2 increases

oceanic primary production. Nature 388, 527.

[15] Heine, J .N. 1983. Seasonal productivity of two red

algae in a Central California Kelp forest. J Phycol.,

19: 146-152.

[16] Hinga, R. 1992. Co-occurrence of dinoflagellate

blooms and high pH in marine enclosures. Mar.

Ecol. Prog. Ser. 86: 181-187, 1992.

[17] Levitt, GJ. And J.J. Bolton 1990. Seasonal primary

production of understory Rhodophyta in an

upwelling system. J. Phycol., 26: 214-220.

[18] Littler, M.M., S. N. Murray and K.E. Arnold 1979.

Seasonal variations in net photosynthesis

performance and cover of intertidal macrophytes.

Aquatic Bot., 7: 35-46.

[19] Low-DECarie, E., Fussmann, G.F., Bell, G.,

(2011). The effect of elevated CO2 on growth and

competition in experimental phytoplankton

communities. Global Change Biology 17, 2525-

2535.

[20] Masojidek, J., Koblizek, M., Torzillo, G. (2004).

Photosynthesis in microalgae. In: Handbook of

microalgal culture (Ed. Richmond A.) pp. 20-39.

Blackwell Publishing Company, 2121 State

Avenue, Ames, Iowa, USA.

[21] Nelson DM, Treguer, Brzezinski MA, Leynaert

A,Queginer (1995). Production and dissolution of

biogenic silica in the ocean: Revised global

estimates comparison with regional data and

relationship to biogenic sedimentation. Global

Biogeochem cycles 9: 359-372.

[22] Nimer, N.A., M.X. Ling,C. Brownlee and M.J.

Merrett (1999). Inorganic carbon limitation,




exofacial carbonic anhydrase activity and plasma

membrane redox activity in marine phytoplankton

species. J. Phycol.35: 1200-1205.

[23] Nimer, N.A., Iglesias-Rodriguez, M.D., Merrett,

M.J., (1997). Bicarbonate utilization by marine

phytoplankton species. Journal of Phycology 33,

625-631.

[24] Nielsen, L. T., Hallegraeff, G.M., Wright, S.W.,

Hansen, P.J., (2012). Effects of experimental

seawater acidification on an estuarine plankton

community. Aquatic Microbial Ecology 65, 271-

285.

[25] Nielsen, L.T., Jakobsen, H.H., Hansen, P.J., (2010).

High resilience of two coastal plankton

communities to twenty-first century seawater

acidification: evidence from microcosm studies.

Marine Biology Research 6, 542-555.

[26] Raven JA, Falkowski PG (1999). Oceanic sinks for

atmospheric CO2. Plant Cell Environ 22: 741–755.

[27] Riebesell, U., Tortell, P.D., (2011). Effects of ocean

acidification on pelagic organisms and ecosystems.

In: Gattuso, J.P., Hansson, L. (Eds.), Ocean

Acidification. Oxford University press, NewYork,

pp. 291-311.

[28] Riebesell, U., (2004). Effects of CO2 enrichment on

marine phytoplankton. Journal of Oceanography

60: 719–729.

[29] Riebesell,U., Wolf-Gladrow, D.A., Smetacek, V.,

(1993). Carbon dioxide limitation of marine

phytoplankton growth rates. Nature 361, 249-251.

[30] Reinfelder, J.R., Kraepiel, A.M.L., Morel, F.M.M.,

(2000). Unicellular C4 photosynthesis in a marine

diatom. Nature 407, 996-999.

[31] Round, F. E., R.M. Crawford & D.G. Mann.,

(1990). The Diatoms Biology and morphology of

the genera. Cambridge University press,

Cambridge: 747 pp.

[32] Santhanam, P and P. Perumal., (2008). Marine

plankton in Indian waters. Training Manual on GIS

and Marine Biodiversity (ed. John Milton), Loyola

College Pub., 1-12pp.

[33] Schippers, p., Lurling, M., Scheffer, M., (2004).

Increase of atmospheric CO2 promotes

phytoplankton productivity. Ecology Letters 7, 446-

451.

[34] Sobrino, C., M. L. Ward & P. J. Neale, (2008).

Acclimation to elevated carbon dioxide and

ultraviolet radiation in the diatom Thalassiosira

pseudonana: effects on growth, photosynthesis, and

spectral sensitivity of photo inhibition. Limnology

and Oceanography 53: 494–505.

[35] Spolaore. P, C. Joannis-Cassan, E. Duran A.

Isambert, commercial applications of microalgae, J.

Biosci. Bioeng., 2006,101, 87–96.

[36] Strickland, J.D.H., and T.R. Parsons. (1972). A

practical handbook of seawater analysis. Fish. Res.

Board Can. Bull. 167, 2nd ed. 310pp.

[37] Subrahmanyan, R. (1946). A systematic account of

the marine plankton diatoms of the Madras Coast.

Proc. Indian Acad. Sci., 24: 85-197.

[38] Torstensson, A., Chierici, M., Wulff, A., (2012).

The influence of increased temperature and carbon

dioxide levels on the benthic/sea ice diatom

Navicula directa. Polar Biology 35, 205-214.

[39] Tortell, P.D., Morel, F.M.M., (2002). Sources of

inorganic carbon for phytoplankton in the eastern

subtropical and equatorial Pacific Ocean.

Limnology and Oceanography 47, 1012-1012.

[40] Tortell, P.D., DiTullio, G.R., Sigman, D.M., Morel,

F.M.M., (2002). CO2 effects on taxonomic

composition and nutrient utilization in an equatorial

pacific phytoplankton assemblage. Marine Ecology

progress Series 236, 37-43.

[41] Tortell, P.D., J.R.Reinfelder and F.M.M. Morel,

(1997). Active uptake of bicarbonate by diatoms.

Nature 390: 243-244.

[42] Tortell PD, Rau GH, Morel FMM (2000). Inorganic

carbon acquisition in coastal Pacific phytoplankton

communities. Limnol Oceanogr; 485–1500.

[43] Venugopalan, V.K., and Paulpandian, A.L. (1989).

Methods in Hydrobiology. CAS in Marine biology,

Annamalai University, Parangipettai, India.

[44] Vonshak, A., Maske, H. (1982). Algae: growth

techniques and biomass production. In: Techniques

in bioproductivity and photosynthesis. (Ed.

Coombs, J., Hall, D.O.), p.66. Pergamon press Inc.,

Maxwell House, Fairview park, Elmsford, New

York, U.S.A.

[45] Wu, Y.P., Gao, K.S., Riebesell, U., (2010). CO2-

induced seawater acidification affects physiological

performance of the marine diatom Phaeodactylum

tricornutum. Biogeosciences 7, 2915-2923.

[46] Yoo, K I. (1991). Population dynamics of

dinoflagellate community in Masan Bay with a note

on the impact of environmental parameters. Mar.

Pollut Bull. 23: 185-188.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1055-1063


A Study on the Early UCC Strength of Stabilized Soil Admixed with

Industrial Waste Materials

JIJO JAMES AND P KASINATHA PANDIAN Department of Civil Engineering, Tagore Engineering College, Chennai – 127, Tamil Nadu, India


Abstract: The present study deals with the effect of additives on the early strength gain of soil stabilized with

conventional stabilizers like lime and cement. The development of UCC strength of stabilized soil admixed with

industrial wastes over a curing period of 2 hours, 3 days and 7 days was studied. The industrial wastes used in the

present work were ceramic dust and press mud obtained from pulverization of ceramic tiles from construction debris

and sugar industry respectively. Ceramic dust was used as admixture for cement stabilization and press mud was

used as admixture for lime stabilization of soil. The results of the study indicated that addition of press mud resulted

in increase in early strength when sufficient quantity of lime was available for reaction. The addition of ceramic dust

resulted in enhanced early strength in case of cement stabilization for all combinations studied but the maximum

strength gain was achieved for a specific percentage of cement and ceramic dust which can be considered as the

optimum dosage for the soil used in the study. The results of the study strengthen the fact established by earlier

studies that industrial wastes can be effectively used for soil stabilization purposes as additives to conventional

stabilizers.

Keywords: Cement Stabilization, Lime Stabilization, Press Mud, Ceramic Dust, Early strength.

Abbreviations:

UCC – Unconfined Compression

ICL – Initial Consumption of Lime

OLC – Optimum Lime Content

TRL – Transportation Research Laboratory

CBM – Cement Bound Material

RHA – Rice Husk Ash

Introduction:

In the present age, the only available land that man can

use for construction rarely suits for construction as most

of the so called good land has already been built upon.

As Civil Engineers, the onus has been thrust on us for

sustainable development of man conserving resources

for future generations as well. This has led to the

development of a branch of study called as Ground

Improvement Techniques for utilization of unusable

land. The basic principles of the ground improvement

have been acknowledged since mankind started

construction on and in the ground (ASCE, 1977).

Mitchell (1981) described a variety of ground

improvement technologies under six categories based on

principles. It is more appropriate to classify ground

improvement techniques under the following headings

viz. replacement, densification, consolidation /

dewatering, grouting, admixture stabilization, thermal

stabilization, reinforcement and miscellaneous methods

(Terashi and Juran, 2000).

With industrialization, another major problem that came

to the fore was pollution and solid waste production.

With rising turnover, industrial solid wastes produced

soon blew up to huge proportions. To cite an example,

the flyash production in India was 131.09 million tons

in the year 2010-11 (Central Electricity Authority,

2011). With research, one of the avenues for the

utilization of these waste materials came out to be their

use in soil improvement. The utilization of several

waste materials have improved over the years, the total

utilization of flyash produced in India stands at 73.13

million tons which is 55.79% of the total production in

the country (Central Electricity Authority, 2011). These

industrial by-products produce even better results when

combined with other materials like lime and cement that

have been used in soil improvement for long time

(Kavak et. al., Guleria and Dutta, 2011, Ramirez et. al.,

2012, Okonkwo et. al., 2012, Rahmat and Ismail, 2011).

Ordinary Portland cement is one of the most commonly

used stabilizers for soil stabilization. The addition of

cement to a material, in the present case soil, produces

hydrated calcium silicate and aluminate gels in the

presence of moisture, which crystallise and bond the

soil particles together. Most of the strength of a cement-

stabilised soil comes from the physical strength of the

matrix of hydrated cement (Al Rawas et. al., 2002).

Cement stabilized soils can be classified into three


1056 A Study on the Early UCC Strength of Stabilized Soil Admixed with




types, Soil Cement, Cement Bound Material (CBM) and

Lean Concrete (TRL, 2003). Soil cement usually

contains less than 5% cement (Lay 1986). CBM uses

granular material like crushed rock or gravel instead of

soil (Croney, 1988). Lean concrete has higher cement

content when compared to CBM and is more like

concrete rather than CBM (TRL, 2003). Lime

stabilization has been extensively studied by earlier

researchers. Lime basically reacts with medium,

moderately fine, and fine-grained soils to produce

reduced elasticity and swell and increased workability

and strength. Such improvement in soil properties are

the result of three basic chemical reactions: 1. Ion

exchange and flocculation; 2. Pozzolanic reaction; and

3. Carbonation (Fang 1991). Lime is a general term that

covers the following three types, Quick Lime (CaO -

Oxide of Calcium), Hydrated or Slaked Lime (Ca(OH)2

– Hydroxide of Calcium) and CaCO3 – (Carbonate of

Calcium). Conventionally only quick lime and hydrated

lime are used in stabilization of roads (TRL, 2003).

Utilization of calcium carbonate in soil stabilization

results in improvement through physical changes. Use

of calcium carbonate in the form of egg shell powder

resulted in increase in the UCC strength of the soil (Jijo

James and Kasinatha Pandian, 2013). Hydrated lime is

used for the stabilization of soil with high clay content

where the main advantage is in raising the plastic limit

of the clayey soil. Small quantities in the range of 1 -

3% are used to reduce the plasticity of the clay (TRL,

2003). In this study, an attempt has been made to study

the effects of such additives on the performance of

conventional soil stabilizers like lime and cement.

The primary objective of this work is to study the effect

of addition of admixtures on the early stabilization

potential of conventional stabilizers.

Materials and Methods:

The materials that were used in this study included

Lime and Cement that were used as the primary

stabilizers and Press Mud and Ceramic Dust, which

were industrial wastes used as respective admixtures.

The soil used for the study was sourced from a farm

land in Thatthamanji Village in Thiruvallur District of

Tamil Nadu, India in the month of January 2012. The

properties of the soil were tested in the laboratory and

the results are summarized in table 1.

Table1: Properties of Soil

S.

No. Property Value

1 Liquid Limit 68%

2 Plastic Limit 27%

3 Plasticity Index 41%

4 Shrinkage Limit 10%

5 Specific Gravity 2.76

6 %Gravel 0

7 %Sand 2.5

8 %Silt 60.5

9 %Clay 37

10 Maximum Dry Density 1.56g/cc

11 Optimum Moisture

Content 25%

12 UCC Strength 343.8kN/m2

13 Soil classification CH

The lime adopted for the study was commercial

laboratory grade lime with 95% purity. The main reason

for adopting laboratory grade lime rather than

commercial grade lime despite the latter being used in

the field was to have a good degree of control over the

test results, which is difficult with the latter. The cement

used for the study was commercially available ordinary

portland cement.

Ceramic Dust:

The ceramic dust is not available in the ready form to be

used in soil stabilization. Construction debris from a

demolition site was collected and the broken ceramic

tiles were carefully segregated keeping in mind only to

pick clean fragments without any plaster sticking to it.

The segregated ceramic tiles were then crushed and

ground to a powder using Deval’s abrasion testing

machine available. The fractions which were finer than

710 micron but coarser than 600 microns were used in

the study. The specific gravity of the ceramic dust was

determined in the laboratory as 2.23. This is in

agreement with the specific gravity of ceramic tile

aggregate given by Kamala and Krishna Rao (2012) and

Sekar et. al., (2011).

Press Mud:

Sugarcane press mud is the residue of the filtration of

sugarcane juice. The clarification process in the

sugarcane mill separates the juice which is sent for

processing whereas the residue or the mud collects at

the bottom. The mud is filtered to separate the

suspended matter, which includes insoluble salts and

fine bagasse. The world output of press mud is

estimated to be 30 million tonnes. It is mainly used as

soil conditioner, fertilizer and for wax production (Tran,

2012). In this study, an attempt has been made to study

its effect on the stabilization potential of lime. The

general composition of press mud is given in table 2

(ICAR, 2013). The press mud used in this study was

obtained from Jawaharlal Nehru Sugar Mills Ltd.

Eraiyur, Perambalur District, Tamil Nadu, India. The

Specific gravity of press mud as determined in the

1057 JIJO JAMES AND P KASINATHA PANDIAN



laboratory came out to be 1.11. A similar value was

given by Carmen Baez-Smith (2008).

Table2: Composition of Press Mud

S. No. Component %

1 Organic carbon 20 - 24

2 Nitrogen 1.26

3 Phosphorus Pentoxide 3.85

4 Potassium oxide 1.46

5 Calcium oxide 11

6 Magnesium 1.6

7 Sulphur 0.23

8 Copper, Zinc, Iron and

Manganese

Miniscule

amounts

Methodology:

The soil from the site was prepared in accordance with

IS2720:1983 Part 1. The soil was air dried, cleaned,

pulverized and then sieved to requirement. The soil was

then tested for its properties and classified (Table 1).

The trial percentages for stabilization were fixed, in the

present case the stabilizers being lime and cement. In

case of lime, two trial percentages were used namely the

optimum lime content (OLC) and the initial lime

consumption (ICL). The OLC was arrived at by

conducting UCC test (IS2720:1991 Part 10) on soil

samples with increasing lime content cured for 2 days.

The ICL was determined from the Eades and Grim pH

Test (ASTM D6276:1999a). The UCC samples were

prepared at the maximum dry density and optimum

moisture content obtained for stabilized soil. For

cement, the percentages were fixed on a trial basis in

increments of 1%. UCC tests on soil with stabilizers but

without any admixtures were taken as control tests.

Trial percentages of admixtures were then fixed

randomly and UCC specimens were prepared with

stabilizers and admixtures. Samples were prepared and

cured for a period of 2 hours, 3 days and 7 days in air

tight sealed polythene covers to study the early strength

gain.

Results and Discussion:

The lowest percentage of lime in soil that gives a pH of

12.4 is the approximate lime percentage for stabilizing

the soil. There may be some soils in which the pH is

greater than 12.4. If this occurs, the lowest percentage

of lime where the higher pH value does not rise for at

least two successive test samples at increasing lime

percentages is selected (ASTM D6276:1999a). Based

on this, the value of initial lime consumption as

determined from the Eades and Grim pH test for the

said soil was 5.5% as shown in figure 1.

0

2

4

6

8

10

12

14

0 1 2 3 4 5 6 7 8

pH

% Lime

Eades and Grim pH Test for Initial Lime Consumption

Figure1: Eades and Grim pH test for ICL

The OLC was determined from UCC tests with

increasing lime content in soil cured for a period of two

days. The lime content corresponding to the highest

UCC strength was taken as the OLC for the soil, which

came out to be 7% as shown in figure 2.

0

200

400

600

800

1000

1200

1400

0 1 2 3 4 5 6 7 8 9 10

UCC

Str

engt

h (k

N/m

2 )

% Lime

Determination of Optimum Lime Content

Figure2: Optimum Lime Content from UCC Tests

Effect of Press Mud on Early Strength of Lime

Stabilized Soil:

The effect of early strength of stabilized soil was

studied by conducting unconfined compression tests on

soil stabilized using stabilizers and admixtures cured

over a period of 2 hours, 3 days and 7 days.

0

200

400

600

800

1000

1200

1400

1600

0 1 2 3 4 5 6 7

UCC

Str

engt

h (k

N/m

2 )

% Press Mud

Variation of UCC Strength with % Press Mud on 5.5% Lime Stabilized Soil

2 Hours 3 Days 7 Days

Figure3: Variation of UCC strength with % press mud

on 5.5% lime stabilized soil





Figure 3 shows the variation of UCC strength with press

mud on 5.5% lime stabilized soil for three different

curing periods viz. 2 hours, 3 days and 7 days. Addition

of press mud as an admixture to 5.5% lime stabilized

soil shows a loss in strength with increasing press mud

content. The same holds good for all three curing

periods. However, at 2% addition of press mud, the

strength of the stabilized soil is higher than that of pure

lime stabilized soil when cured for a period of 7 days.

Similar studies have been carried out by various

researchers with other industrial waste materials.

Addition of RHA to lime resulted in enhanced strength

gain compared to pure lime stabilization. 7% lime gave

better performance compared to 5% and 3% lime

contents. At 7 days curing, 7% lime with 12% and 18%

RHA gave a UCC strength of around 0.7MPa (Jha and

Gill, 2006). In the present study as well, higher lime

content of 7% gave better strength compared to 5.5%

lime stabilization. The addition of 2% press mud with

7% lime in soil stabilization produced UCC strength

close to 2 MPa at 7 days of curing. Silica fume gave

excellent early strength to lime stabilized soils. The

strength at 7 days of curing jumped from 11.8 kPa to

around 45 kPa for addition of 15% silica fume-lime mix

to soil in the ratio of 80:20 (Yarbasi et. al., 2007).

Figure 4 shows the variation of UCC strength with press

mud on 7% lime stabilized soil for three different curing

periods viz. 2 hours, 3 days and 7 days. From figure 4, it

can be seen that a similar trend as noticed in 5.5% lime

stabilized soil is also seen here with the exception that

the strength gain at 2% press mud addition is seen at 3

days of curing itself when compared to 7 days in the

previous case for the same.

0

500

1000

1500

2000

2500

0 1 2 3 4 5 6 7

UCC

Str

engt

h (k

N/m

2 )

% Press Mud

Variation of UCC Strength with % Press Mud on 7 % Lime Stabilized Soil


Figure4: Variation of UCC strength with % press mud

on 7% lime stabilized soil

However, one significant point that stands out in this

case is that the strength of 7% lime stabilized soil is

lower than 5.5% lime stabilized soil in the presence of

press mud at 2 hours curing. The combination lime–

natural pozzolana produced higher strength values than

lime or natural pozzolana alone (Khelifa Harichane et.

al., 2012). Strength gain at early curing periods was

significant due to the addition of natural pozzolana to

lime.

Figure 5 shows the variation of early strength of 5.5%

lime stabilized soil with curing period for the four

combinations under investigation viz. 5.5% lime, with

2%, 4% and 6% press mud as admixture. From figure 5,

it can be seen that, the general trend is that of strength

gain with curing period.

0

200

400

600

800

1000

1200

1400

1600

0 1 2 3 4 5 6 7 8

UCC

Str

engt

h (k

N/m

2 )

Curing Period (Days)

Variation of UCC Strength with Curing on 5.5% Lime Stabilized Soil

5.5% Lime 5.5% Lime + 2% Press Mud

5.5% Lime + 4% Press Mud 5.5% Lime + 6 % Press Mud

Figure5: Variation of UCC strength with curing period

for 5.5 % Lime stabilized soil

All curves show an upward trend with age when seen

individually. However, if the curves are compared

among themselves, there is a significant drop in strength

on addition of 6% press mud. Even on addition of 2%

press mud results in strength of 1415.78kN/m2, a

meagre gain of 31.7kN/m2, which is only 2.3% more

than that of 5.5% lime stabilized soil at 1384.08kN/m2.

0

500

1000

1500

2000

2500

0 1 2 3 4 5 6 7 8

UCC

Str

engt

h (k

N/m

2 )

Curing Period (Days)

Variation of UCC Strength with Curing on 7% Lime Stabilized Soil

7% Lime 7% Lime + 2% Press Mud

7% Lime + 4% Press Mud 7% Lime + 6 % Press Mud

Figure6: Variation of UCC strength with curing period

on 7% lime stabilized soil

Figure 6 shows the variation of strength of 7% lime

stabilized soil with curing period for the various

combinations under investigation. When compared to

5.5%lime stabilized soil with press mud, 7% lime

stabilized soil with press mud performs significantly

better. As seen in the earlier case, the general trend is

that of increasing strength with curing period

individually and reduced strength on increasing press

mud addition with the exception of 2% press mud

addition. The early strength depends upon initial water




content, stabilizer content and curing period (Wang Zhe

et. al., 2012). In the present work, all samples were

moulded with same water content and use admixtures

along with stabilizers. But still it is in agreement with

the report that strength increases with stabilizer content

as 7% lime performed better compared to 5.5% lime

stabilization. The curve for 2% press mud addition

significantly diverges from the other set of curves from

3 days of curing. The strength gain at 3 days curing for

2% press mud addition is 283.08kN/m2 which is 31.3%

more than the strength of 7% pure lime stabilized soil at

905.295kN/m2 for the same curing period. At 7 days of

curing, this gain increases to 509.24kN/m2, a 34.75%

increase over lime stabilized soil at 1465.16kN/m2.

Kavak et. al., (2011) reported addition of 3.33% Ground

Granulated Blast furnace Slag to 5% lime resulted in

better performance compared to 5% pure lime

stabilization. The presence of GGBS in lime

stabilization resulted in a strength gain of approximately

36% at 7 days of curing. Addition of waste cement dust

to lime gave better early strength of lime stabilized

black cotton soil when compared to pure lime

stabilization of the same. It resulted in a strength gain of

50% at 3 days curing and 25% gain at 7 days curing for

8% addition of waste cement dust-lime combination in

1:1 ratio. (Oza and Gundalia, 2013).

The reason behind the better performance of 7% lime

case, over 5.5% lime case may be due to the fact that

ICL lime content is used to raise the pH of the soil to

12.4 thereby producing favourable conditions for

stabilization but results in no extra lime available for the

actual stabilization. When 7% lime is added, there is

additional lime available beyond the initial consumption

of 5.5% for the pozzolanic reactions to take place

between soil and lime. For strength gains to occur, the

chemical reactions require a high pH (>12.4) to be

maintained, which is the ICL value (TRL, 2003). It is

well documented in literature that lime addition in

excess of ICL results in pozzolanic reactions leading to

strength gain (Nazrisar et. al., 2008, 2010). A possible

argument can be put forward to explain the poor

strength of 7% lime stabilized soil admixed with press

mud when compared to 5.5% lime stabilized soil

admixed with press mud at 2 hours curing. It may be

possible that the products of reaction between lime and

press mud may have poor rate of strength gain. In 5.5%

lime case, there isn’t enough lime available to react with

press mud beyond initial consumption, thereby resulting

in minimum interaction between lime and press mud.

Therefore early strength is better but gain is minimal. In

the 7% lime case, there is additional lime available

beyond initial consumption, which can lead to lime

press mud reaction products. This can explain poor

early strength but significant gain in strength with time.

However, the veracity of this argument can only be

confirmed upon detailed investigation. Chen and Lin

(2009) reported that for small quantities of sewage

sludge ash added to cement in soil stabilization, the

early strength gain was small but strength increased

with curing period due to continuation of pozzolanic

reactions. Similar trends are seen in the present study

with strength of 7% lime with press mud stabilized soil

having low strength at early age but gaining

significantly as curing increased.

Effect of Ceramic Dust on Early Strength of Cement

Stabilized Soil:

Usually, for soils with plasticity index greater than 20%,

cement stabilization will be only marginally effective

(TRL, 2003). The soil under study has a plasticity index

of 41% and hence plain cement stabilization may give

only marginal results. Hence, it was aimed to increase

the effectiveness of cement stabilization with

admixtures on such soils. Addition of inert filler,

resulting in greater bulk would aid the distribution

process so that the same amount of active cement would

be available throughout the material (TRL, 2003).

Hence, ceramic dust was used as an admixture

principally to add bulk. Particle size coarser than 600

microns but finer than 710 microns was used for the

study. This range was a balance that was necessitated to

provide bulk while reducing interference with

performance of index tests.

Soil cement consists of less than 5% cement (Lay,

1986). Several studies have been carried out with

additives to cement in soil stabilization. Rice husk ash

as an admixture to cement stabilization of soil resulted

in greater strength at 7 days of curing compared to plain

cement stabilization. It can be used to reduce the content

of cement used in stabilization due to the increased

strength achieved (Basha et. al., 2005). Increase in

ceramic dust content to expansive soil resulted in

increase in UCC strength of the soil up to 30% content

(Sabat, 2012). In the present work, ceramic dust was

used as an admixture rather than a stabilizer as done by

Sabat (2012).

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

0 1 2 3 4 5 6 7 8 9

UCC

Str

engt

h (k

N/m

2 )

% Ceramic Dust

Variation of UCC Strength with % Ceramic Dust of 1% Cement Stabilized Soil


Figure7: Variation of UCC strength with % Ceramic

dust for 1% cement stabilized soil





Cement stabilization was carried out by adopting trial

percentages starting from 1% to 4% in increments of

1%. Ceramic dust admixture was added in increments of

2% to the soil. The effect of addition of ceramic dust to

1% cement is shown in figure 7.

0

1000

2000

3000

4000

5000

6000

0 1 2 3 4 5 6 7 8 9

UCC

Str

engt

h (k

N/m

2 )

% Ceramic Dust





From figures 7 to 10, it can be seen that the addition of

ceramic dust to cement resulted in the increase in the

early strength of the soil. It can be seen in all four cases

that the gain in strength is significant after three days of

curing. At lower cement content, viz. 1% and 2% there

is a dip in performance at 4% addition of ceramic dust

whereas at higher cement content, viz. 3% and 4%,

there is dip in performance at 6% addition of ceramic

dust at 7 days of curing. Egg shell ash admixed cement

stabilization of soil performed better than cement

stabilization after 7 and 14 days of curing. UCC

Strength at 6% and 8% cement content increased

considerably from (370KN/m2 and 471KN/m

2) for 7

days curing period and (432KN/m2 and 655KN/m

2) for

14 days curing period to (614KN/m2 and 687KN/m

2)

and (680KN/m2 and 988KN/m

2) respectively with

increase in eggshell ash content from 0% to 10%

(Okonkwo et. al., 2012). The strength of 2% cement

stabilized soil when admixed with 6% ceramic dust

resulted in the doubling of early strength for 7 day

curing period for the soil under study in the present

work.

0

1000

2000

3000

4000

5000

6000

0 1 2 3 4 5 6 7 8 9

UCC

Str

engt

h (k

N/m

2 )

% Ceramic Dust





The initial strength gain between 2 hours and 3 days of

curing was poor across all four cement content cases.

But out of all combinations of cement, the greatest gain

in strength was observed for 2% cement with ceramic

dust as admixture. The average gain in strength between

3 days and 7 days across all combinations of ceramic

dust addition was 3351.92kN/m2 for 2% cement

stabilization which is a staggering 611% gain on the

average.

0

500

1000

1500

2000

2500

3000

3500

4000

4500

0 1 2 3 4 5 6 7 8 9

UC

C S

tre

ngt

h (k

N/m

2)

% Ceramic Dust





In comparison, 1% cement with ceramic dust had a gain

of 168%, 3% cement with ceramic dust had a gain of

384% and 4% cement with ceramic dust had a gain in

strength of 322%. Addition of calcined paper sludge to

cement in soil stabilization produced similar strength to

that of cement stabilized soil at higher cement dosage at

7 days curing (Amaia Lisbona et. al., 2011). In this

study, addition of ceramic dust to cement has resulted in

significant gain in early strength of stabilized soil.

Hence, it can be inferred that for achieving a particular

strength of stabilized soil, the addition of ceramic dust

to cement can help in achieving the same strength at

lower cement content due to the contribution of ceramic

dust in strength gain.

Figures 11 to 14 show the gain in strength with curing

period for various levels of admixtures for cement

content varying from 1% to 4%. For 1% cement

content, addition of 8% ceramic dust resulted in a

strength gain of 2294.22kN/m2 when compared to pure

cement stabilization, which is a gain of 94%.




0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

0 1 2 3 4 5 6 7 8

UCC

Str

engt

h (k

N/m

2 )

% Ceramic Dust

Variation of UCC Strength with Curing for 1% Cement Stabilized Soil

1% Cement 1% Cement + 2% Ceramic Dust

1% Cement + 4% Ceramic Dust 1% Cement + 6% Ceramic Dust

1% Cement + 8% Ceramic Dust

Figure11: Variation of UCC strength with Curing for

1% Cement stabilized soil

For 2% cement content, the gain was 2778.45kN/m2, a

jump of 105% at 6% ceramic dust addition, for 3%

cement, an increase of 156% corresponding to

3035.22kN/m2 at 8% ceramic dust and 4% cement

content addition had a maximum gain of 2105.4kN/m2

which is 106% rise at 8% ceramic dust addition.

-1000

0

1000

2000

3000

4000

5000

6000

0 1 2 3 4 5 6 7 8

UCC

Str

engt

h (k

N/m

2 )

% Ceramic Dust







Increase in sewage sludge ash content in cement

stabilization of soil resulted in increase in the early

strength gain of the stabilized soil at 7 days curing. At

lower ash content, the strength gain was delayed but at

higher ash content the compressive strength gain was

better due to air moisture (Chen and Lin, 2009). This is

similar to the present work where the strength at lower

levels of ceramic dust dosage resulted in delayed

strength gain across most of the combinations.

0

1000

2000

3000

4000

5000

6000

0 1 2 3 4 5 6 7 8

UC

C S

tre

ngt

h (k

N/m

2 )

% Ceramic Dust







0

500

1000

1500

2000

2500

3000

3500

4000

4500

0 1 2 3 4 5 6 7 8

UC

C S

tre

ngt

h (k

N/m

2 )

% Ceramic Dust







Though the maximum early gain in strength with

respect to initial strength was achieved by 3% cement

with 8% ceramic dust combination, the absolute

maximum strength gain was achieved by 2% cement

with 6% ceramic dust. In all four combinations, it can

be seen that higher content of ceramic dust produced

better strength.

This may be due to the fact that as stated earlier in

literature, higher ceramic dust content may have

resulted in providing bulk ensuring more even

availability of cement throughout the soil. This in turn

would have produced better hydration of cement and

reaction between cement and soil particles.

Conclusions:

Several researchers have studied the use of industrial

wastes as additives and stabilizers. They have found in

general that several of the industrial wastes can be used

as an admixture to conventional stabilizers like lime and

cement. They have also reported that such industrial

wastes enhance the performance of the stabilizers by

several times. Some of the industrial wastes can be also

used as standalone stabilizers as well. This work also





aims at studying two of such industrial wastes, viz.

ceramic dust and press mud in their use as admixtures

for cement and lime respectively.

The results of the tests on stabilized soils admixed with

the aforementioned admixtures leads to the following

conclusions. Addition of press mud to lime enhances the

strength of stabilized soil at early age when compared to

lime stabilization alone. But, the gain in strength is

significant only when sufficient lime is available

beyond the ICL. However, more in-depth study needs to

be done on the performance of the additive under

rigorous conditions to recommend the suitability of

press mud as an additive to lime for all situations.

Addition of ceramic dust to cement in soil stabilization

resulted in significant strength gain at an early age. The

gain in strength is better at higher ceramic dust content

for all cement contents tested, which may be due to it

adding bulk, resulting in an even distribution of active

cement. However, more detailed investigations need to

be performed to really understand the changes that are

being brought about by the addition of ceramic dust to

cement at the micro level.

Acknowledgements:

The authors would like to thankfully acknowledge the

suggestions of the reviewers from Sri Sairam

Engineering College, Anna University, B.S.Abdur

Rahman University and Sathyabama University in the

betterment of this research paper. The authors would

also like to thank Prof. Dr. D. V. Reddy, Editor-in-

Chief, IJEE for kindly considering our work for

publication.

Reference:

[1] Akshaya Kumar Sabat (2012), “Stabilization of

Expansive Soil Using Waste Ceramic Dust”,

Electronic Journal of Geotechnical Engineering,

Vol. 17 Bund. Z pp. 3915 – 3926

[2] Amaia Lisbona, Iñigo Vegas, Javier Ainchil and

Carolina Ríos, “Soil Stabilization with Calcined

Paper Sludge: Laboratory and Field Tests”, Journal

of Materials in Civil Engineering, doi: 10.1061/

(ASCE) MT.1943-5533. 0000437

[3] Amer Ali Al-Rawas, Ramzi Taha, John D. Nelson,

Thamer Beit Al-Shab, and Hilal Al-Siyab, (2002),

“A Comparative Evaluation of Various Additives

Used in the Stabilization of Expansive Soils”,

Geotechnical Testing Journal, GTJODJ, Vol. 25,

No. 2, pp. 199–209.

[4] ASCE (1977) “Soil Improvement History,

Capabilities, and Outlook”. ASCE Geotechnical

Special Publication No. 57, 177 p.

[5] ASTM D6276:1999a - Standard Test Method for

Using pH to Estimate the Soil-Lime Proportion

Requirement for Soil Stabilization

[6] Aydin Kavak, Gamze Bilgen, and Omer Faruk

Capar, (2011) “Using Ground Granulated Blast

Furnace Slag with Seawateras Soil Additives in

Lime-Clay Stabilization”, Journal of ASTM

International, Vol. 8, No. 7, Paper ID JAI103648

[7] Basha, E. A., R. Hashim, H.B. Mahmud and A.S.

Muntohar (2005), “Stabilization of residual soil

with rice husk ash and cement”, Construction and

Building Materials 19: 448–453

[8] Carmen Baez-Smith, P.E. (2008), “Production of

Bioenergy Using Filter Cake Mud in Sugar Cane

Mill Factories”, Sugar Processing Research

Institute Conference, September 29th

to October 1st,

2008, Delray Beach, Florida.

[9] Central Electricity Authority, (2011), Report on Fly

Ash Generation at Coal/Lignite Based Thermal

Power Stations and Its Utilization in The Country

For The Year 2010-11, Government of India.

[10] Croney, D and Croney, P. (1998) the Design and

Performance of Road Pavements. 3rd Edition,

Published by McGraw-Hill, UK, 1998

[11] Fang, H. Y., 1991, Foundation Engineering

Handbook, 2nd ed., Van Nostrand Reinhold

Publishing Company.

[12] Guleria, S. P. and Dutta, R. K. (2011) - Technical

Note, “Unconfined Compressive Strength of Fly

Ash–Lime–Gypsum Composite Mixed with

Treated Tire Chips”, Journal of Materials in Civil

Engineering, August 2011, pp. 1255-1263

[13] IS2720:1983 - Part 1 Preparation of Dry Soil

Samples For Various Tests

[14] IS2720:1991 - Part 10 Determination of

Unconfined Compressive Strength

[15] Jha, J. N. and Gill, K. S. (2006), “Effect Of Rice

Husk Ash On Lime Stabilization”, Journal of the

Institution of Engineers (India), Volume 87, pp. 33-

39

[16] Kamala, R. and Krishna Rao, B. (2012), “Reuse of

Solid Waste from Building Demolition for the

Replacement of Natural Aggregates”, International

Journal of Engineering and Advanced Technology,

Volume 2, issue 1, pp. 74 - 76

[17] Khelifa Harichane, Mohamed Ghrici and Said

Kenai (2012), “Effect of the combination of lime

and natural pozzolana on the compaction and

strength of soft clayey soils: a preliminary study”,

Environ Earth Science, 66:2197–2205

[18] Lay, MG. (1986) Handbook of Road Technology.

Includes Chapter 10: Stabilisation (redrafted 1988).

Published by Gordon & Breach, London

[19] Li Chen and Deng-Fong Lin (2009), “Stabilization

treatment of soft subgrade soil by sewage sludge

ash and cement”, Journal of Hazardous Materials

162 : 321–327




[20] Mitchell, J.K. (1981) “State of the Art – Soil

Improvement”. Proceedings of the 10th ICSMFE.

Stockholm, Vol. 4, pp. 509-565.

[21] Mohamad Nidzam Rahmat and Norsalisma Ismail,

(2011), “Sustainable stabilisation of the Lower

Oxford Clay by non-traditional binder”, Applied

Clay Science 52 pp. 199–208

[22] Nasrizar, A.A., Muttharam, M. & Ilamparuthi, K.

(2008), “Effect of curing temperature on strength of

lime stabilized expansive soil”, Indian Geotechnical

Conference (IGC) 2008, Bangalore, Vol 2:28-30

[23] Nasrizar, A.A., Muttharam, M. & Ilamparuthi, K.

(2010), “Role of Lime Content on Soil-Lime

Reaction under Thermal Curing”, Indian

Geotechnical Conference – 2010, GEOtrendz, pp.

595 – 598

[24] Necmi Yarbası, Ekrem Kalkan, and Suat Akbulut

(2007), “Modification of the geotechnical

properties, as influenced by freeze–thaw, of

granular soils with waste additives”, Cold Regions

Science and Technology 48: 44–54

[25] Okonkwo, U. N, Odiong, I. C. and Akpabio, E. E.

(2012), “The Effects of Eggshell Ash on Strength

Properties of Cement-Stabilized Lateritic”,

International Journal of Sustainable Construction

Engineering & Technology (ISSN: 2180-3242) Vol

3, Issue 1, pp. 18 – 25

[26] Oza, J.B. and Gundaliya P.J. (2013), “Study of

black cotton soil characteristics with cement waste

dust and lime”, Procedia Engineering 51: 110 – 118

[27] Rafael Alavez-Ramirez, Pedro Montes-Garcia,

Jacobo Martinez-Reyes, Delia Cristina Altamirano-

Juarez, Yadira Gochi-Ponce, (2012), “The use of

sugarcane bagasse ash and lime to improve the

durability and mechanical properties of compacted

soil blocks”, Construction and Building Materials

34 pp. 296–305

[28] Sekar, T., Ganesan, N. and Nampoothiri, NVN.

(2011), “Studies on Strength Characteristics on

Utilization of Waste Materials as Coarse Aggregate

in Concrete”, International Journal of Engineering

Science and Technology, Volume 2, No. 7, pp.

5436 - 5440

[29] Sugarcane Breeding Centre, Indian Council of

Agricultural Research,

http://www.sugarcane.res.in/index.php/component/

content/article/15-faq/322-faq-organic-cultivation-

of-sugarcane, Last Retrieved on 28/11/2013

[30] Terashi M, and Juran I, (2000) Ground

improvement: State of the Art, GeoEng 2000,

Melbourne, Australia

[31] Tran G., 2012, Sugarcane Press Mud,

Feedipedia.org. A Programme by INRA, CIRAD,

AFZ and FAO.

http://www.feedipedia.org/node/563, Last

Retrieved on 28/11/2013

[32] Transportation Research Laboratory, UK, (2003),

“Literature Review Stabilised Sub-Bases for

Heavily Trafficked Roads”, PROJECT REPORT

ORIGIN PR/INT/202/00

[33] Wang Zhe, Xu Si-fa and Wang Guo-cai (2012),

“Study of Early Strength and Shrinkage Properties

of Cement or Lime Solidified Soil”, Energy

Procedia 16: 302 – 306.

http://www.sugarcane.res.in/index.php/component/content/article/15-faq/322-faq-organic-cultivation-of-sugarcane



http://www.feedipedia.org/node/563

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1064-1071


An Analysis of Earthquake Information Extraction based on GIS

and RS

ZUOWEI HUANG1, 2

, WEI HUANG1 AND ZOU YU

2

1School of Geosciences and Information-Physics, Central South University, Changsha 410083, China

2School of Architecture and urban planning, Hunan University of Technology, Zhuzhou 412008, China

Email:[email protected]

Abstract: Earthquake, being bursty, transitory and powerful is thought as one of the greatest natural disasters in the

worl, it is very important for earthquake emergency to quickly acquire disaster information. In order to extract the

information of earthquake damage effectively and timely, based on the Spatial information technology, the paper,

take Yushu earthquake as example, explored a new method of information extraction, which uses Moravec operator

to extract feature points and uses RANSAC algorithm to reject mismatching points, the images before and after the

earthquake are processed as classification targets, proposed 3D wavelet transform and 3D curvelet transform based

multi-resolution 3D seismic image data fusion method, it can be better reflect the correlation of geological formation

in seismic interpretation. Finally establishment of the information service system by means of Arc IMS and

distributed database technology. The results show that the designed method is of the high robustness and

information extraction precision from multi-source images. it can be better applied to the assessment of the

earthquake damage ,emergency management and rescue mission than before.

Keywords: information extraction, ArcIMS, image fusion, earthquake disaster.

1. Introduction:

As we all known earthquake is one of the major natural

disasters in world. all of a sudden the devastating

disaster caused by a strong earthquake will bring great

harm to life and property. But the occurring time and

location of the earthquake still cannot be predicted

timely and accurately before the earthquake, Due to

lower accuracy of seismic disaster information

extraction, singularity of seismic disaster recognition

target and lack of complete seismic emergency response

software platform, so that active defense measures

should be taken to reduce the disaster degree[1,2].

GIS and Remote sensing technology can provide a

quick effective approach to access the disaster

information and loss after the earthquake, owing to its

advantage of the quick dynamically all-weather

monitoring with large amount of information and a short

update cycle [3-5]. Especially with the development of

the high-resolution satellite sensors and aerial remote

sensing technology (including altitude unmanned aerial

vehicles and other platforms), remote sensing

technology has substantially enhanced its capacity of

the rapid access to disaster information. So rapid and

comprehensive access to disaster information after the

earthquake is of great significance to carry out rescue

actions and reduce disaster losses [6, 7]. It was the first

time of the rapid production to seismic intensity

envelope line with human-computer interaction and

Geographical Information System (GIS). With the

emergence of high resolution remotely sensed imagery

and development of remote sensing information

extraction technology in recent years, RS became an

effective means for acquiring seismic disaster

information rapidly, emergency response and seismic

disaster assessment. For example, in Japan Nioki and

Furmio (2000) used aerial remote sensing image to

investigate the damage of Kobe Earthquake [9,10]. In

Turkey Earthquake, Athens Earthquake in Greece,

Sumatra Earthquake Tsunami in Indonesia, The satellite

remote sensing data were firstly used to monitor and

evaluate the earthquake damage situation from the Ms

7.6-magnitude earthquake at Nantou County, Taiwan, in

September 2l, 1999 (Wei et al., 2000).and Haiti

Earthquake, remote sensing technology were used to

investigate and assess the damage situation and the

losses after the earthquake. in modern times with the

rapidly development of the computer network

technology and high-resolution satellite sensors, remote

sensing image processing technology has substantially

enhanced its capacity of the rapid access to disaster

information[11,12].

2. Study Area and Technical Flow:

At 7:49 am on April 14 2010, an earthquake happened

(33.2°N, 96.6°E) near Jiegu Town, Yushu Tibetan

1065 ZUOWEI HUANG, WEI HUANG AND ZOU YU



Autonomous Prefecture, Qinghai Province, China.

According to the three elements of Yushu Earthquake

published by China Earthquake Administrator, Detailed

information for the adopted data is: SPOT-5 ortho-

image and “HJ1B-CCD” image data, 1:10000 scale

fundamental geographic data, field work data from State

Bureau of Surveying and Mapping and Ministry of Land

and Resources, population data, social economic data,

hydrological data, geological data, etc. Its focal depth

was 14 km below the ground surface. It killed more than

24000 people.

Based on these geographical spatial data, it establish

uniform data integration framework, so that we can

conduct data format conversion, the projection

transformation, geography association, remote sensing

image enhancement, establishing spatial index

relationship. In order to process remote sensing image

rapidly, shorten the data processing time, based on GIS

and RS the technical flow is constructed.

3. Method:

3.1. RS Image matching Approach Based on Improved

RANSAC Algorithm:

The basic idea of RANSAC is: during the processing of

parameter estimation, it differentiated treatment for all

possible input data, first design a judgment criterion

against the specific problem and use the judgment

criterion to rule out the data those are not consistent

with the estimated parameter in iteration way. Using

RANSAC algorithm to remove the abnormal pixels in

two image overlap, thus find out the best linear

regression with maximized data points to support the

linear transformation model parameter. through the

correct input data to estimate the parameters. It calls for

a certain confidence probability (p), error rate of data

(ε), meet the following relations between the minimum

sample number M an P (P > ε).

1 (1 (1 ) )m Mp (1)

Where the ε represents erroring rate of data, m

represents the matching point number of model

parameters.

In order to improve the ability of anti-interference in the

matching of image matching, a novel matching method

is put forward based on improved RANSAC algorithm

and SIFT. The basic matrix occupies a very important

position in the field of computer vision. Firstly it using

SIFT to extract invariant feature from images and pre-

matching. Secondly, the mismatching is wiped off by

improved RANSAC algorithm.

Given a pair of corresponding points m = [u, v, 1], m' =

[u', v', 1], it meet the following geometric constraints

relation: 0FmmT (2)

Where basic matrix

333231

232221

131211

FFF

FFF

FFF

F (3)

Based on Basic matrix F, for a pair of corresponding

matching points m and m', the Sampson distance is;

2

2

2

1

2

2

2

1 )'()'()()( MFMFFmFm

Fmmd

TT

T

(4)

The image matching process is as follows:

1) Input two images respectively (imag1 and image2),

Detect key points based on image1 and image2 and

get the set of matching points (mi↔m'i,i=1,2,...n),

then normalize the coordinate of points.

2) From the set of matching points, select eight

matching points at random, calculate its basic

matrix Fi based on RANSAC algorithm.

3) Based on the Sampson distance and t (the given

threshold value in advance), searching for all points

which meet the requirement (d<t) from the set of

matching point (mi↔m'i,i=1,2,...n), then view them

as inliers and record the number of inliers points

cater to the constrain of Fi.If the number of inliers

points is greater than the given threshold value,

then keep it, otherwise give up.

4) Repeat step (2) and (3) k times, record the number

of inliers points every time.

5) Selecting FL corresponding to the largest inliers,

taken it as the final inliers which also meet the

requirement of FL constrain. As for points which

cannot meet the requirement (d<t), then view them

as mismatching points and be wiped off.

3.2. 3D image fusion algorithm based on Wavelet

Transform:

Because the factors in the time direction to be more

considered in 3D seismic image fusion compared to the

2D seismic image fusion algorithm, so that the seismic

fusion image can be better reflect the correlation of

geological formation in seismic interpretation. How to

extract features in each 3D image, the main purpose is

preserves the main features. Because in the main texture

image of the earthquake, mainly including river, fault,

and some geological profile. These features in the high

frequency part, how to fusion effectively and get the

best effect of fusion, it resort to Wavelet Transform

,decomposition the 3D seismic image data in multi-scale

at first; Then using different fusion operator in low

frequency and high frequency respectively; finally, get

the 3D fusion image after reconstruction.

Specific decomposition for three-dimensional data

requires several steps, as follows: first of all, conduct

one-dimensional wavelet decomposition for each

1066 An Analysis of Earthquake Information Extraction based on GIS and RS



column of the three-dimensional body and get two 3d

data volume: low frequency L and high frequency data

H. Second, each row data, from the decomposed image

data volume, conducts one dimensional wavelet

Decomposition then get four three-dimensional data:

LL, LH, HL, HH. The main processing flow is shown in

Fig.1.

3D GIS technology is useful to the earthquake relief

and also a powerful assisting tool for remote sensing

image interpretation, The main drawback of the

traditional quadtree data organization model of 3D GIS

lies in its close coupling between data organized way

and data storage way. On the contrary virtual quadtree

brings it separate. It only build quadtree organizational

framework of spatial data and not care about the data

storage way and storage location. By means of

mounting the spatial data adhere to virtual quadtree, so

it set up the pyramid model of multilevel spatial

data.VQT has two types of node:

(1) Information node (Iqnode): the backbone of VQT

and not include specific spatial data.

(2) Data node (Dqnode): specific data storage space. it

mounts to iqnode nodes by means of registeration,

combine into the VQT structure, then forms the

pyramid model of multilevel massive spatial data.

In order to build the quadtree structure and mount

spatial data, Iqnode provides a series of properties and

methods. The main properties and methods are as

follows:

Scale (): it record the scale and resolution of the current

in mounting iqnode data.

Data Location (): it record the storage path of the

current mounting iqnode data.

Visit (): it used to access the spatial data of data

Location attribute.

Load (): it implements the Mounting way of spatial data

move to the iqnode node. When there is spatial data

move to iqnode, Iqnode call Load method and then the

spatial data storage path will be registered to iqnode

node. When some system access iqnode nodes data, it

calls Visit methods to Visit the spatial data of data

Location route.

Unload (): it implement the unload method of spatial

data.

Dqnode data node may be a data node, or it may be a

centralized Quadtree. Aim to the dqnode integrated into

the VQT Seamlessly, dqnode needs to provide a unified

data format and data access methods, such as WMS,

WFS and so on.

Virtual Quad Tree registration mechanism make it

easier when deal with subtree mounting of quadtree and

double tree Fusion algorithm, it need not data

replication and will not damage the original quadtree

structure due to Data coverage.

For Example: If we think Subtree td from B point of

VQT-II node fusion and replace Subtree sd from A

point of VQT-I node. It only need to unload subtree sd

from A node and bring the subtree of B node mount

tonode A. Double tree fusion algorithm as shown in

Fig2.

3.3. Overlapping display of image data with vector

data in different coordinate systems:

If the image data with vector data can be overlapping

display, then the information of seismic damage area

will be display more vivid and visualized, it provided

for rapid disaster assessment and emergency command

ability. If it was used the same projection coordinate

system, the overlapping display is no problem, if not,

we must conduct projection coordinate transformation,

but there are many kinds of projection pattern, and Arc

Fig2: Two trees merge chart of quadtree

Fig1: three dimensional discrete wavelet

decomposition flows




Map, image and vector overlapping display tools,

automatically loading the map projection to the pattern

of the first loaded after map converting, as there is

drawback of network delay, Arc IMS is not displayed in

accordance with the Arc Map, but the true projection of

the original map. To solve this problem, the image map

service and vector map service is relatively independent

of the application program, on the map window we

submitted two map services to the Arc IMS Application

Server at the same time, when displaying vector map

overlay on above of the image map, setting background

color of the vector map to transparent: Map Vector. Set

Background ("255.255.255"), you can see the lower

layer image, while computing the four corners

coordinate of the image map, and then transformed into

the the same coordinates of the vector coordinate

system, so that screen coordinates corresponding to

image coordinates and vector coordinates In this way, it

not only realize overlapping display of image data with

vector data at the web browser-side, but also ensure the

consistency of the coordinate system.

3.4. Network image processing technology based on

ION:

IDL, the product of RSI company ,based on Internet

scientific visualization development tools, has a lot of

analysis toolkit, map projection and software switch

package, adopting high speed graphic image processing

technology, which can quickly realize image processing

function and image classification function. ION Script

adopts CGI network realization mode, in general it is

comprised by ION Script and ION Java, and can be

used alone, also can combine to generate IDL

application in interactive environment. Take advantage

of HTML language and Java script language; it can be

easily developed network information service system.

Through Script interpreter in the ION server it will

convert ION script into HTML language, Script

application is a dynamic output HTML page, the

network image processing flow is shown in Fig.3.

4. Results and Analysis:

4.1. Analysis Platform and System Structure:

The selected platform must support large amount of

information releasing and well-stability aim to tackle

amounts of information duiring the time of information

analysis and processing. Considering the requirement of

earthquake damage assessment timeliness, the system

should possess database management, query function,

spatial analysis and the display function of spatial

characteristics. Web GIS platform: ArcIMS9.0; Servlet

engine: ServletExec5.0 ISAPI; Remote sensing image

processing: ENVI/IDL, database management platform:

Oracle9i; Web Server platform: IIS6.0; The spatial data

engine: ArcSDE9.0 for Oracle9i.according to the terrain

data, text data by means of simulation modeling

software. After the completion of conversion process,

and the successful implementation between all kinds of

heterogeneous systems, will measure the data through

GIS converted to create fire virtual environment

required for the format of the data, such as digital

elevation model data DEM, digital culture data DFD,

texture data TIFF or JPG format.

ION

interpreter

ION Script

document HTML

Database

IDL

Image

server

HTML

Client-

side

Server-side

Request

return

Fig3: Network image process flow of ION Script




4.2. Application and Analysis:

Tthe whole information system is illustrated in Fig.4. In

general, the client do not interact with the database

directly, but to establish a connection through the

COM/DCOM communication with the middle layer, by

means of middle layer processing the application

program put the task such as operation rules, data

access, legality verification into practice. The three-tier

architecture emphasizes the stability, ductility and

efficiency which reduce the network load effectively

and enhance the database response speed, it facilitate

management and maintenance of the system. In a

distributed system construction of B/S computing

environment, the users through the browser access to

server-side map service, through the TCP/IP, Arc XML

communications between the Web Server, Application

Server and Database Server, it achieves map service and

data. Share ins a distributed environment. Based on

Erdas/Idl and and Arc GIS engine ,using B/S structure,

the service system integrates information display, query,

comparison and earthquake monitoring and analysis,

which achieves two-dimensional map and three-

dimensional scene display of Yushu Earthquake area

based on network environment, The system can provide

information services for seismic damage assessment and

information releasing quickly. In the form of thematic

maps, reports and some other forms the results reported

to the decision and command department timely, It

provides the important basis for emergency

management and rescue mission. Creating maps service

Connect oracle database

Check up the completeness of vector and image

Calculate quadrangle point coordinates of vector and image

Create the AXL file according to menu

Create XML command file according to AXL file

Start map service

Quit

Point out nonexistence in file

Fig5: Flow chart of map service dynamically

YES

NO

IIS6.0

ArcIMS9.0

ServletExec5.0 ISAPI

HTTP

TCP/IP ArcXML

DHTML JavaScript Presentation layer

Microsoft .Net

WebServer

ArcIMS Connector

ArcIMS spatial server

MetaData

Editor Application Server Spatial Server

Data layer

Database server Oracle9i

.XML

.htm

Fig4: structure chart of the system

Middle layer




automatically is one of the key step in order to realize

the seismic damage assessment rapidly and earthquake

damage information releasing in real-time, it can realize

the dynamic update of map service. To ensure the

stability of the system and data security this system

using C/S structure model. It adopted flow chart (Fig.5)

to realize create Web mapping service function

automatically. Using for manage the map service, Arc

IMS application server have a series of command. To

perform these commands, in general it should have

XML files of recorded Admin CMD command and

batch files or script file. And then adding commands of

start, stop, delete ArcIMS service of XML files, created

a command file, and then run the file. After create a new

map service chuanjian. XML documents, it can write

the file dynamically in the automatic service program,

then add the following command code: ServiceCMD = "

\ ProgramFiles \ ArcIMS \ Jre \ bin \ Java. Exe Com.

Esri. Aims. Admin. CMD. Exec http://webserver admin

file "and chuanjian. XML. Run the command, new map

service start.

The 3D interface after earthquake is shown in Fig.6.

When the user selects the real-time trajectory tracking,

firstly it select data communication protocol according

to the type of GPS(Fig.7), and then enter the trajectory

of import restrictions and time interval of data detection,

it can choose whether GE automatically path browser,

Finally, click on the "start" button, it can conduct

trajectory tracking in real time(Fig.8).

Based on current received data Policymakers can further

amend the right rescue course, and the corrected results

send to the rescue team in a GPS device, so as to realize

real-time interaction and save rescue time greatly.

The collapsed housing in Jiegu Town is shown in Fig.9.

In the western region of Victory Road, especially in the

southern section (Fig.9(a)), the damage was so serious

that almost all of the civil engineering structures

housing fell destroyed completely. Some of the brick

building was also collapsed .Many of the uncollapsed

ones suffered serious damage to the structure, which are

difficult to repair. For frame structure, the office

buildings for government and the services buildings

along the provincial and national highway were almost

in good condition(Fig.9(b). the collapse rate of the

housings on an alluvial fan in the west side of Jiegu

Town (Fig.9(c))reached up to 86%.

Fig8: Navigation aid route

Fig7: GPS module

Fig6: The 3D interface layout after

earthquake




There might be two reasons: (1) In that alluvial fan after

years of weathering, the surface was relatively flat and

solid, but the lower soil had been eroded, so that the

structure was not stable enough. (2)The region was near

the fault zone. Therefore, for the post-disaster

reconstruction and the future residential land, we should

keep away from the alluvial fan. the interpretation map

of seismic deformation can be obtained (Fig.10).From

the interpretation map, it can be found that the

earthquake caused different degrees of co-seismic

ground deformation between 32°48'N—33°20'N and

96°20'E—97°12'E.The surfface rupture zone from the

Longbao town in the northwest to the Jiegu town in the

southeast reached 70 km long. The overall trend was

119°.The nearer from the Yushu- Ganzi surface fault

zone, the greater the ground deformation. The

deformation gradient was larger in the northeast

direction and was relatively smaller in the southwest

direction.

5. Conclusion:

As one of the greatest natural disasters in the world,

earthquake is very important for emergency to quickly

acquire disaster information. In order to extract the

information of earthquake damage effectively and

timely, On the basis of the practical work in Yushu

Earthquake disaster situation monitoring, this paper puts

the emphasis on the methods and technical flow of

image processing in the application of monitoring and

assessment of earthquake disaster situation. The existing

problems are also analyzed. In modern times remote

sensing technology provides a quick effective approach

to access the disaster information and loss after the

earthquake, owing to its advantage of the quick

dynamically all-weather monitoring with large amount

of information and a short update cycle. It can not only

make full use of resources satellite images, can also

show much detailed image information related to the

earthquake, shorten the development cycle,

popularization of earthquake professional knowledge,

improve the work efficiency, for the ground at the same

time the emergency rescue mode innovation also has a

certain reference significance. It provides services

directly to the professional sectors for the earthquake

analysis and evaluation. This data are processed and

then are used in the information service system

construction to realize the interpretation, assessment and

service for the comprehensive disaster situation

information. The system plays an important role in

disaster relief actions.

References:

[1] Wei C J, Liu Y L, Wang S X, Zhang L F and

Huang X X. 2008. Investigation and assessment of

damage in earthquake Wenchuan Sichuan Quake

based on remote sensing. Journal of Remote

Sensing, 12(5): 673—682.

[2] Liu W G, Cui J T, Zhou L H. Subpixel registration

based on interpolation and extension phase

correlation [J]. Journal of Computer - aided Design

and Computer Graphics 2005, 17 (6): 1273 -1277.

[3] Sun J B, Liang F and Shen Z K.2008. InSAR

deformation observation and preliminary analysis

of the Ms 8 Wenchuan earthquake. Seismology and

Geology, 30(3): 787—795.

[4] Chen S R, Ma H J, Fan Y D, Xu F and Lian J.2008.

Road damage assessment from high resolution

satellite remote sensing imagery in Wenchuan

Fig10: Interpretation map of seismic ground

deformation

Fig9: collapsed housing in Jiegu Town

(a)Region of the south of Jiegu Town; (b) Building

of frame construction along the Provincial and

county roads; (c) Collapse of housing on the

alluvial fan structure housings collapsed




Earthquake. Journal of Remote Sensing,

12(6):949—955

[5] Wen X J, Xu X W, Zheng R Z, Xie Y Q and Wan

C.2003.Average slip-rate and recent large

earthquake ruptures along the Ganzi- Yushu fault.

Science in China (Series D), 33(Z1): 199—207.

[6] Zhu B Q, Wei C J and Zhang Y Z.1998.The quick

extracting of earthquake disasterr data by aerial

remote sensing survey. Journal of Natural

Disasters, 7(1): 34—39

[7] Qiao X J and Guo L M.2007. Study on Jiashi strong

earth quakeswarm area by In SAR. Journal Of

Geodesy And Geodynamics, 27(1):7—13

[8] Sun J B,Liang F and Shen Z K.2008. In SAR

deformation observation and preliminary analysis

of the Ms 8 Wenchuan earth-quake. Seismology

and Geology, 30(3):787—795

[9] Ge L, Zhang K and Alex N G.2008.Preliminary

results of satellite radar differential interferometry

for the co-seismic deformation of the 12 may 2008

MS8.0 Wenchuan earthquake. Geographic

information sciences, 14(1):12—19.

[10] Wei C J, Liu Y L, Wang S X, Zhang L F and

Huang X X.2008.Investigation and assessment of

damage in earthquake Wenchuan Sichuan Quake

based on remote sensing. Journal of Remote

Sensing,12(5): 673—682

[11] Wang S X, Zhou Y, Wei C J, Shao Y and Yan F L.

2008. Risk evaluation on the secondary disasters of

dammed lakes using remote sensing datasets, in the

‘Wenchuan Earthquake, Journal of Remote

Sensing, 12(6): 900—907.

[12] Tahayt A, Feigl KL, Mourabit T, Rigo A, Reilinger

R, McClusky S,. Fadil A Berthier E, Dorbath L,

Serroukh M, Gomez F and Ben Sari D. 2009. The

Al Hoceima (Morocco) earthquake of 24 February

2004, analysis and interpretation of data from

ENVISAT ASAR and SPOT5 validated by ground-

based observations. Remote Sensing of

Environment, 113(2): 306—316.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1072-1081


Flexural Behaviour of Stiffened Cold-Formed Steel Rectangular

Hollow Sections

PRABOWO SETIYAWAN1, MOHD HANIM OSMAN

2, A. AZIZ SAIM

2, AHMAD BAHARUDDIN

ABD.RAHMAN2 AND IMAN FARIDMEHR

2

1Dept. of Civil Engineering, Faculty of Engineering, Sultan Agung Islamic University, Semarang, Indonesia.

2Dept. of Structure and Material, Faculty of Civil Engineering, Universiti Teknologi Malaysia, 81310, Skudai,

Johor. Malaysia.


Abstract: Full-scale cold-formed steel beams stiffened with splice plates have been experimentally studied to

investigate the strengthening effect to the section classification, flexural stiffness and moment capacity of the

sections. The study was done to investigate the increasing capacity of the section. The beams were hollow

rectangular section built up of C channel section and the strengthening of section was performed by adding the

thickness of the compression flange and web using splice hot-rolled steel plates connected using self-drilling screws.

The beam specimens were tested in bending under two point loads. The result was compared to the analytical study.

It was concluded that theoretically, the stiffened beam sections can be classified as plastic section especially for the

specimen with small screw spacing, assuming that a perfect composite action is achieved by the connection between

the cold-formed section and the stiffener plate. In addition, the strengthening has significantly increased the moment

capacity particularly to the beam stiffened with closest screw spacing.

Keywords: Cold-formed steel, section strengthening, rectangular hollow section, class of section, flexural stiffness,

moment capacity.

1. Introduction:

Recently, cold-formed steel C sections have begun to be

used as portal frame members in industrial building.

Combining two C sections to form either rectangular

hollow section or I section to increase the capacity of

the section is often performed. These combined sections

are remain slender, therefore, the methods to improve

the class of the section need to be considered. The

analytical study on behaviour of cold-formed steel

closed section as portal frame rafter had been carried-

out by Setiyawan et al. (2010). The results showed that

theoretically closed section rafter can achieve a longer

span compared to the open section. However, both the

sections failed in “combined bending and compression”

mode of failure. In preventing the combined bending

and compression mode of failure which causing the

section fail in buckling, adding the thickness of

compression elements of the section may be considered

as it significantly affected the section buckling resistant.

Strengthening method of steel sections is often applied

by constructing a compound/plated beam by welding

two equal plates to the flanges of the common beams.

This may increases the area of the sections so enlarging

the beam capacity (Joannides and Weller, 2002, Lam et.

al., 2004). The width of the stiffener plates may be

narrower or wider than the original flanges width. When

the capacity of the compound beams is not adequate to

withstand the applied load then a plate girder is

commonly used. Plate girder is a mostly typical built-up

section encountered in practical use. This section is

constructed by welding two flange plates and a web

plate to form a section. The commonly plate girder

sections found are I and box sections (Fig. 1).


1073 PRABOWO SETIYAWAN, MOHD HANIM OSMAN, A. AZIZ SAIM, AHMAD BAHARUDDIN

ABD.RAHMAN AND IMAN FARIDMEHR



Figure 1: Typical stiffened sections (Joannides and Weller, 2002; Lam et al., 2004)

For cold-formed steel sections, the strengthening is

usually provided by attaching a stiffener at loading

points bolted to the web of channel sections. This is

intended to prevent local buckling as shown in Fig. 2

(Walker, 1975).

Figure 2: Stiffener plate of cold-formed steel C section

(Walker, 1975)

Methods of section strengthening on cold-formed steel

sections had been studied experimentally as well as

numerically by the previous researchers in order to

increase the capacity and observe the structural

behaviour (Salem et al., 2004; Serrette, 2004; Stone and

LaBoube, 2005; Young and Chen, 2008;. Xu et al.

2009). The methods was carried out by constructing

built-up sections or by attaching batten plate. Even

though the methods could increase the section capacity

of the cold-formed steel sections, however they cannot

prevent the local buckling as the section slenderness

especially for the compression elements was not

changed.

All the proposed methods of the section strengthening

however do not increase the class of the section as the

width to thickness ratio of the elements is still large.

Therefore, it is considered to strengthen the sections by

adding the compression element and web thicknesses.

This paper aims to investigate the increasing capacity of

the section when it is stiffened.

2. Research Methodology

2.1 Beam Specimens

The specimens used and the parameter details are

summarized in Table 1 and shown in Fig. 3.

Table 1: Types of beam specimen

No. Code of Specimens Thickness

(mm)

Number of Screw Rows Longitudinal Screw

Spacing (mm) Flange Web

1

2

3

CRBS-T1.5

CRBS-T1.9 CRBS-

T2.4

1.5

1.9

2.4

4

4

4

2

3

3

150

75

37.5

(b) Plate girder sections (a) Compound beams

Stiffener plate

P

1074 Flexural Behaviour of Stiffened Cold-Formed Steel Rectangular Hollow Sections



Figure 3: A typical screw arrangement of beam

2.2 Material Properties

Design strength of the sections and the stiffener plate

were obtained from tensile test are shown in Table 2.

Table 2: Design strength

No. Code of

Specimens

Design Strength py

(N/mm2)

1

2

3

CRBS-T1.5

CRBS-T1.9

CRBS-T2.4

450

550

400

2.3 Test set-up

All specimens were tested in bending under two point

loads. This loading condition was commonly used in

experimental testing of beam tests Osman et al (2007).

The test set-up is shown in Fig. 4.

The specimens were simply supported at both ends.

Preventing crushing at the supports and loading

locations, an internal stiffener of plywood board was

inserted into the section. The load was gradually applied

through a load actuator which was distributed into two

point load using transfer beam.

2.4 Loading procedure

The specimens were tested subsequently in the

following procedure. The beam was supported at both

ends. Point load, P was applied through a load actuator

at mid-span and transferred into two load points of 800

mm distance using transfer beam. A pin support was

used as base to reduce bearing pressure at the each

loading point. Deflection at 3 locations i.e. under the

two point loads and at mid-span as well as rotation at

both ends of the beams were measured using LVDT and

Inclinometer, respectively.

Figure 4: Schematic drawing of bending test

The load was gradually applied until the beam failed

which was indicated by sudden increase of the

deflection and rotation. Deformation and rotation

corresponding to the applied load was observed and

recorded while the load-deflection relationship was

printed using data logger.

90

0

P/2

Transfer beam

P/2

Load P

LVDT 1 LVDT 2 LVDT 3

Inclinometer 1 Inclinometer 2

400 400

L = 2600 mm

Stiffener

90

0

(a) Screw rows on flange and web (b) Longitudinal screw spacing

65

65

152

203

40 40

s s s s s s





3. Results and Discussion

3.1 Moment-Rotation Curve

The result of the test, the moment-rotation diagrams, is

shown in Fig. 6 to Fig. 8. M’c and Mp are the

theoretical section capacities based on the Assumptions

A1, B1 and C1 (Setiyawan et.al. 2010). These

assumptions were taken in calculating the capacity of

the stiffened sections as it is not mentioned clearly in

the current design standard, BS 5950 Part 5:1998. The

assumptions are described as follows:

A1 : the thickness and dimensions of the stiffened

section is assumed equal to the thickness and

dimensions of the original section (t, B, D);

(ts = t l = B, D)

A2 : ts = t l = ax, ay

B1 : ts = tp l = B, D

B2 : ts = tp l = ax, ay

C1 : ts = t + tp l = B, D

C2 : ts = t + tp l = ax, ay

B B

D

t

ax

ax ay

B B

D

t

tp

D

B B

tp

ax

ay D

B B t

tp

D

B B

t

tp

ax

ay D

B B




D1 : t = te t < te < t + tp l = B, D

D2 : ts = te t < te < t + tp l = ax, ay

Where,

t = thickness of original section;

ts = assumed thickness of stiffened section;

B, D = breadth and depth of original section;

ax, ay= screw rows spacing;

tp = thickness of stiffener plate;

te = equivalent thickness of stiffened section;

l = length of section element.

The ultimate moments and the corresponding mid-span

deflection and rotation of each specimen are presented

in Table 2.

The rotation of the beam is determined using the

formula in Equation 1 and referring to Figure 5.

1 2 Eqn. (1)

Where,

= beam rotation;

1 = rotation of beam end 1;

2 = rotation of beam end 2.

Figure5: Derivation of rotation

Table 3: Results of the test and theoretical moment capacities

Code of

Specimens

Ultimate

Moment

Mu (kN.m)

Midspan

Deflection

Δ (mm)

Beam

Rotation

𝝓 (º)

Theoretical

Moment Capacities

M’c (kN.m) Mp (kN.m)

CRBS-T1.5

CRBS-T1.9

CRBS-T2.4

55.76

80.00

95.00

23.5

24.4

37.9

2.4

2.8

4.3

15.3

27.5

33.0

22.8

43.9

46.1

te

D

B B

te

ax

ax ay

B B

D





Figure 7: Moment-rotation curve of specimen CRBS-T1.9

M’c = 105(t = ts + tp)

Mp = 136 (t = ts + tp)

Mp = 43 (t = ts)

Mp = 79 (t = tp)

M = 80

Figure 6 Moment-rotation curve of specimen CRBS-T1.5

Mp = 92 (t = ts + tp)

M’c = 80(t = ts + tp)

Mp = 59 (t = tp)

Mp = 22 (t = ts)

M = 55




3.2 Section Plasticity

The experimental results show that the tested moment

capacities are greater than the theoretical plastic

moment on the assumption that the section thickness is

equal to the thickness of the original section (ts = t), as

shown in Table 3. This means that all the specimens can

be classified as plastic section particularly based on

such limited assumption.

It is indicated that the applied section strengthening has

significant contribution on increasing the section

capacity which exceeds the capacity of the original

section (Mmax > Mp in which ts = t). For the plastic

moment of ts = t, the plastic moment is slightly

increased from specimen CRBS-T1.9 to specimen

CRBS-T2.4 as the thickness affecting the value more

significant than the design strength. Meanwhile, for the

other assumptions the plastic moment is more affected

by the design strength.

3.3 Flexural Stiffness

The experimental flexural stiffness of the specimens can

be determined based on the load-deflection data. For

beams subjected to two symmetrical point loads,

deflection at the midspan may be determined using the

following formula (Mikhelson, 2004):

(2)

Table 4: Maximum moments and theoretical plastic moments (l = B, D)

Code of

Specimens

Maximum Moment

Mmax

(kN.m)

Theoretical Plastic Moment

Mp

ts = t

(kN.m)

ts = tp

(kN.m)

ts = t + tp

(kN.m)

CRBS-T1.5

CRBS-T1.9

CRBS-T2.4

55.76

80.00

95.00

22.43

43.86

46.10

59.42

79.84

59.74

92.53

136.17

110.85

Figure 8 : Moment-rotation curve of specimen CRBS-T2.4

Mp = 110 (t = ts +tp)

M = 95

Mp = 59 (t = tp)

Mp = 46 (t = ts)

M’c = 108 (t = ts + tp)





Figure9: Determination of experimental flexural rigidity

Taken the value of P as equal to half of applied

maximum load on each beam, a = 900 mm, L = 2600

mm and E = 205,000 N/mm2, the flexural stiffness of

the specimens can be determined. The results are

compared to the theoretical values as shown in Table 4.

The theoretical values are calculated based on three

different assumptions as explained before.

It is shown that the theoretical flexural stiffness values

are closer to the experiment values if the thickness of

the stiffened section is assumed equal to the thickness of

the original section (t = ts). Comparing with the flexural

stiffness of the equivalent section, the experimental

values are equal to the same values on various

thicknesses as shown in Table 5.

The thickness of the equivalent sections is compared to

the total thickness of each stiffened section. Specimen

CRBS-T1.5 and T1.9 reach the theoretical flexural

stiffness similar to the experimental flexural stiffness

when the thickness is higher than the thickness of the

original section (ts > t). Meanwhile for specimen CRBS-

T2.4, the value is achieved when the thickness equals to

the thickness of the original section (ts = t).

It can be concluded that the strengthening applied to the

section does not significantly improve the flexural

stiffness of the beams as the value is equal to the

theoretical flexural stiffness for the thickness lower than

the total thickness of the stiffened sections.

Table 5: Experimental and theoretical flexural stiffness values

Code of

Specimens

Experimental

Flexural Stiffness, EI

(kN/mm2)

Theoretical

Flexural Stiffness, EI (kN.mm2)

(t = ts) (t = tp) (t = ts + tp)

CRBS-T1.5

CRBS-T1.9

CRBS-T2.4

3.6 x 109

4.9 x 109

5.3 x 109

1.9 x 108

2.5 x 108

3.8 x 108

4.8 x 108

4.8 x 108

5.0 x 106

7.6 x 108

8.2 x 108

9.0 x 108

Table 6: Equivalent thicknesses

Code of

Specimens

Section

Thickness

t

(mm)

Experimental

Flexural Stiffness,

EI

(mm4)

Stiffener

Thickness

tp

(mm)

Total

Thickness

(ts + tp)

(mm)

Equivalent

Thickness for

EIth = EIexp.

teq

(mm)

CRBS-T1.5

CRBS-T1.9

CRBS-T2.4

1.5

1.9

2.4

3.6 x 109

4.9 x 109

5.3 x 109

3.0

3.0

3.0

4.5

4.9

5.4

2.3

3.0

2.4

3.4 Moment Capacity

The experimental moment capacities of the specimens

were determined by calculating the maximum moment.

In these tests, the maximum moment is equal to bending

moment, Mmax = 900 x P, where P is half of the

maximum P of the test result for each tested specimen.

The experimental moment capacity of the specimens is

then compared to the theoretical value as shown in

Table 6.

It is shown that the experimental moment capacity

exceed the theoretical values for Assumption A1 and B1

in which the thickness is assumed equal to the thickness

of the original section or the stiffener plate (Tables 7 (a)

and 7 (b)). In addition, the specimen CRBS-T2.4 is the

only specimen which the experimental moment capacity

a a

L

P P

Δ




exceeding the theoretical value on Assumption C1 in

which the thickness is assumed equal to the total

thickness of the original section and the stiffener plate

(Table 7 (c)), This means that the applied section

strengthening method is more effective than the other

specimens. Comparing to the moment capacity of the

equivalent section, the experimental moment capacity

values are equal to the values on various thicknesses as

shown in Table 8.

The equivalent thickness was compared to the total

thickness of each stiffened section as shown in Table 8.

t is obtained by trial and error on the calculation of

moment capacity until the value closed to the

experimental moment capacity (Assumption D1) is

derived.

It is indicated that the strengthening applied on the

specimen CRBS-T2.4 significantly improves the

moment capacity. The value is equivalent to the

theoretical moment capacity in which the thickness

larger than the total thickness of the original section and

stiffener plate. This does not occur on the specimens

CRBS-T1.5 and CRBS-T1.9.

Table 7: Experimental and theoretical moment capacities

Code of

Specimens

Experimental

Moment Capacity

(kN.m)

Theoretical Moment Capacity

ts = t

(kN.m)

ts = tp

(kN.m)

ts = t + tp

(kN.m)

CRBS-T1.5

CRBS-T1.9

CRBS-T2.4

55.76

80.00

95.00

15.29

27.55

33.00

49.19

57.81

44.64

80.25

105.41

87.82

Table 8: Moment capacities of experimental and equivalent sections

Code of Specimen Experimental Moment Capacity

(kN.m)

Theoretical Moment Capacity

(kN.m)

CRBS-T1.5

CRBS-T1.9

CRBS-T2.4

55.76

80.00

95.00

55.58 (teq = 3.3 mm)

80.90 (teq = 3.9 mm)

94.47 (teq = 5.8 mm)

Table 9: Ratios of equivalent to total thicknesses

Code of Specimens Section Thickness

(mm)

Total Thickness

(mm)

Equivalent Thickness

(mm)

CRBS-T1.5

CRBS-T1.9

CRBS-T2.4

1.5

1.9

2.4

4.5

4.9

5.4

3.3

3.9

5.8

4. Conclusion

Three specimens of full-scale cold-formed steel beams

stiffened with 3 mm plates have been tested. The

theoretical flexural stiffness and moment capacity had

been calculated and compared with the experimental

results.

The main conclusions of this investigation can be drawn

as follows:

i. The stiffening applied on cold-formed rectangular

hollow sections can improve the class of the

sections. Comparing to the thickness is assumed

equal to the thickness of the original section.

ii. The strengthening significantly increases the

flexural stiffness of the beams as the values are

higher than obtained theoretically for all the

assumptions.

iii. Moment capacity of the beams increases when the

longitudinal screw spacing decreases.

Generally, it is concluded that the strengthening applied

on cold-formed rectangular hollow sections can

improves the flexural behaviour of the sections.

5. Acknowledgements

The material required for the specimen fabrication was

generously donated by BlueScope Ltd. Financial

assistance was provided by Universiti Teknologi

Malaysia (UTM) and Sultan Agung Islamic University

(UNISSULA), Indonesia. The authors are grateful to Dr.

Joanna M. Dulinska for revising the article and also

acknowledge Dr D. Venkat Reddy, Editor- in -Chief -

International Journal of Earth Sciences & Engineering –

IJEE, for facilitating the publication process.

6. Reference:

[1] Beng, K. F. C. (1993). Building Design using Cold-

Formed Steel Sections: Work Examples to BS

5950: Part 5: 1987. Berkshire: The Steel

Construction Institute.





[2] British Standards Institution (BSI). (1998). BS

5950: Structural Used of Steelwork in Building Part

5: Code of Practice for Design of Cold Formed

Section. London: The Steel Construction Institute.

[3] British Standards Institution (BSI). (2000). BS

5950: Structural Used of Steelwork in Building Part

1:2000: Code of Practice for Design in Simple and

Continuous Construction: Hot Rolled Section.

London: The Steel Construction Institute.

[4] International Textbook Company. (1975). Design

and Analysis of Cold-Formed Sections, edited by

A.C. Walker. London: International Textbook

Company.

[5] Joannides, F., Weller, A. (2002). Structural Steel

Design to BS 5950: Part 1. London: Thomas

Telford Publishing.

[6] Lam, D., Ang, T. C., Chiew, S. P. (2004).

Structural Steelwork Design to Limit State Theory

Third Edition. Oxford: Elsevier Butterworth-

Heinemann.

[7] Mikhelson, I. (2004). Structural Engineering

Formulas. New York: McGraw-Hill.

[8] Osman, M.H., Saim, A.A., Saleh, A.L. (2006).

Strength Test on Stainless Steel Purlins (Final

Report) for Yick Hoe Metal Industries Sdn Bhd.

Malaysia: CETU UTM Skudai.

[9] Osman, M.H., Saad, S., Saim, A.A., Keong, G.K.

(2007). Structural Performance of Composite Beam

with Trapezoid Web Steel Section. Malaysian

Journal of Civil Engineering 19 (2): 80 – 95.

[10] Serrette, R.L. (2004). Performance of Edge-Loaded

Cold-Formed Steel Built-Up Box Beams. Practice

Periodical on Structural Design and Construction,

ASCE. 9 (3): 170-174.

[11] Setiyawan, P. Osman, M. H. Saim, A. A. (2009).

Stiffening Of Cold-Formed Steel Section at Plastic

Hinge Formation. Monograf Persidangan

Kebangsaan Kejuruteraan Awam Kelima. USM,

Malaysia: 173-182.

[12] Setiyawan, P. Osman, M. H. Saim, A. A. (2010).

Stability of Cold-Formed Open and Closed Sections

Rafter in Portal Frame Design. Proceedings of the

Third International Graduate Conference on

Engineering, Science and Humanities. UTM,

Malaysia: 65

[13] Stone, T. A. and LaBoube, R. A. (2005). Behaviour

of Cold-Formed Steel Built-Up I-Sections. Thin-

Walled Structures. 43 (2005): 1805-1817.

[14] Sum, N. Z. (2006). Performance of Stainless Purlin

in Bending, These. Malaysia: Civil Engineering

Department UTM Skudai.

[15] Xu, L., Sultana, P., Zhou, X. (2009). Flexural

Strength of Cold-Formed Steel Built-Up Box

Sections. Thin-Walled Structures. 47 (2009): 807-

815.

[16] Young, B. and Chen, J. (2003). Design of Cold-

Formed Steels Built-Up Closed Sections with

Intermediate Stiffener. Journal of Structural

Engineering, ASCE. 134 (5): 727-737.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1082-1088


Response and Reliability Analysis of Pile Foundation under Strong

Excitation

YONGFENG XU, HAILONG WANG, LIQUN ZHANG AND JIANLIN HU Hebei Institute of Architecture and Civil Engineering Zhangjiakou City, Hebei Prov, China

Email: [email protected], [email protected], [email protected], [email protected]

Abstract: With the rapid development of high speed railway in china, more and more bridges or viaducts have been

constructed. Due to the vast area and complex terrain, many viaducts are needed in many cases. In addition, natural

condition is quite complex, so reliability of the viaducts is quite important to provide people’s safety. Usually, pile

foundation plays important role in keeping the good performance of viaducts. In order to improve the reliability of

giant pile group foundation, proper model has been used to simulate the response under strong excitation. For the

reason of uncertainty of natural condition, random process is used in the analysis. For a specified project, strong

excitation has been applied in the structure. In the calculation, equivalent method has been used to replace the

nonlinear load. The results show that biggest stress and bending moment is on the pile top and upper part of pile.

After paying attention to the design of pile, reliability can meet the demand of safety. Besides, the results also verify

the affectivity of the method.

Keywords: pile foundation, strong excitation, equivalent method, reliability.

1. Introduction:

With the rapid development of high speed railway in China, more and more bridges or land bridges have been constructed. Due to the vast area and complex terrain, many viaducts are needed in many cases. Take Shang hang high speed railway as the example, percentage of bridge and tunnel is up to 72.3%. In fact, percentage of bridge and tunnel in many other high speed railways is also more than 70%. This means a lot of viaducts with high reliability is needed. Due to the bridge is distributed in all of the country, complex natural condition in each place has to be taken into consideration, such as rain [1-3], wind [4-6], earthquake [7-11] and so on. Except natural condition, viaducts also influenced by other non-natural conditions, for example, train passing with different speed [12-16], heavy machinery, etc. Both the natural and non-natural conditions have to deeply study.

Pile foundation as supporting structure has been for thousands of years. But with the time developing, pile group has been used with the bridge becomes much larger. At the same time, pile with different work mechanism has its own classification, such as friction pile, supporting pile and so on.

Earthquake is a main cause to make bridge and pile failure. So, mechanism research of earthquake on bridge and pile is quite valuable. Due to the complexity of pile, we can’t apply correct load on the structure in the experiment. In fact, we can’t construct the same size of model as the same of actual size. So, simulation method is quite effective way to determine the parameters of the

structure. Simulation algorithm, with results of enough accuracy, is quite important to develop. Many studies have already been developed under different conditions [17-21]. At the same time, FEM method also has vast of research [22-25].

Due to the importance of pile, in this paper, response of

pile under strong excitation has been mainly studied.

We simulate the earthquake and study the response of

pile and pile group. Strong excitation is a random

process. The main contribution of the paper is the

proposition of an effective simulation method. The

remainder of the paper is organized as follows: random

excitation and reliability analysis method is described in

section 2. Parameters are listed in section 3. Analysis is

shown in section 4 and conclusion is in section 5.

2. Random excitation and reliability analysis

method:

2.1. Random vibration under strong excitation:

Assume that bedrock vibration is a random process,

)(txg is the bedrock acceleration time history with the

seismic risk analysis. The maximum of acceleration

value is 1.2 2/ sm , as shown in figure 1. If we take

)(txg as a sample function of random process, from

stochastic process theory, we can get:

TAEsgg x

Tx /]|)([|lim)( 2

(1)



1083 YONGFENG XU, HAILONG WANG, LIQUN ZHANG AND JIANLIN HU



Where, )(gxA

is Fourier spectrum amplitude of

sample function )(txg ; )(

gxs is power spectral

density of bedrock acceleration time history )(txg ;

]|)([| 2gxAE

is expectation of square of Fourier

spectrum amplitude; T is the duration time. From

equation (1), we can know that if T is big enough,

power spectral density function can be obtained by

Fourier spectrum.

Fig1: Time history of bedrock motion

Based on the ergodic assumption, power spectral

density function of bedrock vibration can be obtained by

equation (1):

TAEsgg xx /]|)([|)( 2 (2)

Figure 2 shows the power spectral density.

Fig2: Power spectral density of bedrock

Figure, images and corresponding text should be clear.

2.2. Random excitation response calculation method:

Using the complex reaction technology, under reaction

of earthquake, motion function of pier-pile-soil system

under excitation can be described as the following:

gX xIMxKxM }]{[}{][}]{[ * (3)

In the equation, ][M and *][K are the mass matrix

and stiffness matrix of interaction system, respectively;

}{x and }{x are nodal acceleration and relative

displacement vector; }{ XI is influence coefficient

vector of motion; gx is duration time of bedrock

vibration.

From equation (2), structure harmonic excitation for

each discrete frequency point : yi

jxg

j

geStx

)()( , combined with equation

(3), we can get:

)(}]{[}(){][][( *2jxXjj g

SIMiXKM (4)

Equation (4) can be solved directly, then, displacement

response vector can be obtained as the following: yi

jxjj

j

geSiHtX

)()}({},({ (5)

In the equation (5), )}({ jiH are discrete values. The

values represent relative displacement transfer function

vector. From the displacement response in equation (5)

we can calculate the stress and strain on frequency point

j and power spectral density of other response

magnitude.

2.3. Maximum response estimation:

In the paper, biggest reaction distribution function are

assumed for Markov process based on counting process

of crossing times [26]. As shown in equation (6).

1)2

exp(

)/2/exp(1exp{)]

2exp(1[)(

2

2

2.1

02

2

x

x

x

qTF

(6)

Where, // 020 is average zero-crossing

rate; 0 x The standard deviation;

2021 //1 q

parameters of bandwidth;

,...)2,1,0()(0

idSX

i

i

is the

thiSpectral moment.

According to equation (6), formula of expectation and

standard deviation of maximum response are given [27]:

xee TTE )ln2/5772.0ln2()( (7)

1.265.0

1.2,])ln2(13

4.5

ln2

2.1[

2.3

T

TTT

ex

ex

ee

(8)

Where,

1084 Response and Reliability Analysis of Pile Foundation under Strong Excitation



69.0

69.0,)38.063.1(

0

045.0

q

qq

e

e

(9)

2.4. Nonlinear dynamics of soil:

In the condition of strong excitation, nonlinear dynamic

performance of soil is studied by equivalent linear

method. In the equivalent linear iteration calculation,

equivalent shear strain of each soil element can be

determined by [28]:

0.22 eq (10)

Where, eqis the equivalent shear strain; is

standard deviation of random process of shear strain

response; 0, is zeroth order spectrum moment in

random process of shear strain response. With

equivalent shear strain, equivalent modulus and

equivalent damping ratio can be determined. Then

iterate the calculation till meet the accuracy

requirement.

2.5. Failure criterion of strength:

If mark yield shear of pile body yS as strength control

index, strength failure probability of pile foundation

under strong excitation is:

njmiSSPtP Yijf ,...,2,1,,...2,1},]{max[)( (11)

Where, ijS is shear response of pile

jat the depth i .

Using first-passage failure mechanism and maximum

overpassing yield shear events of pile body shear

response follows Poisson assumption. For stationary

random process with zero mean, probability of

maximum pile body shear force less than yield shear

can be described as:

)]2

exp(2

exp[)(2,

2

,

,

ijS

y

ijS

ijS

yij

STSSP

(12)

Where, ijS and ijS

are standard deviation of shear

process and derivative reaction process of pile j

at the

depth of i in group piles foundation respectively.

Accordingly, group pile foundation is in beyond

boundary for the first time, namely that first time failure

probability can be expressed as:

)(1, yijijf SSPP (13)

By equation (13) and biggest shear reaction process of

each pile at different depth in pile group, we can get

strength damage based failure probability, which can

evaluate the overall reliability of the foundation of pile

group.

3. Parameters:

T For a specified pile group structure, covering layer

with shear wave velocity m/s 500 < V , which has a

thickness of about 100 m, is composed of 10 kinds of

soil. Property of soil is shown in figure 1. Soil shear

wave velocity is tested in elastic wave velocity

experiment. And maximum shear modulus of soil is

determined by shear wave velocity. Figure 3 shows the

shear wave velocity profile. Relationship among ground

soil dynamic shear modulus ratio and damping ratio has

also been acquired by experimental method [29].

Ground soil is divided into 50 layers in the calculation

and thickness of each layer is smaller than 1/5

wavelength. Among of the all layers, maximum

thickness is 3.6 m and minimum thickness is 1.0m.

Considering the finite element boundary effect, set the

width of field soil layer is 5 times than that of the

foundation.

Table1: Gravities of soils

Soil layer Gravity/

3 mkN

Soil layer Gravity/

3 mkN

(1) silt 18.5 (6) fine sand 20.7

(2) fine sand 19.7 (7) coarse and 21.0

(3) silt 18.9 (8) medium sand 20.5

(4) silt 20.1 (9) gravel sand 22.4

(5) silt 20.5 (10) clay 21.3

elev

atio

n/m

soil shear velocities/m·s-1

Fig3: Section of soil shear velocities

Pile group is composed of 125 bored piles with diameter

of 2.5m and length of103m . Plane size of pier is

m113.5m 5.48 and thickness is m9 . Distance

between pier bottom and ground is m20 . Reinforced

concrete structure is simplified into equivalent mass




effect on the pier. In the calculation, pile group is

simplified as equivalent 30 root piles. Pile body is

adopts beam element. Table 2 shows the main element

parameters of pile. Whole bridge pier structure is

simulated with block elements and mass of each block

elements iskg81038.3

. Moment of inertia is 2111056.5 mkg

Table2: Parameters of beam elements

Elasticity

modulus/

kPa107

Shear

modulus/

kPa107

Gravity/-3mkN

Cross-

sectional

area/

2m

Effective

shear

area

coefficient

Inertia

moment

/4m

3.21 1.62 24.4 4.9 0.9 1.92

4. Analyses:

Damage of foundation structure can cause damage of

upper bridge structure. This is one of main reasons to

lead to destroy. Base damage may be caused by damage

of foundation soil or inadequate strength of itself. Based

on characteristics excitation, bridge pile foundation and

foundation soil response are analyzed in detail. Figure 4

shows the system of pier - pile – soil. Pile is at depth of

83.0 m and distance between two piles is 3.375 m. In

the figure, #1 and

#29 are two side pile, and #15 pile

is in the middle. A ~ G represents x = 67.0, 85.0, 132.0,

177.0, 235.0, 301.0, 40.0 m in the soil profile.

Pier

x

A

B

C D E F G

y

Soil

29#

7# 1#

15#

11#

Fig4: Pier-pile group-soil system

4.1. Displacement Response:

Figure 5 shows the average reaction of largest

displacement of pile and soil. We can see that pile body

displacement reaction has relation with field soil

engineering properties and upper structure of bridge

pier. There are some remarkable characteristics: first,

due to large integral stiffness of pile group foundation,

and foundation soil assumed to transverse isotropy, pile

body displacement of side pile and middle pile are

almost uniformly; second, since the friction pile

foundation, pile end is in a state of freedom. During the

excitation, foundation of pile group may produce some

overall lateral movement. In the test, displacement of

pile end moves about 2.5 cm. In addition, due to the soft

soil layer from ground and depth of 37 m has weak

constraint to the pile body displacement and pull effect

of inertia of the upper structure of bridge pier, body

displacement distribution along the pile depth is in

triangular. From one end of the pile, displacement

increases gradually. Pile body displacement would get

the peak at about 30 m below the ground. At this time,

the displacement is 14.3 cm. Then, pile body

displacement decreases. Displacement of pile body at

the ground is about 7.2 cm and the displacement of pile

top is about 2.7 cm.

soil

soil

soil

15# pile

1# pile

Fig5: Pile displacement

soil

soil

soil

soil

soil

soil

soil

μx/m

H/m

Fig6: Displacement of soil

As shown in figure 6, for the foundation soil

displacement response, foundation soil away from the

base has a decreasing displacement with the increase of

depth. Displacement response at the surface is the

strongest. Displacement response of soil surrounding

foundation is close to pile displacement response at both

aspects of size and depth. As shown in figure 5. In

addition, Displacement response of the foundation soil

on both sides is related to the distance between soil and

foundation. This embodies the effect of pier-pile-soil

interaction on soil displacement reaction. If both sides

of the soil is divided into near field and far field, we can

know that pier-pile-soil interaction has significant effect

on displacement reaction of near field soil, and

displacement response of far field soil has almost close




to that of free field. Effects of interaction are quite

weak. It can be concluded that near field and far field

can be roughly divided with F section. Influence range

of interaction effect of pier-pile-soil on foundation soil

displacement is about 1.7 times of foundation width on

both sides.

4.2. The internal force response of pile foundation:

1# pile

7# pile

11# pile

15# pile

Fig7: Bending moment of pile

1# pile

7# pile

11# pile

15# pile

Fig8: Shear force response of pile shaft

Results analysis of pier - pile - soil system shows that

response of bending moment and sheer force of pile

above ground is very big. Maximum bending moment

is mkN101.15 4 , and maximum shear is

kN101.23 3 . This is because upper bridge piers have

huge inertia force to the basis under the strong

excitation. This may make the connection part between

upper structure of bridge pier and pile foundation is in a

range of plastic zone. Pile top would be probably

broken. Figures 7 and 8, shows the response of average

maximum bending moment and sheer force of pile body

underground. It can be seen from the figure, bending

moment and shear force response of pile body with 40m

depth are quite strong. Bending moment of side pile is

slightly larger than that of middle pile. Shear force

reaction of side pile is slightly bigger than that of

middle pile. Pile group effect has great influence on

shear response. In addition, we still can see the shear

force at depth of 5m is higher than that of side pile top.

We can conclude that when the pile is suffer great

excitation, pile top may be broken for the great force

and pile body also may be broken with the great force.

4.3. Reliability analysis of pile foundation:

In the paper, four yields shear stress are selected as

strength damage standard of pile foundation. The four

stress is kNSy 3000,2500,1500,1000 respectively. Figure

9 shows the reliability analysis results of side pile and

pile. We can see from the figure that side pile top and

pile body near the ground occurs with strength damage

under the low intensity level. This is due to the seismic

shear. With a relative high intensity of damage standard,

reliable probability is more than 99%. It basically

satisfies the reliability demand.

For middle pile, pile body stress is relatively small and

probability of failure is also small. But pile top and pile

body near ground are both the seismic vulnerability.

Overall, for large bridge seismic design of pile

foundation to meet the safety demand under the

condition of little excitation, and combined with

perspective of economy, pile top and upper part of pile

of pile group must take necessary measures into

consideration.

reliability

(a) Side pile

reliability

(b) Middle pile

Fig9: Reliability of pile in different places

It should be noted that strength damage standard of pile

foundation is arbitrary selected. In actual engineering,

strength damage standard of pile foundation should be

determined by experiment.

5. Conclusion:




Due to the vast area and complex terrain and natural

condition, viaducts, with enough safety, are needed to

construct. Pile foundation is base of the bridge and

safety of it has great relationship with safety of viaduct.

For the reason of uncertainty of natural condition,

random excitation is used in the analysis and strong

excitation has been applied in the pile. In the

calculation, equivalent method has been used to replace

the nonlinear load.

In the result, we can obtain some useful conclusion. Due

to the inertia effect and weak constraint of soft soil on

pile body, under the condition of strong excitation,

maximum displacement appears at upper part of the pile

body and pile top displacement is smaller. If use pile

group, displacements of side pile and middle pile are

almost the same. Displacement reaction has close

relationship with distance between soil and foundation.

Under strong excitation, side pile would bear larger

bending moment and shear force.

Simulation method can use simple model instead of

complex model. It transfers the actual problem into

mathematical description of the problem. It can simulate

the actual situation by solving the mathematical model.

With the development of random function, model can

easily get realistic boundary conditions and get more

accurate result, which is close to the actual situation.

The simulation method and random function used in the

simulation of the pile foundation can get more accurate

results and reduce the design cost.

6. Reference:

[1] Xin Da-Bo, Wang Liang, Duan Zhong-Dong, Ou

Jin-Ping, Li Hui, Li Zhong - Hua. Static

characteristics of bridge deck sections of long-span

bridges under simultaneous actions of wind and

rain. Journal of Jilin University (Engineering and

Technology Edition), 41(SUPPL. 2): 174-179.

(2011)

[2] Gu Ming, Xu Shu - Zhuang. Effect of rain on

Flutter derivatives of bridge decks. Wind and

Structures, An International Journal, 11(3): 209-

220. (2008)

[3] Xin Dabo, Li Hui, Wang Liang, Ou Jinping.

Experimental study of rain effects on vortex

shedding of long Span bridge girders. Advances in

Structural Engineering, 15(10):1793-1799. (2012)

[4] Xin Dabo, Li Hui, Wang Liang, Ou Jinping.

Experimental study on static characteristics of the

bridge deck section under simultaneous actions of

wind and rain. Journal of Wind Engineering and

Industrial Aerodynamics, 107-108: 17-27. (2012)

[5] Denoël Vincent, Hortmanns Michael, Sedlacek

Gerhard. A semi-empirical method for the design of

bridge cables and bridge hangers in order to avoid

rain-wind induced vibrations. Stahlbau, 76(2):126-

130. (2007)

[6] Casas Joan R, Aparicio Angel C. Rain-wind-

induced cable vibrations in the Alamillo cable-

stayed bridge (Sevilla, Spain). Assessment and

remedial action. Structure and Infrastructure

Engineering, 6(5):549-556. (2010)

[7] Xia Xiushen, Chen Xingchong, Li Xiaoyang.

Analysis of bridge earthquake damage and

reasonable aseismic systems. Advanced Materials

Research, 433-440:4782-4787 (2012)

[8] Marshall Justin D, Coulston Paul J. Comparison of

bridge earthquake resisting system design in a

moderate seismic zone. Practice Periodical on

Structural Design and Construction, 18(2):83-91.

(2013)

[9] Kuribayashi Eiichi, Iwasaki Toshio. PROGRESS

REPORT OF RESEARCH WORKS IN BRIDGE

EARTHQUAKE ENGINEERING AT THE

PUBLIC WORKS RESEARCH INSTITUTE,

JAPAN. Turk Natl Comm on Earthquake Eng,

Ankara, 9:49-64. (1980)

[10] Feng X. Bridge earthquake protection with seismic

isolation. Technical Council on Lifeline Earthquake

Engineering Monograph, (16):267-275. (1999)

[11] Carden Lyle P. Itani Ahmad M., Buckle Ian G.

Composite action in steel girder bridge

superstructures subjected to transverse earthquake

loading, Transportation Research Record, (1814):

245-252. (2002)

[12] Xia Chao-Yi, Lei Jun-Qing, Zhang Nan. Coupled

vibration analysis for train and simply-supported

bridge system subjected to floating-ice collision.

Journal of Vibration and Shock, 31(13): 154-158.

(2012)

[13] Wang Shao-Qin, Xia He, Guo Wei-Wei, Du Xian-

Ting. Coupling vibration analysis of wind-train-

bridge system considering geometric nonlinearity

of bridge. Engineering Mechanics, 30(4):122-128.

(2013)

[14] Zhang Zhi-Chao, Zhang Ya-Hui, Lin Jia-Hao.

Influence of train speeds on random vibration of

train-bridge systems. Journal of the China Railway

Society, 33(4): 93-99. (2011)

[15] Zhu Yan, Li Xiaozhen. Stochastic space vibration

analysis of train-bridge time-varying system.

ICLEM 2010: Logistics for Sustained Economic

Development - Infrastructure, Information,

Integration - Proceedings of the 2010 International

Conference of Logistics Engineering and

Management, 387:1197-1204. (2010)

[16] Li Qi, Wu Dingjun, Li Jun. Train-bridge vibration

analysis of mix- marshalling freight trains'

traversing medium and small span bridges. Journal

of Tongji University, 35(2)171-175. (2007)




[17] Li Xiao-Peng, Duan Ze-Liang, Li Tao, Wen Bang-

Chun. Dynamic simulation of a vibratory pile

driving process. Journal of Vibration and Shock,

31(7): 24-26+38. (2012)

[18] Li Xiao-Peng, Zhao Guang-Hui, Ju Xing, Liang

Ya-Min. Dynamic simulation analysis of the

vibratory sinking pile system based on AMESim.

Journal of Northeastern University, 34(5): 683-686.

(2013)

[19] Liu Cheng-Wei. Han Xuan. Method and parameters

of numerical simulation on single pile. Chinese

Journal of Geotechnical Engineering, 32(SUPPL.

2): 204-207. (2010)

[20] Chen Ya-Dong, Wang Xu-Dong, Zai Jin-Min. Soil

displacement field visualization research in model

test of pile groups and numerical simulation.

Journal of Sichuan University (Engineering Science

Edition), 43(5):45-51. (2011)

[21] Zhang Qing-Song, Li Shu-Cai, Liu Song-Yu, Du

Guang-Yin. Study on test and numerical simulation

of soil disturbed around pile by dry jet mixing pile

construction. Rock and Soil Mechanics, 27(n

SUPPL.):127-130. (2006)

[22] Wang Wei-Dong, Li Yong-Hui, Wu Jiang-Bin.

Pile-soil interface shear model of super long bored

pile and its FEM simulation. Rock and Soil

Mechanics, 33(12):3818-3824+3832. (2012)

[23] Tan, S.A. Bui, T.Y. Interpretation of pile load test

failure using FEM simulation. Proceedings of the

6th European Conference on Numerical Methods in

Geotechnical Engineering - Numerical Methods in

Geotechnical Engineering, 607-611. (2006).

[24] Lu Qun, Yuan Jianbo, Li Huixia. 3D FEM

simulation and analysis of compacting effects on

underground pipeline by construction of single

jacked pile. ICPTT 2011: Sustainable Solutions for

Water, Sewer, Gas, and Oil Pipelines - Proceedings

of the International Conference on Pipelines and

Trenchless Technology 2011, 1429-1437. (2011)

[25] Lo Kuo- Feng, Ni Sheng-Huoo, Huang Yan-Hong,

Lehmann Lutz. Free field vibrations due to pile

driving using a coupled fem with SBFEM method.

Journal of Geoengineering, 7(2):69-73. (2012)

[26] Vanmarcke E H. Properties of spectral moments

with applications to random vibration. Journal of

Engineering Mechanics, ASCE, 98(EM2): 72－88.

(1972)

[27] Kiureghian A D. Structural response to stationary

excitation. Journal of Engineering Mechanics,

ASCE, 106(EM6): 875－884. (1980)

[28] CHEN Guo-xing, XIE Jun-fei, ZHANG Ke-xu. A

method of 2-D random response analysis of earth

dam earthquake performance. Earthquake

Engineering and Engineering Vibration, 14(3): 81-

90. (1994)

[29] YANG Wei-lin, CHEN Guo-xing, WANG Zhi-hua.

2D and 3D seismic response analyses of supergiant

bridge pier-group pile-soil interaction system on

deep overburden layers. Earthquake Engineering

and Engineering Vibration, 26(5): 172－183.

(2004).

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1089-1095


Settlement algorithm research of pile foundation under load impact

YONGFENG XU, LIQUN ZHANG, HAILONG WANG AND MINFENG LI Hebei Institute of Architecture and Civil Engineering, Zhangjiakou City, Hebei Prov, China

Email: [email protected], [email protected], [email protected], [email protected]

Abstract: Bridge served as an important part of the infrastructure, is widely used in the design of highways and

high-speed railways. As bridges of large and complex, such as road and railway dual-use Bridge, bridge security

design is also extremely important. Pile is the base of bridge, and its stability and reliability must meet high

requirements. As one of the requirements, pile foundation settlement characteristics are very important. In order to

study characteristics of pile foundation settlement, a new algorithm of load-settlement curve, which is based on

dual-line load transfer function, has been developed. The method is then used to analyze the influence of

surroundings soil condition on the pile bearing capacity. According to instance validation, the method proposed in

this paper is effective and simple. This method can be used to promote in engineering practice.

Keywords: pile foundation, settlement, load impact, bearing capacity

1. Introduction:

Pile foundations, as the basis of common type, have

been widely used in high-rise buildings [1-4], high-

speed railways [5-7], highways [8-10], bridges [11-15],

special ports [16-17], large structures, and so on. In

order to meet the safety demand of buildings, ultimate

bearing capacity of pile has to be determined [18-22].

Under the working condition, deformation of the pile

would influence the settlement, so, we must understand

the deformation of the pile and settlement

characteristics under load impact [23-25].

In actual condition, pile load settlement is affected by

many factors [26], including pile length, material,

shape, and size. Besides, characteristics of interaction

between foundation soil and pile tip is also included.

Currently, common method used in the engineering to

determine the bearing capacity of pile is through static

load tests to get the relation between pile top load and

settlement [27-28]. The curve is named as p-s curve, as

shown in Figure 1. Static loading test result to determine

the bearing capacity of the pile is certainly reliable. But

it has to spend a lot of money and time, especially for

different sizes of piles under different soil conditions.

With the development of computers, analytical method

and finite element method have a great development.

Especially the finite element method, which can convert

complex practical problems to a series of simple

calculation models, has been widely applied in scientific

activities and engineering. However, too much

simplified conditions and Solver simpler make the finite

element method can’t get more accurate results. So, in

settlement problem of pile foundation, a number of

dedicated solutions have been developed, such as one-

way compression layer summation method, resonance

method and mechanical impedance method and so on.

One-way compression layer summation method is a

typical method of soil mechanics [29-30]. This method

is based on e-p curve obtained by confined compression

test, and amended by monitoring data in engineering

practice.

Pa

s

12

3

a

b

Pb Po

Fig.1: P-s curve

In the paper, a new settlement method based on double

lines hardening model is established to calculate the

relation between axial loading-settlement of piles. And

characteristics of pile-surrounding soil effects are taken

into account. Case analysis shows the availability of the

method and main contribution of this paper is

developing a new algorithm based on double broken

line model. The remainder of the paper is organized as

follows: analysis of pile settlement curve with axial load

is shown in section 2. Relationship between initial static

stiffness and ultimate bearing capacity is shown section

3. Analysis of test results is shown in section 4 and the

conclusion is shown in section 5.



1090 Settlement algorithm research of pile foundation under load impact



l

o

P

Pl

PB

Pl

ul u

λ1

λ2

PB

ub u

k1

k2

(a) Pile (b) Soil surrounding pile (b) Pile Subsoil

Fig2: Framework of generalized multivariate system

2. Pile of the axial load settlement curve analytic

formula:

The mechanical model of pile and soil used as shown in

figure 2, which assumes that the cross section of pile is

homogeneous and elastic rods. Effect of pile and pile

subsoil is approximately expressed with double broken

line load transfer function. 1 and 2 are corresponding

shear stiffness coefficient of before and after pile

surround soil limit displacement lu . 1k and 2k are

corresponding compressive stiffness of before and after

pile surrounding soil elastic limit displacement. In

homogeneous foundation, when the pile top is bearing

axial stress, axial force and settlement of each section of

pile is decreasing with the increasing of depth.

Therefore, with the increasing of the pile top load, pile

soil is gradually into plastic hardening stage until

destruction from shallow to deep, gradually.

2.1. Pile surrounding soil at elastic stage

For illustration purposes, P and S are used to represent

pile top load and settlement, while BP and BS represent

reaction and sedimentation beard by pile bottom.

BP and BS are parameters in the formula below. When

the load is small, all of pile soil is in elastic state. The

displacement )(xu of pile section should satisfy the

following equation:

Blx

Blx

Su

Pdx

udEA

udx

udEA

|

|

012

2

(1)

Where, E and A are elastic modulus and sectional

area of respectively. If we mark EAbi / ,

1,2i . By equation (1), the ratio of pile top load and

settlement or pile stiffness can be calculated.

)()/(

)(/

11

111

lbthSPEAb

lbthEAbSPEAb

S

PK

BB

BB

(2)

For a given pile soil system, whether K in equation is

constant is determined by the relative size of lu and

bu . There are two conditions as the following:

(1) bl uu . In this condition, while luS0

and 1kS

P

B

B keeps constant, K is also a constant

of eoK . If we mark ji, as the following:

2,1,,)(

)(

1

1

ji

lbthkEAb

lbthEAbk

ji

ij

ij (3)

Then, we can describe eK as

EAbKe 111 (4)

eK is the slope of the s-axis at initial line segment on

the p-s curve.

(2) If bl uu . In this case, pile subsoil is possibly in

plastic hardening stage earlier than surrounding soil.

Let’s suppose 1crS to be the corresponding settlement

under the condition of bB uS . From equation (1),

we can get 1crS to be described as the following:

01

1

111 )]()([ bcr ulbsh

EAb

klbchS (5)

If 11 uScr , then subsoil pile will go into plastic state.

And under the condition of 10 crSS ,

Relationship of P-S is obviously a straight line segment

with slope of eK . But in lcr uSS 1 , the relation

becomes the following.

bulbshlbchkkEASbP )]()()[( 112121112 (6)

It means a straight line with slope of EAb112 and

point ),( 11111 crcr SEASb on the line. For a given pile

1091 YONGFENG XU, LIQUN ZHANG, HAILONG WANG AND MINFENG LI



soil system, there generally are 21 kk ,

so 1112 . It means that if settlement of pile top is

greater than the 1crS , stiffness of the pile began to

decrease.

If lcr uS 1 , surrounding soil of pile would be in

plastic state earlier, relationship of P-S can be shown by

a line with slope of ek and with original point on the

line under the condition of luS 0 .

2.2. Part of surrounding soil into the plastic hardening

stage

When the settlement of pile top is greater than lu ,

continue to increase pile load will make soil from

shallow to deep gradually into the plastic state. Plastic

length 2l in the following figure 3 is used to illustrate

the algorithm of P-S curve of pile top. For mechanics

model shown in figure3, if first list two types of

differential equations of pile section, and then according

to continuous conditions of force and displacement on

the boundary surface, the solving process is very

complicate. In the paper, we drop the method, and

introduce a new kind of method. The new method can

be used to calculate the relationship between load

settlement and force of homogeneous cross section and

layered foundation

l

o

P

Pl

PB

x

C

l1l2

Elas

tic s

ectio

nPl

astic

sect

ion

B

Fig3: Shallow soil around pile into plastic hardening

state

First, use axial force cP and settlement cS to represent

influence of cross section under section C in figure 3.

Obviously, under the condition

of ll 20 , lc uS , as shown in figure 3, due

to BC section of pile soil in the elastic stage, equations

(4) and (5) the l is replaced by )( 2ll , then

corresponds cP to 2l can be calculated conveniently by

the above formula. Then, by figure 3, with differential

equation and boundary conditions of OC section as the

following:

clx

lx

ll

Su

Pcdx

duEA

uuudx

udEA

2

2

|

|

0)(212

2

And we can get pile top load and sedimentation value

when pile surrounding soil enters the plastic state.

lc

x

lcx

ulbshEAb

KlbchuS

ulbchklbshEAbdx

duEAP

)]1()()()1[(|

)]()()1[{|

22

2

220

222220

(7)

Where,

l

c

u

S ,

2

2

1

2

1 )(b

b

,

l

cc

u

PK . For

the reason that in condition of ll 20 , lc uS ,

so 1 .

2.3. All pile surrounding soil enters the plastic

hardening state

On this section, 3crSS , lB uS . In figure 3,

section C overlaps with pile bottom B . Obviously,

relationship between P-S of pile is still available to

determine with equation (7). But, some relative

parameters has to be changed as the

following: )1(l

B

u

Sa , ll 2 ,

l

Bc

u

PK .

Due to the calculation of BP has relationship with

relative size of BS and bu , we should give the explicit

expression of P - S relationship as the following two

conditions:

(1) lb uu . Because of bB uS , equation (7)

can be described as the following:

l

b

ulbchlbshEAb

ulbshlbchkkEASbP

])(*)([)1(

)](*)()[(

2222222

222221222

(8)

This means relationship of P-S is a line with slope of

EAb222 in condition of 3crSS .




(2) lb uu . When bBl uSu , BB SkP 1 .

We can get the sediment 4crS of pile under condition

of bB uS .

lbcr ulbchulbshEAb

klbchS ]1)()[1()]()([ 22

2

124

(9)

Then, we can also get the description of relationship

between P-S under condition of 43 crcr SSS as

the following:

lEAublbchlbshEASbP 2212212221 ])()()[1( (10)

So, we can see the relationship is a line with slope of

EAb221 . Obviously, when bB uS or 4crSS ,

relationship of P-S can follow equation (8). For a

special pile system, 2221 , so, stiffness is higher

when 43 crcr SSS than that of 4crSS .

At this point, as shown in figure 1 in this paper, the

mechanical model with a set of complete analytical

formulas is established for calculating the pile

settlement curve with axial load. If the pile

soil 02 , then the application must have some

changes.

l

c

lc

ulbEA

lKS

uKlP

)2

1(

)(

22

21

2

21

(11)

Then, in equation (8) and (10), EAb222 and

EAb221 would be instead by EAlk

k

/1 2

2

and

EAlk

k

/1 1

1

respectively.

It is easy to see from the derivation process and its

results described above, pile soil and pile elastic limit

displacement of the subsoil and relative size of pile's

axial load settlement curve shape can produce certain

effect. From calculation of 1crS , we can see only on the

condition

ofEAb

k

u

u

EAb

klbsh

b

l

1

12

1

11

21)

2()( ,

bottom soil of pile would enter plastic stage earlier than

surrounding soil.

3. Relationship between initial static stiffness and

ultimate bearing capacity

It is easy to see from (3) ~ (11) that axial load-

settlement curve is influenced by 9 parameters besides

parameter 2l . If the curve is drawn in coordinate system

ofl

S

EA

P , the curve would be a function of 6

dimensionless parameters described as the following:

lb11 , 21 / , luu ll / ,

EAlkk /11 , 21 / kk , luu bb /

From the physical and mechanical sense, 1 and 1k

reflect the ratio of pile shear stiffness coefficient and

pile subsoil compressive stiffness of pile on stiffness of

pile body. and represents the nonlinear level of

surrounding soil and subsoil of pile. So, load -

settlement curve is not only associated with the

nonlinear of soil, but also controlled by relative

characteristics between pile body and foundation soil.

3.1. Ultimate bearing capacity of pile

The ultimate bearing capacity of pile can be determined

by second inflection point on measured P-S curve.

Under this method, ultimate bearing capacity uP of

model shown in figure 2 is theoretically defined in two

conditions:

(1) Top load when pile surrounding soil in plastic stage

under condition of lb uu ;

(2) Top load when subsoil in the plastic stage under

condition of lb uu .

With the definition described above and equation (7),

dimensionless ultimate bearing capacity of pile can be

determined as the following:

lbbl

lbbluu

uuuchkshshu

uuchukkuchksh

EA

PP

)()1(

)()(

212222

2212222

(12)

Where 12 , 12 kk . From

equation (12) we can see know some important factors.

Firstly, ultimate bearing capacity of pile is not only

influenced by physics and elasticity parameters of pile

and soil, but also influenced by elastic displacement

limit and nonlinear parameters of soil. Secondly, when

took the ultimate bearing capacity as top load of pile,

degrees of intensity of soil and subsoil are different.

This is different from that we can see from theory.

When lb uu and 021 , equation (12)

can be described as the following:




lu ukP )( 221 (13)

So, only in this case, the ultimate bearing capacity of the

pile can be determined by side friction and reaction of

pile subsoil.

3.2 Relationship between initial static stiffness and

bearing capacity of pile

For model shown in figure 1, if used dimensionless

parameters and pile stiffness K , eK can be described

as the following:

1

111

111

*

*

thk

thk

K

KK e

e

(14)

So, we can see that eK is function of just

1 and1k .

eK increases nonlinearly with the increasing of 1k .

But when 21 10k ,

eK is almost not influenced

by1k . At the same time,

eK increases with 1

increases, and after about 5.21 ,1eK . This

means that, if the ratio of pile length to diameter is too

large or pile soil is hard, 1 may bigger than 2.5. Then,

all kinds of piles may have nearly the same initial static

stiffness. In fact, this is impossible. So, we can just

know that initial stiffness and ultimate bearing capacity

of pile would has no one to one correspondence

relationship when 5.21 .

For 5.21 , compared with equation (12) and (14),

we can find the quite complex relationship between

initial stiffness and ultimate bearing capacity of pile.

Therefore, with pile stiffness and correlation coefficient

to calculate bearing capacity is a kind of approximate or

experience method. As a special case, under the

conditions of lb uu and 0221 kk ,

analytic relations between pile bearing capacity and

initial stiffness can be derived as the following:

e

l

u Kth

uP )(

1

1

(15)

We can see that even in the simplest case, correlation

coefficient uR between pile bearing capacity and initial

static stiffness is also not a constant. It is also influenced

by pile size and elastic modulus. As shown in figure 4.

So, a conclusion can be acquired that correlation

coefficient is different under different size or strength

level of pile even if the properties of foundation soil are

exactly the same. This can be approved at the other side

that using stiffness to conclude ultimate bearing

capacity of the pile is difficult.

2

0 2 4 6

4

6

Fig.4: Coefficient eu KP of association for pure

friction pile

4. Analysis of test results

We would like to list some examples to verify the

affectivity of algorithm.

(1) Example 1

In this example, parameters of the pile are given:

mD 6.0 in diameter, ml 85.20 in length,

elastic modulus MPaE 4102.3 . Shear

modulus is about kPa41023.1 . Tested relationship

of P - S is described as solid lines in figure 6. It can be

approximately as a straight line at the beginning and the

end of load. According to the solid line and theoretical

formula available, we can obtain the result of

mkNk /1002.6 42 . Dotted lines in figure 5

represent the theoretical curve according to the model

proposed in the paper. Calculation parameters are

obtained by fitting the measured and selected data.

241 /1012.4 mkN 02

mmul 12.2

mkNk /1077.2 51

mmub 50.5 1

(2) Example 2

Pile soil is mainly clay. Pile bottom enters into strongly

weathered rock. Test pile has parameters

of cmcm 3535 , length ml 19 and elastic

modulus MPaE 4109.2 .

Based on the quick loading method, purpose of

engineering test is to determine the ultimate bearing

capacity of the pile. Solid line in figure 6 is the

measured P - S curve of pile, and dotted line is the

theoretical curve. Calculation parameters are described

as the following:




231 /100.3 mkN , 02

mmul 0.8

mkNk /103.2 51

mkNk /108.1 42 mmub 7.1

According to the definition method proposed in this

paper, the theoretical value of ultimate bearing capacity

of pile is 960kN

15000

5

10

15

3000 4500

P/kN

S/m

m

Test result

Simulation result

Fig.5: P-S curve of example 1

4000

5

10

15

800 1200

P/kN

S/m

m

Test result

Simulation result

Fig.6: P-S curve of example 2

5. Conclusion

Pile foundations is common type of basis and widely

used in various conditions. In order to meet the safety

demand of buildings, load – settlement relation has to be

determined. In the paper, we propose a method for the

determination. And, following conclusions are based

and limited to the extent of the data presented. The

method proposed in the paper has been verified by cases

and main conclusions are listed as the following:

(1) Based on the double lines hardening model, a set of

complete model is established to calculate the

relation between axial loading-settlement of piles.

The method can also easily calculate the load

settlement of piecewise homogeneous section.

(2) With the selected mechanical model, the

discriminant of pile subsoil prior to pile

surrounding soil into plastic state has been given.

Analysis shows that under axial load, pile

surrounding soil would enter plastic state earlier

than subsoil. Intensity level of the two kinds of soil

is different when the pile top is bearing ultimate

load.

(3) Analysis in the paper indicate that when the pile

soil system is in condition of value of 1 more than

2.5, there is almost no correlation between initial

stiffness and ultimate bearing capacity of pile;

when in the condition of value of 1 no more than

2.5, there would be some relation between them,

but its quantitative relation very complex. In

addition, friction pile is no exception. So, if we

calculate the ultimate bearing load with the pile

stiffness, we should give more attention.

5. Reference:

[1] Li Jun-Cai, Ji Guang-Qiang, Song Gui-Hua, Zhang

Qiong, Wang Zhi-Liang, Yan Xiao-Min. In-situ

measurement and analysis of sparse pile-raft

foundation of high-rise building. Rock and Soil

Mechanics, 30(4):1018-1022. (2009)

[2] Guo Zhong-Xian, Wang Zhan-Lei, Yang Zhi-Hong.

Field study of reactions of rigid pile composite

foundation for high-rise building. Rock and Soil

Mechanics, 30(1):163-168. (2009)

[3] Fan Cunxin, Chen Rong, Xue Songtao. Effect of

pile-soil-structure interaction on downwind

response of highrise building to wind-induced

vibration. Chinese Journal of Rock Mechanics and

Engineering, 23(12): 2078-2084. (2004)

[4] Huang Shaoming, Yue Jianyong, Liu Feng, Lu

Yunian. Comparative study on settlement bending

deflection of pile foundation mat of a certain super

high-rise building in Shanghai between

computation and observation. Journal of Building

Structures, 33(9):81-86. (2012)

[5] Ding Mingji. Numerical simulation on the post-

construction settlement of CFG pile-slab composite

foundation in high-speed railway. China Railway

Science, 29(3):1-6. (2008)

[6] Zeng Jun-Cheng, Zhang Ji-Wen, Tong Xiao-Dong,

Tu Yong-Ming. In-situ test of stress characteristics

of CFG pile composite foundation of high-speed

railway. Journal of the China Railway Society,

33(5):91-96. (2011)

[7] Zheng Gang, Jiang Yan, Han Jie, Liu Yuan-Feng.

Performance of cement-fly ash-gravel pile-

supported high-speed railway embankments over

soft marine clay. Marine Georesources and

Geotechnology, 29(2):145-161. (2011)




[8] Wang Cheng. Study of effective length of super-

long pile group in highway bridges. Rock and Soil

Mechanics, 31(5):1569-1573. (2010)

[9] Chakraborty Sanjoy, Montgomery Michael W.

Louisiana Highway 1 pile load test program.

Geotechnical Special Publication, 178:58-65.

(2008)

[10] Tomisawa K. Egawa T., Miura S. A study on the

design of Highway Bridge pile foundations in

volcanic ash ground. Advances in Transportation

Geotechnics II - Proceedings of the 2nd

International Conference on Transportation

Geotechnics, ICTG 2012, 635-642. (2012)

[11] Lin Cheng, Bennett Caroline R., Parsons Robert L.,

Han Jie. Analysis of bridge response considering

water-soil-pile-structure interaction under a scoured

condition. Geotechnical Special Publication, n 205

GSP, 73-80. (2010)

[12] Roling Matthew, Abdelsalam Sherif, Sritharan Sri,

Suleiman Muhannad. Load and resistance factor

design calibration for bridge pile foundations.

Transportation Research Record, (2204):233-241.

(2011)

[13] Mylonakis George, Nikolaou Aspasia, Gazetas

George. Soil-pile-bridge seismic interaction:

kinematic and inertial effects. Part I: soft soil.

Earthquake Engineering & Structural Dynamics,

26:337-359. (1997)

[14] Bergado D.T., Lorenzo G.A., Phien-Wej N., Lin

S.S., Voottipruex P. Compression mechanism of

DMM pile in subsiding soft ground under

embankment loading with application to bridge

approach embankment. Proceedings of the 16th

International Conference on Soil Mechanics and

Geotechnical Engineering: Geotechnology in

Harmony with the Global Environment, 3:1149-

1153. (2005)

[15] Xiao Y., Guo Y.R., Zhu P.S., Kunnath S., Martin

G.R. Networked pseudo dynamic testing of bridge

pier and precast pile foundation. Engineering

Structures, 38:32-41. (2012)

[16] Varun, Assimaki Dominic. Integrating soil-

structure interaction analyses of pile-supported

wharfs in seismic risk management of port systems.

Geotechnical Special Publication, (181). (2008)

[17] Lai Cheng, Jaradat Omar A., Weismair Max,

Arulmoli K. Seismic design criteria for pile-

supported wharves at the port of long beach.

TCLEE 2009: Lifeline Earthquake Engineering in a

Multihazard Environment, 357:86. (2009)

[18] Gao Xiao-Juan, Zhu Xiang-Rong. Forecasting

ultimate bearing capacity of single squeezed branch

pile by hyperbola method. Rock and Soil

Mechanics, 27(9):1596-1600. (2006)

[19] Zai Jin-Min, Lu Zhou, Huang Guang-Long.

Reliability analysis of design method of composite

pile foundation according to ultimate bearing

capacity of single pile. Rock and Soil Mechanics,

25(9):1483-1486. (2004)

[20] Liu Yong-Jian. Study of prediction method of

vertical ultimate bearing capacity of single pile

based on genetic algorithm and neural network.

Rock and Soil Mechanics, 25(1):59-63. (2004)

[21] Yao Zeliang, Jian Zhen, Bai Guoliang. Study and

application on computational methods for ultimate

bearing capacity of single pile. Key Engineering

Materials, 400-402:329-334. (2009)

[22] Yang Yuwen. Estimation of ultimate bearing

capacity of single pile with golden section. Chinese

Journal of Rock Mechanics and Engineering,

28(SUPPL. 2):3577-3584. (2009)

[23] Zhang Lei, Gao Guangyun, Gao Meng, Gu Baohe.

A method of determining the deformation modulus

of large diameter belled pile. Journal of Building

Structures, 31(SUPPL. 2): 285-291. (2010)

[24] She Yuexin, Chen Yadong, Wang Chengwu.

Bearing capacity and deformation behaviour of

branch pile. Applied Mechanics and Materials, 90-

93: (1355-1358). (2011)

[25] Qian Shuangbin, Zhang Lihua. Research of load-

bearing and deformation characteristics of long-

short pile composite foundation. Applied

Mechanics and Materials, 256-259(PART 1):592-

595. (2013)

[26] Stelzer D, Andersland O.B. Creep parameters for

pile settlement equations. Journal of Energy

Resources Technology, Transactions of the ASME,

111(4):258-263. (1989)

[27] Dong Tian-Wen, Liang Li. Solution of load-

settlement function of single screw pile under axial

pressure. Chinese Journal of Geotechnical

Engineering, 29(10):1483-1487. (2007)

[28] Zhang Zhan-Tao, Feng Zhi-Yan. Research on

bearing capacity and settlement property of

pressed-in pile in Xi'an loess area. Chinese Journal

of Rock Mechanics and Engineering, 24(14):2548-

2553. (2005)

[29] Cao Wen-Gui, Liu Hai-Tao, Zhang Yong-Jie. Study

on layer-wise summation method of settlement

computation for composite foundation with friable

material piles. Journal of Hydraulic Engineering,

41(8):984-990. (2010)

[30] Stolyarov V.G. Calculation of Bed Settlement by

the method of Layer-by-Layer Summation in

Accordance with Building Code 51-101-2004 and

the method of Integration. Soil Mechanics and

Foundation Engineering, 48(2):67-72. (2011).

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1096-1103


Empirical Study on the Co-integration Relationship between Urban

Construction land, Economic Growth and Urbanization

Development of Jiangxi Province

WEI LIU1,2

, YANG AN BAO1 AND CHANG XIN XU

1

1Hohai University, School of Business, Nanjing, Jiangsu Province, CHINA

2East China Jiaotong University, School of Civil Engineer and Architecture, Nanchang, Jiangxi Province, CHINA


Abstract: This paper studies the dynamic relationship among urban construction land, economic growth and

urbanization development of Jiangxi based on Statistical Yearbook 1997-2002. It applies the long-term and short-

term relationship to correlation analysis, co-integration test, Granger cause-and-effect test, impulse response and

variance decomposition of dynamic econometric analysis. Results show that (1) there is a co-integration relationship

in the long term among urban construction land, economic growth and urbanization development; (2) error

correction model suggests that the departure in the short term from the long-term urbanization development will be

adjusted to a equilibrium state by 7.8%; (3) impulse response function and variance decomposition show that urban

construction land is easily influenced by self-impact which is owing to the lagging behind; urban construction land

plays an important role in promoting the economic growth and land expansion accelerates the urbanization process.

Keywords: Urban construction land, economic growth, urbanization development, co-integration relationship.

1. Introduction:

As Jiangxi’s economy grows steadily and rapidly, the

urbanization is proceeding in depth with the proportion

increasing from 25.32% in 1997 to 47.51% in 2002 [1-2]

.

In the near future, Jiangxi will become a highly

urbanized region. The shifting of population from rural

areas to urban areas and the transformation of rural

areas into urban areas have brought about more urban-

utilized land and urban construction land. At the same

time, non-agricultural economic elements are gathered

in urban areas. They are the driven force of the

economic growth which further results in the increase of

the urbanization rate featured by more urban

construction land [2]

.

Current researches about urbanization and economic

growth mainly focus on the mechanisms or the

empirical studies of their relationships. Zhou Xiaogang

and Chen Dong (2008) conducted an empirical study of

relationship between urbanization rate and per capita

GDP of Jiangxi and found out the co-integration

relationship[2]

. Li Jinchang and Cheng Kaiming (2009)

adopted econometric methods and co-integration theory

of two variables to test the dynamic relationship

between economic growth and urbanization

development based on time sequence data [3]

. Gao Wei,

Min Jie and Zhang Anlu (2010) studied the dynamic

relationship among agricultural urban land, economic

growth and urbanization development [4]

.Research

group of Chinese economic growth and macroeconomic

stability did research on the relationship among

urbanization, industry efficiency and economic growth [5]

. Ba Shusong (2010) thought that urbanization could

generate a permanent effect on Chinese economic

growth and shift of mode by influencing aggregate

supply and demand [6]

.

Liu Aiying, Yao Lifen and Li Qingchen (2010) pointed

out the relationship between level of urbanization and

GDP by empirical study based on co-integration theory,

error correction model and Granger cause-and-effect

test. They also did comparative study on the relationship

between tertiary industry and level of urbanization.

Meng Junfeng (2013) based his study on vector

regression model, variance decomposition, impulse

response function and regression model as well as

urbanization rate and per capita GDP for Shaanxi

Province between 2000 and 2010. The study showed

that the long-term co-integration relationship between

urbanization and regional economic growth did exist

and that regional economic growth was more of the

driven force of urbanization than the vice versa [8]

.

As for researches about urban construction land, they

are mainly focused on the driven force and the

contribution to the economic growth under the method

of AHP and multiple linear regression model of

econometric analysis [9-13]

. Based on data from 1981 to

2007, Zhao Ke and Zhang Anlu (2011) adopted co-

1097 WEI LIU, YANG AN BAO AND CHANG XIN XU



integration test, impulse response function and variance

decomposition to test and analyze the relationship

between urban construction land, economic growth and

urbanization development [14]

.

However, researches mentioned above did not integrate

urban construction land with economic growth and

urbanization development. This paper, based on

Statistical Yearbook 1997-2002, studies the long-term

and short-term relationship through co-integration test,

Granger cause-and-effect test, impulse response and

variance decomposition of dynamic econometric

analysis.

2. Theoretical framework, methodology and data

processing：

2.1. Theoretical framework：

The relationship among urban construction land,

economic growth and urbanization development is

complicated. According to Sims’ theory, if a group of

variables is simultaneous, they should be treated

equally. There should be no distinction of endogenous

from exogenous variables in advance [15]

. Therefore,

urban construction land, economic growth and

urbanization development are explainable variables.

Given the influence of hysteresis value on these

variables, the econometric model is described as:

1 1 1 1 1 1t t k t k t k t k t k t k tUCL UCL UCL GDP GDP UR UR

1 1 1 1 1 1t t k t k t k t k t k t k tGDP GDP GDP UCL UCL UR UR (1)

1 1 1 1 1 1t t k t k t k t k t k t k tUR UR UR UCL UCL GDP GDP

In expression (1),UCL , GDP ,UR stands for urban

construction land Urban construction land economic

growth and urbanization development;

、、、、、、、、

are regression

coefficients of each variable; t t t 、、 are random

disturbance terms of each equation respectively; and the

variables are not relevant to hysteresis values.

2.2. Methodology：

Traditional econometric method is based on economic

theory to describe the relationship between variables.

But such theory fails to provide an explanation for their

dynamic relationship. Endogenous variables may show

up at the left side of the equation or the right side, which

makes the estimation more complicated [16]

. VAR model

is good enough to solve this problem and easy to

operate.

2.3. Variables selection:

(1) Urban construction land (UCL ). Urban construction

land refers to land for housing, public facility, industry,

storage, transportation, road and square, municipal

public facilities, green land and special lands. With the

industrialization and urbanization process and under

strong economic growth, more and more agricultural

land will be transformed to urban construction land

which is the carrier of urban economy. Urban economy

is in the dominant place of the national economy. In this

paper, the area of urban construction land is the research

object. The unit is km2.

(2) Gross regional domestic product ( GDP ). GDP is

used to measure regional economic growth. National or

regional GDP refers to the total output of goods and

labor generated by a country or a region within a certain

period of time (one quarter or one year). It is regarded

as the best measurement of national or regional

economy. The unit is 100 million yuan.

(3) Urbanization (UR ). It is also called urbanization rate,

a quantity measurement of level of urbanization.

Usually, the urbanization rate is calculated by the

proportion of urban residents to total population. The

unit is %.

2.4. Data source and processing:

Data for UCL , GDP and UR are from Statistical

Yearbook of Jiangxi and the Statistical Yearbook of

China. Urban construction land of Jiangxi, GDP and

urbanization rate are shown in Table 1 and Figure 1.

Moreover, heteroscedasticity is eliminated to avoid its

influence and make research result more practical. Get

the natural logarithm of UCL , GDP and UR and apply

them to Eviews6.0 [16]

.

Table 1: Urban Construction Land of Jiangxi (UCL ), Gross Regional Domestic Product ( GDP ), Urbanization rate

(UR ) – Annual raw data

Year UCL (km2) GDP (100 million Yuan) UR (%)

1997 478 1605.77 25.32

1998 502 1719.87 26.05

1098 Empirical Study on the Co-integration Relationship between Urban Construction land,

Economic Growth and Urbanization Development of Jiangxi Province



1999 529 1853.65 26.79

2000 559 2003.07 27.69

2001 580 2175.68 30.41

2002 602 2450.48 32.20

2003 624 2807.41 34.02

2004 647 3456.70 35.58

2005 671 4056.76 37.00

2006 759 4820.53 38.67

2007 825 5800.25 39.79

2008 848 6971.05 41.36

2009 888 7655.18 43.19

2010 966 9451.26 44.06

2011 986 11702.82 45.70

2012 1034 12948.88 47.51

Based on data from the Statistical Yearbook of Jiangxi and the Statistical Yearbook of China.①

0

4,000

8,000

12,000

16,000

98 00 02 04 06 08 10 12

GDP

400

600

800

1,000

1,200

98 00 02 04 06 08 10 12

UCL

25

30

35

40

45

50

98 00 02 04 06 08 10 12

UR

Figure1: Line Graphs of GDP-UCL-URI

3. Econometric analysis result:

3.1. Correlation analysis:

According to Table 1, carry out the correlation analysis

of urban construction land (UCL ), gross regional

domestic product ( GDP ) and urbanization rate(UR ) with

Eviews6.0. The results are shown in Table 2.

Table 2: Correlation Analysis

Correlation GDP UCL UR

GDP 1.000000

UCL 0.970392 1.000000

UR 0.928668 0.978630 1.000000

( , ) 0.970392r UCL GDP , ( , ) 0.928668r UR GDP ,

( , ) 0.978630r UR UCL

It can be concluded that UCL , GDP and UR have strong

correlation. The urban construction land, economic

growth and urbanization development are driven forces

of each other. Next, apply the three to stability test,

Johansen co-integration test, VEC error correction and

Granger cause-and-effect test.

3.2. Stability test:

This paper uses ADF and PP to test the stability of the

time sequence. Results show that lnUCL , ln GDP

and lnUR are unstable under 10% significant level. But

the first order difference sequences of lnd UCL ,

lnd GDP and lnd UR are stable. Thus, time sequences

of urban construction land, economic growth and

urbanization development are all first order single

whole sequence, namely, lnUCL - 1I

, ln GDP -

1Iand lnUR -

1I. Root test results are shown in

Table 3.

Table3: Root test results of variables

Variable ADF test PP test

t-statistic threshold Conclusion t-statistic threshold conclusion

lnUCL -2.203552 -3.342253(10%) unstable -1.879086 -3.324976(10%) unstable

lnd UCL -3.344689 -3.098896(5%) stable -3.625193 -3.098896(5%) stable




ln GDP -2.536906 -3.342253(10%) unstable -2.392840 -3.342253(10%) unstable

lnd GDP -2.417891 -2.409889(5%) stable -2.304453 -2.298896(5%) stable

lnUR -1.329978 -3.342253(10%) unstable -1.096321 -3.324976(10%) unstable

lnd UR -2.809617 -2.690439(10%) stable -2.809617 -2.690439(10%) stable

3.3. Johansen co-integration test:

This paper uses Johansen co-integration test to see the

co-integration relationship among urban construction

land, economic growth and urbanization development.

The sequence selected has linear trend. But the co-

integration equation has only intercept. Table 4 shows

that the trace statistics and the Max-Eigen statistic are

below significant level of 5%, which means there is a

co-integration relationship.

Table 4: Johansen co-integration test results of variables

Hypothesized

No. of CE(s) Eigenvalue

Trace Statistic Max-Eigen Statistic

statistic 5%Critical

Value Prob.** Statistic

5%Critical

Value Prob.**

None * 0.952260 55.82577 29.79707 0.0000 42.58779 21.13162 0.0000

At most 1 0.596316 13.23798 15.49471 0.1064 12.69971 14.26460 0.0870

At most 2 0.037718 0.538268 3.841466 0.4632 0.538268 3.841466 0.4632

According to the test result, turn the co-integration

equation into a mathematical expression and make it

equal to VECM. There is:

- 0.0609* - 0.644* 1.161VECM LNUR LNGDP LNUCL (2)

Use ADP and PP to test the unit root of VECM. It is

found to be stable and thus gives testimony to the co-

integration relationship. Expression (2) shows the long-

term equilibrium relationship among urban construction

land, economic growth and urbanization development.

The urban construction land expands as the economy

grows and the level of urbanization rises. The elasticity

of urbanization development to economic growth is

0.0609, which means if the economy grows by 1%, the

urbanization rate will increase by 0.0609%. The

elasticity of urbanization development to economic

growth is 0.644, which means if the urban construction

land expands by 1%, the urbanization rate will increase

by 0.644%. And it is concluded that urbanization

development is mainly owing to the expansion of urban

construction land rather than the economic growth. It

indicates that the urbanization development of Jiangxi is

reliant on urban construction land. And urbanization

development is an unavoidable result of the expansion

of urban construction land and the economic growth [16]

.

3.4. Error correction model:

Co-integration relationship is a long-term stable

relationship between variables. VEC is introduced to

reflect the short-term relationship. VEC model is a VAR

model with co-integration restriction [16]

. The co-

integration relationship among three variables, namely,

lnUCL , ln GDP and lnUR , is expressed by VEC as:

int 1 - 4.472 * ln 2.729* ln 11.269Co Eq dnlGDP d UR d UR

The VEC model can be written as:

1 1

0.189 -0.142 1.147 0.549 0.082

0.196 int -0.629 0.419 -0.290 0.131

-0.078 0.166 -0.234 0.227 0.022

t t ty Co Eq y

;

ln

ln

ln

d GDP

Y d UCL

d UR

The overall test result of VEC model shows that the

simulation effect is good. The error correction positive

coefficient is -0.078, which fits the reversion correction

mechanism. The short-term result will be departure

from the long-term equilibrium and will be adjusted to

the equilibrium state by 7.8%. The error correction

positive coefficient of economic growth and urban

construction land is positive, which means as time goes

by, it requires more efforts to adjust to the equilibrium

state.

3.5. Granger causality test:

lnUCL , ln GDP and lnUR are time sequences and have

co-integration relationship. Based on Granger theory

and VEC model, the cause-and-effect relationship

among urban construction land, economic growth and

urbanization development of Jiangxi is shown in

Table5.





Table 5: Granger causality test results based on VEC model

Null Hypothesis: Obs F-Statistic Prob.

LNGDP does not Granger Cause LNUR 14

0.21541 0.8102

LNUR does not Granger Cause LNGDP 8.68700 0.0079

LNUCL does not Granger Cause LNUR 14

0.33468 0.7241

LNUR does not Granger Cause LNUCL 1.39407 0.2969

LNUCL does not Granger Cause LNGDP 14

1.12442 0.3665

LNGDP does not Granger Cause LNUCL 4.27237 0.0496

From test results, it is clear that economic growth is the

Granger cause of urbanization development and urban

construction land is the Granger cause of economic

growth. The results are in line with the anticipation [16]

.

3.6. Impulse response function:

The impulse response function is used to measure the

influence of a standard deviation of the random

disturbance item on endogenous variables and the future

value [14]

. Table 2 shows the response of lnUCL , ln GDP

and lnUR to the standard deviation impact. The

horizontal axial stands for the lag period of the impact.

The vertical axial stands for the response degree of

dependent variable to explainable variables. The real

line is the impulse response function and the dotted line

stands for twice the positive or negative as the standard

deviation belt.

First consider the response of lnUCL to the standard

deviation impact of each variable. Figure 2 Line 1

shows a positive response of urban construction land to

ln GDP . It rises up between period 0-4, reaches the

maximum value at period 4 and decreases from period 4

to 10. It indicates that the economic growth has a

positive influence on urban construction land. The

response to self-standard deviation impact during the

lag period is also positive and reaches the maximum in

period 1 and reduces as a whole in latter periods. The

impact of lnUR to lnUCL is positive and continues to

increase.

Then, analyze the response of ln GDP to standard

deviation impact of each variable. Figure 2 Line 2

shows a positive response of ln GDP to self-standard

deviation impact during the lag period. But it is

downward generally. The response is positive with a

decrease first and an increase after that. It reaches the

minimum value at period 4, indicating that urban

construction land expansion has a great influence on the

economic growth. The response of lnUR to

ln GDP impact is positive and lasts for a long time under

a stable state since the second period. It suggests that

the urbanization development has a big influence on the

economic growth.

Next, analyze the impulse response function of lnUR

to standard deviation impact of each variable. Figure 2

Line 3 shows that the impact of lnUR has little

influence, though positive. It is a downward trend.

When lnUR suffers from lnUCL standard deviation

impact, the response is to increase first and maintains

stable from period 4 to 10. The impact of lnUCL to

lnUR is positive all the time. But the impact of lnUR

to itself shows a downward trend, though positive. It

becomes weaker with the maximum at period 1.

Generally speaking, the urban construction land

expansion, economic growth and urbanization

development have strong correlation. In particular, the

other two present a phase-after-phase relationship with

the economic growth. Economic growth will promote

the expansion of urban construction land and transform

agricultural land to urban construction land. However,

given that the land is the scare resource, the land area

will decrease as the economic grows and under

government control. Urbanization development will also

call for urban construction land.

The increase of urban construction land will also

increase the GDP. This is why local governments have

put efforts to increase the construction land in order to

stimulate the GDP growth. It is worthy of noticing that

government also controls the construction land

according to real situation. In recent years, many

countries have pursued urbanization development by

increasing the investment and input, which resulted in

the rapid growth of GDP. Besides, rural population

begins to dwell in urban areas. It will also promote the

urbanization development [14]

.




-40

0

40

80

120

1 2 3 4 5 6 7 8 9 10

Percent LNUCL variance due to LNGDP

-40

0

40

80

120

1 2 3 4 5 6 7 8 9 10

Percent LNUCL variance due to LNUCL

-40

0

40

80

120

1 2 3 4 5 6 7 8 9 10

Percent LNUCL variance due to LNUR

-40

0

40

80

120

1 2 3 4 5 6 7 8 9 10

Percent LNGDP variance due to LNGDP

-40

0

40

80

120

1 2 3 4 5 6 7 8 9 10

Percent LNGDP variance due to LNUCL

-40

0

40

80

120

1 2 3 4 5 6 7 8 9 10

Percent LNGDP variance due to LNUR

-40

0

40

80

120

1 2 3 4 5 6 7 8 9 10

Percent LNUR variance due to LNGDP

-40

0

40

80

120

1 2 3 4 5 6 7 8 9 10

Percent LNUR variance due to LNUCL

-40

0

40

80

120

1 2 3 4 5 6 7 8 9 10

Percent LNUR variance due to LNUR

Variance Decomposition ± 2 S.E.

Figure 2: The responses of lnUCL , ln GDP and lnUR to corresponding standard deviations

3.7. Decomposition of Variance:

Variance decomposition is to evaluate the importance of

different structure impact by analyzing the contribution

degree of each structure impact to endogenous variables [12]

. Table 6, 7 and 8 are the results of variance

decomposition of urban construction land, economic

growth and urbanization decomposition.

Table 6 shows that the influence of ln GDP on lnUCL

grows fast in the previous 4 periods with an average

growth rate at 7% and the biggest contribution degree of

27.75%. However, the influence of lnUR on lnUCL is

the fastest in the previous 6 periods and becomes stable

at period 7 and 8 with only 26% and continues the fast

growth after that. The change of lnUCL is mainly due to

self-impact. Its contribution degree is as high as 58%.

Table 7 suggests that ln GDP is mainly influenced

by lnUCL . The contribution rate in the previous 6

periods grows very fast, from 28.79% to 41.63%. It

becomes stable after that, with the maximum

contribution rate of 52.98%. This indicates that urban

construction land is the most important factor to the

economic growth. The urbanization development is also

important as its biggest contribution degree is 40.67%.

Table 8 shows that lnUR is mainly influenced by

lnUCL . As construction land is more in urban areas,

many agricultural workers find jobs and settle in cities.

This may facilitate the urbanization process. The

contribution degree of the economic growth to the

urbanization is the largest in the second period and

drops down rapidly after that with the minimum value

of 1.31% at period 5. It doesn’t present a stable trend

within the predicted period [16]

.

Table 6: Variance decomposition results of urban construction land

Period S.E. LNGDP LNUCL LNUR

1 0.032093 0.000000 100.0000 0.000000

2 0.033841 5.935571 87.43660 6.627825

3 0.041851 19.75958 68.21250 12.02792

4 0.051755 27.74684 61.66603 10.58713

5 0.063839 22.06610 65.73274 12.20116

6 0.084157 17.77496 60.29260 21.93244

7 0.100443 18.35312 55.24070 26.40618

8 0.114052 15.17527 57.92360 26.90113

9 0.129123 11.90760 59.57147 28.52093

10 0.142173 10.64629 58.06309 31.29062





Table 7: Variance decomposition results of economic growth


1 0.015517 71.21130 28.78870 0.000000

2 0.024283 72.43622 27.28603 0.277746

3 0.028878 56.59993 22.35259 21.04748

4 0.030977 52.09657 15.67714 32.22629

5 0.034854 36.54180 27.48478 35.97342

6 0.039294 21.08367 41.62746 37.28886

7 0.041909 15.21326 44.38843 40.39830

8 0.046108 12.03310 47.29354 40.67336

9 0.052235 9.388398 51.28810 39.32351

10 0.055870 7.748685 52.98804 39.26328

Table 8: Variance decomposition results of urbanization development


1 0.017965 2.169130 35.48921 62.34166

2 0.024629 3.201246 37.46880 59.32995

3 0.033451 2.085268 53.24939 44.66534

4 0.041809 2.001316 58.69138 39.30731

5 0.045731 1.676429 58.48093 39.84264

6 0.048274 1.504576 59.37822 39.11720

7 0.051012 1.457351 60.69081 37.85184

8 0.052813 1.359713 60.67422 37.96606

9 0.053876 1.364532 60.47869 38.15678

10 0.055003 1.309878 60.75556 37.93456

4. Main conclusion and policy suggestion:

4.1. Main conclusion:

Based on time sequence data of Jiangxi between 1997 to

2000, mainly conclusions are:

(1)There is a co-integration relationship or long-run

equilibrium relationship among urban construction land

lnUCL , economic growth ln GDP and urbanization

lnUR .

(2)Error correction model shows that the urbanization

development is a departure from the long-term

equilibrium in the short run and will be adjusted to the

equilibrium state by 7.8%.

(3)The impulse response function and variance

decomposition results show that urban construction land

is mainly influenced by self-impact and has much to do

with the lag value. Urban construction land plays an

important role to the economic growth and the

expansion of construction area will promote the

urbanization process.

4.2 Policy suggestion:

From the abovementioned analysis, it is concluded that

urban construction land is an unavoidable result of

economic growth and urbanization. In the near future,

the construction land will further expand. Much

agricultural land will become high-qualified cultivated

land that may threat the food security of Jiangxi

Province. Therefore, local government should strictly

comply with urban planning and land planning to

control urban growth and clear illegal construction land.

Currently, there is a large amount of illegal agricultural

construction land in the city periphery. The local

government should take an effective use of the land and

increase the intensive degree of land utilization. It

should also transform the economic growth mode from

the extensive mode of development to the one featuring

intensive, innovative and environment-friendly, so as to

stimulate the growth of Jiangxi in central China [2]

.

Note: ① Based on data from the Statistical Yearbook of

Jiangxi and the Statistical Yearbook of China (1997-

2012). ②This paper is sponsored by Graduate

Research and Innovation Project for Institutions of

Higher Learning in Jiangsu Province (2013B24214).




References:

[1] China statistical yearbook newsroom, Statistical

Yearbook of Jiangxi (1997-2012), China Statistics

Press, 1997-2012.

[2] Zhou Xiaogang and Chen Dongyou, Economic

Growth and the Rise of Jiangxi (A Talk on Paper)

,Jiangxi Social Science, Volume No8. Issue No11.,

PP32-36., 2008.

[3] Li Jinchang and Cheng Kaiming, Dynamic

Econometric Analysis of China’s Urbanization and

Economic Growth, Journal f Finance and

Economics, Volume No32. Issue No9., PP19-

30.,2006.

[4] Gao Wei, He Jie and Zhang Anlu, Econometric

analysis of the dynamic relationship between the

circulation of farmland city, urbanization and

economic growth- a Case Study of Hubei Province

from 1980 to 2003, Resources Science, Volume No

32. Issue No39., PP 564-572., 2010.

[5] Research Group on China’s Economic Growth and

Macro Stability, Urbanization, industrial efficiency

and economic growth, Economic Research, Volume

No10., PP 4-21.,2009.

[6] Ba Shusong, Xing Yujing and Yang Xianling,

Impetus for rbanization and Economic growth: a

Long-term Perspective, Reform and Strategy,

Volume No26. Issue No2., PP 16- 19.,2010.

[7] Liu Aiying, Yao Lifen and Li Qingchen. Co-

integration Analysis on the Relation between

Urbanization and Economic Growth in China,

Journal of Anhui Agricultural Sciences, Volume

No39. Issue No14., PP 8710-8712.,2011.

[8] Meng Junfeng. Study on the relationship between

urbanization and Regional economic growth,

Northwest University, PP 5-50., 2013.

[9] Liu Rui, Zhu Daolin and Zhu Zhanqiang, Analysis

of the Driving Forces of Urban Sprawl in Dezhou

City Based on a Logistic Regression Model,

Resources Science, Volume No31.Issue No11., PP

1919-1926., 2009.

[10] Bao Liping and Wang Jinggang, motives of China’s

Urban Construction Land Expasion, China Land

Sciences, Volume No23. Issue No8., PP 68-

72.,2009.

[11] Chen Ligen, Chen Huiguang and Qu Futian,

Economic growth industrial structure adjustment

and urban construction land scale- a Case Study of

Maanshan City, Resources Science, Volume No26.

Issue No6., PP137-144., 2004.

[12] Jiang Hai, Xia Yanrong and Qu Futian,

Contribution of Urban Construction Land

Expansion to economic growth and Regional

differences, China Land Science, Volume No23.

Issue No8., PP4-8., 2009.

[13] Jiang Hai and Qu Futian, Contribution and

Response of Urban Construction Land Expansion

on Economic Growth at Different Development

Stages, China Population Resources and

Environment, Volume No19. Issue No1., PP70-

75.,2009.

[14] Zhao Ke and Zhang Anlu, An Econometric

Analysis of the Relationships Among Urban

Construction Land, Economic Development and

Level of Urbanization, China Population Resources

and Environment, Volume No21. Issue No1., PP 7-

12.,2011.

[15] [US]Damodar N.Gujarat, Basic Econometrics (the

Fourth Edition), China Renmin University Press,

PP 67-91., 2005.

[16] Li Yanyi, Liu Rong and Ding Weidai, Eviews

Statistical Analysis and Application (Revised) ,

Publishing House of Electronics Industry, PP 41-

188., 2013.

http://www.cnki.com.cn/Journal/D-D1-AHNY-2011-14.htm



http://cdmd.cnki.com.cn/Area/CDMDUnitArticle-10697-2013-1.htm

http://cdmd.cnki.com.cn/Area/CDMDUnitArticle-10697-2013-1.htm

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1104-1109


Discussion on Empirical Formula of Vertical Bearing Capacity of

Single Press-in Pipe Pile

YUZHENG LV1, JIAFU YANG

1, KUN LIU

1 AND ZHAOJUN CHU

2

1Beijing Canbao Institute of Architectural Design, Beijing, China

2Logistical Engineering University of PLA, Chongqing, China


Abstract: The plugging effect and skin or tip compacting effect are produced because the soil at the bottom of

press-in pipe pile is compacted into the inside or outside of pipe pile and lower pipe pile bottom. They can affect the

vertical bearing capacity of single pipe pile, so the bearing mechanism of the press-in pipe pile is more complicated

than solid pile or closed pipe pile. The empirical formula of vertical bearing capacity of single pipe pile is obtained

in the current specifications in which only plugging effect is considered and the skin or tip compacting effect is not

considered. The effects on the vertical bearing capacity of single press-in pipe pile from the plugging effect and skin

or tip compacting effect are devoted in this paper. The outside skin friction resistance and tip resistance of the

empirical formula in the current specifications is revised and the empirical modified formula is established both

considering the plugging effect and skin or tip compacting effect.

Keywords: Press-in pipe pile; plugging effect; compacting effect; vertical bearing capacity.

1. Introduction:

With the advantages of good stability, high bearing

capacity and small settling volume, the pile foundation

is widely applied in high-rise buildings, super high-rise

building and long-span bridges engineering and is an

important base form of engineering constructions in our

country[1]. In recent years, the negative effects caused

by piling activities have attracted more and more social

attentions. The traditional construction methods of

piling (such as pile-driving method by hammer and pile

driving method by vibration) will produce noises,

vibration, pollution and spent clays and furthermore, the

applied mechanical equipments are cumbersome and

greatly rely on the people and lead to large limitation on

the field environment. With continuous improvement of

science and technology, the traditional construction

methods fail to meet the demands of development of

today’s society. Thus Japanese Giken Company put

“press-in” principle into practice and developed

pollution-free piling technology--- “press -in

construction method” [2]. The principle is to utilize the

pull-out resistance generated by several piles which

have been pressed into the ground (completed piles) as

the counter force and press the next pile into the ground

with static load at the top of press-in piles [3].

The press-in method will not cause significant ground

vibration and noises in the course of pile sinking. The

applied machines are convenient with high degree of

automation. The advantages in the aspects of

environmental protection property, economical

efficiency, speed ability and security are obvious

comparing with the traditional methods. The press-in

methods will play an important role in infrastructure

construction in our country [2-5]. But the construction

cost for pile foundation is relatively high and how to

reasonably determine the vertical bearing capacity of

single press-in pile has an important significance. In the

process of pile sinking, the press-in pipe pile will

produce compacting effect and plugging effect both of

which may impact vertical bearing capacity of single

pile of pile foundation. The bearing mechanism is more

complicated

[6-13]. As the static load test method,

dynamic testing method and in-situ testing method are

difficult and the costs are high and will waste time and

energy, the empirical formula method is a widely

applied determination method of bearing capacity of

pile foundation. The empirical formula of vertical

bearing capacity of single pipe pile in Technical Code

for Building Pile Foundations (JGJ94-2008) [14]

determined by empirical relations between index of

physical property of soils and bearing capacity

parameters of foundation only considers the impacts of

plugging effect and neglects compacting effect of pile

sides on the basis of JGJ94-94 [15], but both of them

fail to take the impacts of compacting effect of pile top

into consideration.

This article analyzes the impacts of compacting effect of

pile sides and pile top as well as plugging effect on the

vertical bearing capacity of press-in single pipe pile and

1105 JIAFU YANG, YUZHENG LV, KUN LIU AND ZHAOJUN CHU



amends the empirical formula of vertical bearing

capacity of single pipe pile in current pile foundation

standards.

2. Analysis of compacting effect and plugging effect

of press-in pipe pile

2.1 Formation mechanism of compacting effect

The pile sinking process of press-in pipe pile is a

process where the pile body is formed by squeezing the

soil body under the pile tip downward or toward the side

direction. The process of static pile loading can be

deemed as uniform linear motion, i.e. the velocity of

pile sinking is regarded as unchanged. When the pile tip

enters the soil body, original stress state of soils will be

destroyed and the soil body starts to compact and

generate deformation until punching shear. With

increase of pressing strength on the pile, shearing

strength of the soil body at the pile tip is insufficient to

bear the pressing strength, so the deformation of soil

body increases sharply and plastic failure occurs. The

soil body around the piles will form the plastic zone and

meanwhile, the pile body of press-in pipe pile starts to

sink and pipe pressing resistance on the pile body (i.e.

sum of frictional resistance at the pile side and

resistance at the pile tip) increases. As the pressing

strength on the piles continues to increase, when it is

larger than the pipe pressing resistance, the pile body

will continue to sink and plastic failure will occur from

top to bottom gradually until the press-in pipe pile

completes the pile sinking.

When the pile sinking is carried out in clayey soil body

for the press-in pipe pile, the surficial soil body will

hunch upward and the soil body under the pile tip will

squeeze out around the pile to destroy the soil body

structure of near soil body on one hand; on the other

hand, it will squeeze toward the downside of the pile tip.

The soil-squeezing subareas of soil body around the

outside piles are shown in figure 1. A small part of soil

mass near the pile sides is fully disturbed and rebuilt

owing to radial squeezing action of pile sinking and the

instantaneous strength reduces. But as time goes on, the

shearing strength will increase gradually and a “hard

shell” which clings to the outside surface of piles will

form (subarea 1A in figure 1). The soil mass at area B is

disturbed (the degree is smaller than that of area A)

owing to radial squeezing action of pile sinking. The

strength will finally surpass the original state due to

effects of solidification and thixotropy.

Area C is far from the soil body and the soil mass will

not be disturbed by the radial squeezing action of pile

sinking and its strength will remain unchanged.

There is certain difference between press-in pipe pile

sinking in sandy soil mass and clayey soil mass. At the

earth’s surface, the sandy soils will sink. As the excess

pore water pressure of sandy soils is low and the

dissipation rate is fast, the impacts on the soil strength

surrounding the piles are strengthening effect and

relaxation effect. The ranges of compaction area and

disturbance area of soil mass around the pile during pile

sinking differ greatly from those of clayey soils

2.2 Formation mechanism of plugging effect

In the course of press-in pipe pile sinking, the pile tip

will punch the soil mass and the soils at the down side

of pile tip will be squeezed into pipe pile. With increase

of depth of pile pressing, soils inside the pipe pile will

continually increase and plugging height will increase.

Soil plugging and frictional resistance of walls inside

the pipe pile also increase. When soil plugging and

frictional resistance of walls inside the pipe pile reach

the limit, the open steel pipe pile and closed steel pipe

pile are similar and this effect is called as “plugging

effect”.

The plugging effect should be divided into two states:

when the plugging can make the pipe pile completely

similar to the closed pile, it will be referred to as

complete plugging; conversely, it will be incomplete

plugging. The strength of plugging effect is closely

related to foundation soil category, inner diameter of

pipe pile, pressing depth and depth of bearing stratum,

but mainly depends on the inner diameter of pipe pile. If

the pile specification, soil conditions and pile pressing

method are same, the strength of plugging effect is

inversely proportional to inner diameter of pipe pile. In

the static pile loading, if the inner diameter of pipe pile

is larger and plugging diameter is larger, the tip

resistance of plugging and plugging height will be larger

and it will be hard to form complete plugging state; if

the soil mass of pile tip is softer, plugging resistance is

smaller, the frictional resistance of inner pipe is larger,

it will easier to form plugging state. The steel pipe pile

is pressed in the layered soil and the plugging height is

closely related to hardness of soils and sort sequence.

1106

Discussion on Empirical Formula of Vertical Bearing Capacity of Single Press-in Pipe Pile



3. Analysis of vertical bearing capacity of single

press-in pipe pile

3.1 Analysis of transfer process of vertical load of

press-in pipe pile

The ultimate bearing capacity of press-in pipe pile is

composed of frictional resistance of pile side and

resistance of pile tip. The vertical load on the tip will be

transferred to soils at surrounding pipe pile through

frictional resistance of pipe sides and gradually

transferred downward, leading to soil compaction of soil

mass at the bottom of pile tip, thus producing pile tip

resistance. The transfer process of vertical load of press-

in pipe pile is a process where the pile tip load is

undertaken by the frictional resistance at pile side and

pile tip resistance.

When the vertical load is gradually posed on the pile

tip, the upper pile body will produce compaction and

deformation and downward displacement relative to the

soils at the surrounding pile. The outside frictional

resistance will gradually play a role from top to bottom

and the pile load will be transferred to the soils around

the pile to reduce the pile load and pile compaction and

deformation with increase of depth. The outside

frictional resistance at original state of load imposing is

approximately linear to displacement. With increase of

vertical load of pile tip, the amount of compression and

relative displacement increase and the outside frictional

resistance at the bottom of pile will gradually play a

role. As the pile body and plugging form a whole and

there is no relative displacement, the inside frictional

resistance fails to play a role. When the outside

frictional resistance approaches the limit, the whole pile

body (including plugging) will sink with increase of

vertical load at pile tip and the relative replacement will

increase. The frictional resistance outside the pile plays

a role and the soils at the bottom of pile produced pile

tip resistance caused by compaction and react upon the

net area of pile tip. Meanwhile, the soils at the bottom

of pile squeeze into the pipe pile, causing relative

displacement between soil plug and inner wall of pipe

pile. The frictional force inside the pile plays a role

from top to bottom gradually. When the frictional

resistance outside the pile reaches the limit, the relative

displacement between pile and soil continues to

increase. Frictional resistance outside the pile remains

unchanged. The increment of vertical load at pile tip is

undertaken by pile tip resistance and frictional

resistance inside the pile. With increase of vertical load

at pile block, the frictional resistance inside the pile

gradually plays its role until reaching the limit and

remains unchanged. If the vertical load continues to

increase, the increment will be assumed by the pile tip

resistance until it reaches the limit or the soil body

deform and is not suitable for bearing the load. Then the

load borne by the pipe pile is the vertical ultimate

bearing capacity.

When the vertical load capacity acts upon the pile top,

the frictional resistance outside the pile and frictional

resistance inside the pile as well as pile tip resistance do

not play their roles simultaneously. Owing to soil plugs

inside the pile pipe, frictional resistance outside the pile

will reach the ultimate limit state before frictional

resistance inside the pile which reaches the limit state

prior to the pile tip resistance.

3.2 Impacts of plugging effect on vertical bearing

capacity of press-in pipe pile

Based on the analysis of formation mechanism of

compacting effect and plugging effect during press-in

pipe pile sinking, the soils at the press-in pipe pile tip

will move toward three directions: inwards the pipe pile,

outwards the pipe pile and downward the pipe tip.

The pipe tip soils squeezing into the pipe pile will form

soil plug, producing “plugging effect”. Without regard

to plugging effect and considering plugging effect, the

compositions of vertical bearing capacity of single

press-in pipe pile are shown in figure 2 (a) and (b).

(a) Without regard to

plugging effect

Fig. 1 The vertical bearing capacity of press-in

pipe pile

As shown in figure 2, no matter when the plugging

effect is considered or is not considered, the pile top

load is balanced by frictional resistance outside the

press-in pipe pile 1R , frictional resistance inside the

press-in pipe pile 2R and tip resistance of net area bpR ,

i.e. vertical bearing capacity of single press-in pipe

pile uR is composed of frictional resistance outside the

press-in pipe pile 1R , frictional resistance inside the

press-in pipe pile 2R and tip resistance of net area bpR .

However, there is significant difference between the

cases where frictional resistance inside the press-in pipe

pile is considered and not considered. When the

plugging effect is not considered, the frictional




resistance inside the press-in pipe pile is frictional

resistance from inner walls and soils inside the pipe pile

and plays its role with frictional resistance outside the

press-in pipe pile; when the plugging effect is taken into

consideration, the frictional resistance inside the press-

in pipe pile is frictional resistance from inner walls and

soil plug and plays its role when the frictional resistance

outside pipe pile approaches the limit.

When the plugging effect is not considered, the

calculation formula of vertical bearing capacity of

single press-in pipe pile is shown in formula (1):

paisiaisia

ppaisiaisia

bpu

qddlqdlqd

Aqlqulqu

RRRR

)(4

2

12

1

21

21

(1)

Therein: 1u and 2u peripheral length—peripheral length

outside the press-in pipe pile and peripheral length

inside the press-in pipe pile;

siaq —ultimate collateral resistance of soils at i

layer;

paq —ultimate tip resistance of bearing stratum;

pA —pile tip net area;

il —thickness of soils at i layer;

d、1d —outer diameter and inner diameter of

press-in pipe pile.

When the plugging effect is considered, the calculation

formula of vertical bearing capacity of single press-in

pipe pile is shown in formula (2):

pasaisia

ppasaisia

bpu

qddhqdlqd

Aqhqulqu

RRRR

)(4

2

12

1

21

21

(2)

Therein: saq —weighted collateral resistance of soil

plug;

h —height of soil plug

3.3 Impacts of compacting effect on vertical bearing

capacity of press-in pipe pile

When soils at pile tip squeeze outwards the pile,

compacting effect will occur. The soils surrounding the

pile will produce compaction displacement and become

denser and have more strength. The frictional resistance

outside the pile increases. When soils at pile tip squeeze

downward the pile tip, the pile tip compacting effect

will occur. The soils under the pile tip will produce

compaction displacement and become denser and

impose compacting effect on the foundation soils to

cause larger bearing capacity on the foundation. The

pile tip resistance will also increase. Therefore, when

calculating vertical bearing capacity of press-in pipe

pile, we should not only take plugging effect into

consideration, but also consider the compacting effect.

In order to consider the impacts of compacting effect

on vertical bearing capacity of single press-in pipe pile,

pile side compacting effect coefficient 1 and pipe tip

compacting effect coefficient 2 are introduced. The

frictional resistance outside the pile in formula (2) and

pile tip resistance are revised as shown in formula (3):

pasaisia

ppasaisia

bpu

qddhqdlqd

Aqhqulqu

RRRR

2

2

12

11

2211

21

)(4

(3)

There are multiple factors impacting height of soil

plug h , mainly including inner diameter of pipe pile,

characteristic value of pipe tip resistance, characteristic

value of soil plug and inner wall of pipe pile and depth

of pile tip in bearing stratum. According to analysis of

formation mechanism of plugging effect during press-in

pipe pile sinking, height of soil plug h is in direct

proportion to characteristic value of pipe tip resistance,

characteristic value of soil plug and inner wall of pipe

pile and depth of pile tip in bearing stratum and is

inversely proportional to characteristic value of

frictional resistance inside the soil plug. After referring

to related materials, the value of height of soil

plug h can be obtained as formula (4):

sa

pap

q

qdh

4

1 (4)

Therein: p —pile tip plugging effect coefficient and is

valued as 1 for closed-end pipe pile and is

valued according to formula (5) for open-

end pipe pile.

5,8.0

5,16.0

1

1

dh

dhd

h

b

bb

p

(5)

Therein: bh — the depth of pile tip in bearing stratum

Substitute formula (4) into formula (3) and the vertical

bearing capacity of single press-in pipe pile considering

plugging effect and compacting effect is shown in

formula (6):

1108

Discussion on Empirical Formula of Vertical Bearing Capacity of Single Press-in Pipe Pile



pappaisia

sa

papsapaisia

pasaisiau

qd

qddlqd

q

qdqdqddlqd

qddhqdlqdR

4)(

4

4)(

4

)(4

2

12

2

12

1

112

2

12

1

2

2

12

11

(6)

After analysis of formula (6), the vertical bearing

capacity of single press-in pipe pile is made of three

contents: frictional resistance outside the press-in pipe

pile when the pile side compacting effect is considered;

tip resistance at pile tip net area when the pile tip

compacting effect is considered; tip resistance at pile tip

exposure area converted by frictional resistance inside

the press-in pipe pile when the plugging effect is

considered.

3.4 Comparison with empirical formula of vertical

bearing capacity of single concrete hollow pile in

Technical Code for Building Pile Foundations

Technical Code for Building Pile Foundations (JGJ94-

2008) stipulates that: when vertical bearing capacity of

single concrete hollow pile is determined by empirical

relation between physical property index of soils and

bearing capacity parameters, the calculation can be

carried out according to formula (7):

pappaisia

ppjpaisia

psu

qd

qddlqd

AAqlqu

RRR

44

)(2

121

2

1

(7)

Therein: u —peripheral length outside the press-in pipe

pile;

jA —pile tip net area of concrete hollow pile;

1pA —exposure area of concrete hollow pile.

Comparing formula (6) with formula (7), the calculation

of vertical bearing capacity of concrete hollow pile

according to formula (7) only considers impact of

plugging effect and neglects impact of compacting

effect.

In JGJ94-2008, empirical formula of vertical bearing

capacity of single concrete hollow pile is shown in

formula (8):

pappaisia

ppjpaisia

psu

qd

qddlqd

AAqlqu

RRR

44

)(2

121

21

1

(8)

Comparing formula (6) and (7) with formula (8), the

calculation of vertical bearing capacity of single

concrete hollow pile based on formula (8) considers

impact of pile side compacting effect but neglects

impact of pile tip compacting effect on the basis of

formula (7).

4. Conclusion

During press-in pipe line sinking, the soils at pile tip

will move toward three directions: one is inner side of

pipe pile, forming “plugging effect” and impacting the

frictional resistance inside the pile; the other one is

outside pile, forming “compacting effect” and impacting

frictional resistance outside the pile; the last one is at the

bottom of pile tip, forming “compacting effect” and

impacting pile tip resistance. Therefore, when

calculating vertical bearing capacity of single press-in

pipe pile, it is necessary to consider the impacts of

plugging effect and compacting effect. This article took

both plugging effect and compacting effect into

consideration and introduced pile side compacting effect

coefficient 1 , pile tip compacting effect coefficient 2

and pile tip plugging effect p into consideration and

empirical formula of vertical bearing capacity of

concrete single hollow pile in Technical Code for

Building Pile Foundations is revised.

The values of coefficients of 1 , 2 and p are

researched according to model test, field test, theoretical

analysis and numerical modeling. But there are multiple

reasons impacting plugging effect and compacting

effect in the course of pipe pile sinking, such as size of

outside and inside diameters of pile, depth of bearing

stratum, physical and chemical properties of bearing

stratum, pressing velocity, roughness of inner wall of

pile. It will be difficult to consider impacts of the

following factors dividually. So it has an important

practical significance to find a way to determine values

of 1 , 2 and p efficiently.

5. Reference:

[1] ZENG Guoxi, Feng Guodong et al. Pile Foundation

Engineering Manual [M]. Beijing: China Building

Industry Press, 1995.

[2] Akio Kitamura. Construction Revolution [M]

Japan: Giken Seisakusho Co., Ltd., 2006: 22-35.

[3] Yukihiro Ishihara. The Press-in Method: A

Construction Solution for Urban Development [J]

Chinese Journal of Underground Space and

Engineering, 2010, 6(2): 1707-1708. (In Chinese)

[4] Advantages of Press-in Method [EB/OL].

www.giken.net. cn/st/index.html, 2010-11-20

[5] Akio Kitamura. Silent Piling Technologies [M]

Japan: Giken Seisakusho Co., Ltd., 2008.

[6] CHU Zhaojun, LU Xin, DING Zhenzhou. Analysis

on Deformation of Soil at Surrounding Pile Due to

Press-in Piling in Layered Ground[J]. Journal of

http://www.giken.net.cn/st/




logistical engineering university, 2012, 28(1): 16-

21. (In Chinese)

[7] Du Laibin. Brief Analysis of Plug Effect of PHC

Pipe Piles [J] Industrial Construction, 2005, 35

(Supplement): 590-594. (In Chinese)

[8] LIU Zengrong, DU Lihu. Analysis on Ultimate

Bearing Capacity of Single Pipe Pile of Open-end

Prestressed Concrete Pipe Pile [J] Concrete, 2012,

(3): 120-123. (In Chinese)

[9] LI Qi, LI Song, FEI Kang. Analysis on Soil

Plugging Effect of PHC Pipe Pile [J] Journal of

Water Resources and Architectural Engineering,

2009, 7(2): 45-47. (In Chinese)

[10] SHI Feng. Experimental Research on Load Transfer

Mechanism of Pretensioned High Strength Spun

Concrete Piles [J]. Chinese Journal of Geotechnical

Engineering, 2004, 26(1): 95 99. (In Chinese)

[11] FEI Kang, LIU Hanlong, GAO Yufeng, et al. Load

Transfer Mechanism for Field Pour Concrete Thin

Wall Cased Pile (PCC) [J]. Rock and Soil

Mechanics, 2004, 25(5): 764-768. (In Chinese)

[12] Luo Zhanyou, Tong Jianer, Gong Xiaonan. Study

on Compacting Effects of Jacked pile in the

Condition of Prebored Hole and Tubular Pile [J].

Ground Improvement, 2005, 16(1):3-8. (In

Chinese)

[13] PENG Jie, SHI Jianyong, HUANG Gang. Analysis

of Bearing Capacity of Pile Foundation in

Consideration of Compaction Effect [J] Journal of

Hehai University, 2002, 30(2): 105-108. (In

Chinese)

[14] JGJ94-2008, Technical Code for Building Pile

Foundations[S]. Beijing: China Building Industry

Press, 2008.

[15] JGJ94-94, Technical Code for Building Pile

Foundations[S]. Beijing: China Building Industry

Press, 1994.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1110-1117


Research and Experiment of Sticking Dust Isolation Curtain for

Dust Reduction

YANG XIUDONG AND JIN LONGZHE University of Science and Technology Beijing, Civil and Environmental Engineering School, Beijing, China


Abstract: Coal mine working place dust has been a major hazard of underground accidents and also harms the

health of workers. To solve difficulty in the control of underground coal mine working face dust, dust sticking agent

formula is researched and polyhedral balls are used to conduct related experiments in coordination with the dust

sticking agent, thus finally forming the sticking dust isolation curtain dust reduction technology used for

underground tunnels. As shown from experiment results, the sticking dust isolation curtain technology used for

underground working face can reach the efficiency of 15% to 20% to greatly improve operating environment of

working face, thus providing a low-cost solution for dust control of underground operation space.

Keywords: Dust curtain, Dust technology, Coal mine working place, Migration regularity, Sticky agent.

1. Introduction:

China is a major coal production country. Coals account

for about 70% of primary energy consumption in China.

For a long period of time, coals will continue to be very

important strategic resources. However, China is three

times other coal production countries in the total death

toll of coal mine. Coal mine production safety situation

is still very severe. A vast majority of major coal mine

accidents are caused by coal dust involved in explosion

or coal dust explosion [1, 2]. In addition, mine dust also

can cause a great deal of occupational hazards on the

health of coal mine workers. With improvement of coal

mining mechanization and production efficiency, dust

production capacity also increases [3, 4].

In order to improve underground operating environment

and ensure coal mine production safety, coal production

countries have taken a series of dust control measures

[5, 6]. Currently, coal mines mainly take win and water-

based integrated dust control measures and adopt water

injection for coal cutter dust, atomizing for dust

reduction, foam dust removal and other new

technologies and equipment to greatly improve

underground operating environment. But these

technologies are subject to a number of problems such

as huge investment, high energy consumption and

electrical safety hazards [7-9]. Therefore, it is necessary

to develop a new convenient and low energy

consumption technology used for underground coal

mining.

Sticking dust isolation curtain is proposed under such

idea and background. According to research on dust

migration law of coal mine working face, dust sticking

agent formula is proposed and polyhedral balls used for

relevant experiments with such dust sticking agent to

finally develop the sticking dust isolation curtain

technology used for underground tunnels. The

technology can improve special operating environment

of the whole working face and play a positive role in

prevention of coal dust explosion, so it has important

practical value and practical significance on protecting

heath of underground operating personnel and

promoting coal mine production safety.

2. Research on Dust Migration Law:

2.1. Dust Migration Law Experiment:

The experiment was conducted at No.1 Mine of

Yangshita Company under Huineng Group. FC-4 dust

sampler is used to take sample of dust concentration for

working face of underground intake airway and return

airway. Dried samples are weighted byusing a electronic

balance to get dust concentration for sampling points to

obtain dust migration law of intake airway, working

face and return airway in the mine.

2.2. Dust Migration Law Analysis:

No.1 Mine of Yangtashi Company adopts mining way

of one in service and two in standby and coals are

conveyed completely via belts. Therefore, the research

on dust migration law is mainly measurement and

analysis of intake airway dust, working face dust and

return airway dust.

Dust concentration distribution of intake airway and

return airway of working face is shown in Figure 1.

1111 YANG XIUDONG AND JIN LONGZHE



Dust Migration Law Diagram

0

50

100

150

200

250

Air intake

200m

from

air intake

400m

from

air intake

600m

from

air intake

Coa

l cutter h

ead

Coa

l cutter w

orking

face

Coa

l cutter tail

500m

from

air ou

tlet

400m

from

air ou

tlet

300m

from

air ou

tlet

200m

from

air ou

tlet

Distance

Dust

concentr

ation(mg/m3)

```

Figure1: Dust Concentration Distribution of Intake

Airway and Return Airway of 6299 Working Face.

The dust concentration in all position of the intake

airway is 10 to 200 mg/m3. Air flows of the intake

airway are largely fresh, so the dust concentration is

relatively low. The dust concentration will be increased

when the distance from the working face becomes

smaller. It will follow a reduction trend when the

distance becomes larger. The return airway has a higher

dust concentration than intake airway.

2.3. Dispersity Experiment for Return Airway Dust:

According to dust concentration situation in the

experiment site, three measuring points in the position

of 500m, 400m and 200m in the return airway are

selected to conduct dust dispersity experiment.

2.3.1. Experiment principle: The dust mainly consists

of coal dust while the coal dust can be dissolved in butyl

acetate solutions which can be made into microscope

slide to observe number of coal dust in different grain

sizes by using a microscope. Through statistics,

dispersity of dust on the filterable membrane can be

further determined.

Figure2: Experimental Apparatus

Figure3: Dust Particle Size Distribution in the

Microscope

2.3.2. Dispersity analysis: In the laboratory experiment

of dispersity, the dust dispersity is measured at three

measuring points. According to statistics of dust grain

size on the photo, average percentage of different grain

sizes for dust on each microscope slide is calculated, as

shown in Table 1.

Table1: Dust Dispersity of Return Airway

Sampling

Position of

Filterable

Membrane

Microsco

pe Slide

No.

Average Percentage of Dusts in

Different Particle Sizes

＜2μm 2-

6μm

6-

10μm

＞10

μm

500m of

Return

Airway

A-1 99.00

%

0.48

%

0.25

%

0.26

%

A-2 98.12

% 1.45%

0.34%

0.10%

Mean 98.56

%

0.97

%

0.30

%

0.18

%

400m of Return

Airway

B-1 94.26

% 4.69%

1.02%

0.03%

B-2 98.33

%

1.47

%

0.12

%

0.08

%

Mean 96.30

% 3.08%

0.57%

0.06%

200m of

Return

Airway

C-1 97.35

%

1.18

%

0.46

%

1.01

%

C-2 98.94

% 0.87%

0.19%

0.07%

Mean 98.15

%

1.03

%

0.33

%

0.54

%

As shown from the table, for return airway dust

measured in the underground experiment of No.1 Mine

of Yangtashi Company, dust grains in a small size take

a larger share. The reason may be that water spraying

plays a better effect on dust reduction of coal mine

working face. It reduces concentration of dust in the

airflow, especially number of large-size dusts, so that

only small-size dusts can enter the return airway with

the airflow. For the reason, dust dispersity measured in

the experiment is relatively low.

1112 Research and Experiment of Sticking Dust Isolation Curtain for Dust Reduction



3. Research on Main Formula of Sticking Agent:

Dust sticking agent is mainly made of surfactants. The

surfactant refers to the substance which has fixed

hydrophilic and lipophilic groups arranged directionally

on the solution surface to significantly reduce surface

tension [10, 11]. Therefore, surfactant solution can be

used to reduce water surface tension and generate

foaming effects. It can be made into dust sticking agent

and stunk on the dust isolation curtain to bond

underground dusts (mainly consisting of coal dust), thus

significantly reducing dusts.

According to actual needs and market availability of

surfactants, six kinds of common surfactants are

selected for research, including OP-10, triton X-100,

dodecylbenzene sulfonic acid (LAS), sodium dodecyl

benzene sulfonate, sodium lauryl sulfate and sodium

alcohol ether sulphate (AES).

3.1. Surfactant Wetting Experiment:

Wetting performance of surfactant solution on coal dust

is measured in the coal dust sedimentation experiment

mainly by sedimentation method specified in the

MT506 - 19965 Mine Dust Reduction Agent

Performance Measurement Method.

Wetting performance of surfactant solution on coal dust

is measured mainly in the coal dust sedimentation

experiment. If the coal dust sedimentation time is

shorter, wetting performance of coal dust in the

surfactant solution will be better; otherwise, wetting

performance of coal dust will be reduced.

Figure4: Coal Dust Sample Sedimentation

As shown from Figure 4, sedimentation time of coal

dust in the surfactant solution of 1% is measured. Its

statistics and analysis are shown in Figure 5.

Average sedimentation time of coal dust in surfactant solution of 1%

Aver

age

sedim

enta

tion tim

e /s

Sodium

alco

hol e

ther su

lpha

te (A

ES)

DOde

cyl b

enze

ne su

lfona

te

Sodium

dod

ecyl su

lfate

OP-

10

Triton

X-1

00

Dod

ecylbe

nzen

e sulfonic a

cid (L

AS)

Surfactant solution of 1%

Figure5: Sedimentation Experiment for Coal Dust in

Surfactant Solution of 1%

As shown from Figure 5, sedimentation time for coal

dust in surfactant solution of 1% follows the following

sequence: OP-10> sodium alcohol ether sulphate

(AES)> dodecylbenzene sulfonic acid (LAS)> dodecyl

benzene sulfonate> triton X-100> sodium dodecyl

sulfate. Therefore, triton X-100 and sodium dodecyl

sulfate can provide the better wetting performance.

3.2. Determination of Surfactant Surface Tension:

When dust suppressant can wet coal dusts, its surface

tension will be lower than water surface tension. Its

wetting performance will get better if the surface

tension is reduced. Surface tension for surfactant

solution of 0.1% is determined mainly by using BZY-

201 automatic surface tensiometer.

Determination results are shown in Figure 6.

As shown from Figure 6, all surfactant solutions of

0.1% have an ideal surface tension value of 20 to 40

mN•m-1, indicating surfactant solutions can wet coal

dusts to get the better dust suppression effect.

The experiment shows that, triton X-100, sodium lauryl

sulfate and sodium alcohol ether sulphate (AES) can be

taken as auxiliary materials for dust sticking agents to

provide a better effect; triton X-100 can provide the bets

effect, followed by OP-10, sodium dodecyl benzene

sulfonate and sodium alcohol ether sulphate (LAS).

Surface tension of surfactant solution of 0.1%

Surf

ace tensi

on m

N/m

Deion

ized

water

OP-

10

Dod

ecylbe

nzen

e sulfonic a

cid (L

AS)

Sodium

dod

ecyl su

lfate

Dod

ecyl ben

zene

sulfo

nate

Surfactant solution of 0.1%

sodium

alco

hol e

ther su

lpha

te (A

ES)

Trito

n X-100

Figure6: Determination of Surface Tension for

Surfactant Solution of 0.1%




As shown from Figure 6, all surfactant solutions of

0.1% have an ideal surface tension value of 20 to 40

mN•m-1

, indicating surfactant solutions can wet coal

dusts to get the better dust suppression effect.

The experiment shows that, triton X-100, sodium lauryl

sulfate and sodium alcohol ether sulphate (AES) can be

taken as auxiliary materials for dust sticking agents to

provide a better effect; triton X-100 can provide the bets

effect, followed by OP-10, sodium dodecyl benzene

sulfonate and sodium alcohol ether sulphate (LAS).

4. Experimental Research on Field Dust Removal

Efficiency of Dust Isolation Curtain:

4.1. Field Experiment of Dust Isolation Curtain:

Dust control principle of the dust isolation curtain is that

dust particles are collected by effective collision

between dust particles and polyhedral ball blades and

dust sticking agent adsorption to reduce dusts. For this

reason, dust removal efficiency of the dust isolation

curtain may be associated with specific surface area of

polyhedral balls (size), curtain layer number and relative

position, dust sticking agent and other factors.

In order to study influencing factors of dust removal

efficiency for dust isolation curtain, a field experiment

of the dust isolation curtain is conducted in the return

airway of underground working face in the mine of

Yangshita Company. Polyhedral balls in two sizes are

selected for this experiment. The polyhedral ball in a

larger size is at the diameter of 50mm, specific surface

area of 236m2/m

3 and porosity of 85%; the polyhedral

ball in a smaller size is at the diameter of 38mm,

specific surface area of 300m2/m

3 and porosity of 90%,

as shown in the Figure 7.

Figure7: Scattered Small Polyhedral Balls

Polyhedral balls are collected with an engineering line

to form a dust isolation curtain. According to

measurement experiment of roadway at the sectional

dimension of 3.3m high and 5.2m wide, a small ball

curtain section needs about 120 small ball strings and

each string needs about 90 small balls; a large ball

curtain section needs about 80 large ball strings and

each string needs about 68 large balls, as shown in

Figure 8.

Figure8: Preliminarily Connected Curtain Strips of

Dust Isolation Curtain

Four sections are suspended in the experiment, of which

two are made of small balls and the other two made of

large balls. Balls are connected on the ground, packed in

sacks for loading and then transported to underground

return airway experiment position. It is required to hook

a section of curtain string by using a prefabricated

curved hook, lift it on the roadway roof anchor net and

then vertically hang it. No.1 large ball curtain, No.2

large ball curtain, No.1 small ball curtain and No.2 ball

curtain are hung successively along airflow direction at

the interval of 20m and wait for some time before

conducting the experiment. The dust isolation curtain

hung is shown in Figure 9.

Figure9: Dust Isolation Curtain Hung

After four dust isolation curtains are hung, successively

select sampling points along the airflow direction.

Sampling points are selected in a position 10m from the

large ball curtain in the upwind and downwind. Two

experimenters simultaneously start dust samplers to take

samples for 5min with a preset flow rate of 20L/min to

get the dust concentration before and after the dust runs

through a layer of large ball curtain, thus obtaining dust

removal efficiency for a layer of large ball curtain by

calculation. In such a way, the dust removal efficiency

can be calculated for curtains of one small ball layer,

two large ball layers, two small ball layers, one large

ball layer and one small ball layer, two large ball layers




and one small ball layer, one large ball layer and two

small ball layers as well as two large ball layers and two

small ball layers.

After the above experiment comes to an end, select

appropriate time to uniformly spray dust sticking agents

on the dust isolation curtain and then conduct an

experiment for sticking dust isolation curtain

technology. The dust sticking agent is that previously

used for similar experiments. After the dust sticking

agent is sprayed, conduct an experiment and compare

the result with that of the experiments without dust

sticking agents to obtain dust removal efficiency of ball

curtains coated with dust sticking agents.

4.2. Result of Dust Removal Efficiency Experiment for

Dust Isolation Curtain:

In the experiment, the dust concentration is measured

respectively for roadways in the upstream and

downstream of the dust isolation curtain. The dust

concentration of the upstream roadway is represented by

C1 and that of the downstream roadway is represented

by C2 to calculate dust removal efficiency E of the dust

isolation curtain.

1 2

1

100%C C

EC

(1)

According to the data from the experiment, make

calculation to obtain values in Table 2.

Table2: Dust Sticking Agent Influence on Dust Removal Efficiency of Dust Isolation Curtain

Number of dust isolation curtain

layer

Efficiency of dust isolation curtain uncoated with dust sticking

agent

Efficiency of dust isolation curtain coated

with dust sticking agent

Efficiency improvement

Average efficiency

improvement

Efficiency improvement

percentage

Average improvement

percentage (excluding one layer of large

balls)

Average efficiency

improvement

percentage

One large ball layer

7.93% 10.70% 2.77%

10.22%

34.96%

287.45% 255.89%

Two large ball layers

8.89% 19.73% 10.84% 121.97%

Two large ball layers and one small ball layer

4.14% 13.64% 9.49% 229.22%

Two large ball layers and two

small ball layers 3.61% 17.45% 13.84% 382.86%

One large ball layer and two

small ball layers 4.79% 14.60% 9.81% 204.81%

Two small ball layers

3.51% 20.63% 17.12% 487.89%

One small ball layer

3.01% 11.98% 8.97% 297.79%

One large ball layer and one

small ball layer 3.11% 12.04% 8.93% 287.59%

As shown from Table 2, dust removal efficiency of the

dust isolation curtain is significantly improved after the

dust sticking agent is sprayed. One large ball layer has a

relatively small improvement of dust removal efficiency

while the other layers improve the dust removal

efficiency by more than 100%. Two small ball layers

provide the most significant improvement of dust

removal efficiency up to nearly 500%, indicating the

dust sticking agent can provide good dust removal

effects. According to gas-solid two phase flow theory,

the dust sticking agent can increase surface tension of

ball liquids to enhance dust particle absorption.

5. Analysis on Influencing Factors of Dust

Removal Efficiency for Dust Isolation Curtain:

5.1. Influences of Layer Number on Dust Removal

Efficiency:

1) Comparison between curtains made of the same

balls: It is required to compare different layers of

large ball section and small ball section coated

with the dust sticking agent. One small ball layer,

two small ball layers, one large ball layer and two

large ball layers respectively have a dust removal

efficiency of 11.98%, 20.63%, 10.70%, 19.73%

on the gross section coated with the dust sticking

agent. Two small ball layers are two times one

small ball layer while two large ball layers are

also two times one large ball layer in the dust

removal efficiency, indicating the curtain made

of the same balls will have the higher dust




removal efficiency if its layers are more and the

efficiency is approximately proportional to the

number of layers.

2) Comparison between curtains made of different

balls: It is required to compare curtains made of

different balls in the dust removal efficiency on

the section coated with the dust sticking agent.

Two-layer curtain made of one small ball layer

and one large ball layer, three-layer curtain made

of two large ball layers and one small ball layer,

three-layer curtain made of two small ball layers

and one large ball layer and four-layer curtain

made of two large ball layers and small ball

layers respectively reach the dust removal

efficiency of 12.04%, 13.64%, 14.60% and

17.45% and the efficiency increases with the

number of layers.

Dust removal efficiency of the dust isolation curtain

increases with the number of layers, because more

layers can improve coal dust collection probability to

increase collisions between coal dust and balls.

5.2. Influences of Ball Size on Dust Removal

Efficiency:

Curtains made of different balls in the same layers are

compared in the dust removal efficiency as follows:

For one-layer curtains, large ball curtain and small ball

curtain respectively reach the dust removal efficiency of

10.70% and 11.98%, indicating the small ball curtain

can provide a better effect than the large ball curtain.

For two-layer curtains, large ball curtain, small ball

curtain and combined curtain respectively reach the dust

removal efficiency of 19.73%, 20.73% and 12.04%,

indicating the small ball curtain can provide a better

effect than the large ball curtain. For three-layer

combined curtains, the combined curtain made of two

large ball layers and one small ball layer and that made

of two small ball layers and one large ball layer

respectively reach the dust removal efficiency of

13.64% and 14.60%, indicating the combined curtain

provides the higher efficiency.

5.3. Change of Sticking Agent Dust Removal

Efficiency with Time:

The dust sticking agent can improve dust removal

efficiency of the dust isolation curtain, but it will be

reduced when absorbing coal dusts. Its dust removal

efficiency will decline with the time. For the reason,

dust sticking agents shall be timely added within a

certain time but frequent addition will cause cost

increase. Therefore, appropriate adding interval shall be

selected to reach the best level of cost and efficiency.

In this experiment, curtains hung for different days are

selected to calculate the dust removal efficiency in order

to study the change of dust removal effect with the time.

Dust reduction experiment is conducted for the curtain

made of two large ball layers, that made of one small

ball layer, that made of two small ball layers, that made

of two large ball layers and one small ball layer, that

made of two large ball layers and two small ball layers,

that made of one large ball layer and two small ball

layers and that made of one large ball layer and one

small ball layer just coated with the dust sticking agent

and those coated with the dust sticking agent for one

day. Its data is processed to calculate the dust removal

efficiency of curtains just coated with the dust sticking

agent and those coated with the dust sticking agent for

one day, as shown in Table 3.

It can be seen that, dust removal efficiency of dust

sticking agent coated curtains gradually declines with

the time. The dust removal efficiency of all curtains can

decline by more than 10% in one day. The curtain made

of two small ball layers provide the largest decline up to

nearly 50% while the curtain made of two large ball

layers and one small ball layer provide the smallest

decline of 10%.

Then, dust removal efficiency of curtains uncoated with

dust sticking agent and that of those just coated with

dust sticking agent are used to calculate the decline of

dust removal efficiency from coating to loss of dust

sticking agent effect. Calculation formula is shown as

follows:

Table3: Change of Dust Removal Effect of Dust Sticking Agent Coated Dust Isolation Curtain with Time

Number of dust isolation curtain layer

0 day One day

Dust reduction

effect decline

Average dust reduction effect

decline

Dust reduction effect decline percentage

Average dust reduction effect

decline percentage

Two large ball layers 12.99% 10.70% 2.29%

4.50%

17.63%

26.32%

Two large ball layers and one small ball layer

19.73% 11.95% 7.78% 39.44%

One large ball layer and one small ball layer

13.64% 12.27% 1.37% 10.02%

One large ball layer and two small ball layers

12.04% 10.13% 1.91% 15.87%

Two small ball layers 14.60% 10.19% 4.41% 30.21%

One large ball layer 20.63% 11.39% 9.24% 44.77%




1 2

1

( )100%

E E

E

(2)

Where, E1 is the dust removal efficiency of dust

isolation curtain just coated with dust sticking agent, E2

is the dust removal efficiency of dust isolation curtain

uncoated with dust sticking agent and ω is the decline of

the dust removal efficiency at loss of dust sticking agent

effect.

According to dust removal efficiency decline at loss of

dust sticking agent effect and that of curtains coated for

one day, effective period of dust sticking agent can be

calculated by the following formula:

1

D

(3)

Where, ω is dust removal efficiency decline at loss of

dust sticking agent effect and ω1 is that of curtains

coated with the dust sticking agent for one day.

Combined sticking dust isolation curtains have a

different effective period. The dust sticking agent can be

effective in the maximum time of seven days and

minimum time of one day, indicating effectiveness of

the dust sticking agent is unstable. The possible reason

is that the dust sticking agent is coated on each layer

and four layers are mutually independent in the

effectiveness of the dust sticking agent, but the

experiment adopts the method of mixed layer number

and measurement in groups for measurement of dust

removal efficiency. It causes a difference of effective

period between combined sticking dust isolation

curtains.

On the whole, average effective period of combined

sticking dust isolation curtains is 3.2 days. It can be

indicated that, dust isolation curtains coated with the

dust sticking agent can reach average effective period of

three days. After three days, dust isolation curtains need

to be cleaned and recoated with the dust sticking agent

to re-obtain high dust removal efficiency.

6. Conclusion:

Feasibility of sticking dust isolation curtain technology

is studied according to the research on migration law of

dusts from the upstream to the downstream of the

working face in No.1 Mine of Yangtashi Companny,

experimental determination of wetting and surface

tension for six kinds of surfactants and field experiment

for sticking dust isolation curtain technology. The

conclusion is reached as followed:

1) A shown from experiment research on six kinds of

surfactants including OP-10, triton X-100,

dodecylbenzene sulfonic acid (LAS), sodium

dodecyl benzene sulfonate, sodium lauryl sulfate

and sodium alcohol ether sulphate (AES), triton X-

100 and sodium lauryl sulfate can be taken as

auxiliary materials for dust sticking agents to

provide a better effect; triton X-100 can provide the

bets effect, followed by OP-10, sodium dodecyl

benzene sulfonate and sodium alcohol ether

sulphate (LAS).

2) According to field sampling measurement and

analysis, the dust concentration near the coal mine

working face in No.1 Mine of Yangtashi Company

is relatively low on the whole and it can be up to

200mg/m3. The dust concentration will be higher

when getting closer to the working face; it will

decline when getting away from the working face;

the return airway is higher than the intake airway in

the dust concentration.

3) According to field experiment of the dust sticking

agent, if no dust sticking agent is coated, small ball

curtains can provide relatively high dust removal

efficiency and the dust removal efficiency gradually

improves with increase of layers; if the dust

sticking agent is coated, small ball curtains are

higher than large ball curtains by 8.27% in the dust

removal efficiency.

4) According to analysis on the change of dust

removal effect with the time, coated dust sticking

agent can effectively improve the dust removal

efficiency and effective period of the dust sticking

agent can be up to three days. After three days, dust

isolation curtains need to be cleaned and recoated

with the dust sticking agent to re-obtain high dust

removal efficiency.

7. Reference:

[1] Dastidar Ashok, Amyotte Paul, Going John and

Chatrathi Kris, “Inerting of coal dust explosions in

laboratory-and intermediate-scale chambers”, Fuel,

vol. 80. no. 11., pp. 1593-1602., 2001.

[2] Sapko Michael J, Weiss Eric S, Cashdollar Kenneth

L, Zlochower Isaac A, “Experimental mine and

laboratory dust explosion research at NIOSH”,

Journal of Loss Prevention in the Process

Industries, vol. 13. no. 3., pp. 229-242., 2000.

[3] Page Steven J, Reed Randy, Listak Jeffrey M, “An

expanded model for predicting surface coal mine

drill respirable dust emissions”, International

Journal of Mining, Reclamation and Environment,

vol. 22. no. 3., pp. 210-221., 2008.

[4] Pollock D. E., J. D. Potts and G. J. Joy,

“Investigation into dust exposures and mining

practices in mines in the southern Appalachian

Region”, Mining Engineering, vol. 1. no. 2., pp. 44-

49., 2010.

[5] Belle Bharath K. and Du Plessis Jan, “Recent

advances in dust control technology on South

African underground coal mines”, Journal of the




Mine Ventilation Society of South Africa, vol. 55.

no. 4., pp. 138-144., 2002.

[6] Ren Ting X, Brian Plush, Naj Aziz, “Dust Controls

and Monitoring Practices on Australian

Longwalls”, Procedia Engineering, vol. 26., pp.

1417-1429., 2011.

[7] Belle Bharath K, “Recent developments in the

control of respirable dust concentrations following

the 12m directive”, J Mine Vent Soc S Afr, vol. 56.

no. 3., pp. 85-90., 2003.

[8] Swanson J.-G, Agasty A, Langefeld O, “Wetting

the coal face for dust control in longwall mining at

high ventilation air speeds”, SME Annu. Meet.

Exhib., SME, Meet. Prepr, pp. 536-540., 2012.

[9] Kim J. and J. C. Tien, “Effect of operating

parameters of a liquid spray system on coal dust

suppression”, CIM bulletin, vol. 93. no. 1040., pp.

72-75., 2000.

[10] Dixon-Hardy, Darron William, Beyhan, Sunay,

Ediz I., Göktay and Erarslan Kaan, “The Use of Oil

Refinery Wastes as a Dust Suppression Surfactant

for Use in Mining”, Environmental Engineering

Science, vol. 25. no. 8., pp. 1189-1196., 2008.

[11] Jin_Longzhe, Jiang_Nannan, Chen_Shaojie,

“Experimental study on the impact of complex

surfactant in coal-water contact angle”, Progress in

Mine Safety Science and Engineering II, pp. 175-

177., 2014.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1118-1126


Predicting frost penetration of high-speed railway subgrade in

seasonally frozen regions based on empirical method

ZHANG YU-ZHI1,2

, DU YAN-LIANG1 AND SUN BAO-CHEN

1

1Structural Health Monitoring and Control Key Laboratory of Hebei Province, Shijiazhuang Tiedao University,

Shijiazhuang, China 2School of Mechanical and Electronic Control Engineering, Beijing Jiaotong University, Beijing, China


Abstract: The frost penetration is important to choose the anti-freezing measures of the high-speed railway (HSR)

in seasonally frozen regions. However, at present, the effective prediction model is lacking. Based on the three-year

field temperature data captured in Harbin-Dalian HSR subgrade in China, we proposed to develop suitable empirical

formulas of the frost penetration depths with air and subgrade surface freezing indices. The steps toward its

development are detailed. Firstly, the frost penetration and surface temperature with time for further research could

be obtained from a temperature semi-empirical estimation formula continuous in time and space on the basis of

temperature monitoring data. And the air freezing index was also appropriately reduced through the comparison of

ratios between freezing and thawing indices during different freezing-thawing cycles. Secondly, correlations

between frost penetrations and freezing indices, including air and subgrade surface, were developed using empirical

methods. The surface n factors of different sections were presented and analyzed to identify the main influence

factors of frost penetration. Finally, prediction empirical formulas with a set of correction coefficients, which allows

conversion between the site-specific frost penetration and those associated with other conditions, were put forward.

The procedure and results provide reference for the design, construction and maintenance of similar engineering in

seasonally frozen regions.

Keywords: Frost penetration, Empirical method, freezing index, High-speed railway subgrade, seasonally frozen

region.

1. Introduction:

Frost penetration, reflecting the soil freezing capacity, is

the main basis of the anti-freezing measures for the

engineering experiencing frost problems in frozen

regions (Smith and Rager, 2002; Martel, 1988;

McCormick, 1990; Baïz et al., 2008). Besides, the frost

penetration in frozen regions directly influences the

engineering stability on it. The soil frost heaving and

thaw subsidence within frost penetration will cause the

irregularity of rail track; even influence the traffic

safety. (Baïz et al., 2008)

There are two kinds of models to predict frost

penetration, including theoretical and empirical models.

Theoretical models, incorporating meteorological

conditions and soil properties, can describe the physical

meaning. Yet the parameters included are rather

difficult to determine or complicated equations and

correlations are very hard to solve. (Smith and Rager,

2002; Martel, 1988; McCormick, 1990; Xu et al., 2010;

Xu et al., 2011a; Xu et al., 2011b) Since the empirical

model of frost penetration is generated in the form of

simple equations with the correction coefficients created

from the long-term historical records of frost effects, the

revised empirical formulas are usually used to predict

the frost penetration of the frozen regions engineering

(Baïz et al., 2008; McKeown et al., 1988; Shin E et al.,

2013; Bing et al., 1995; Khalili et al., 2007; Li and

Chen, 1992). In fact, the frost penetration depths of

these engineering vary from each other because the

different structure layers on it have largely changed the

soil thermal conduction. The new engineering structural

style in cold regions may produce the frost penetration

quite different from the predesigned.

Harbin-Dalian high-speed railway (HD HSR) is the first

HSR constructed in seasonally frozen regions in China

and the first HSR put into operation in the world.

Therefore, in the case of lacking plenty of monitoring

data, frost penetration prediction formulas for ordinary

railway were used at the design stage, relevant

theoretical research of the anti-freezing measures were

based on the former similar research, estimating the

frost penetration under supposed initial and boundary

conditions (Xu et al., 2010; Xu et al., 2011a; Xu et al.,

2011b). The soil layers being disturbed in construction

need certain time to form new balance thermal


1119 ZHANG YU-ZHI, DU YAN-LIANG AND SUN BAO-CHEN



equilibrium state, so even the computation using less

than two-year monitoring data will be not consistent

with the field monitoring.

Air temperature, wind velocity and direction, solar

radiation density, subgrade height and route direction

are all influence factors of subgrade temperature field

which all aggravate the complexity of heat transfer

process between subgrade and ambient environment

(Liu et al., 2013). In general, it is very hard to obtain all

data about the factors, and the measurement of many

parameters, especially for the heat change between the

soil and atmosphere, has low accuracy. Air temperature

is the main factor which governs frost penetration depth

of soil (Khalili et al., 2007). Therefore, the empirical

prediction formula of frost penetration is often analyzed

according to air freezing index or surface n factor

method (Baïz et al., 2008; McKeown et al., 1988;

Cheng 2003).

Since the field temperature data are usually influenced

by kinds of fluctuant factors, frost penetrations during

different freezing-thawing cycles are definitely diverse.

In view of this, on the basis of three-year monitoring

data from four different subgrade sections of HD HSR,

the semi-empirical temperature estimation forumula is

put forward to determine the frost penetration and

surface freezing/thawing temperature with time for

further research. Ratio between air freezing and thawing

indices are compared, and the reduction method of a

selected freezing index with time is advised. Thus,

generalized correlations of the frost penetration with

freezing indices, including air and subgrade surface, can

be develpoed in a simple form. The emprical prediction

formulas of frost penetration suitable for HSR subgrade

in seasonally frozen regions are then discussed. During

the progress, main influence factors of frost penetration

are analysed and considerd.

2. Semi-empirical Temperature Estimation Formula

and Air Freezing Index Reduction Method

2.1 Subgrade Monitoring Profile:

The monitoring sections are in Shenyang-Harbin

division of HD HSR, as presented in Table 1.

Table 1: Monitoring Sections

Number Section Style Subgrade

Height

Geographic

Location

Local Maximum

Frost penetration

Section 1

DIK503+580 Embankment 3.179m

The Kaiyuan stati-

on adjacent to

Shenyang in Liao-

ning Province

1.37m Section 2

DIK503+745

Embankment-culvert

transition 3.679m

Section 3

DK883+330 Embankment 5.433m

The Shuangcheng

station adjacent to

Harbin in Heilo-

ngjiang Province

1.85m Section 4

DK883+400

Embankment-culvert

transition 3.082m

The subgrade width is 13.6m. The upper 0.4m of it is

subgrade top layer, composed of well-graded broken

stone. The insulting layer, just below broken stone, is

composed of 0.05m sand cushion and composite

geomembrane. The lower part under the insulting layer

is 1.0m anti-freezing layer with Ⅰ-class non-frost A/B

fillings (high quality and good filling materials in China

railway subgrade design code). And then, under the

above layers, the rest subgrade is composed of normal

A/B fillings. The foundation soil is reinforced by CFG-

piles.

Under the subgrade top layer of embankment-culvert

transition, the well-graded broken stone stabilized with

3-5% cement addition is used in the form of right-

angled trapezoid. Upper line of the trapezoid is 2m and

gradient is 1:2. Well-broken stone of roadbed top layer

in 20m range are added with 3-5% cement, too.

The temperature monitoring points were arranged as

follows: 0.5m intervals between 0.8m and 3.8 m, 1.0m

intervals between 3.8m and 10.8 m under subgrade

surface. Three measuring bore holes, 14 points for each

one, are in east shoulder, central line and west shoulder.

The main part of temperature sensors is thermistor.

Measurement range is from -40°C to ＋60°C, accuracy

is ±0.03°C between -20°C and＋20°C. The monitoring

data, from 1 August 2010 up to 31 July 2013, show that

frost penetration depths at different locations of the

subgrade section are rather different because of different

covering layers, solar radiation and so on (Zhang et al.,

et al., 2014). In general, the measured maximum depths

of frost penetration during the three-year (Table 2) are

larger than the local maximum depths of frost

penetration. The orders of frost penetration depths at

different locations are the same: west shoulder, east

shoulder, and central line.

2.2 Semi-empirical Temperature Estimation

Formula

A temperature estimation formula based on statistical

analysis here is advised to process the three-year data

from the four monitoring sections.

1120 Predicting frost penetration of high-speed railway subgrade in seasonally frozen regions

based on empirical method



Considering the time phase difference, supposing the

ground is homogeneous soil with constant state, ,T z t

varying with the time t and the depth z , at a certain

subgrade section location in a specific area, can be

estimated as (Zhang et al., 2014; ANDERSLAND and

LADANYI, 2011):

2

, exp sinm s

u u

tT z t T A z z

a p p a p

(1)

where is the annual average temperature at depth z,

unit is degree centigrade, °C, As is the annual

temperature amplitude of subgrade surface, unit is °C, z

is the depth from the subgrade surface, unit is m, is

the average soil thermal diffusion coefficient ignoring

the earth interior heat flux, unit is m2/day, p is the

vibration period, 365 days,

is the initial phase angle ,

uz a p

is the time lag of the depth z from the

surface, t is the interval from the beginning date, unit is

day,

It is noticed that in the reference (Zhang et al., 2014;

ANDERSLAND and LADANYI, 2011) is adopted

uniformly using the average temperature of the

subgrade surface, which is suitable when the

temperature envelope is zero gradient type, but not

suitable when the temperature envelope is the positive

or negative type. So this paper adopted fitting functions

describing the average temperature with depth.

Original series were obained through averageing the

three-year monitoring annual average temperature and

temperature amplitude respectively. Then the functions

of annual average temperature and temperature

amplitude with depth can be fitted using least square

method. Through the initial value at depth 0.8m

comparasion with maximum value of the first cycle, the

initial phase angle could be determined. The thermal

diffusivity difference of broken stone and A/B fillings is

not considerd. The estimation formula could not only

eliminate the dispersibility of the field monitoring data,

but also provide a tool for temperature calculation

continuous in space and time. And then, supposed the

soil freezes at 0 °C, the frost penetration with time, a

series of [ , ] can be caculated from the

equations , 0T z t . The maximum depth of frost

penetration calculated (Table2) are very consistant with

the meastured results, whether the depths and the arrival

dates. The subgrade surface temperature with time could

be attained by the ,T z t at the depth z=0, then the

subgrade surface index could be computed.

Table 2: Measured and estimated maximum depths of frost penetration and arrival dates

Location

2010-2011 2011-2012 2012-2013 Estimated Model

Measure

d (m) Date

Measure

d (m) Date

Measure

d (m) Date

Estimate

d (m) Date

Section

1

East

shoulder 2.05 26 Feb 2.75 6 Mar 2.18 1 Mar 2.32 1 Mar

Central line 1.63 21 Feb 2.10 1 Mar 1.97 2 Mar 2.09 2 Mar

West


Section

2

East


Central line 1.55 6 Mar 2.10 6 Mar 2.25 6 Mar 2.25 5 Mar

West

shoulder 2.00 6 Mar 2.85 6 Mar 2.75 6 Mar 2.83 5 Mar

Section

3

East



West

shoulder 2.02 16 Mar 2.94 1 Apr 2.45 1 Apr 2.85 20 Mar

Section

4

East



West


2.3 Air Freezing Index Reduction Method

Corresponding to the frost penetration with time

calculated from the above estimation model, the

freezing index with time should be determined from the

three-year air temperature data. Through comparing the

ratio between air freezing and thawing indices, the

reduction method of air freezing index is proposed.

The surface thawing and freezing indices and their ratio

are the important indicators of the subgrade thawing

subsidence and frost heaving in frozen regions (Cheng

mT

sA

deptht depthz




et al., 2003). Referring to the ratio mentioned above, the

ratio between air freezing and thawing indices (FTR) is

obtained:

a f atr T T (2)

where a fT and atT are annual air freezing and

thawing indices, unit is degree-day centigrade, °C.days,

the sum of temperatures below and above 0°C in the

cold period using Norwegian method (Khalili A, 2007).

The FTR reflects the specific value between annual

cooling and heat quantity imported into the subgrade

and reveals the subgrade freezing capacity of one

freezing-thawing cycle to some extent. The ratios of the

three freezing and thawing cycles from 2010 to 2013 are

presented in Table 3. Generally, compared with

measured frost penetration (Table 2), the larger the FTR

is, the deeper the frost penetration is.

On account of the parameters of temperature estimation

formula obtained from the three-year monitoring data

using statistical analysis, the frost penetration depths

estimated are considered to correspond with the average

FTR respectively. In addition, to compare with the long

term meteorological statistical data, the FTR derived

from monthly average temperature (Cheng et al., 2003)

are presented in Table 4.

It is concluded: though there are some differences

between the two methods results, we can see that the

ratio between 2011-2012 FTR and the average FTR

(named as R) is almost most close to 1. Therefore,

subsequent calculation chooses the 2011-2012 freezing

indices as standards, which is reduced by the R’

calculated from the daily mean temperature. Compared

with 30 years statistic data, the computational results

should be accepted.

Table.3: FTR and R from the Air Daily Mean Temperature

Area Period Freezing

index(°C.days)

Thawing

index(°C.days) FTR R

Kai-yuan

2010-2011 -948.0 3684.5 0.2573 0.78

2011-2012 -1338.5 3866.5 0.3462 1.05

2012-2013 -1445.5 3780.0 0.3824 1.16

Shuang-

cheng

2010-2011 -1444.5 3485.0 0.4145 0.86

2011-2012 -1600.5 3560.0 0.4496 0.93

2012-2013 -2065.5 3543.0 0.5830 1.21

Table.4: FTR and R from the Air Monthly Average Temperature

Area Period

Maximum monthly

average

temperature

(°C)

Minimum monthly

average

temperature

(°C)

FTR

analytical

solution

R

Kai-

yuan

2010-2011 24.35 -14.98 0.4660 0.98

2011-2012 25.13 -15.63 0.4746 1.00

2012-2013 24.34 -15.35 0.4852 1.02

30 years

statistic data 23.80 -13.40 0.4058 0.85

Shuang-

cheng

2010-2011 23.89 -18.38 0.6623 0.99

2011-2012 24.68 -17.77 0.5969 0.89

2012-2013 24.08 -20.00 0.7470 1.12

30 years

statistic data 23.20 -18.10 0.6771 1.01

3 Correlations between Frost Penetration and

Freezing Indices, including Air and Subgrade

Surface

3.1 Correlation between Frost penetration and Air

Freezing Index

From a design point of view, the parameters of frost

penetration prediction formula developed should be as

less as possible, e.g., the correction coefficients and the

freezing index. The formula 0.5( )afz a T b is used,

where a fT , air freezing indices of 2012-2013, is

reduced as above, whose absolute value increases with

time gradually. The parameters a, b and test indicators

of fitting precision are presented in Table 5.





Table.5: Fitting Function Parameter and Test Indicator

Number Location Parameter

a b R-squared

Section 1

East shoulder 0.0811 0.210 0.998

Central line 0.0835 0.761 0.999

West shoulder 0.0832 0.061 0.998

Section 2

East shoulder 0.0844 0.282 0.998

Central line 0.0893 0.818 0.999

West shoulder 0.0871 0.154 0.998

Section 3

East shoulder 0.0796 1.032 0.999

Central line 0.0828 1.330 0.999

West shoulder 0.0868 0.834 0.999

Section 4

East shoulder 0.0971 1.027 0.998

Central line 0.0897 1.320 0.998

West shoulder 0.0996 0.861 0.996

a of different location in different areas varies within a

narrow range, between 0.08-0.10 basically, but b within

a rather large range. If the parameters are as less as

possible, supposing that b is a definite value in different

area, it would bring a large error.

Therefore, this paper selects a formula uniformly: 0.50.09( )afz T c . The parameters c and test

indicators are presented in Table 6. The frost

penetration depths and its fitting function with time are

presented in Figure 1(a) to (d).

Table.6: Fitting Function Parameter and Test Indicator


c R-squared

Section 1

East shoulder 0.4062 0.986

Central line 0.9205 0.993

West shoulder 0.2100 0.992

Section 2




Section 3




Section 4




The accuracy of this formula is relatively high, which

can describe the correlations of frost penetration and air

freezing index well. The parameters of two sections in

Kaiyuan are very close to each other. However, even the

two sections in Shuangcheng are adjacent; there is more

significant difference of the fitting function parameters.

The results may be due to the difference of embankment

height, subgrade fillings and statistical error, etc.




0 200 400 600 800 1000 1200 14000.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Air Freezing Index reduced/°C.daysF

rost

pen

etr

ati

on

/m

DEPTH1EO

DEPTH1EF

DEPTH1CO

DEPTH1CF

DEPTH1WO

DEPTH1WF

(a) Section 1

0 200 400 600 800 1000 1200 14000.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Air Freezing Index reduced/°C.days

Fro

st P

en

etr

ati

on

/m

DEPTH2EO

DEPTH2EF

DEPTH2CO

DEPTH2CF

DEPTH2WO

DEPTH2WF

(b) Section 2

0 200 400 600 800 1000 1200 1400 1600 18000.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5


Fro

st

Pen

etr

ati

on

/m

DEPTH3EO

DEPTH3EF

DEPTH3CO

DEPTH3CF

DEPTH3WO

DEPTH3WF

(c) Section 3

0 200 400 600 800 1000 1200 1400 1600 18000.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5


Fro

st

Pen

etr

ati

on

/m

DEPTH4EO

DEPTH4EF

DEPTH4CO

DEPTH4CF

DEPTH4WO

DEPTH4WF

(d) Section 4

Figure 1: The depths of frost penetration obtained from

semi empirical formula and fitting function verse air

freezing index reduced

DEPTH: the frost penetration depth; 1-4:section1-4; E:

east shoulder, C: central line, W: west shoulder; O: frost

penetration obtained from semi empirical formula; F:

frost penetration obtained from fitting function.

3.2 Correlation between Frost penetration and

Subgrade Surface Freezing Index-surface n factor

method

A formula 0.5( )afsz T is used to describe the

correlation of frost penetration and subgrade surface

freezing index (McKeown et al., 1988). afsT Computed

from the temperature semi-empirical formula. The

parameters and test indicators are presented in

Table7.

Table.7: Fitting Function Parameters and Test Indicators


R-squared

Section 1




Section 2




Section 3




Section 4








The parameters are rather close to each other, almost

around 0.09. The value is rather consistent with the

Canadian studies (McCormick, 1993), in which is

0.040 – 0.083 for grounds below cement pavements.

The n factor, that is the ratio between subgrade surface

freezing/thawing index and air freezing/thawing index

reduced, are presented as Table. 8

Table.8: n factor of sections

Number Location Factor

n+ n

- n

-/ n

+

Section 1

East shoulder 1.10 0.72 0.65

Central line 1.00 0.51 0.51

West shoulder 1.06 0.85 0.80

Section 2




Section 3




Section 4




n- : the freezing indices ratio between air and subgrade surface, n

+: the thawing indices ratio between air and

subgrade surface, n-/ n

+: the specific value of the two above ratios.

Several results can be summarized as follows:

1. The covering layers of different locations are

different, n- varies in a larger range than n

+. However, n

factors of different sections, including n+, n

- , n

-/ n

+,

show the similar rules. The orders of n+ from high to

low are: east shoulder, west shoulder, and central line.

But the orders of n- and n

-/ n

+ are: west shoulder, east

shoulder, and central line. n+, n

- , n

-/ n

+ could be

considered as the subgrade indicators of heat quantity,

cooling quantity and freezing capacity. The shoulders

are influenced by climate factors more serious than

central line, providing evidences for the frost

penetration depths differences at different locations of

one section.

2. n factors at the same locations of two sections in

Kaiyuan are very close. It is shown that the heat

quantity imported into and emitted from the subgrade is

very similar although the embankment heights and

fillings are little different. Two sections in Kaiyuan and

section 4 in shuangcheng have similar heights, n factors

are very similar too.

3. n factors of two sections in Shuangcheng are very

different. The n+ of section 3 are larger than section 4,

however, n- and n

-/ n

+ are smaller and the difference

between east and west shoulders is slightly bigger. The

results indicate that embankment which has a larger

height is conductive to thermal storage and frost

penetration depth decrease, but at the same time, may

make the transverse thermal difference more evident.

We can conclude that n factors are mainly related with

the embankment height and section locations in the case

of the same subgrade route direction and covering layer.

4. Predicting Frost penetration based on Empirical

Method

Based on the above, this paper attempts to put forward

the empirical frost penetration prediction formula with

correction coefficients for HSR subgrade in seasonally

frozen regions.

4.1 Empirical Formula according to Atmosphere

Freezing Index

The parameter a can be determined as 0.10

conservatively. Then considering the main influence

factors, the empirical formula for design frost

penetration is defined as:

0.5

0.10 f r dz T k k b (3)

Where kr: the correction coefficient of inherent

conditions, such as subgrade height and filling. In

general, the higher the height is, the larger the value will

be.

kd: the influence coefficient of route direction.

Generally, the sunny slope of east-west (EW) direction




has the maximal value, and that of south-north (SN)

direction keeps the minimal value. For the shady slope,

the above rules are on the contrary (Li and Chen, 1992;

Liu, et al., 2013).

fT is the air freezing index on the basis of

meteorological data during a long-term, that is, at least

30 years.

b is a comprehensive coefficient of regional influence

factors which include Air temperature, topography,

geothermic, soil property and water, etc. The specific

areas, that the main influence factors are similar, could

be classified according to the criteria. Regional b series

could be obtained through statistical analysis of the field

monitoring data using the method in this paper.

According to the above data, b could be selected as 0.80

for Shuangcheng, 0.10 for Kaiyuan. In the case of

relative fixed subgrade conditions, b is mainly related

with regional climate.

4.2 Empirical Formula according to Subgrade

Surface Freezing Index-n Factor Method

The parameter is selected as 0.10 conservatively.

Then considering the main influence factors, the

empirical formula for maximum frost penetration is

defined as:

0.5

0.10 r d fz r r n T (4)

Where n: the freezing indices ratio between atmosphere

and subgrade surface.

fT is as same as above.

The variation trends of coefficients are contrary and

value range selected may be different.

rr: the correction coefficient of inherent conditions, such

as subgrade height and filling. In general, the higher the

height is, the smaller the value will be.

rd: the correction coefficient of route direction.

Generally, the sunny slope of EW direction has the

minimal value, and that of SN direction keeps the

maximal value. For the shady slope, the above rules are

on the contrary (Li and Chen, 1992; Liu, et al., 2013).

For the frost penetration of a certain subgrade location, z

can be corrected by the coefficients ks and kt. ks contain

influences of section form, gradient, solar exposure, etc.

In general, the value of shady slope is bigger. kt

represents the factor of distance from shoulder, the

larger the distance is, the smaller the value is.

5. Conclusions:

The frost penetration is an important basis of anti-

freezing measures in frozen regions. Correlations

between the frost penetration and air, subgrade surface

freezing indices are studied using empirical method, and

the computation steps for the prediction model are

detailed.

During the analysis process, some conclusions can be

obtained:

1. The FTR reflects the freezing capacity of a certain

cycle in a specific area to some extent. n factors,

including n+, n

- , n

-/ n

+ , could clearly show the rules of

cooling quantity, heat quantity and freezing capacity at

subgrade different locations of different areas. Main

influence factors of frost penetration can be discussed

through comparisons of FTR and n factors.

2. The main influence factors of frost penetration are:

air temperature, subgrade height and section transverse

location, etc. The larger the local freezing index and the

smaller the height are, the deeper the frost penetration

will be. The frost penetration of different section

locations are different, central line is smallest, then the

east shoulder, the west shoulder is largest.

3. The frost penetration is strongly related with air and

subgrade surface freezing indices. So the uniform

empirical formula could be attained to describe the

correlations between them. The influence factors can be

considered through the parameters and correction

coefficients of the formula, which can be determined

using the field monitoring data.

In practice, the modified empirical formula can be

useful for estimating the frost penetration at different

sites along the railway that have similar geotechnical

properties. What is more, the process and conclusions

about the frost penetration in this paper will provide

some references for the design and construction of

similar engineering, the maintenance and stability

analysis of HSR in seasonally frozen regions.


This work was supported by the

National Key Technology R&D Program, the Ministry

of Science and Technology of the People's Republic of

China [Grant No: 2012BAG05B01]; and the National

Basic Research Program, the Ministry of Science and

Technology of the People's Republic of China [Grant

No: 2012CB723301].

2. Reference:

[1] ANDERSLAND O B, LADANYI B, 2011. Frozen

ground engineering. 2nd ed. Translated by Yang R,

Li Y., Beijing: China Architecture and Building

Press.

[2] Baïz S, Tighe S L, Haas CT, Mills B, Perchanok,

M, 2008. Development of frost and thaw depth

predictors for decision making about variable load

restrictions. Transportation Research Record:

Journal of the Transportation Research Board,

2053(1): 1-8.





[3] Bing WS, Zhou J, Wang XF, 1995. Calculation

methods research of road frost depth, Journal of

Harbin university of Architecture and Engineering,

28: 11-16.

[4] Cheng GD, Hao J, Wang KL, Wu QB, 2003.

Freezing and thawing index on the embankment in

permafrost regions. Journal of Glaciology and

Geocryology, 25: 603-607.

[5] Fischer C. J., Zubeck H, 2013. Comparison of

Theoretical Freezing Depths of Permeable and

Conventional Pavements. In ISCORD 2013@

sPlanning for Sustainable Cold Regions, 447-458.

ASCE.

[6] Khalili A., Rahimi H., Shariatmadari Z. A, 2007.

Validation of air freezing index (AFI), for

determination of frost penetration depth in typical

arid and semi-arid zones of Iran. DESERT

(BIABAN), 12:23-31.

[7] Li AG, Chen QH, 1992. Determination of frost

penetration in canal design, Water Resource and

Water Engineering, 3: 1-1.

[8] Liu ZY, Chen JB, Jin L, Zhang YJ, Lei C, 2013.

Roadbed temperature study based on earth-

atmosphere coupled system in permafrost regions

of the Qinghai-Tibet Plateau. Cold Regions Science

and Technology, 86: 167-176.

[9] Martel CJ, 1988. Predicting freezing design depth

of sludge-freezing beds. Journal of cold regions

engineering, 2: 145-156.

[10] McCormick G, 1990. Soil temperatures and

freezing indices at depth. Canadian Geotechnical

Journal, 27(6): 749-751.

[11] McCormick G, 1993. Frost penetration beneath

cleared pavements. Frost in Geotechnical

Engineering. Balkema, Rotterdam, 117-126.

[12] McKeown S., Clark J. I., Matheson D, 1988. Frost

penetration and thermal regime in dry gravel.

Journal of cold regions engineering, 2(3): 111-123.

[13] Shin E C, Park J J, Lee J S, 2013. An Analysis of

Frost Penetration Depth with Field Temperature

Data of Paved Road in Korea. In ISCORD 2013@

sPlanning for Sustainable Cold Regions, 367-378.

ASCE.

[14] Smith G M, Rager R E, 2002. Protective layer

design in landfill covers based on frost penetration.

Journal of geotechnical and geoenvironmental

engineering, 128(9): 794-799.

[15] Xu J, Niu FJ, Li AM, lin ZJ, 2010. Analysis of the

prevention effect of thermal-insulation method on

frost heave of railway subgrade in seasonal frozen

regions. Journal of the China Railway Society, 32:

124-31.

[16] Xu J, Niu FJ, Niu YH, lin ZJ, Xu ZY, 2011a. Study

on the temperature field of insulated subgrade with

frost resistant berm on seasonal frozen region,

Journal of the China Railway Society, 33: 84-90.

[17] Xu J, Niu FJ, Niu YH, Liu H, 2011b. Analysis on

the effect of replacing-soil method on inhibiting

frost heave of railway subgrade in seasonal frozen

soil region, China Railway Science, 32: 1-7.

[18] Zhang YZ, Du YL, Sun BC, 2014. Temperature

distribution research of high-speed railway roadbed

in seasonally frozen regions, Chinese Journal of

Rock Mechanics and Engineering, 33: 1286-1296.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1127-1134


Comprehensive Evaluation of the Development Quality of New-

Type of Urbanization of Jiangxi Province Based on Ecological Views

WANG YONG XIANG East China Jiaotong University, School of Civil Engineering and Architecture, Nanchang, Jiangxi Province, CHINA

Email:[email protected]

Abstract: New-type of urbanization is more than a shift of rural population to urban population. More importantly,

it sets up a direction to resource conservation, environment protection, science development and social harmony.

However, there are few efforts to evaluate the quality of new-type of urbanization. Based on previous researches,

this paper constructs a comprehensive evaluation system for evaluating the development quality of new-type of

urbanization and conducts a comprehensive evaluation through statistical methods such as cluster analysis and factor

analysis. Results show that following the scientific outlook on development and establishing the comprehensive

evaluation system with ecological ideas are significant to improve science and culture, enhance the quality of

ecological urbanization and protect our homeland. Under the system, more efforts will be made to beauty and enrich

Jiangxi.

Keywords: Ecology, Jiangxi Province, New-type of urbanization, Comprehensive evaluation.

1. Introduction:

The 18th

National Congress of the Communist Party of

China proposed to accelerate the pace of building new-

type of urbanization. Urbanization in the future will

become an important carrier of building a moderately

prosperous society as well as the driven force of

leveraging domestic demand. The rapid development of

urbanization is becoming a powerful engine for

economic growth and social development. "A resource-

conserving and environment-friendly society" is

scientific, economic, social and ecological civilization.

New-type of urbanization is an integration of urban and

rural areas, an interaction between production and city

and being economical, intensive, livable featuring

harmonious development. It aims at coordinated

development and mutual progress among large, middle-

sized cities and small cities and towns [1-9]

.

As a revolutionary base, big agricultural province and

tourism province, Jiangxi’s urbanization develops very

slowly. According to statistics reported in 2013 in

Jiangxi Province, by the end of 2012, the level of

urbanization in Jiangxi Province was 29.09% (see

Figure 1) [10]

; current and the next 20 years will be a

crucial period of rapid development of urbanization in

Jiangxi; Jiangxi’s urbanization will become the main

theme of economic and social development; new

urbanization has become a national development

strategy in central China.

However, there are still problems presenting in the

course of urbanization of Jiangxi. For example, the scale

of the city is small. The spatial distribution is far from

being reasonable. Some cities and towns fail to tailor to

their local situation. Urban planning is not reasonable.

Enclosure brings about a waste of land resources.

Infrastructures such as transportation, water and utility,

energy and communication are backward. Environment

deterioration has brought much pressure on the

development of Jiangxi. Therefore, it is necessary to

make the economic construction a center and reach a

balance between economic development, social

harmony and ecological protection [11]

.

Over the past 20 years, domestic scholars have achieved

fruitful results on the study of urbanization. But for a

long time, the proportion of urban population to total

population in the country is always used to measure the

urbanization level. Such method is the simplest one for

single indicator method. However, there are also many

disadvantages and limitations. In recent years, more and

more scholars began to focus on problems brought by

urban development and evaluate the quality of

urbanization from a multi-dimensional perspective

including population, economy, resources, environment

even the whole society rather than economic alone. Bai

Xianchun et al (2004) established an evaluation system,

determined the target value according to certain

standards and made a thorough study on coordination

development of urbanization based on econometric

model [12]

.

Research group of Fujian on City Investment (2005)

constructed the urbanization quality evaluation system

mainly from the core carrier and regional carrier of

1128 Comprehensive Evaluation of the Development Quality of New-Type of Urbanization

of Jiangxi Province Based on Ecological Views



urbanization. It used the indicator method to calculate

quality indicators as a quantitative study and yielded

fruitful results [13].

Liu Yanjun et al (2006) also

emphasized the importance to put in place an indicator

system and did in-depth research on the level of

urbanization. He analyzed the advantages and

disadvantages of urbanization, commented on its

development level and provided suggestion to future

development [14]

. Yuan Xiaoling et al (2008) and He

Wenju et al (2009) proposed the evaluation system for

urbanization quality from the perspective of material

civilization, spiritual civilization and ecological

civilization according to relevant regional development [11]-[15]

.

Under the instruction of the scientific outlook on

development and the vision of establish a “resource-

conserving and environmental-friendly society”, this

paper bases itself on previous researches and conducts a

comprehensive evaluation of urbanization quality of

Jiangxi while taking the real situation of Jiangxi into

consideration. It aims at promoting the development of

Jiangxi Province.

0

10

20

30

40

50

Nan

chan

gJi

ngdezh

enPin

gxian

gJi

ujiang

Xinyu

Yingta

nG

anzh

ou

Ji'a

nYic

hun

Fuzhou

Shangra

o

The urbanization

rate(%)

Fig1: Urbanization level of Jiangxi Province

Data from: Statistical Yearbook of Jiangxi Province2013.①

2. Constructing comprehensive evaluation system of

the development quality of new-type of urbanization

based on ecological views:

Many factors needed to be considered in constructing

the comprehensive evaluation system of the

development quality of new-type of urbanization. The

urbanization levels vary from each other under different

classification methods. Based on the systematic,

operable and comparable principles, 11 prefecture-level

cities of Jiangxi Province were selected as the research

objects. Material civilization, spiritual civilization and

ecological civilization were taken as three evaluation

indicators. The evaluation system is shown in Table 1 [15]

.

Statistical Yearbook of Jiangxi Province 2013 was the

data source of this paper. After standardization, the gap

of variables in scale could be eliminated so that data

were comparable. Standardization was conducted by

SPSS and the results were saved in a new variable with

the initial capital of z [16]

, there is:

iZ-SCORE= /x x SD

In the expression, ix is the original observation value;

x

is the average value and SD is the standard

variance. The average value of new variables after

standardization is 0. The standard variance is 1. If the

variable is below 0, it suggests that this variable is

below the average and vice versa.

Table 1: Comprehensive Evaluation Index System of the New-type of Urbanization of Jiangxi Province

Note: (1) Statistical Yearbook of Jiangxi Province 2013 serves as the database for factor analysis; (2) Engel

coefficient, proportion of spending on technology in total fiscal expenditure and Urban employment in total

employment were calculated based on original indicators. (3) The statistical software is SPSS17.0.

Index Unit

material

civilization

fiscal revenue 100 million Yuan

per capita GDP Yuan

Proportion of the tertiary industry in total GDP %

total retail sales of consumer goods 100 million Yuan

total export-import volume 10,000 Dollars

1129 WANG YONG XIANG



property investment 10,000 Yuan

Per capita disposable income of urban residents Yuan

grain output 10,000 Tons

Engel’s coefficients of urban residents %

spiritual

civilization

Proportion of spending on technology in total fiscal

expenditure %

Urban employment in total employment %

Telecom service income 100 million Yuan

Private Car Ownership Number of cars

fulltime teachers in institutions of higher learning Number of teachers

Number of beds Number of beds

ecological

civilization

Per capita public green areas ㎡

Public facilities coverage to total construction land %

delivering quantity of household reguse 10,000 tons

Sanitation vehicles Number of cars

Coverage of city gas %

Treatment rate of domestic sewage %

Household Water Use 100 million m³

3. Comprehensive evaluation method of the

development quality of new-type of urbanization:

Existing evaluation methods include principal

component analysis method, factor analysis, AHP,

fuzzy comprehensive evaluation method, data

envelopment analysis, artificial neutral network

evaluation method, grey correlation cluster and relative

gap method [17]

. This paper mainly uses factor analysis

to do the evaluation.

Factor analysis was derived from principal component

analysis which was proposed by Hotelling in 1933.

Dimensionally reduction was the main thought of

principal component analysis in which many relevant

indicators were rearranged into a new group of

irrelevant indicators [18]

.

Factor analysis is a diversified statistical method

starting from the reliance relationship of correlation

matrix of indicators. Complicated variables were turned

into irrelevant factors. The purpose is to screen several

significant variables from the mass. The basic thinking

is to group the variables and compare the correlation of

variables within the group. Each group of variables

represents a basic structure, namely, the common factor [19]

.

3.1. Cluster analysis:

Cluster analysis aims at identifying the system

characteristics according to the observation of sample

indicators by grouping indicators or samples. The

essence lies in establishing a classifying method which

enables sample data or variables to classify themselves

automatically according to their correlation without

much pre-knowledge. The object of R-type cluster

analysis is the indicator. As social and economic

indicators are much related to each other, it is suggested

to apply ordinal indicators to cluster analysis [16]

.

3.2. Factor analysis:

It is necessary to give weight to indicators by Delphi

method. Besides the effect of multicollinearity on the

analysis, the amount of work and the subjectivity of

scoring are not easy to be addressed. Factor analysis

focuses on how to select several variables from the mass

with the least loss of information and how to produce

explainable variables [19]

.

3.3. Main steps:

Main steps of this research are described as follows:

(1) Apply indicators to R-type cluster analysis

(2) Standardize the data samples

(3) Calculate correlation matrix R of samples

(4) Calculate the characteristic root and characteristic

vector of correlation matrix R

(5) Determine the number of main factors according to

the accumulative contribution rate. Usually main factors

with an accumulative contribution rate of 85-95% are

selected. Explain those main factors.

(6) Calculate factor load matrix A

(7) Decide the factor model

(8) Analyze the system based on calculation results

(9) Carry out in-depth analysis based on the scores

4. Empirical analysis of the comprehensive

evaluation of development quality of new-type of

urbanization of Jiangxi Province:

Based on Statistical Yearbook of Jiangxi Province 2013,

this paper uses R-type cluster analysis in SPSS to





cluster variables in which indicators of common

characteristics are put together and representative

indicator in each group is selected to reduce the number

of variables [19]

. Cluster analysis can reflect the

relationship between samples through the vertical bar

graph automatically produced by SPSS.

Subject 9 indicators of material civilization, the first

classification, in the comprehensive evaluation system

for development quality of new-type of urbanization of

Jiangxi. The first classification can have three sub-

categories, namely, Engel coefficient, proportion of the

tertiary industry in total GDP and the rest seven.

Correlation coefficient matrix of 7 indicators is shown

in Table 2.

Table 2: Correlation Coefficient Matrix

Zscore(fis

cal

revenue)

Zscore

(per

capita

GDP)

Zscore

(total retail

sales of

consumer

goods)

Zscore

(total

export-

import

volume)

Zscore

(property

investment

)

Zscore (Per

capita

disposable

income of

urban

residents)

Zscore

(grain

output)

Zscore(fiscal

revenue)

Pearson

Correlation 1 .048 .961

** .768

** .911

** .294 .463

Sig(2-tailed) .889 .000 .006 .000 .380 .152

N 11 11 11 11 11 11 11

Zscore (per

capita GDP)

Pearson

Correlation .048 1 .121 .370 .155 .845

** -.651

*

Sig(2-tailed) .889 .722 .263 .649 .001 .030

N 11 11 11 11 11 11 11

Zscore (total

retail sales of

consumer

goods)

Pearson

Correlation .961

** .121 1 .784

** .977

** .386 .353

Sig(2-tailed) .000 .722 .004 .000 .241 .287

N 11 11 11 11 11 11 11

Zscore (total

export-

import

volume)

Pearson

Correlation .768

** .370 .784

** 1 .772

** .478 .031

Sig(2-tailed) .006 .263 .004 .005 .137 .928

N 11 11 11 11 11 11 11

Zscore

(property

investment)

Pearson

Correlation .911

** .155 .977

** .772

** 1 .371 .311

Sig(2-tailed) .000 .649 .000 .005 .261 .352

N 11 11 11 11 11 11 11

Zscore (Per

capita

disposable

income of

urban

residents)

Pearson

Correlation .294 .845

** .386 .478 .371 1 -.482

Sig(2-tailed) .380 .001 .241 .137 .261 .133

N 11 11 11 11 11 11 11

Zscore (grain

output)

Pearson

Correlation .463 -.651

* .353 .031 .311 -.482 1

Sig(2-tailed) .152 .030 .287 .928 .352 .133

N 11 11 11 11 11 11 11

**. Correlation is significant at the 0.01 level (2 - tailed).

*. Correlation is significant at the 0.05level (2 - tailed).

Select a representative indicator in the third sub-category and calculate the average number of its correlation

coefficient (the square of correlation coefficient) with other indicators. There is:




2 2

1

/ 1m

ij

i

R r m

Per capita GDP is the most representative indicator and thus is chosen. Therefore, indicators of population index are

simplified as Engel coefficient, proportion of the tertiary industry in total GDP and per capita GDP. Then apply the

same way to first-class indicators and get the most representative indicator in each category [13]

. 22 original

indicators are simplified to be 9 comprehensive indicators, as is shown in Table 3.

Table 3: Simplified Index System

Category Index

economic development

per capita GDP

Proportion of the tertiary industry in total GDP

Engel coefficient or urban residents

living quality of urban residents

Proportion of spending on technology in total fiscal expenditure

Urban employment in total employment

Private Car Ownership

Environmental carrying capacity

Per capita public green areas

delivering quantity of household reguse

Treatment rate of domestic sewage

This paper uses factor analysis to select several irrelevant indicators to reduce the number of variables while trying

its best to ensure the completeness of the information. The number of factor and factor rotation are shown in Table4.

Table 4: Total Variance

Indicator

Original characteristics value Load the extraction of sum of

square

Load the rotation of sum of

square

Total Varianc

e %

Accumula

tion % Total

Variance

%

Accumula

tion % Total

Variance

%

Accumul

ation %

1 4.568 50.759 50.759 4.568 50.759 50.759 3.757 41.749 41.749

2 2.519 27.994 78.753 2.519 27.994 78.753 3.043 33.813 75.561

3 1.006 11.182 89.935 1.006 11.182 89.935 1.294 14.374 89.935

4 .446 4.951 94.886

5 .347 3.853 98.739

6 .065 .717 99.457

7 .043 .481 99.937

8 .004 .042 99.979

9 .002 .021 100.000

Extraction method: Principal Component analysis.

From Fig.2, we can see the characteristics value of three indicators is more than 1. Then only analyze these three.

Fig.2: Scree plot

Afraction

Eig

enva

lue





Among contribution rate factor accumulative variance in Table 4, if three common factors extracted can describe

89.935 % of total variance of original variables, it indicates that 85% information is covered and these factors can

reflect most of the information. Then rotate the factor load matrix by varimax.

Table 5: Rotating Component Matrixa

Evaluation index ingredient

1 2 3

Zscore(per capita GDP) .894 .297 .216

Zscore (proportion of the tertiary industry in

total GDP) .139 .854 -.250

Zscore (Engel coefficient of urban residents) .649 .365 .407

Zscore (Proportion of spending on technology in

total fiscal expenditure) .349 .883 .185

Zscore (Urban employment in total employment) .894 .319 .202

Zscore (Private vehicle ownership) -.829 .485 -.158

Zscore (per capita green areas) .940 .121 -.120

Zscore (delivering quantity of household reguse) .033 .974 .009

Zscore (Treatment rate of domestic sewage) .155 -.105 .951

Extraction method: principal component analysis.

Rotation method: Orthogonal rotation method with

Kaiser standardization.

a. Rotation became convergence after 4 iterations.

For the first factor, per capita GDP, urban employment

in total employment and per capita green areas have

larger load. For the second factor, proportion of the

tertiary industry in total GDP, proportion of spending on

technology in total fiscal expenditure and delivering

quantity of household reguse have larger load. For the

third factor, Engel coefficient and treatment rate of

domestic sewage have larger load. Then, calculate the

coefficient of factor score function by regression

method and calculate the comprehensive score of three

factors.

COMPUTE score= 1 2 30.50759 * F + 0.27994 * F + 0.11182 * F

5. Conclusion:

Subject indicator value to linear system of equation,

rank in descending order and get the comprehensive

evaluation as Table 6 [10]

. The population urbanization

rate is also put into the Table for comparison.

Table 6: Comprehensive Evaluation of the development quality of new-type of urbanization of Jiangxi Province

City F1 F2 F3 F Ranking

Population

urbanization

rate

Rank by

urbanization

level

Nanchang 0.56268 2.66891 0.65429 1.11 1 46.13 1

Xinyu 1.82702 -0.17779 0.12013 0.89 2 34.83 3

Jingdezhen 1.29031 -0.44937 -1.65535 0.34 3 38.96 2

Pingxiang 0.50661 -0.33438 0.49066 0.22 4 31.13 4

Yingtan 0.4844 -1.02747 0.70866 0.04 5 28.2 5

Jiujiang -0.36713 0.10297 0.7656 -0.07 6 27.27 7

Ji’an -0.67359 -0.41021 0.53957 -0.4 7 22.45 9

Wuzhou -0.47306 -0.63646 -0.76393 -0.5 8 23.67 8

Yichun -0.86399 -0.53375 0.71764 -0.51 9 27.74 6

Shangrao -1.12524 0.01866 0.40607 -0.52 10 19.22 11

Ganzhou -1.168 0.77889 -1.98333 -0.60 11 20.41 10

Data from: Statistical Yearbook of Jiangxi

Province2013.①

From the result it is clear that:




(1) In terms of comprehensive factor of new-type of

urbanization, Nanchang ranks the first. The rank of

urbanization quality is in line with the rank of

population urbanization rate. This has something to do

with its geographical advantage and policy

environment. But indicators such as per capita green

areas and treatment rate of domestic sewage still need

improvement.

(2)The rank of comprehensive factor of new-type of

urbanization of Xinyu, Jingdezhen, Pingxiang, Yingtan

and Jiujiang is in line with the rank of population

urbanization rate. But problems for each city in terms of

urbanization quality are different. Xinyu, Jingdezhen

and Pingxiang have some advantages in urbanization

quality. Efforts have been made on urban ecology in

recent years, making them green and civilized cities.

Urbanization quality in Jingdezhen and Pingxiang are

not satisfying and needs more efforts. Ji’an, Wuzhou,

Yichun, Shangrao and Ganzhou don’t stand out in

urbanization quality. Their indicators are relatively

weak and thus a comprehensive improvement is needed.

(3) The rank of comprehensive factor of Yichun is the

9th

, lagging behind Jiujiang (the 6th

). It only outperforms

Jiujiang in terms of population urbanization rate. This

indicates that it is not scientific to evaluate the

urbanization level simply by population urbanization

rate.

(4)If the urbanization level is evaluated by population

urbanization rate, Jingdezhen, Pingxiang, Yingtan and

Jiujiang are doing pretty well. But if evaluated by

urbanization quality, their ranks are much behind.

Jiangdezhen, Pingxiang and Yingtan are industrial cities

while Jiujiang is a tourist city. They have some

prevalences in urbanization development and are

suggested to take the opportunity of establishing a

“resource-conserving and environment-friendly” city to

upgrade traditional industries and put in place an

integrated and interactive industry clusters so as to

realize the effective use of resources and a positive

cycle of ecology. The urbanization quality will be

improved as a result.

(5)The score of urbanization quality of Nanchang and

Xinyu is higher than that of population urbanization

rate. These two cities have done a good job in

urbanization process. But their resources and

environment bearing capacity are not adapted to the

new-type of urbanization level.

Therefore, following the scientific outlook on

development and establishing the comprehensive

evaluation system with ecological ideas are significant

to improve science and culture, enhance the quality of

ecological urbanization and protect our common

homeland. Under the system, more efforts will be made

to beauty and enrich Jiangxi.

Notes: ① Based on data from the Statistical Yearbook

of Jiangxi Province 2013.

2. Reference:

[1] Luo Jianrong, Study on the Scientific Outlook on

development and Urbanization, Haitian Press, PP.

5-31, 2005.

[2] United Nations Human Habitat, The State of the

World's Cities Report 2001, Un-habitat, 2002.

[3] Urban Indicators Guidelines, United Nations

Human Settlements Programmer, Augest, 2004.

[4] Bloom DE, Canning D and Fink G, Urbanization

and the Wealth of Nation. Science, Volume 319.

Issue 1., PP772-775., 2008.

[5] USEAP, Pollution Prevention 1997-A National

Progerss Rpeort, EAP-742-R-97-000, 1997.

[6] USEAP, Innovation at the Environment protection

Agency, Decade of Progress, EPA 100-R-00-20,

2000.

[7] Segnstam L., Indicators of Environment and

Sustainable Development Theories and Pratical

Experience, Environmental Economics Series,

Paper No.98, The World Bank.2002.

[8] Williams O.O.David W., Natural resource and

environmental poli-cy trade-offs:a CGE analysis of

the regional impact of the Wetland Re-serve

Program. Land Use Policy, Volume19, PP231-241.,

2002.

[9] Young W.J., Lam D.C.L. and Ressel V., I,

Development of an environ-mental flows decision

support system. Environmental Modelling & Soft-

ware, Volume15., PP257-265., 2000.

[10] China Statistical Yearbook Newsroom, Statistical

Yearbook of Jiangxi Province2013, China

Statistical Publishing House, 2013.

[11] Yuan Xiaoling, Wang Xiao, He Weiwei and Chen

Yue, Comprehensive Evaluation of the quality of

Urbanization, Urban Development Research,

Volume2., PP38-45., 2008.

[12] Bai Xianchun, Ling Kang and Guo Cunzhi,

Synthesis Evaluation Of The Quality Of City

Development, China Population Resources and

Environment, Volume6., PP91-95., 2008.

[13] National Research Group, Fujian Team,

Constructing System of Quality Evaluation of

Urbanization of China and the Experimental

Analysis, Statistical Research, Volume7., PP15-19.,

2005.

[14] Liu Yanjun, Li Chenggu and Sun Di,Research On

Urbanization Integrated Level Of Urban Regional

Centers Evaluation--A Case Study On 15 Vice-

Provincial Cities Of China, Economic Geography,

Volume3., PP225-229., 2006.

[15] He Wenju, Deng Bosheng and Yang Zhimei,

Urbanization Quality from the two-oriented society





perspective-a Case Study of Hunan Province,

Financial Theory and Practive, Volume30., PP118-

121., 2009.

[16] Yang Weizhong and Zhang Tian, SPSS Statistical

Analysis and Industry Applications (2nd

Edition),

Tsinghua University Press, PP75-190., 2013.

[17] Wang Hui and Chen Li, Multi-index

Comprehensive Evaluation Model and the Selection

of Weight Coefficient, Journal of Guangdong

College of Pharmacy, Volume5., PP33-36., 2007.

[18] Bai Huiqiang, The Application of Principal

Component Analysis in SPSS- An Example of the

Herbosa of Wenyu River, Sci-Tech Information

Development & Economy, Volume9., PP173-177.,

2009.

[19] Du Qiang and Jia Liyan, Mastering SPSS Statistical

Analysis, Posts and Telecom Press, PP248-305.,

2011.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1135-1140


Simulation study of dense coherent tower sintering flue gas

desulfurization system

QIAN JIA1, DONGHUI ZHANG

1,2, CUNYI SONG

1, ZHENSONG TONG

1 AND BAORUI LIANG

1

1School of Civil and Environmental Engineering, University of Science and Technology Beijing, 100083,

Beijing, China 2Beijing Guodian Longyuan Environmental Engineering Co., Ltd, Beijing, 100039


Abstract: Using Fluent software, dense coherent tower simulation flue gas rectification system and delta wing plate

on the flue gas flow field and turbulence, circulating ash trajectory and the system pressure loss. The results showed

that: in the simulated conditions of flue gas rectification system optimum deflector arrangement of the C delta wing

plate can significantly increase the tower near the region of turbulent kinetic energy, multilayer delta wing panels

can increase the high-turbulence intensity region and mitigationits decay rate, the conditions to install the multi-

layer delta-wing boards, and more desirable. Dense coherent tower flue full access to all the cloth and the gas-solid

most of the particles involved in the inner loop, a small part into the dust collector, the system pressure drop of about

300 pa.

Keywords: Coherent tower; sintering flue gas; desulfurization; numerical simulation; flow field; turbulence;

particle trajectory; pressure loss.

1. Introduction:

Sintering flue gas with flow and low SO2 concentration,

high temperature and extremely unstable, leading to a

mature applied to other industrial flue gas

desulfurization technology cannot be directly applied to

the sintering flue gas desulfurization [1]

. The most

widely used industrial flue gas desulfurization limestone

gypsum wet FGD systems due to system complexity,

covers an area of great features, not suitable for steel

mills did not reserve enough space [2]

. Sintering flue gas

SO2 concentration in the lower features semidry

desulfurization technology feasibility. According to the

different characteristics of circulating fluidized semi-dry

FGD reactor can be divided into gas-solid flow

upstream bed (riser) and gas-solid and shed row bed [3]

.

The former has a high gas-solid slip velocity, the better

"three and one against" performance in industrial flue

gas desulfurization project in certain applications, such

as circulating fluidized bed (CFB Circulating Fluidized

Bed), due tosintering flue gas flow fluctuations to

enhance the flue gas velocity in the pipe is very

unstable, the collapse of the bed "phenomenon occurs

when the lower than the superficial gas velocity, which

severely restricted its stability [4]

. Gas-solid line and

shed bed plug-flow characteristics of its cis gravity

level, the minimum gas velocity, pressure loss, stable

running, on behalf of the technology is: rotating spray

drying method (SDA, Spray Dryer Absorber) method,

the drawback the flue gas turbulence intensity is low,

gas-solid slip velocity is low, and is not conducive to

the reaction [5]

.

Dense coherent tower flue gas desulfurization

technology (DFA-of FGD, Dense Flow Absorber Flue

Gas Desulphurization) is on the basis of gas-solid and

shed row bed to optimize the structure of the tower

body, the flue gas rectification system and chain

blender, adjust the flue gas flow field, and strengthen

the "three pass against" successfully applied to the

sintering flue gas desulfurization project [6]

. In this

paper, Fluent software simulation of the flue gas

rectification system and chain agitator on the gas flow

in the gray cycle track and the reaction zone turbulence

intensity, explore, and to seek a reasonable working

conditions, and provide a reference for the actual

project.

Fluent is a special simulation and numerical analysis for

flow and heat in the complex geometric region of the

exchange of fluent simulation software, using C

language development, the support of UNIX and

Windows and other platforms. Fluent provides a flexible

grid characteristics, can use the structured grid and

unstructured grid is convenient for various kinds of

complex regional division. For the two-dimensional

problem, can generate triangular element meshes and

quadrilateral element mesh; for the three-dimensional

problem, provide grid unit comprises a tetrahedron,

hexahedron, pyramid, wedge and hybrid grids. It is

widely applied in the field of aviation, automobile,

1136 Simulation study of dense coherent tower sintering flue gas desulfurization system



machinery, electronics, power, water conservancy,

building design, material processing, processing

equipment, environmental protection, and shows good

applicability.

1 Research Contents and Methods

Dense coherent tower set flue gas rectifier system, the

upper part of the desulfurization tower and the tower

desulfurization reaction zone to install a chain blender.

A steel desulfurization tower geometry parameters and

run the basic parameters, such as shown in Table 1; In

this paper, flue gas rectifier system and the role of the

chain blender principle, to seek a reasonable layout and

operation mode, the study is divided into two aspects:

First, study three different types of flue gas rectification

system for gas flow in the study of the influence of the

chain agitator on the flue gas turbulence, and gas flow

in the last simulation and optimization of dense coherent

tower, circulating ash trajectory summary of the secret

coherent design and operation methods of the main

components of the tower law.

Table 1: mil coherent Taki this parameter

Table 1 Basic parameters of DFA

Parameters Value

Section size of the flue / (m) 3.2×3.2

Tower body size / (m) 8×5×18

Average flue gas flow rate /

(million Nm3 • h-1) 45

Entry SO2 concentration /

(mg • m-3) 1000～2000

Inlet flue gas temperature /

(℃) 100～160

Circulating ash concentration

/ (g • m-3) 400～800

Circulating ash moisture

content / (%) 3～5

Calcium sulfur ratio 0.9～1.6

Circulation ratio 20～100

1.1 flue gas rectifier systems research methods

Dense coherent tower flue gas rectification system of

gas flow in the main factors is the size, number and

location of the deflector. By adjustment of the spoiler

arrangement, with a smaller pressure loss tower

effective reaction zone flue gas flow evenly, to prevent

the deposition of circulating ash, and to prevent

"Karman vortex resonance in the following department

effect [7]

. A, B, C three deflector arrangement, for

example, the flue gas rectifier system simulation

studies, the relevant dimensions and the coordinates

shown in Figure 1, distribution of effect characterization

of flue gas through two methods: First, the simulated

flue gas flow lines, with intuitive qualitative flue gas

distribution uniformity; two quantitative comparison of

different height plane at different positions of the flue

gas flow rate in the Y plane and Z be 0 X different value

compared its speed, the difference between the smaller

flue gas distribution is more uniform.

Fig1 Geometric model of flue gas rectifier system

1.2 delta wing plate of research methods

The delta wing plate is one hassel within the

components installed in the effective reaction zone close

coherent tower desulfurization system, aims to raise the

tower two-phase flow turbulence intensity, the effective

reaction zone of gas-solid slip velocity to achieve the

purpose of improving the efficiency of the reaction, the

system does not produce too high pressure losses. Delta

wing plate analog array arranged in spiral blades, each

blade is at right angles to the edge 1m isosceles triangle,

its side of the axis of a right angle, tilt angle 60 °, each 8

leaves to form a helical blade, a layer of non-drivers

within the component layout from the dense tower at the

top of 5.5m, the second layer of delta-wing boards in the

first layer of the downstream 2 m at the third layer of

1137 QIAN JIA, DONGHUI ZHANG, CUNYI SONG, ZHENSONG TONG AND BAORUI LIANG



the delta wing board layout in the first layer of the delta

wing plate downstream of 4m as shown in Figure 2.

This article first analog to install only the first layer of

the delta wing plate turbulent kinetic energy distribution

in the different sections of the dense coherent tower to

study the chain stirrer speed systems Y plane average

impact of the turbulence intensity, and then simulate the

multi-layer delta wing plate turbulent kinetic energy

indense coherent the distribution of the different

sections of the tower, of layers on a different system, the

Y plane average impact of the turbulence intensity.

Fig:2 geometry model of Delta wing plate

1.3CFD calculation

1.3.1 Meshing and model selection

Numerical simulation for the close coherence tower

with gas-solid flows using the Gambit software to

generate a dense coherent tower geometry, in order to

improve the stability and accuracy of the calculation,

the mesh in the numerical simulation of a hybrid grid

technology. Numerical simulation region in the

geometry of the rules, the use of six-side body structure

grid; in the irregular geometry of the region, using the

tetrahedral unstructured grid. Set in the value of the

number of simulated Chinese Communist 85140

computational grid, and 176 383 surface, 17 538 nodes.

Fluent software, gas standard k-ε turbulence model

closed the NS equations, based on Euler-Lagrange

method of Discrete Phase Model (DPM, Discrete Phase

Model) to calculate the particle phase orbit, the particle

size parameters based on the engineering of the actual

graythe sample size distribution data fitting, the median

diameter of 263μm, the distribution coefficient of 1.035,

the minimum size of 20μm, the maximum particle size

of 1000μm. Calculations using the coupled solver,

discrete format using the format of the second order

accuracy, pressure and velocity coupling the SIMPLE

algorithm.

1.3.2 The boundary conditions

Coherent dense tower reactor inlet velocity inlet

boundary condition, the speed to take the constant

value, the position before long-distance flue, it can be

regarded as fully developed pipe flow, for the entrance

of the turbulent kinetic energy dissipation rate, compiled

by Fortune Calculator for solving the estimation of

turbulence intensity and Reynolds number. In this

simulation, the speed at 18 m • s-1

, the turbulence

intensity to take 2.46%, the head diameter to take 3.2 m;

reactor outlet pressure outlet boundary conditions, the

reaction of the outlet and bag filter entrance connected

to the relative position static pressure fluctuations in the

-800 ~ -900 Pa, the numerical simulation value to -850

Pa; flue, reactor shell deflector and delta wing wall

boundary conditions for the restriction of fluid and solid

regions, which chain stirrer is set to slip wall around its

axis of rotation at a certain speed, and other wall

boundary using the default non-slip boundary

conditions.

2 Results Analysis

2.1 flue gas rectifier systems for flue gas distribution

Coherent dense tower flue gas rectification system

deflector arrangement under simulated flue gas stream

line in Figure 3; the tower at different heights plane flue

gas velocity distribution shown in Figure 4; different

tower cross-section of the turbulent kinetic energy

gradient, such asshown in Figure 5.

Fig3 Flue gasstreamlines on different arrangements of

baffle

Can be seen from Figure 3, in the role of the flue gas

rectification system, the flue gas are not found in the

fluidized bed easy backmixing [8], but the A, B, in the

emergence of a different position of the flue gas flow

deadarea, and C in the flue gas are more evenly

distributed and no obvious flow of the dead zone. Can

be seen from Figure 4, the deflector added to make the

flue gas flow rate, there was greater volatility in the

rectification of (Y = 6 plane) into the desulfurization

main reaction zone, A, B, in the flue gas flow are likely

to left low right trend, the minimum flow rate is less

than 1m / s, while the maximum flow rate has reached

more than 8.1m / s, C figure tower flue gas flow rate is




more uniform, the flow rate fluctuations in the 4 ~ 6m /

s, these consistent with the gas flow in the line of Figure

3. Fluent calculation of the pressure loss of the A and B,

C, three kinds of arrangement in the form were 32Pa,

36Pa, 47Pa, visible fabric of the uniform gas is to

increase the pressure loss at the expense of You can see

from Figure 5, additional deflector rectification zone

flue gas turbulent kinetic energy is significantly higher,

which is due to the air flow around the obstacle to a

Karman vortex phenomenon, which is also Figure 4 of

the Y = 6 curves coincide; can also see the rectifying

action of the deflector of the main desulfurization

reaction zone to reduce the turbulent kinetic energy, flue

gas distribution more homogeneous turbulent kinetic

energy, the lower the deflector cannot be alone at the

same time flue gas rectification and enhance the purpose

of the turbulent kinetic energy. Taking into account the

uniform distribution of flue gas desulfurization reaction

is of great significance, the flue gas pressure loss of the

rectifier system is not large, the flue gas of the turbulent

kinetic energy can be increased by other means, this

article selects C for flue gas rectification system

optimum deflector layout. , With the flue gas parameters

and flue shape change in the actual project, the deflector

arrangement of the flue gas rectifier system has to do to

be adjusted accordingly. In addition, although the

diversion board in consideration when designing a

circulating ash angle of repose, and other factors, but

may still run in the local sticky gray, so the maintenance

of flue gas rectifier system checks in the system

maintenance.

Fig4 Flue gas velocity distributions in different Y

planes（Z=0）

Fig5 Turbulent kinetic energygradients in different

sections

2.2 Simulation Study of the delta wing plate

2.2.1 The single-layer delta-wing board on the close

coherence tower with turbulence

Installation of single-layer delta-wing board, the

gradient of the turbulent kinetic energy of the close

coherence tower with different cross sections shown in

Figure 6; layer delta wing plate is installed, the tower Y

plane of the average turbulence intensity shown in

Figure 7.

Fig6: Turbulent kinetic energy gradients in different

sections

Fig7: Average turbulence intensities in different Y

planes

Can be seen from Figure 6, the addition of the first layer

of the drive within the component, dense tower

turbulent kinetic energy has undergone significant

changes, especially the delta wing plate 3 ~ 4m

downstream, the turbulent kinetic energy with no

1139 QIAN JIA, DONGHUI ZHANG, CUNYI SONG, ZHENSONG TONG AND BAORUI LIANG



additional components compared to increases of 2 to 3

times. Curve from Figure 7 we can see that the addition

of delta wing panels can not only change the

downstream turbulence intensity and its upstream flow

turbulence intensity. A turbulence intensity of the peak

area occurred in the area set the deflector, very

desulfurization cycle ash and flue gas mixing, and

turbulence intensity in the first layer of components

within the drive downstream 0.5m at a second the peak,

very conducive to the collision reaction of

desulfurization agent particles and SO2 from flue gas.

We can see from the figure, by adding a layer of delta

wing board, dense tower turbulent kinetic energy has

increased, but only extended to the downstream 3 ~ 4m

regional, and therefore we consider the addition of

multi-storey triangular wing.

2.2.2 multilayer delta wing plate of dense coherent

tower turbulence

Due to the limitations of the dense the Tata body size,

each delta wing plate cannot occupy the entire dense

phase the Tata body cross-section. Without layout of the

region of the delta wing plate, the flow pressure loss is

smaller, the flue gas flow from the middle, while the

short circuit flow phenomena, which is detrimental to

the dense tower of desulfurization. Set of two layers of

delta wing board, dense tower of smoke flow axial

velocity showed intermediate and high and low on both

sides of the distribution of additional multi-layer delta-

wing boards, and more conducive to the balanced flue

gas flow pressure loss. Layout of the first layer

downstream of the delta wing plate 2m, 4m and 6m at.

Close coherence of the different cross-section of the

tower mount multilayer delta wing plate, turbulent

kinetic energy gradient as shown in Figure 8 below; not

mounted delta wing panels and installation of single-

layer, four-story delta wing plate tower Y plane, the

average turbulence intensity in Figure9 shows.

Fig8: Turbulent kinetic energy gradients in different

sections

Fig9: Averageturbulence intensities in different Y

planes

Can be seen from Figure 8, after the addition of four-

layer delta-wing board, the new addition of a single set

of delta wing plate on the flue gas flow redistribution

and rectification, coupled with its own role to spin, the

dense phase tower internal effective reaction zone of

turbulent kinetic energy is further increased, the

downstream turbulent kinetic energy region and the

dense tower of export of high turbulent kinetic energy

region is connected into one. Can be seen from Figure 9,

after the addition of four-layer delta-wing boards, dense

tower of five high turbulence intensity peak, and four

other turbulence intensity peak occurred then ABCD

four-story-free drive component location; additional

four-story-free drive components, dense tower of

average turbulence intensity is increased to not add

components in case of a preset 3 to 4 times. Add a four-

layer delta-wing boards, dense towers of different

heights section of flue gas flow axial velocity gradient

is, we can see from Figure 10; additional four-story

delta wing panels, the tower flue gas flow improvement

in the flue gas flow is relatively uniform. Turbulence

intensity of the component after an additional four-story

drive within the region increased to 2.2 times the

original, gas-solid slip velocity significantly increased

the formation of a desulfurization reaction enabling the

effective reaction zone, the effective reaction zone The

pressure loss is about 80Pa, dense tower overall

pressure loss is less than 300Pa.

Figure 10: a high degree of cross-section on the flue

gas flow axial velocity

2.3 optimized dense coherent tower simulation study




The optimized gas flow in the dense coherent within the

Tata line, recycling gray trace shown in Figure 11.

Fig11: Flue gas streamlines in DFA and trace line of

circulating ash

Can be seen from Figure 11, the optimized dense

coherent tower flue gas flow line and the gray loop can

be evenly distributed and full access to the delta wing

plate at the same time increase the flue gas turbulent

kinetic energy, there is no impact on mainstream smoke

the uniform distribution did not produce flue gas back

mixing on the desulfurization reaction is favorable.

Delta wing plate can to some extent, to change the

trajectory of circulating ash particles, the particle size

the greater the impact is more obvious. Semidry

desulfurization process a larger common problem is

when the cycle of gray humidification uneven, there

will be partial reunion, which led to the sticky wall and

even harden a serious threat to the stable operation of

the system. Another important role of the delta wing

plate is possible reunion cycle of gray broken to prevent

the sticky wall knot and other problems occur [9]

. Can

also see from the figure, most of the circulating ash will

be deposited into the ash bucket, a small part of the

small particle size of circulating ash to be flue to carry

out into the bag filter. Delta wing plate on the one hand

to strengthen the inner loop, the other hand, can

effectively reduce the load of the bag filter [10]

. Fluent

computing the entire dense coherent tower pressure loss

of about 300 pa flue gas rectifier system and two layers

of chain blender produce about 80 Pa and 180 Pa of

pressure loss. Visible even at the within component,

dense coherent tower smooth gravity flow, pressure loss

is still small.

3 Conclusions

Dense coherent tower flue gas rectifier systems in Table

1, conditions optimum deflector arrangement for C, flue

gas to be able to achieve uniform distribution of the

flow rate fluctuations within 4 to 6m / s, the pressure

loss of 47 Pa; delta wing board to increase the density of

coherent turbulent kinetic energy of the tower,

especially the effect on its nearby area, the turbulent

kinetic energy increases by 3-4 times; to enhance the

local turbulence intensity of the flue gas to a certain

extent, but to enhance the effect on the overall

turbulence intensity not obvious; mount multilayer delta

wing plate can increase the coverage area of high

turbulence intensity, and compared with the single-layer

turbulence intensity is not easy to attenuation.

Simulation results show that under optimal conditions:

uniform distribution of dense coherent tower of smoke

and gas-solid full access to the delta wing plate a

significant impact on the movement of large particles

circulating ash, most of the circulating ash particles

settlement involved in loop, only a small part of the

small particles of flue gas brought into the dust

collector, close coherence tower pressure loss of

approximately 300 pa.

4. Reference:

[1] Dang Y H, Qi Y H, Wang H F. Technology of flue

gas desulfurization, Journal of Iron and Steel

Research, 2010, 22(5): 1-6

[2] Feng L. Analysis of wet type desulphurization

method of sintering exhaust fume, Mining

Engineering, 2011, 9(3): 66-68

[3] Zhao Y Z, Cheng Y, Jin Y. CFD-DEM simulation

of clustering phenomena in riser and downer,

Journal of Chemical Industry and Engineering,

2007, 58(1): 44-53

[4] Zhou Y G, Peng J, Zhu X, et al. Hydrodynamics of

gas–solid flow in the circulating fluidized bed

reactor for dry flue gas desulfurization, Powder

Technology, 2011, 205(1-3): 208-216

[5] Wang F, Lu M. The study on spray dryer absorber

flue gas desulfurization for industrial boiler ,Fuel

and Energy Abstracts, 2002, 43 (4): 274

[6] Chang G Q, Song C Y, Wang L. A modeling and

experimental study of flue gas desulfurization in a

dense phase tower, Journal of Hazardous Materials,

2011, 189(1-2): 134-140

[7] Hou H C. Preliminary Analysis and Application on

karman Swirl, Journal of Qinghai Normal

University(Natural Science Edition), 2005, 3:23-25

[8] Wang J C, Wang X Y, Wang S D, et al. The

research on gas-solid flow in high particle flux CFB

riser, Boiler Technology, 2010, 41(1): 36-39

[9] Zheng X Y, Liu B Q. Mechanism research on

particle clustering and crushing in dense flue gas,

Electric Power Environmental Protection, 2006,

22(1): 55-57

[10] Zhu J L, Chen X J, Zhang T, et al. Computational

fluid dynamics simulation of hydrodynamics in an

internal-loop fluidized bed reactor with a funnel-

shaped internal, Acta Scientiae Circumstantiae,

2011, 31(6): 1212-1219.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1141-1150


Vehicle Information Compression and Transmission Methods Basis

on Mixed Data Types

YANG JINGFENG1,2,3

, LI YONG4**

, ZHANG NANFENG1,5

, HE JIARONG6 and XUE YUEJU

1*

1South China Agricultural University, Guangzhou, 510642, China

2Guangzhou Public Transport Data Management Center, Guangzhou, 510620, China

3Guangzhou Transport Information Control Center, Guangzhou, 510620, China

4Guangzhou Institute of Geography, Guangzhou, 510070, China

5Guangzhou Entry-exit Inspection and Quarantine, Guangzhou, 510623, China

6South China University of Technology, Guangzhou, 510641, China

Email: [email protected], [email protected], liyong@ gdas.ac.cn

Abstract: For the unbalanced coverage of mobile communication base station, data loss and incomplete data are

prone to emergence when a large amounts of real-time data transmission through 3G/GPRS network in the poor

communication environment of subregion, which is difficult to achieve large-scale farmland information collection

and transmission. Therefore, optimized Huffman coding compression algorithm for text-oriented information

(including location information) ,optimized iCAM06 compression algorithm for image, MPEG-4 compression

algorithm combined with DCT differential modulation algorithm for video are promoted in this paper. Through

corresponding compression method for vehicle information mixed data types of text, image, video respectively, this

paper designed data transmission strategies of different transmission priority and the data uniqueness checking

method at the same time. The whole process test results of data exchange and decompression show that, the data

compression algorithms can effectively realize data (text, image, and video) compression. The data sampling period

is relatively smaller and its adjacent data is closer, the higher compression rate can be obtained by corresponding

compression algorithms, and it can the basically ensure the decompression data without distortion. This will improve

the efficiency of information transmission of transportation, as well as to ensure the integrity of information, which

is of great significance to realization of transportation energy conservation.

Keywords: Vehicle Information; Data Compression; Priority Strategy Transmission Mode; Mixed Data Types.

1. Introduction:

The key operational vehicle information safety

supervision has been continuously strengthened along

with the national, provincial governments at all levels.

The standard terminal with satellite positioning function

were forcibly mounted on passenger bus, tourist charter,

dangerous goods vehicles, coaches, trucks, heavy

trucks, buses, taxis and so on. Some cities also require

passenger bus, tourist chartered buses, taxis and other

installation equipment using a video camera in order to

enhance information security regulatory standards.

Regardless the using of satellite positioning terminal or

video camera equipment, it must bear a lot of the cost of

wireless data transmission. If using new energy

vehicles, such as electric buses, hybrid buses, electric

taxis, etc., due to the need to upload more types of data

to the regulatory status of the platform, its wireless

transmission costs will rise at the same time. How to

meet the requirements of the competent authorities, and

saves the cost of wireless data transmission has become

an important problem to be solved at present industry.

The compressed data transmission is one of the

important ways to solve the problem, and it can ensure

the real-time and data quality of data transmission in the

city communication environment, many scholars have

conducted extensive research in this areas. Vehicle

information is now covered by the text, pictures, video

and other types of information. In order to improve data

transmission quality, many researchers have proposed a

variety of compression and transmission methods, for

example, Hashemian[1]、 Sharma[2]、 Shi[3]、

Meng[4] et al proposed for the Huffman coding

technology for wireless transmission of GPS, including

GPS coordinate, records and other text data

compression and transmission problems; Hu[5]、

Jou[6]、 Wang[7]、 Barr[8] et al put forward the LZW

compression algorithm which can be optimized by

establishing a method to quickly find a dictionary, and it

can greatly reduce the time of data compression to

achieve optimal compression, and improve transmission

efficiency; Wavelet compression[9] method for wireless

transmission of compressed gradually become a hot spot


1142 Vehicle Information Compression and Transmission Methods Basis on Mixed

Data Types



in recent years. In terms of traffic information, based on

GML data compressions, Zhang[10] achieved a traffic

emergency data and realized a framework of urban

traffic emergency system architecture; Through random

matrices with restricted isometric conditions of the

original high-dimensional data onto a low dimensional

space, Li[11] realized the data fast and efficient

compression, and via a convex optimization algorithm

to complete the data decompression in traffic

information processing terminal after data transmission;

According to the characteristics of traffic data, Zhao[12]

using its research and comparative data compression

methods based on principal component analysis (PCA)

and independent component analysis (ICA) method,

respectively. Scholars' research results could furnish

some useful suggestions in the transmission terminal

data to a certain extent. However, the problem can not

be ignored is, because of the unbalanced nature of the

distribution of urban mobile communication base

stations, urban communication environment vary

greatly, the data transmission capacity is in greater

differences in different parts of the city, part of the

region exists communication blind area. The method

proposed by scholars rarely used in acquisition and

transmission terminal data on one hand, also rarely take

into consideration differences the problem in the larger

urban communication environment for real-time

transmission of data, that is, how to realize text,

pictures, video and other information integrated

transmission while coverage of communication base

station is uneven with large difference between the

communication environment in urban areas. In addition,

the large flow of data transmission needs to consume

more energy; low-power transmission[13] has important

significance to prolong the length of the various devices

work. Therefore, based on priority principle approach

for simple algorithms, short time-consuming,

minimizing overhead of data collection terminals and

the backend server, the vehicle operating data

information compression and transmission methods

based on priority strategies is proposed in this paper..

Data collection vehicle information compression and

transmission method can be achieving the optimized by

encryption, compression, transmission, reception,

decryption process, which is important to improve the

transmission efficiency of vehicle information to ensure

the integrity of information and transportation industry

to achieve energy saving.

2. Vehicle Information Compression and

Transmission Method

According to national standards, Ministry of

Transportation and the provincial technical standards

requirements, the current standard of vehicle

information including position, speed, altitude, license

plates, start/flameout, lights, driving status, drivers and

other types of text, plus photos and video that collected

through the data collection terminals. If the vehicles for

new energy vehicles, it needs to extending battery

status, temperature, and overall and the cell voltage,

current and other battery management information. The

data with different format, size, and type will be

uploaded via the data exchange and the wireless

communication module of the vehicle terminal.

The whole process from the point of view of data

transmission, namely, data collection, transmission,

reception, storage throughout the process point of view,

when compared to a large number of simultaneous

transmission vehicle terminal especially, the main

pressure is focused on how the backend server receives

a large number of short, transmission, integration of

data transmitted by the terminal.

The way to solve the problem of large amounts of data

transmission and access mainly includes three. First,

data optimization in vehicle terminal, extract the key

significance with application data, the remaining data is

temporarily stored in the terminal memory device.

Second, data are identified in the data receiving

platform to extract of critical data, but it must guarantee

the reliability of data transmission. The last, data are

compressed by vehicle terminal firstly, and

decompressed on the platform.

The compression-decompression method can ensure the

integrity of data transmission, but also can reduce the

data communication flow, which can save

communication costs, but its disadvantages are also very

obvious. Optimization transmission of terminal is with

less disadvantages, the main disadvantage is it requires

the performance processing each group of data

compression, which would lead to transmission time

delay. Real-time upload of vehicle information is

demanding on safety supervision, except pictures and

video information in the daily management which does

not affect the actual data application with reasonable

delaying. Moreover, with higher hardware configuration

of vehicle terminal at present, the capable of handling

large amounts of data vehicle terminal hardware

compression technology is promoted. Data collecting,

data compressing and transmitting could be performed

at the same time, which could not be simultaneously

affected.

2.1. Data Compression Algorithms

Due to the limitations of the vehicle terminal hardware

conditions, the applications for data transmission

compression by vehicle terminal were less in the past

for a long period of time in the practical. Images, video

and other data that takes up a lot of storage space are

usually collected in offline mode, namely, images, video

and other data in not considering the real-time

requirement when used by off-line way.

1143 YANG JINGFENG, LI YONG, ZHANG NANFENG, HE JIARONG and XUE YUEJU



Although the vehicle terminal hardware technology

continues to improve, but data collection, compression

and transmission operate at the same time will occupy

the terminal necessarily overhead occupy the terminal.

Therefore, the selection of data compression algorithms

in this paper compared with traditional data

compression algorithms, the main principles include

that, it must reach the compression effect and the

algorithm should not be too complicated in limited

bandwidth resources, which would avoid the influence

of compression and transmission efficiency, and not

occupy too much memory(Taking up too much memory

will affect data collection, as well as other applications

occupy a larger memory), and use a common

transmission protocol.

Based on the requirements of practical applications, we

use a method to realize data compression by

compressing respectively and data transmission by

sending a unified package. Aiming at the localization of

GPS information and the other text data to be

transmitted, Optimized Huffman coding techniques [10,

11] for lossless compression of data effectively is

promoted. The current high-profile vehicle terminal is

equipped with high dynamic range image collection

equipment; HRD image data can be captured. Since

HRD image requires for greater data storage relative to

the general image. Taking into account the use of high

dynamic range images to facilitate analysis of

emergency, we used iCAM06 algorithms to realize the

compression of HRD image. The iCAM06 algorithms

can extend brightness level from dark to bright visual,

avoid flare occurs, and it can accurately predict a large

range of image brightness level of complexity human

visual attributes, which is of great significance to

capture the complexity vehicle information in different

areas of the city. For video data, we use generic MPEG4

video compression algorithm to realize the video data

compression.

Based on the operating environment consideration of

various commercial vehicles, the selection of

compression algorithm for operating state of the

commercial vehicle in this paper must be followed

principles of less calculation, fast compression, simple

algorithm and easy to implement. The algorithms must

to be reached an optimum compression effect which can

not be too complicated to avoid affecting the

compression efficiency of the transmission under the

limited bandwidth resources to meet the requirements of

the terminal hardware and communications

environment. Text data is the main data for the

commercial vehicle terminal collecting(including

GPS/Beidou positioning information).Optimized

Huffman coding technique[1,15] for effectively lossless

compression is proposed for the GPS/Beidou position

data and the other text of commercial vehicles operating

state to be transmitted in this paper.

Huffman coding principle is: Set consisting of the text is

represented as a set of characters with

1 2 3, , nI I I I I , where xI represents different

characters of text. Assuming the frequency of a

character xI is xF , the code length is xL . To make the

total length of code source text file of the shortest, you

need to determine the coding mode. Making the

minimum value of

0

n

x x

i

F L

, this encoding is called

Huffman coding [3]. Huffman code based on Huffman

tree, Huffman tree construction steps are shown as

follows [3,4].

Step 1. Set the given n weights 1 2 3, , , , nw w w w

according to n binary trees, where every tree xT has

only one weight xw is the root of 1, the left and right

sub-tree are empty;

Step 2. Select the root node of the binary trees in F

with the minimum weights of sub-tree as to construct a

new binary tree, and the root of the new lien binary is

the value of its left and right sub-tree root node the sum

of the weights.

Step 3. Delete the two trees in F , while the new binary

was added to F .

Step 4. Repeat Step 2 and Step 3 until F set up with

only one tree, this tree is the Huffman tree.

Typically, a large number of repeated characters are

existed, such as GPS data. GPS and the other repeated

characters text data included by vehicle information can

be regarded as redundant information to be removed.

On the basis of Huffman compression coding table for

rapid processing after compression by contrasting data,

then stored in the storage buffer for post-processing.

Among them, Huffman compression coding tables are

the number of times pre-generated by the background

server for text data in the statistical characters, and pre-

stored in the Flash of vehicle terminal. The Realization

of the process is shown in figure 1.

Removed the redundant

information

Contrast Huffman compression

coding tables

Fast Huffman encoding and compression

Pre-stored in the storage

buffer

Text data of vehicle terminal

Fig.1 The Compression Flow Chart Of Vehicle

Terminal Text Information


Data Types



Huffman encoding is completely according to the

characters of probability to construct coding, there is no

error protection function. The algorithm needs to

calculate the source character of probability statistics

and obtains the probability distribution of the source

symbols. This process requires a large cache using the

hardware implementation of the algorithm. Generally, it

believes that the decoding is complexity, and it's not

conducive to the realization of hardware implementation

[15,16]. Huffman coding and arithmetic coding is a

typical probability model, many scholars proposed the

dictionary model for optimizing the original model, the

typical optimizing algorithms such as LZW algorithm

[4,15], etc.

For compression and decompression speed, LZW

algorithm is fast, and requires only one scan of the

compressed data to be compressed. The compression

ratio of the algorithm is very good for the continuous

input stream to be compressed with high rate repetition.

But the main drawback of this algorithm is the

adaptability is rather poor, and the compression rate is

less than Huffman algorithm. Moreover, the LZW

algorithm usually needs to combine with other

algorithms to reach the expected aim for some file with

very low complexity. For vehicle operation status data,

the data type apparently does not apply to LZW

algorithm.

Huffman coding method has certain limitation, such as

it need to scan flow of input symbols for two times, be

stored or transmitted by Huffman tree before the storage

or transmission of Huffman coding results. However,

the algorithm has high compression rate, simple and

practical, and all the coding has uniqueness while

decoding, very suitable for vehicle information data that

has high identification requirements. In order to

overcome the practical problem of cache, high

complexity caused by the Huffman coding, we

improved the structure of Huffman tree based on the

original. The main steps are as follows:

Step 1. Initialize the established binary Huffman tree,

one of the root node has a weight of 0;

Step 2. To encode without the new characters of newly

generated two nodes, the connection of the parent node

weights add 1, and another node, which is defined as

node weights of 0 nodes to define new weight of 0;

Step 3. The new characters have been encrypted through

search was coded character location, in the case of the

characters of the weight add 1 with coding and

compared with the original weight of the same node;

Step 4. Repeat Step 2 and Step 3 until all characters are

encoded.

The improved way of Huffman coding by weight set to

0, the original requires two symbols flow is simplified

to a scan. The algorithm only needs to scan a probability

of 1 characters, and it can only exchange the node

between the binary tree needs to node numbering, which

reduces the occupation of the cache, and the complexity

of the algorithm is decreased. Its deficiencies are mainly

embodied in the complete lost of coding error protection

function, and the data requirements are relatively high.

Popular image compression algorithm at present include

equalize histogram, local color adaptation algorithm,

tone compression method, adaptive algorithm, iCAM

algorithm, iCAM06 algorithm[17] et al. Through the

compression of base layer of large scale and layer

relatively complete reservation details, iCAM06

algorithm can obtain the histogram shape close to

original HDR images, and the distribution of different

pixel HDR image has good applicability, it is the best

performing algorithm [17]. This paper uses iCAM06

algorithm to compress the image information of vehicle

terminal in the driving process, and then realizes data

transmission.

ICAM06 algorithm converts the data into a XYZ signal,

using the edge preserving filter decomposes image into

a base layer (contains only large-scale change

information and detail layer) and detail layer (edge and

texture detail) to layered double filter processing

[18,19]. In view of the human visual system to the

characteristics of the local contrast is more sensitive

than the global contrast, the compression operation is

realize in high dynamic range scene and preserve local

details in the scene global dynamic range[18,19].

Layered double filtering processing algorithm is shown

as follows:

1( ) ( )

( )s p s p

p

J f p s g I I Ik s

Where, ( ) ( ) ( ), ( )p s

p

k s f p s g I I f p s

is

the Gaussian filtering function in the spatial domain,

( )p sg I I is the Gauss filter function intensity

domain, pI and sI are pixel intensity respectively,

represents the whole image.

Layered double filtering processing can adopt fuzzy

mathematics theory to deal with HRD images

effectively while keeping the edge sharpening in good

condition. Layered double filtering processing can also

prevent artificial halo phenomenon, and accurately

predict human visual properties in complex images

environment of a large range of brightness levels. This

proposed algorithm can better predict the color

appearance of complex wide range scene, and it's more

suitable for HDR image for cross-media copying [17,

18, 19].




The nonlinear compression effect of iCAM06 algorithm

is to realize the high dynamic range image gradual

compressing into a low dynamic range. Due to cut off

some color range by compression process, it may lead to

appear unrealistic rendering. The reason for this is, color

has been tonal mapping into general in the middle range

image, image color was scaling and tailored to coloring

is directly mapped to a specific profile. When the output

range compared to the input range greatly, the

specifically profile problem is especially obvious. In

this case, one specific performance is, if the image

scaling to smaller scale directly, many colors in the

output will become too dim to be observed, gradient

between colors can also be lost while the output

incompatible colors of the excess. Another specific

performance is, cutting operation is likely to cut off the

middle range of color, or to cut off a wide range of color

and to combine with the color of less zoom range.

The main method to solve this problem is to make the

color gamut in each direction being more uniform,

namely, to increase the brightness and improve

saturation. The specific approach is shown as follows.

Settings the average values of the image brightness

as avrJ . Setting the average brightness value for all

colors on the image with the sort brightness value

greater than 99%, denoted by 99J ; the average

brightness value for all colors on the image with the sort

brightness value less than or equal to 1%, denoted

by 01J . If 99 01

avr

J JK

J

( K is the proportional

coefficient, usually take 2K ), then continuing to

sorting statistics brightness value is greater than 98%

and less than or equal to 2% of average sort brightness

mean value, and remember to 98J and 02J respectively.

Calculating 98 02

avr

J J

J

, if 98 02

avr

J JK

J

, then further

calculating according to the extent of 1% recursive until

100i i

avr

J JK

J

, while iJ can be defined as the

maximum brightness value, and 100 iJ is the minimum

brightness value.

Cutting operation by iCAM06 algorithm will be done

after the maximum brightness value and minimum

brightness value has been got. Removing the color

beyond the range of the maximum and the minimum

limit, and scaling color between maximum brightness

and minimum brightness values, then the image

compression operations could be realized.

For the compression of video data, this paper uses

MPEG-4 video compression algorithm. Video image

can be divided into macro block by Mpeg-4 algorithm,

while DCT combined with differential modulation

methods for compression. The specific calculation of

MPEG-4 algorithm has the separable characteristics,

which can be fast compression at the same time and

provide compressed in different ways according to the

different object. That will not only provide a higher

compression ratio but also a better visual quality.

2.2. Data transmission method and Priority Strategy

Transmission Mode

The uploaded data of vehicle information according to

the file size of text data (including GPS positioning

data), image and video. According to the vehicle data

information for wireless transmission is of simple

structure, single function and small data redundancy

tolerance, all data would be uploaded by the

international standard NMEA-0183 GPS data

transmission as well as a variety of on-board GPS

monitoring terminal transmission protocol in this paper.

On the deployment of mobile base station, the

deployment of the city is not balanced, signal fade area

exist in the villages and overpass nearby due to various

reasons. Before the data is transmitted it usually

requires multiple handshaking to establish wireless

communication network, while communication is easy

to failure when handshaking. In expanding the data

acquisition range of the vehicle terminal, it needs to

transmit a large flow of vehicle information such as text,

pictures, video, and various exchanged sensor data,

which is easy to lose in the data transmission process. It

is difficult to ensure the integrity of the vehicle

information. Incomplete data may also be difficult to

parse while backend server receives the data, the

calculations pressure of background server for

completeness checking on the vehicle information will

also be increasing at the same time.

Based on the actual condition of vehicle information

and requirements, priority transmission principles and

requirements must be developed in the first time to meet

the different requirements of different transmission

situations. In the default setting or in this case of

receiving terminal without particular choice by users,

default the following principles for data transmission.

(1) Give priority to compress and transmit text data,

followed by the image, final the video in a better

wireless transmission or communications environment.

Make it a priority to compress and transmit the text data

such as GPS, temperature automatic collected or

exchanged by the receiving module or the sensor data

without manual input when receiving terminal did not

put forward the image, video transmission requirements,

then the manual input data, and the image, video etc.


Data Types



(2) Give priority to compress and transmit text data,

followed by the final image, video in a better wireless

transmission or communications environment. Give

priority to compress and transmit text data in a poor

wireless transmission or packet loss environment. Image

and video data will save by the Flash embedded in the

terminal. Make it a priority to compress and transmit the

automatically collected text data, and the data

supplementary transmission will be no longer reserved

GPS data compression and transmission requirements in

the transfer process. Manual input data can be

selectively compressing transmission.

The main approach of transmission priority strategy

mode is, different types of data with data coding mode

corresponding to the data when uploading, uploading

the smaller compressed data packets of data such as text

data firstly. As to the images, video and other data, they

will be compressed by vehicle terminal and transmitted

by residual bandwidth if they’re necessary to real-time

upload, and transmission priority of images is higher

than the video. Uploading strategy and permission

assignment algorithm can ensure that the requirements

of different data transmission each priority.

Transmission priority strategy in particular is, meeting

the uploading requirements of text data has been

compressed packet firstly, then the compressed images

through remaining wireless transmission bandwidth, and

the video last. Generally, images and video would be

retained in the memory of vehicle terminal while it has

been met compression and transmission requirements of

text data. Vehicle information are collected by sensor,

and transmitted by wireless channel, which is difficult

to ensure the quality in the wireless communication

environment. Code missing is very frequent while in the

poor communication environment and it must carry on

the correction processing corresponding to the vehicle

terminal software as well as the background. Software

error correction method is mainly used in this paper.

The reason is, vehicle information collecting has certain

continuity, location and other information cannot mutate

in a short time. It can determine the new data whether it

is reasonable according to the collected data and correct

them.

2.3. Data Decompression and Parsing

To distinguish the uploaded data from different vehicle

terminal and to ensure uniqueness of uploaded data,

uploaded data must be parsed and recognized. Default

the parsing code for different vehicle terminal before

uploading the data. Read and Match the parsing code by

the reception server to achieve data uniqueness

verifying. The specific processing is shown as follows.

Data compression and parsing code adding by vehicle terminal

Data receiving and parsing code

validation

Data receiving and parsing code

validation

Data decompression and storage

Fig.2 The Flow Chart of Data Parsing

Pictured above, it needs to setting parsing code

corresponding to vehicle information first. Data

compression and parsing code adding are processed by

vehicle terminal. Backend server receives the data

compression package, and matches the SIM card

number, parsing code and other information

corresponding to set parsing codes corresponding table.

The perfectly matched data will be decompressed and

written to the database.

3. Experiments and results

According To meet the requirements of real-time data

transmission of images and video, this paper uses 3G

network as a data transmission channel for data

transmission. In 3G network, the prerequisite to achieve

the theoretical rate is the premise of carrier frequency

for data communication, completely and exclusive to a

user only. Due to the self-interference system affected

by the communication environment, the use of the

number of actual formation, the actual data transmission

rate and theory of data transmission rate is often of large

difference in the vehicle information collection

environment. Data transmission implemented by

transmission priority strategy is also need to be further

confirmed by end-to-end tests in real environment in

this paper.

3.1. Data transmission test

Data transmission test mainly completes two test

including the network delay test and TCP transmission

rate test in the actual environment in city village.

Unicom's 3G networks was selected for communication

channel for testing in city village environment, vehicle

terminal data was collected and uploaded, then received

by server.

Table1: Testing Results of Raw Data Uploading

Serial Number/

Test Items

Maximum Value of

Ping Testing

Minimum Value

of Ping Testing

Average Value of

Ping Testing

Received

Volume

Packet Loss

Volume

Packet

Loss Rate

1(Packets：120) 6154ms 1089ms 2289ms 114 6 5.00%

2(Packets：120) 7123ms 998ms 2441ms 113 7 5.83%

3(Packets：120) 6615ms 1213ms 2221ms 115 5 4.17%

4(Packets：120) 5118ms 936ms 2583ms 117 3 2.50%

5(Packets：120) 5767ms 1007ms 2417ms 116 4 3.33%




Round trip time (RTT) is usually used as a measure of

3G network time delay. Low delay time is of extreme

importance in actual application. Lower delay is

advantageous for the uplink direction to improve

capacity and data throughput, and increase the coverage

of high bit rate, obtain higher transmission bit rate.

Typically, data transmission testing uses PING method,

namely by obtaining response time of sending and

receiving data packets to get connected server response

time. When sending upload packets in requesting, all the

serial number sequence will be added one firstly before

sending, the replying information will also be marked

the corresponding serial number. Then observe the Ping

packet response messages to detect the link for data

transmission delay calculation, such as packet loss,

grouping repetition and fault sequence etc.

The following table shows the data uploaded test results

in a village environment of a city. To be clear, several

uploaded tests were started at 10 in the morning, ICMP

packets for a total of 120 were send every time, the 3G

bandwidth rate of 70-120kbps (city 3G bandwidth rate

is about 700-900kbps), TCP transmission rate is close to

3G bandwidth rate.

The result of data packets uploaded test shows that, the

maximum delay time is 7123ms,the average delay time

of 2382ms, which meets the basic requirements in real

time data uploaded. 3G network takes the average

packet loss rate 4.17% of data uploading while the

network bandwidth is 70-120kbps. Relative to the

GPRS data transmission mode, 3G network has been

able to meet the requirement of actual application in the

setting of upload strategy and selecting transport

protocol, but also laid the foundation for the data

compression test environment. In addition, it appears

from the data uploading test results, village area of the

communication environment is relatively poor, data

compression and transmission will still be of importance

as well.

3.2. Data compression and decompression test

Based on the established vehicle terminal, 3G networks

and backend server system, setting default priority

transmission strategy, compression and transmission can

be started by the collected vehicle information, and the

results of comparing the uncompressed data

transmission with compressed data transmission can be

shown. The data compression testing is shown in Table

2. The main evaluation on the effect of data

compression is mainly based on the compression rate

and transmission time. Compression rate is one of the

most important indexes to measure the data

compression algorithm. Compression rate is a direct

measure of data reduction, the running time reflects the

complexity of the algorithm, and the combination of the

two indexes can be further estimated energy efficiency

of communication. Transmission error rate is the rate

between data length of accurately praised data

(compressed data containing the decompression

process) had not received by backend server and the

data length before compressing. Packet loss rate is the

ratio between the packet numbers of server has not been

received and data packets had been transmitted by

vehicle terminal.

Table2: Testing Result of Data Compression

Serial

Type

of

Data

Data

Collectio

n Cycle

Data

Length

Before

Compres

sion (kb)

Data

Length

After

Compres

sion (kb)

Compres

sion Rate

Transmis

sion Time

Before

Compres

sion(s)

Transmis

sion Time

After

Compres

sion (s)

Transmissi

on Error

Rate

Before

Compressi

on

Transmissi

on Error

Rate After

Compressi

on

Packet

Loss

Rate

Before

Compre

ssion

Packet

Loss

Rate

After

Compre

ssion

1 Text 30s 148 48 32.43% 0.45 0.14 2.21% 2.01% 4.39% 3.89%

2 Text 20s 2156 646 29.96% 6.76 2.02 2.47% 2.15% 4.12% 3.73%

3 Text 10s 5121 1473 28.76% 14.56 4.26 2.24% 2.07% 3.84% 3.22%

4 Text 5s 9892 2563 25.91% 29.28 7.09 2.29% 1.99% 3.67% 3.25%

5 Image 30min 6632 2236 33.72% 17.23 6.76 - - - -

6 Image 20min 13287 4029 30.32% 40.21 12.89 - - - -

7 Image 10min 24659 6321 25.63% 65.03 17.98 - - - -

8 Video 30min 28935 9356 32.33% 78.76 28.44 - - - -

9 Video 20min 53127 16549 31.15% 167.29 50.66 - - - -

10 Video 10min 112789 31080 27.56% 302.26 75.31 - - - -

Experiment result shows that, compressed transmission

effect was more notable. The transmission amount of

compressed data is reduced greatly, accounting for

about between the original amount of data of 1/4-1/3,

and the corresponding transmission time is shortened

into 1/4-1/3. In addition, it can be seen from the cycle of

data collection, data sampling period is relatively

smaller and its adjacent data is closer, all kinds of


Data Types



algorithm could get a higher compression rate. In

particular, the correlation data has obtained the very

good compression rate, which greatly enhances the

communication efficiency of system, but also save the

cost of communication.

In order to further verify the results of data compression

algorithm, this paper compares the original data with the

parsed decompressed data. Through the data

exchanging, vehicle terminal could obtain the speed

sensor data. Exchange data through data compression

method based on the technology of Huffman coding for

data compression, and then the original data file and

compressed data files are transmitted to the backend

server through 3G network. After transmission, read the

received data file for decompression and inverse

quantization, and compare the difference between the

original data and data contrast reduction before and after

data compression.

To be clear, the speed testing data was selected for a

passenger bus driving data from 8:00 am to 16:00. An

average of 10 seconds data collection cycle was set in

the passenger bus vehicle terminal, and it generated for

a total of 2727 (theoretical value should be 2880,

because of various reasons in the process of data

collection, there are 153 errors, lost or not collected data

samples) speed data during the period. Before the data

compression, the original speed changing curve is

shown in Figure 3. In Figure 3, the horizontal represents

the speed of data collection time; the ordinate represents

speed value, where are the same as Figure 4. After the

data compression and transmission, the parsed speed

changing curve is shown in Figure 4. Figure 5 is the

speed value error rate comparison chart of before data

compression and after decompression. The horizontal

represents the speed of data collection time, the ordinate

represents the difference between the speed value before

compressed and the speed value after decompressed.

Fig.3 The Velocity Curve Before Data Compression

Figure 4 shows the decompressed and parsed data which

have been compressed, transmitted, parsed,

decompressed and warehoused to the backend server.

Fig.4 The Velocity Curve After Data Decompression

Fig.5 The Curve of Curve of Velocity Error Values

From the speed curve of the data before and after

compression, the speed error value curve can be seen,

the speed data consistent with original data basically

before compressed through the process of compression,

transmission, parsing, decompression, and data storage.

Coding, compression and other data processing

technology have met the basic requirements in actual

application. To accurately describe the data differences

between situation before and after the data compression

and transmission, this paper calculated the mean

variance to discriminate compression deviation after

transmission of the data. After calculation, the mean

square difference before and after data compression and

transmission is of only 0.2462, which is within a

reasonable range. The main reason for the deviation of

the data exists in data loss, data failing to parse and the

other reason during data transmission.

A total of 2727 data is to accurately have been parsed

during 2880 speed data transmitted by vehicle terminal,

the proportion of accurate transmission and parsing data

is up to 94.69%, which basically meet the requirements

of actual application. 153 errors, loss, and no sample

data contained the error data of 23, accounting for

0.80% of the total, loss of data 74, accounting for 2.57%

of the total loss rate, no sample data of 56 due to various

reasons did not collect, accounting for 1.94% of data

should be collected of the total. The availability testing

experiment results of compression and transmission

algorithm show that, the algorithm has high efficiency

and high transmission integrity.

4. Conclusions

Compression algorithm method is researched in this

paper to the vehicle information including text

information collection types (including position




information), image and video to compress the data by

using Huffman coding technology, iCAM06 algorithm,

MPEG-4 video respectively. On this basis, all kinds of

vehicle information is transmitted based on NMEA-

0183 various kinds of vehicle GPS monitoring

transmission protocol. Through the process of vehicle

information compression, transmission, decompression,

parsing and storage, the compression algorithms are

verified by speed data as an example. The algorithm

achieved great test results in the experiments, the

proportion of accurate data transmission and parse is up

to 94.69%, basically meet the requirements of data

uploading actual applications in different city region for

the urban public transport, passenger transport, taxi and

other industries. Meanwhile, the algorithm play an

important role for enhancing the efficiency of wireless

transmission applications and reducing the cost of data

transmission as well as the time, which has the practical

significance for the deployment of this base station, the

selection of communication mode and the other uses.

The data compression algorithms mentioned in this

paper provides a method with transmission reliability

and minimal distortion. It can provide reliable data

transmission method for different mobile base stations

coverage of the urban environment, which is important

to improve vehicle transmission efficiency of

information, ensure the integrity of information and the

realization of transport energy conservation.


The research is supported by Science Program of

General Administration of Quality Supervision,

Inspection and Quarantine the People’s Republic of

China (2014IK183) and Guangdong Research Projects

Special Funds of Industry-Academy-Research

Cooperation (2012B091100345) and Guangzhou

Science and Technology Project (2014Y2-00044).

6. Reference:

[1] R. Hashemian. Condensed Table of Huffman

Coding, a New Approach to Efficient Decoding,

IEEE Transactions on Communications, Volume

No. 52, Issue No.1, 6-8, 2004.

[2] M. Sharma Compression Using Huffman Coding.

International Journal of Computer Science and

Network Security, Volume No. 10, Issue No.5,

133-141, 2010.

[3] P. Shi, M. Li, T. Yu. Design and Application of an

XML Date Compression Algorithm Based on

Huffman Coding. Journal of Beijing University of

Chemical Technology (Natural Science Edition),

Volume No. 40, Issue No.4, 120-124, 2013.

[4] F.Y. Meng. Research and Application of Huffman

Coding in Environmenta1 Protection Rea1

Monitoring System, Thesis of Degree of Master of

Ocean University of China,2010

[5] P. Hu, J.Z. Zhang. Study and Implementation for

Terminal Data Compression Technologies in

Remote Fault Diagnosis System. Computer

Engineering and Applications, Volume No. 48,

Issue No.4, 130-135, 2012.

[6] J.M. Jou, P.Y. Chen. A fast and efficient lossless

data-compression method. IEEE Transactions on

Communication, Volume No. 47, Issue No.9, 1278-

1283, 2006.

[7] Q. Wang, C. Qi, X.M. Luo. Modified LZW

algorithm and its parameters optimization. Journal

of Chongqing University of Posts and

Telecommunications, Volume No. 17, Issue No.3,

351-355, 2005.

[8] K.C. Barr, K. Asanovic. Energy-aware lossless data

compression. ACM Transactions on Computer

Systems, Volume No. 24, Issue No.3, 250–291,

2006.

[9] S.H.Yang, Y. Jia, S.W. Zhou. Virtual nodes-based

wavelet compression algorithm for wireless sensor

networks, Ruanjian Xuebao/Journal of Software,


[10] H.Y.Zhang. Based on ArcGIS Server Urumuqi

Traffic Emergency System Data Compression

Technique Research, Thesis of Degree of Master of

Xinjiang University, 2011

[11] Q, Q. Li, Y. Zhou, Y. Yue. Compression method of

traffic flow data based on compressed sensing.

Journal of Traffic and Transportation Engineering,


[12] Z.Q. Zhao, Y. Zhang, J.M. Hu. Comparative Study

of PCA and ICA Based Traffic Flow Compression.

Journal of Highway and Transportation Research

and Development, Volume No. 25, Issue No.11,

109-114, 2008.

[13] H.L. Xu, H. Zhang, Y. Shen. Environment

monitoring system for flowers in greenhouse using

low-power transmission. Transactions of the

Chinese Society of Agricultural Engineering

(Transactions of the CSAE), Volume No. 29, Issue

No.4, 237-244, 2013.

[14] X.J. Liu, X.L. Qiu, C.F. Sun. Design and

application of knowledge model and PDA-based

precision farming system. Transactions of the

CSAE, Volume No. 26, Issue No.1, 210-215, 2010.

[15] Q. Wang, C. Qi, X.M. Luo. Modified LZW

algorithm and its parameters optimization, Journal

of Chongqing University of Posts and

Telecommunications, Volume No. 3, Issue No.1,

85-91, 2009.

[16] X.M. He. Lossless data compression and

decompression software and hardware realization,

Thesis of Degree of Master of University of

Electronic Science and Technology of China, 2011.

[17] Z. Liu, D.D. Yand, M. Zhu. Comparison of HDR

Image Compression Algorithms, Imaging Science


Data Types



and Photochemistry, Volume No. 29, Issue No.1,

32-44, 2011.

[18] J. Kuang, G.M. Johnson, M.D. Fairchild. iCAM06:

A Refined Image Appearance Model for HDR

Image Rendering, Journal of Visual

Communication and Image Representation, Volume

No. 18, Issue No.5, 406-414, 2007.

[19] F. Durand, J. Dorsey. Fast Bilateral Filtering for the

Display of High Dynamic Range Images, ACM

Transactions on Graphics, Volume No. 21, Issue

No.3, 257-266, 2007.

[20] A.N. Tang, C.C. Wu, L.H. Zheng. Wireless Data

Transmission Technology for Mobile Agriculture

Terminal, Transactions of the Chinese Society for

Agricultural Machinery, Volume No. 40, Specil

Issue, 244-247, 2009.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1151-1157


Numerical Investigation on Effect of Pile Tip Shape on Soil

Crushing Behaviour

YANG WU

Graduate School of Science and Engineering, Yamaguchi University, Ube 7558611, JAPAN


Keywords: Pile, Particle crushing, Finite element method, Plastic work, Breakage factor.

1. Introduction:

The study of the mechanism for bearing capacity of pile

in sand is complicated by the rearrangement and

degradation of the sand structure under crushing. The

majority of the uncertainty is caused by a lack of

understanding of the physical mechanism that controls

the characteristics of deformation, strain and stress in

soil during pile loading. It was reported by Miura and

Yamamouchi (1977) that significant particle crushing

had been observed around pile tip region where

complicated stress existed. Accurate estimation of the

bearing capacity of pile penetrating into sand ground

relies on the reasonable consideration of strength

reduction of sand around pile tip owing to the particle

crushing. Well understanding of the crushing behaviour

of sand in surrounding of pile is beneficial to capture the

bearing mechanism of pile loading process.

Many previous studies on the relationship between the

crushing of soil and bearing capacity of pile in sand as

well as behaviour of sand surrounding pile tip have been

acquired. Simonini(1996) firstly correlated the breakage

factor Br with the constitutive model for sand. That

model had been employed to represent the mechanical

behaviour of the particle-crushing region around the pile

and to predict the development of the distribution of

breakage factor Br. Yasufuku and Hyde (1995) also

discussed the relationship between the bearing capacity

of pile and the load-displacement relationship in sand

considering the compressibility of sand. Zhang et al.

(2013) adopted a breakage mechanics model accounting

for the evolution of grain size distribution and combined

it with a new formula to simulate the end-bearing

capacity of pile penetrating into crushable soil. Until

now, all the numerical analysis focuses on the straight

pile penetrating into sand. In recent two decades, the

enlarged-base pile (underreamed pile) has gained

popularity in the economical aspect. Wu et al. (2013a),

Wu and Yamamoto (2014) implemented the numerical

analysis and confirmed the effect of pile tip shape on the

development of strain and stress contours around pile

tip. However, limited study has been conducted to

discuss the influence of pile tip shape on the crushing

behaviour of sand around pile.

The aim of this study is to predict the influence of pile

tip shape on evolution of breakage factor around pile

Abstract: Accurate estimation of the bearing capacity of pile penetrating into sand ground relies on the reasonable

consideration of strength reduction of sand around pile tip owing to the particle crushing. Crushing degree of sand is

dependent on not only its inherent features such as hardness, grain diameter size and relative density but also on the

pile tip shape which sand particles interact with. To investigate the influence of pile tip shape on the bearing capacity

of pile, model pile loading test with different pile tip shapes are performed under three surcharge pressure levels. To

understand the deformation of sand and its interactive behaviour with pile, finite element method incorporating the

mixed incremental method for the Updated Lagrangian method is employed to simulate the pile penetrating process.

The mechanical behaviour of sand in model ground and interactive region between sand and pile are represented by a

simple constitutive model for sand considering particle crushing and joint element, respectively. This study presents

a numerical study on the effect of pile tip shape on the deformation and crushing behaviour of sand in surrounding of

the pile. The predicted relationship between normalized bearing capacity and displacement shows good agreement

with the experimental measurements. Breakage factor is correlated with the plastic work based on plenty of

experimental data on granular materials and calculated for piles with different tip shapes respectively. The

distributions of plastic work and breakage factor around pile with variable convergent angle are displayed and

compared under different surcharge pressure levels. Numerical results demonstrate that degree of sand crushing

around pile decreases as the pile tip becomes sharper. Three sand elements, which are selected at different depths

beneath the pile tip, are predicted to exhibit distinctive crushing behaviour.


1152 Numerical Investigation on Effect of Pile Tip Shape on Soil Crushing Behaviour



from the viewpoint of energy consumption. A mixed

incremental method for the UL method is used in the

finite element analysis combined with particle crushing

model to solve the large deformation problem. To better

simulate the interface behaviour, joint element,

simulating the soil-pile interaction, is employed in the

finite element analysis. The distribution of plastic work

and breakage factor are displayed for the model pile

with three different convergent angles respectively. The

predicted breakage factor plotted against the surcharge

pressure of the three elements beneath pile tip at

different depths under variable convergent angles and

surcharge pressures are also represented.

2. Description of model pile test with different pile

tip shapes:

Underreamed piles are usually made by enlarging the

diameter of pile base and hence provide greater bearing

capacity at a more economical cost than a straight-shaft

pile. Yamamoto et al. (2003) performed model pile

loading tests to understand the effect of pile tip shape on

the pile bearing mechanism. Three model piles adopted

in test owned the same shaft diameter as 30 mm and

enlarged-base diameter as 54 mm but differed with

convergent angle in Figure 1. In the model test, the

convergent angle between the axial line and base line

of base is defined to describe the shape of model pile

tip. The smaller the convergent angle is, the shaper the

tip of model pile becomes.

Model ground tank is filled with Toyoura sand, one kind

of uniform fine sand with mainly sub-angular particle

shape. The relative density of dry, dense, uniform sand

is approximately 90%.The mean grain size of Toyoura

sand is approximately 0.2 mm. Its maximum and

minimum densities are 1.332 g/cm3 and 1.646 g/cm

3,

respectively. The specific gravity of Toyoura sand is

2.65.

Model pile was driven into model ground tank by a

displacement-controlled method. The penetrating rate of

the model pile into the sand ground was 0.5 mm/min.

The maximum displacement of the model pile was

extended to a depth of 54 mm. Three surcharge pressure

levels, 200 kPa, 400 kPa and 600 kPa, was directly

applied to the upper surface of the model ground tank to

simulate actual soil stresses at different depths. Teflon

sheets were pasted to the interface between the model

ground tank and the Toyoura sand and between the

model pile and the Toyoura sand using silicon grease to

eliminate the effect of frictional force. The penetration

depth of the model pile and the load cell on the pile

head were recorded by a data logger and computer.

Figure 1: Model pile with different pile tip shape

(a) Model pile (b) Meshing

Figure 2: Outline of model pile test and meshing

It is very challenging work to simulate the pile loading

and penetrating process because it involves both

geometrical and material nonlinearities. More accurate

and reliable numerical solutions are obtained and

provide insight into the essence of the pile loading

problem by large deformation theory. The geometrical

nonlinearity has been solved by the mixed incremental

method for Updated Lagrangian (UL) method (Chen

and Mizuno (1990)) because that the stress and strain

integrations can be performed in a manner similar to

that used for the small deformation problem. The

material nonlinearity is solved by employing the

constitutive model considering particle crushing and

joint element. Pile is generally modeled as an

axisymmetric structure for analytical convenience. The

entire analysis area is the model ground tank, 300 mm

radially and 550 mm high, as shown in Figure 2 (a). The

surcharge pressures acting on the upper surface of the

model ground tank simulate the actual soil stress at

Different pile tip shape

90o 60o 30o

(mm)

300

550

(Joint element)

Toyoura

Sand

(Triangular

element)

Pile

3. Numerical preparation work:

1153 YANG WU



different depths. The initial lateral earth pressure

coefficient Ko is assumed to be 0.5.

To describe the sand crushing behaviour around the pile

tip, a simple constitutive model for sand with particle

crushing, proposed by Yao et al. (2008), is employed in

the finite element analysis. The constitutive model can

predict the positive to negative dilatancy behaviour of

sand under low confining pressures but can only predict

negative dilatancy under high confining pressures. This

constitutive model demonstrates the reduction in the

strength of sand with increasing confining pressures.

The constitutive model can be simply integrated in the

numerical computation. The joint element is applied to

simulate the interface behaviour during the pile loading

process. The joint elements are placed on the pile shaft

and the pile tip. For calculation simplicity, the meshing

is finished using three nodal triangular elements. The

parameters for constitutive model with particle crushing

and joint model are shown in Table 1. The triangular

element meshing for the model pile with convergent

angle as 30 degree is shown in Figure 2 (b).

The numerical calculations are conducted using the

displacement-controlled method. The displacement

control points are at the bottom of the pile tip and all

control points descend simultaneously. The

displacement is added by equivalent increments. The

final displacement is equal to the shaft diameter of the

pile in each calculation step.

4. Numerical results and discussions:

To eliminate the effect of pile diameter size, the

predicted relationships between normalized bearing

stress and normalized displacement are compared with

experimental results in this study. Numerical results by

0.0 0.5 1.00

50

100

150

No

rmal

ized

bea

rin

g s

tres

s

Normalized displacement (S/D)

Test 200 kPa

Predicted 200 kPa

Test 400 kPa

Predicted 400 kPa

Test 400 kPa

Predicted 600 kPa

Figure 3: Relationship between normalized bearing

stress and normalized displacement for pile with

convergent angle as 90 degree

Table 1: The model parameters for sand crushing and

joint elements (Uesugi and Kishida (1986), Yao et al.

(2008))

Joint

model

Crushing model

Triaxial

compression

Isotropic

consolidation

Poisson

ratio

Wu and Yamamoto (2013b) demonstrated that the

confinement effect from the interior surface of model

tank container can be neglected in this study. Herein,

the normalized bearing stress is defined as the ratio of

the current bearing stress at the pile tip to the surcharge

pressure. The normalized displacement is often used in

analyses, as defined by the ratio of the current

displacement (S) to the pile shaft diameter (D). Figure 3

shows the experimental and numerical normalized

bearing stress plotted against the normalized

displacement for model pile with convergent angle as 90

degree.

The predicted values agree well with the experimental

results during the entire loading process under each

surcharge pressure. However, all the predicted values

slightly overestimate the actual experimental results.

The normalized bearing capacity decreases as the

surcharge pressure increases. It is similar to the

phenomena that the peak strength of sand specimen

reduces in triaxial shear test as confining pressure is

improved. The strength reduction can be explained as

that particle crushing occurs near to the pile tip as the

surcharge pressure is enlarged.

4.2 Effect of pile tip shape on the distribution of

plastic work:

The occurrence of particle crushing accompanies with

the interior energy consumption. Particle crushing can

be regarded as the process that increasing surface area

and fine content of specimen are generated owing to the

extremely high input work. The ratio of incremental

surface area by the incremental plastic work dissipation

had been validated by Miura and Yamamouchi (1977)

as an effective index to represent the particle crushing.

However, breakage factor proposed by Hardin (1985)

was a more direct index commonly employed to

describe the degree of particle crushing. This study

intends to quantitatively express the degree of particle

crushing from the viewpoint of energy consumption.

Figure 4, Figure 5 and Figure 6 show the distribution of

plastic work of Toyoura sand around pile with

convergent angles as 90, 60 and 30 degree respectively.

3/sK =300N cm

5 3/cK =10 N cm

= 0.45

tC 0.0044

eC 0.0016

m 0.5

n 0.085

5.8 5cp MPa

M 1.35

v 0.3

4.1 Relationship between the normalized bearing

stress and displacement:




(a) 200 kPa (90 degree) (b) 600 kPa (90 degree)

Figure 4: Distribution of plastic work of sand around

pile with convergent angle as 90 degree







10-2

100

102

104

106

108

0.0

0.2

0.4

0.6

0.8

1.0

Bre

akag

e fa

cto

r

Plastic work Wp/kPa

Br=W

p

0.65/(W

p

0.65+550)

Figure 7: Relationship between breakage factor and

plastic work for Toyoura sand


Figure 8: Contour of breakage factor Br of sand around



Figure 9: Contour of breakage factor Br of sand around



Figure 10: Contour of breakage factor Br of sand

around pile with convergent angle as 30 degree

(J/m3) (J/m3)

(J/m3) (J/m3)

0.10

0.05 0.05

0.10

0.15

0.05

0.10

0.15

0.05

0.15

0.10

0.05

0.10

0.05

0.10

(J/m3) (J/m3)

1155 YANG WU



Predicted results represent that the distributed range of

high-value plastic work around pile significantly shrinks

as the convergent angle decreases. The high-value

plastic work concentrates at the corner point of the pile

tip and extends towards to the central line of model pile.

It is believed that the sand particle around that corner is

subjected to complicated stresses. To examine the

surcharge pressure effect on predicted result, a figure

includes two sub-titles (a) and (b) displaying the

distribution of plastic work for Toyoura sand when the

surcharge pressure are 200 kPa and 600 kPa. The

distributed range also expands in radial and vertical

direction with increasing surcharge pressure by

comparing (a) with (b) in each figure.

The experienced relationship between the breakage

factor and the plastic work is summarized by Hu et al.

(2011) from the experimental data on many kinds of

granular materials. The empirical equation for Toyoura

sand is shown in Figure 7. The breakage factor slightly

increases when plastic work is less than 102 J/m

3/kPa

and drastically increase as plastic work is over 103

J/m3/kPa. Crushing becomes sufficient once the plastic

work beyond 106 J/m3/kPa.

4.3 Effect of pile tip shape on the distribution of

breakage factor:

To investigate the effect of pile tip shape on the

crushing behaviour at different depths beneath the pile

tip, the breakage factor of the elements beneath the pile

tip at different depths are displayed and analyzed. Pile

tip bearing stress is estimated by the factor of the

sand layer beneath the pile tip. Yang (2006)

demonstrated that the influential zone beneath the pile

tip is approximately 2 and 3 times larger than the pile

diameter. Therefore, three elements at one and two

diameter length beneath pile tip are selected in Figure

11. Relationship between the breakage factor and

surcharge pressure for model pile with different

convergent angles as 90, 60 and 30 degree are shown in

Figure 11: Three elements beneath pile tip at different

depths

0 200 400 600 8000.076

0.080

0.084

0.088

0.092

0.096

0.100

Bre

akag

e fa

cto

r B

r

Surcharge pressure (kPa)

90 degree

60 degree

30 degree

(a) Breakage factor of element 1 beneath pile tip

0 200 400 600 8000.076

0.080

0.084

0.088

0.092

0.096

0.100

Bre

akag

e fa

ctor

Br


90 degree

60 degree

30 degree

(b) Breakage factor of element 2 beneath pile tip

0 200 400 600 8000.076

0.080

0.084

0.088

0.092

0.096

0.100

Bre

akag

e fa

cto

r B

r


90 degree

60 degree

30 degree

(c) Breakage factor of element 3 beneath pile tip

Figure 12: Crushing behaviour of sand beneath pile tip

at different depths for pile with three kinds of pile tip

shape

1 2

3

D

D

D

Figure 8, Figure 9 and Figure 10 display the breakage

factor contour of Toyoura sand around pile with

convergent angles as 90, 60 and 30 degree. The

crushing degree of Toyoura sand decreases as

convergent angle of pile becomes small. It is believed

that particles beneath flat base pile have less space to

move or rotate compared those beneath the pencil type

pile. The maximum breakage factor is 0.15 for pile with

convergent angle as 90 at surcharge pressure as 600 kPa

but it reduces to 0.10 for pile with convergent angle as

30. Similar to the distributed pattern of plastic work, the

high-value degree of particle crushing occurs at the at

the corner point of the pile tip. It displays the irregular

bulk shape. The distributed range of breakage factor

expands with increasing surcharge pressure.




Figure 12 (a), (b) and (c) respectively. Element 1 is on

the central line two diameter length beneath the pile tip.

Element 2 and 3 are two and one diameter length

beneath the corner point of pile tip in vertical direction.

It is represented that the predicted breakage factor

decreases as the convergent angle decreases. The

breakage factor reduces from 0.90 to 0.845 for element

1 as the convergent angle varies from 90 degree to 30

degree. Degree of particle crushing is affected by the

pile tip shape even the selected position changes. It is

also noticeable that breakage factor increase with the

increasing surcharge pressure. However, the increasing

tendency is less dependent on surcharge pressure level.

5. Conclusions:

To investigate the effect of pile tip shape on crushing

behaviour of sand around a pile, finite element analysis

has been presented that incorporates the mixed

incremental method for UL method and the

characteristics of particle crushing. The penetration of

model pile with three kinds of pile tip shape under

different surcharge pressures are analyzed respectively.

The distribution of breakage factor is predicted from the

viewpoint of energy consumption. The major findings

of the study are summarized below.

1. The predicted relationship between normalized

bearing stress and normalized displacement of model

piles agrees with the experimental results.

2. The distributed range of plastic work shrinks as the

convergent angle of pile becomes small. The high-value

plastic concentrates at the corner point of the pile tip

and extends towards to the central line of model pile.

The distributed range of plastic work also expands in

radial and vertical direction with increasing surcharge

pressure.

3. The distributed range of breakage factor deceases as

the convergent angle of pile decreases. The high-value

degree of particle crushing occurs at the at the corner

point of the pile tip. The distributed shape of breakage

factor displays the irregular bulk shape.

4. The breakage factor of three selected elements

beneath pile tip all show decreasing tendency as

convergent angle decreases. Degree of particle crushing

is affected by the pile tip shape even the selected

position changes. It is also noticeable that breakage

factor increase with the increasing surcharge pressure.

This study is the preliminary study of the crushing

degree of sand in surrounding of pile. This study can be

improved by some aspects. The particle crushing owing

to the contact between sand and pile should be further

discussed and more advanced interface element should

be employed in further study.

6. Acknowledgement:

This research is supported by the scholarship of the

Ministry of Education, Culture, Sports, Science and

Technology, Japan. The author expresses thanks to Prof.

Haruyuki YAMAMOTO in Hiroshima University for

his kind support. The editor-in-chief Prof. D. V. Reddy

and reviewers are especially thanked for their kind work

and constructive suggestions.

7. Reference:

[1] W. F. Chen and E. Mizuno (1990) “Nonlinear

analysis in soil mechanics: Theory and

Implementation”, Nertherlands, ELSEVIER, pp.

492-504.

[2] B. Hardin (1985) “Crushing of soil particles”,

Journal of Geotechnical Engineering, 111(10), pp.

1177-1192.

[3] W. Hu, Z. Y. Yin, C. Dano and P. Y. Hicher (2011)

“A constitutive model for granular materials

considering grain breakage”, Science China

Technological Sciences, 54(8), pp. 2188-2196.

[4] N. Miura and T. Yamamouchi (1977) “Effect of

particle-crushing on the shear characteristics of a

sand”, Proceeding of Japan Society of Civil

Engineering, 260, pp. 109-118. (In Japanese)

[5] P. Simonini (1996) “Analysis of behavior of sand

surrounding pile tips”, Journal of Geotechnical and

Geoenvironmental Engineering, 122(11), pp. 897-

905.

[6] M. Uesugi and H. Kishida (1986) “Influential

Factors of friction between Steel and Dry sands”,

Soils and Foundations, 26(2), pp. 33-46.

[7] Y. Wu, H. Yamamoto and Y. P. Yao (2013a)

“Numerical study on bearing behavior of pile

considering sand particle crushing”, Geomechanics

and Engineering, 5(3), pp. 241-261.

[8] Y. Wu and H. Yamamoto (2013b) “Numerical

analysis on effects of confinement and surcharge

pressure on the behaviour of sand surrounding

pile”, International Journal of Earth Sciences and

Engineering, 6(6), pp. 1472-1482.

[9] Y. Wu and H. Yamamoto (2014) “Numerical

Analysis of the Effect of Pile Tip Shape on Soil

Behavior around Pile”, Geotechnical Engineering

Journal of the SEAGS & AGSSEA, 45(2), pp.78-

89.

[10] H. Yamamoto, W. Li, K. Tominaga and H. Ogura

(2003) “Experiment study on effects of

overburdening pressures and end shapes for point

bearing capacities of pile”, Structure and

construction of proceeding of Japan Architecture,

49(B), pp. 157-162. (In Japanese)

[11] J. Yang (2006). “Influence zone for end bearing of

piles in sand”, Journal of Geotechnical and

Geoenvironmental Engineering-ASCE, 132(9), pp.

1229-1237.

1157 YANG WU



[12] Y. P. Yao, H. Yamamoto and N. D. Wang (2008)

“Constitutive model considering sand crushing”,

Soils and Foundations, 48(4), pp. 603-608. N.

[13] N. Yasufuku and A. F. L. Hyde (1995) “Pile end-

bearing capacity in crushable sands”,

Geotechnique, 45(4), pp. 663-676.

[14] C. Zhang, G. D. Nguyen and I. Einav (2013) “The

end-bearing capacity of piles penetrating into

crushable soils”, Geotechnique, 63(5), pp. 341-354.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1158-1163


Estimation of Reservoir Capacity Using Remote Sensing Data – A

Soft Classification Approach

JEYAKANTHAN V S Deltaic Regional Centre, National Institute of Hydrology, Siddartha Nagar, Kakinada-533 003,

Andhra Pradesh, India


Abstract: Sediments carried by the rivers are deposited in the reservoirs and cause several detrimental effects,

which include loss of storage capacity, upstream aggradations, effect on water quality and damage or impairment of

hydro-equipments. The deposition of sedimentation not only reduces the capacity but also the water-spread of the

reservoir. Satellite data has long been in use to estimate the water-spread area at different elevations of a reservoir,

which in turn can be used to quantify the capacity of the reservoir. This methodology to estimate the capacity of a

reservoir using remote sensing data involves hard or per-pixel based classification to delineate the water-spread area

at a particular elevation. One of the limitations of this approach is that the border pixels, representing soil class with

moisture, are classified entirely as water pixels, thereby giving inaccurate estimate of the water-spread area. To

estimate the water-spread area of Nagarjuna Sagar reservoir accurately, the sub-pixel or linear mixture model

(LMM) approach has been adopted in this study. The sub-pixel approach uses a linear mixture model of the spectra

to quantify the proportions of water and other classes present in each of the border pixels of the reservoir. IRS-1C

and 1D satellite image data (24m) of eight optimal dates ranging from minimum draw down level (MDDL) to full

reservoir level (FRL) were used to estimate the water-spread area of the reservoir. The extracted water-spread areas

using per-pixel and sub-pixel approaches were in turn used to quantify the capacity of the Nagarjuna Sagar

reservoir. The estimated capacity of the Nagarjuna Sagar reservoir using per-pixel and sub-pixel approaches was

8101.63 Mm3 and 8014.49 Mm3 respectively.

Keywords: Reservoir, Water-spread area, Capacity estimation, Per-pixel and Sub-pixel approaches.

Introduction:

Natural processes, such as erosion in the catchment

area, movement of sediment and its deposition in

various parts of the reservoir, require careful

consideration in the planning of major reservoir

projects. The silt that is deposited at different levels

reduces the storage capacity of the reservoir [1, 2].

Reduction in the storage capacity beyond a limit

prevents the reservoir from fulfilling the purpose for

which it is designed. Periodic capacity surveys of the

reservoir help to assess the rate of sedimentation and

reduction in storage capacity. Conventional techniques

for the estimation of the capacity of a reservoir, such as

hydrographic survey and inflow-outflow approaches,

are cumbersome, time consuming and expensive, and

they involve significant manpower. As an alternative to

conventional methods, the remote sensing technique

provides cost- and time-effective estimation of the live

capacity of a reservoir [3, 4]. Multi-date satellite remote

sensing data provide information on elevation contours,

in the form of water spread area, at different water

levels of a reservoir. The water spread area thus

interpreted from the satellite data is used as an input into

a simple volume estimation formula to calculate the

capacity of a reservoir. Such work has been reported by

[5] for the Malaprabha Reservoir in India, [6] for the

Poondi Reservoir in India, [7] for the Bargi Reservoir in

India and [4] for the Bhakra Reservoir in India

For quantification of the capacity of a reservoir, the only

thematic information that has to be extracted from the

satellite data is the water-spread area at different water

levels of the reservoir [8, 9]. The different approaches

such as maximum likelihood, minimum distance to

mean classification and band threshold method, to

delineate various thematic information from the remote

sensing digital data adopt the per-pixel based

methodology and assign a pixel to a single land cover

type [10, 11] whereas in reality, a single pixel may

contain more than one land cover (known as a mixed

pixel). Mixed pixels are common especially near the

boundaries of two or more discrete classes [12, 13]. The

boundary pixels of the water-spread area that are mixed

in nature, representing soil, vegetation class with

moisture are also classified as water pixels when a per-

pixel based approach is applied, thereby producing

inaccurate estimate of the water-spread area. To

1159 JEYAKANTHAN V S



accurately compute the water-spread area to the

maximum possible extent, thereby reducing the error in

the estimation of capacity of a reservoir, a sub-pixel or

linear mixture model (LMM) approach has been chosen

for classifying the boundary pixels of water-spread area

from different water levels of Nagarjuna Sagar reservoir

located in Andhra Pradesh state of India.

Study Reservoir 1

Nagarjuna Sagar is one of the world's largest masonry

dam built across the river Krishna in Nagarjuna Sagar

town, Nalgonda District of Andhra Pradesh, India,

between 1955 and 1967. The capacity of Nagarjuna

Sagar reservoir at FRL (Full Reservoir Level) during

impoundment was 11,472 million cubic metres. The

dam is 490 ft (150 m) tall and 1.6 km long with 26 gates

which are 42 ft (13 m) wide and 45 ft (14 m) tall. The

geographic region pertains to the reservoir experiences

hot and dry summer throughout the year except during

the South-west monsoon season. The monsoon season is

from June to September and retreating monsoon or the

post monsoon season is during October to November.

The average rainfall in the district is 772 mm. 71% of

the annual rainfall is received during south-west

monsoon. The mean daily maximum temperature is

about 40o

C and the mean daily minimum is about 28o

C. The soils in and around the study reservoir mainly

comprise loamy sands, sandy loams and sandy clay

loams. In the areas of flat topography and alongside the

river Krishna and its tributaries comprises mainly of

black cotton soil.

Satellite Data Used

The image data used in this study were obtained by the

Indian Remote Sensing (IRS) satellites IRS-1C & 1D

(LISS-III sensor) which provides a spatial resolution of

24 m and spectral resolution in four different bands

(0.52-0.59, 0.62-0.68, 0.77-0.86, 1.55-1.70 μm). The

different dates of satellite data used and the respective

water level during the pass of the satellite over the

reservoir are given in Table 1. Reservoir water level

data and the hydrographic survey details have been

collected from the Nagarjuna sagar reservoir authority

responsible for the maintenance and operation of the

reservoir.

Table. 1 Details of satellite data used and the water

level during the pass of the satellite over the reservoir

Sl.No. Name of

the

Satellite

Date of

Satellite

Pass

Reservoir

Elevation

(m)

1. IRS-1C 23.10.2001 175.32

2. IRS-1D 12.12.2001 166.68

3. IRS-1D 24.12.2001 164.74

4. IRS-1D 25.02.2002 157.03

5. IRS-1D 14.09.2002 156.94

6. IRS-1D 11.05.2002 154.72

7. IRS-1D 27.11.2002 153.19

8. IRS-1D 22.12.2002 152.28

Methodology

The changes in water-spread could be accurately

estimated by analyzing the areal spread of the reservoir

at different elevations over a period of time using the

satellite image data [9,1]. Both the per-pixel and sub-

pixel or Linear Mixture Model (LMM) approaches have

been used in this study to extract the water-spread area

of the reservoir. Estimated water-spread areas were used

in a simple volume estimation formula to compute the

storage capacity of the reservoir. Estimation of water-

spread area and computation of capacity of the reservoir

using per-pixel and sub-pixel approaches are discussed

in the following sections.

Geo-Referencing Of Satellite Data

In the IRS-1C and 1D (LISS-III) satellite data the

reservoir water-spread area was free from clouds and

noise in all of the eight images used. The image scene of

23rd

October 2001 was geo-referenced with respect to

1:50,000 survey of India topographic maps. The geo-

referencing was done using ployconic projection and

nearest neighborhood re-sampling technique to create a

geo-referenced image of pixel size 24m x 24m.

Subsequently the other satellite images corresponding to

various water levels were also registered with the geo-

referenced image using the image to image registration

technique. In every image 20 to 25 ground control

points were used, which resulted in a root mean squared

error (RMSE) of 0.15 to 0.18 of a pixel.

3.2 Per-pixel based approach

The most widely used method for extracting

information on the surface cover from remotely sensed

data is image classification. Digital image classification

uses the spectral information represented by the digital

numbers in one or more spectral bands and attempts to

classify each individual pixel (per-pixel approach) based

on this spectral information. A number of commonly

used per-pixel based classifiers exist including the

maximum likelihood, the minimum distance to mean,

mahalanobis distance and the parallelipiped classifiers.

A detailed account of these and the other classifiers can

be found in [14, 15]. The objective of the per-pixel

based classifiers is to assign all pixels in the image to

particular classes or themes (e.g. water, forest, urban,

agriculture). The resulting classified image is comprised

of a mosaic of pixels, each of which belongs to a

particular theme and is essentially a thematic map of the

original image. The per-pixel based classification can be

divided into unsupervised and supervised classification.

The underlying requirement of supervised classification

1160 Estimation of Reservoir Capacity Using Remote Sensing Data –

A Soft Classification Approach



techniques is that the user/analyst has sufficient pixels

(training sets) available for which the true class is

known. From these pixels, representative signatures can

be developed for those classes for subsequent

classification.

The steps involved in supervised classification include

(i) identification and decisions relating to the set of

classes/cover types into which the image is to be

segmented, (ii) choice of the representative pixels for

each of the desired set of classes (i.e., construct the

training data), (iii) estimation of the parameters (eg.

class signatures) of the classification algorithm from

the training data, (iv) use of the classifier to label (or

‘classify’) every pixel in the image into one of the

desired/predetermined cover types, (v) production of

thematic maps which contain the results of the

classification, and (vi) determination of the accuracy of

the classification by comparing the output to ground

truth data. In this study, water-spread area was extracted

using the most commonly adopted maximum likelihood

classification approach. The estimated water-spread was

in-turn used in a volume estimation formula to compute

the capacity of the reservoir.

Sub-pixel based approach:

The basic assumption of linear mixture model is that the

measured reflectance of a pixel is the linear sum of the

reflectance of the components that make up the pixel.

The basic hypothesis is also that the image spectra are

the result of mixtures of surface materials, shade and

clouds, and that each of these components is linearly

independent of the other [16, 17]. Linear unmixing also

assumes that all the materials within the image have

sufficient spectral contrast to allow their separation. In

soft classification, the estimated variables (the fractions

or proportions of each land cover class) are continuous,

ranging from 0 to 100 percent coverage within a pixel.

[18] and [12], proposed a mathematical expression for

linear spectral unmixing. The theory behind this is the

contribution of a series of end-members present within a

pixel to its spectral signature. Hence, the spectral

signature of a pixel would be derived from the sum of

the products of the single spectrum of the end-members

it contains, each weighted by a fraction plus a residue

which would be explained by the following

mathematical model:

Ri = ∑ fk Rik + Ei (1)

Where, ∑ fk = 1 (2)

and 0 ≤ fk ≤ 1 (3)

i = 1. . . m (number of spectral bands)

k = 1,. . ., n (number of endmembers)

Ri = Spectral reflectance of band i of a pixel which

contains one or more endmembers

f k = Proportion of endmember k within the pixel

Rik = Known spectral reflectance of endmember k

within the pixel on band i

Ei = Error for band i (Difference between the observed

pixel reflectance Ri and the reflectance of that pixel

computed from the model).

Equations 1 and 2 introduce the constraints that the sum

of the fractions is equal to one and they are non-

negative. To solve fk, the following conditions must be

satisfied: (i) selected end members should be

independent of each other, (ii) the number of

endmembers should be less than or equal to the spectral

bands used, and (iii) selected spectral bands should not

be highly correlated.

In this study, the linear spectral unmixing is adopted

based on the equations described below to segregate the

actual information within a pixel of an image

R1 = Fwater * R1water + FVeg* R1 Veg + FSoil*R1 Soil + ε1

R2 = Fwater * R2 water + FVeg* R2 Veg + FSoil*R2 Soil + ε2 (4)

R3 = Fwater * R3 water + FVeg* R3 Veg + FSoil*R3 Soil + ε3

Where,

R1, R2 and R3 represent the signal recorded at the

satellite in the green, red and NIR bands of the

LISS-III sensor.

Fwater , FVeg and FSoil are the fraction of the pixel

covered by water, vegetation, and soil.

R1water, R2water and R3water represent the reflectance

of water in each of the three spectral bands.

R1 Veg, R2 Veg and R3 Veg represent the reflectance of

vegetation in each of the three spectral bands.

R1 Soil, R2 Soil and R3 Soil represent the reflectance of

soil in each of the three spectral bands.

ε1, ε2 and ε3 are the error components of band 1, 2,

and 3.

The system of linear equations shown above can be

solved by a least square solution which minimizes the

sum of squares of errors.

The sub-pixel based approach was applied to find out

the proportion or fraction of water class that exits in the

periphery pixels of the reservoir. The first step executed

in the sub-pixel approach was, selection of end-

members. In general the border pixels may contain any

combination and proportions of water, vegetation and

soil classes, therefore these three classes were chosen to

collect the end-members. Scatter plot method was used

to identify the end-members. The identified end-

member spectra were supplied as input to the linear

mixture model (LMM) approach. The output of the

model run contains three images known as water, soil

and vegetation fraction images. Description of the

fraction images is given in section 4.2.




Computation of volume between successive water

Levels

Traditionally the reservoir volume between two

consecutive reservoir water levels, was computed using

the prismoidal formula, the Simpson formula and the

trapezoidal formulae [19]. Of these, the trapezoidal

formula has been most widely used for computation of

volume [20, 21]. The water-spread area estimated using

sub-pixel approach was used as an input in the volume

estimation formula to find out the, volume at different

water levels of the reservoir. In this study the volume

between two consecutive reservoir water levels was

computed using the following trapezoidal formula.

Trapezoidal Formula:

V = H/3 (A1+A2 + (√A1*A2)) (5)

Where V is the volume between two consecutive water

levels. A1 and A2 are the water-spread area at the

reservoir water level 1 and 2 respectively and H is the

difference between these two water levels.

Computation of storage capacity of the reservoir:

The volume computed between different water levels

(i.e from Minimum Draw Down Level to Full Reservoir

Level) was added up to calculate the cumulative or

storage capacity of the reservoir.

Results and discussion

Computation of reservoir capacity by per-pixel

approach

The supervised classification was performed by using

training sets generated for the different landcover

classes in and around the reservoirs. Since this exercise

involves the estimation of pixels that are occupied by

water, it is essential that the training set generation

should involve careful selection of water pixels. It may

note that the image subset of the study area contains

vegetation and soil as the primary land cover classes in

addition to the water class. Therefore only three

feature/landcover classes were considered to generate

training sets for the maximum likelihood classification.

Training sets were selected from the center of the

reservoir, periphery of the reservoir, from the

agricultural fields, shrubs in the peripheral areas and

from barren lands representing the soil class. The choice

of representative pixels (both number and type) was in

accordance with the suggestion of [14]. This procedure

was followed carefully for all the eight images

pertaining to the Nagarjuna Sagar reservoir. It may be

noted that three simple and spectrally distinct classes

namely water, soil and vegetation form the basic input

for the classification process. The extracted water

spread areas and the estimated capacity (8101.63 Mm3)

using per-pixel approach is given in Table 2.

Figure 1 (a) Feature space plot (NIR vs RED), (b) End-member spectra of Soil, Water and Vegetation (c) Fraction

images obtained by spectral un-mixing of image data of Nagarjunasagar Reservoir pertaining to the water level

175.32 m (Due to the recurring nature of the fraction images, only figures pertaining to the highest water level is

presented).

1162 Estimation of Reservoir Capacity Using Remote Sensing Data –

A Soft Classification Approach



Table 2: Capacity estimation of Nagarjunasagar reservoir using the per-pixel and sub-pixel classification

approaches (2002)

Computation of reservoir capacity by sub-pixel

approach

The fraction images (Figure 1) generated using the sub-

pixel approach described in the methodology section

contain a wealth of information about the reservoir.

Each fraction image corresponds to a single land cover

only. For example, the pixels in the water fraction

image provide information only on the proportion or

amount of water it contains. Likewise the vegetation and

soil fraction image provide information on the

proportions of the respective classes only. However, in

this study the interest is only to know about, the amount

of water present in the border pixels of the reservoir.

The value of the pixels in the fraction image ranges

from 0 to 1. A pixel from the water fraction image

having a value of 0 indicates that there is no water at all

in that pixel, whereas a pixel having a value of 0.25

indicates that 25% of the area of the pixel is occupied by

water while a pixel value of 1 indicates that 100% of the

area of the pixel is occupied by water (i.e the pixel is

fully occupied by water). Therefore, for a pixel having a

value of 0.72, the area of water occupied by that pixel is

414.72 m2 (0.72 x 24m x 24m). The pixels representing

the reservoir border which have a minimum value of up

to 0.1 in the water fraction image (i.e a pixel contains a

minimum 10% of area of water) were isolated from the

water fraction image and the area covered by water in

these border pixels were estimated. The number of

pixels that contains 100% of water was also found out.

By summing up the area occupied by these two types of

pixels, the total water spread area, corresponding to a

particular water level of the reservoir was computed.

This exercise was carried out for all the eight images

used in this study. The water-spread area thus estimated

was again used as an input in the trapezoidal formula to

compute the storage capacity or cumulative capacity of

the Nagarjuna sagar reservoir using the sub-pixel

classification approach.

Here it is worth mentioning that, a pixel containing 71%

of water may be labelled as containing 100% of water

by the per-pixel approach. In such case the water-spread

area is over estimated. Conversely if the pixel contains

40% of water, then the entire pixel may not be

considered as a water pixel. Hence, the water-spread

area is under estimated. Such errors due to over

estimation or under estimation do not occur in the sub-

pixel approach. Thus, the sub-pixel approach reduces

the error imposed by the per-pixel approach. The

estimated cumulative capacity of Nagarjuna sagar

reservoir at the water level 175.32 m (Near FRL) using

the sub-pixel approach was 8014.49 Mm3. The capacity

estimated using per-pixel and sub-pixel approaches are

given in Table 2.

Conclusions

High spatial-resolution image data enables accurate

mapping of the terrain features. The use of high spatial

resolution satellite image data, however, is constrained

by factors such as cost and less area covered by the

sensor. Hence, in hydrological applications estimating

water-spread area may be difficult because a reservoir

may not be imaged in a single pass of the satellite and

atmospheric condition would be different from path to

path [22]. An alternative method to overcome such

constrains is the use of the sub-pixel based approach.

The simplest methodology for such an approach is

linear mixture modeling, which has been demonstrated

in this study to estimate the capacity of Nagarjuna sagar

reservoir in India. Though the sub-pixel approach found

to be a better alternative to per-pixel approach, there are

certain limitations such as the spatial location of the

fractions within a pixel is unknown. In addition the sub-

pixel classifier produces more accurate results only with

hyperspectral images. Hence, the use of hyperspectral

image data with higher spatial resolution would have

yielded better results.

Acknowledgement

Sl.

No

Date of

Satellite Pass

Reservoir

Elevation

above m.s.l

(m)

Waterspread

area - Perpixel

approach

(Mm2)

Waterspread

area - Subpixel

approach

(Mm2)

Cumulative

Volume -

Perpixel

approach

(Mm3)

Cumulative

Volume -

Subpixel

approach

(Mm3)

1. 23.10.2001 175.32 238.55 235.15 8101.63 8014.49 2. 12.12.2001 166.68 203.73 197.03 6192.96 6149.90 3. 24.12.2001 164.74 192.89 194.08 5808.28 5770.53 4. 25.02.2002 157.03 173.75 170.43 4395.53 4366.33 5. 14.09.2002 156.94 166.93 167.42 4380.20 4351.13 6. 11.05.2002 154.72 163.47 155.06 4013.46 3993.26 7. 27.11.2002 153.19 156.69 149.02 3768.56 3760.66 8. 22.12.2002 152.28 153.79 144.11 3627.29 3627.29




The author thanks the Nagarjuna Sagar reservoir

authorities for providing the ground based data. The

author also thanks Director, NIH for giving permission

to publish the paper.

References

[1] Smith, L.C., and Pavelsky, T.M, “Remote sensing

of volumetric storage changes in lakes”, Earth

Surface Process and Landforms, Vol.34, pp.1353

– 1358, 2009.

[2] Sreenivasulu, V and Udayabaskar, P, “An

integrated approach for prioritization of reservoir

catchment using remote sensing and geographic

information system”, Geocarto International,

Vol.25, pp.149-168, 2010.

[3] Sabastin, M., Rao, P.P.N., Jayaraman. V., and

Chandrasekhar, M.G, “Reservoir storage loss

assessment and sedimentation modeling through

remote sensing techniques”, Proceedings of 6th

Inter. Symp. River sedimentation on Management

of Sediment – Philosophy, Aims and Techniques,

New Delhi, India, 1995

[4] Jain, S.K., Singh, P., and Seth, S.M, “Assessment

of sedimentation in Bhakra Reservoir in the

western Himalayan region using remotely sensed

data”, Hydrological Sciences Journal, Vol.47,

pp.203-212, 2002.

[5] Manavalan, P., Sathyanath, P., Sathyanarayn, M.,

and Raje Gowda, G. L, “Capacity evaluation of

the Malaprabha reservoir using digital analysis of

satellite data”. Tech. Report no. RC: BO: WR:

001:90, Regional Remote Sensing Service Centre,

Bangalore and Karnataka Engineering Research

Station, Krishnarajsagar, India, 1990.

[6] Jeyakanthan V.S, Sreenivasalu V, Rao Y.R.S.,

Ramasastri K.S, “Reservoir Capacity Estimation

Using Satellite Data”, International Symposium on

Photogrametry and Remote Sensing, Hyderabad,

2002.

[7] Goel, M.K., Jain, S.K., and Agarwal, P.K.,

“Assessment of sediment deposition rate in Bargi

Reservoir using digital image processing”,

Hydrological Sciences Journal, Vol.47(S),

pp.S81-S92, 2002.

[8] Morris, G.L., and Fan Jiahua, Reservoir

Sedimentation Handbook, (New York: McGraw-

Hill Book Co), 1998.

[9] Peng Dingzhi, Shenglian Guo., Pan Liu, Ting Liu,

“Reservoir storage curve estimation based on

remote sensing data”, Journal of Hydrologic

Engineering, Vol.11, pp.165-172, 2006.

[10] Jensen R.J., 1996. Introductory Digital Image

Processing : A remote sensing perspective, 2nd

edition, (London: Prentice Hall International

Ltd.,).

[11] Bastin, L. “Comparison of fuzzy c-means

classification, linear mixture modelling and MLC

probabilities as tools for unmixing coarse pixels”,

International Journal of Remote Sensing, Vol.18,

pp.3629-3648, 1997.

[12] Foody, G. M., and Cox, D. P., “Sub-pixel land

cover composition estimation using a linear

mixture model and fuzzy membership functions”,

International Journal of Remote Sensing, Vol.l5,

pp.619-631, 1994.

[13] Oleson, K. W., Sarlin, S., Garrison, J., Smith, S.,

Privette, L., and Emery, W., “Unmixing multiple

land cover type reflectance’s from coarse

resolution satellite data”, Remote Sensing of

Environment, Vol.54, pp.981-912, 1994.

[14] Mather, P. M., and Koch, M, “Extension of the

technique of Principal Component Analysis for

use in end-member selection”, pp. 346-353, 1997.

[15] Swain, P. H., and Davis, P. H., (Eds.), Remote

Sensing: The Quantitative Approach, New York,

NY: McGraw-Hill, 1978.

[16] Quarmby, N. A., Townsend, J.G.R., Settle.J.J.,

White, K. H., Milnes, M., Hindle, T. L., and

Silleos, N., “Linear mixture modelling applied to

A VHRR data for crop area estimation”,

International Journal of Remote Sensing, Vol.l3,

pp.415-425, 1992.

[17] Borel, C. C., and Gerstl, S. A. W., “Non-linear of

spectral unmixing models for vegetative and soil

surface”, Remote Sensing of Environment,

Vol.47,pp. 403-416, 1994.

[18] Settle, J.J., and Drake, N.A., “Linear mixing and

the estimation of ground cover proportions”,

International Journal of Remote Sensing, Vol.14,

pp.1159-1177, 1993.

[19] Patra, K.C., Hydrology and Water Resources

Engineering, (New Delhi: Narosa Publishing

House), 2001.

[20] Goel, M. K., and Jain, S. K., “Evaluation of

reservoir sedimentation using multi-temporal IRS-

1A LISS II data”, Asian Pacific Remote Sensing

& GIS Journal, Vol.8, pp.39-43, 1996.

[21] Rathore, D.S., Anju Choudhary, and Agarwal,

P.K., “Assessment of sedimentation in Harakud

Reservoir using digital remote sensing technique”,

Journal of the Indian Society of Remote Sensing,

Vol.34, pp.344-383, 2006.

[22] Hung, M.C., and Wu. Y.H., “Mapping and

visualizing the Great Salt Lake landscape

dynamics using multi-temporal satellite images-

1972-1996”, International Journal of Remote

Sensing, Vol.26, pp.1815-1834, 2005.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1164-1170


Experimental Investigation on RC and Retrofitted RC Column under

Cyclic Loading

A MURUGESAN1 AND G S THIRUGNANAM

2

1Department of Civil Engineering, Sona College of Technology, Salem-636005, Tamil Nadu, INDIA

2Department of Civil Engineering, Institute of Road and Transport Technology, Erode-638316, Tamil Nadu, INDIA


Abstract: Columns are important structural elements in a multi storey building it transmitting the entire loads to the

foundation. For the purpose of wind or earthquake engineering, columns may be designed to resist lateral forces. If

the columns are subjected to lateral load due to wind/ earthquake, the load carrying capacity of the member is

substantially reduced. Hence the load carrying capacity of the compression member has to be increased. There are

number of repairing techniques are available like external post tensioning, base isolators, supplementary dampers,

tuned mass dampers, active control system, Adhoc addition of structural support/reinforcement, connections

between buildings and their expansion additions and also retrofitting techniques. In this study, load carrying capacity

is increased by the way of retrofitting the columns. This experimental study aims in assessing the behavior of such

reinforced concrete columns with retrofitting technique, to investigate the ductile behavior and strength properties of

RC and retrofitted RC columns for the effective performance of the same during earthquakes. The RC column and

retrofitted RC column are designated as RCC and RRCC. The size of the column RCC and RRCC are 120 mm X

230 mm X 600 mm and designed for M30 mix concrete. This paper presents the seismic performance of the

reinforced concrete column designed for seismic loads as per IS 1893-2002 with and without retrofitting of FRP and

ductility recommendations are adopted as per IS 13920-1993. Specimen are cast with and without retrofitting and

tested under cyclic loading, simulating earthquake actions. The columns are examined in terms of load carrying

capacity, load deflection behavior, energy dissipation capacity, ductility factor and cracking characteristics. Based

on the test result important conclusions are drawn and the advantages of retrofitting with FRP wrapped reinforced

concrete column have been established.

Keywords: RC Column, RCC, Retrofitted RC Column, RRCC, Cyclic Loading, FRP.

1. Introduction:

Columns are skeletal structural elements, whose cross-

sectional shapes may be rectangular, square, circular, L-

shaped, etc. The size of the column section is dictated,

from a structural viewpoint, by its height and the loads

acting on it-which, in turn, depend on the type of floor

system, spacing of columns, number of storey, etc. The

column is generally designed to resist axial compression

combined with (biaxial) bending moments that are

induced by ‘frame action’ under gravity and lateral

loads.

When a column is subjected to an axial load within

elastic limits, just like any other composite section, the

stresses induced in steel and concrete are in proportion

to their modulli of elasticity, Es and Ec, respectively.

The failure of the tied column occurs suddenly with the

breaking down of concrete cylinder in a compression

test. In the present day, the seismic zones are altered in

India and the quantity of seismic force level has been

increased. It is harmful to the existing building

structures which needs to be strengthened and

retrofitting of existing concrete buildings are most

significance. Several investigators conveyed their

experimental and/or theoretical investigations on

structural elements such as concrete columns and beams

with fibre reinforced polymer composites in order to

carry their effectiveness. Some of the prominent works

are Saadatmanesh and Ehsani1, Naani

2 and

Neelamegam3 .The practical approach of FRP laminates

technically good and practically very effective and

efficient method of strengthening and upgrading

moment resisting RC frame members that are

structurally insufficient and impaired. The RC elements

such as beams, columns etc are in the FRP composite

materials are found special help with engineers and

applicators because of their many advantage. The main

objective of the study given herein was, to examine the

efficiency and effectiveness of FRP in columns.

Seismic deficiency of RC column:

Identification of detailing deficiencies is significant in

selection of mitigation strategies because acceptable


http://en.wikipedia.org/wiki/Wind_engineering

http://en.wikipedia.org/wiki/Earthquake_engineering

1165 A MURUGESAN AND G S THIRUGNANAM



performance often may be achieved by local adjustment

of detailing rather than by adding new lateral force-

resisting elements. The columns which suffer severe

damage during earthquakes lack ductile design and

detailing. Longitudinal reinforcing bars under

compression in columns are prevented from buckling by

the lateral restraint provided by concrete. Under cyclic

loading, that does not involve alternating flexure; the

compression steel in columns does not ordinarily buckle

out of concrete, even at high strains or in the absence of

restraining stirrups and ties. However when covering

concrete subjected to high compressive stresses become

unstable, the restraining effect is reduced and the bar

buckles. Hence to ensure sufficient ductility, code limits

are placed on the ratio of the distance between

transverse reinforcement to the diameter of the

longitudinal reinforcing bar.

Seismic retrofitting:

Strengthening of a building to increase its seismic

resistance before an earthquake is termed as

rehabilitation, whereas strengthening done after an

earthquake is called retrofitting. The purpose of retrofit

is to enhance the structural capacities such as strength,

ductility, stability and integrity, so that the performance

of the building can be raised to the level to withstand

the design earthquake. A decision on whether or not to

retrofit an unsafe building depends as many factors.

Lifeline buildings such as hospitals must necessarily be

retrofitted, in view of their extreme importance.

Experimental investigation:

A Two bay five storey RC moment resisting framed

structure shown in figure 1 has been analyzed by using

STAAD.pro software as per IS 1893-2002.The structure

is assumed to be located in seismic zone – III. The

design is carried out as per IS 456-2000 and detailed as

per IS 13920-1993 recommendations.

Figure1. Two bay frame – Elevation view

Details of specimen:

In this present investigation column with and without

retrofitting were cast and tested under cyclic loading.

The test specimen was reduced to 1/5th scale to suit the

loading arrangement and test facilities. Prototype

specimen having beam dimension of 305 X 460

including slab thickness and column dimension of 305

X 460. The dimension of the beam and column was

fixed as 120 x 170 and 120 x 230 respectively. The

height of the column for test specimen was 600 mm.

Reinforcement details:

The columns were casted in our laboratory for a

dimension of 120 mm X 230 mm X 600 mm. The

column was reinforced with eight numbers of 8 mm

diameter high yield strength deformed (HYSD) bars.

Shear reinforcement are provided with 6 mm diameter

at 40 mm c/c. The clear concrete cover of 20 mm was

ensured for all four sides. The reinforcement details are

shown in figure 2.

Research significance:

The significance of the present investigation is to arrive

at an efficient method of retrofitting the damaged RC

columns. Two specimens were cast and tested. The first

column is loaded in forward loading up to ultimate load.

Second column is loaded up to 80% of ultimate load of

first beam and then the beam is retrofitted and the load

applied up to ultimate load. A detailed experimental

study has been carried out for the Retrofitted RC

Column under forward loading and the various

parameters like load deflection behavior, cumulative

load deflection behavior, ductility factor, cumulative

ductility factor, energy absorption characteristics has

been investigated and compared with RC column

specimen. Based on the relative performance of the

specimens conclusions has to be drawn.

Figure2. Reinforcement detailing of column

1166 Experimental Investigation on RC and Retrofitted RC Column under Cyclic Loading



Casting of specimens:

Material used:

Ordinary Portland cement (53 Grade) confirming to IS:

12269-1987 with specific gravity 3.1 was used for the

investigation.

Fine aggregate used was locally available river sand

passing through 4.75 mm sieve to grading zone III of

IS: 383-1970 and having a fineness modulus of 2.96

with specific gravity 2.70. Crushed blue granite coarse

aggregates of maximum size 20 mm and graded as per

IS: 383-1970 with specific gravity 2.70 and fineness

modulus is 6.69. Potable water was used for mixing and

curing of specimens. For retrofitting the columns

unsaturated polymer resin, fibre mat, fibre glass powder,

accelerator, catalyst and pigment for colour are used.

M30 grade concrete was designed as per IS 10262-2009.

The materials required for one meter cube of concrete as

shown in Table I.

Table1. Materials required for one m3 of concrete

Materials Cement Fine

aggregate

Coarse

aggregate Water

Weight in Kg 446 541 1205 186

Casting and curing:

Casting of cube specimens for compressive strength:

The cubes of size 150 mm X 150 mm X 150 mm were

cast to find out the compressive strength of Control

concrete. A three set of specimens were cast .The fresh

concrete mix was filled in the steel moulds in three

equal layers and each layer was well compacted using

table vibrator. After de-moulding they were cured in

water for 28 days. The samples are taken from curing

tank and dried for one hour before testing. The

Compression Testing Machine of 1000 kN capacity was

used to find compressive strength of specimens.

Casting of Beams for structural behavior: The

specimens were cast with and without retrofitting and

designated as given below. The test specimens were

stripped from the mould at the end of 24 hours and

cured under wet gunny bags for 28 days. All specimens

were white washed and grid lines are marked at the top

and bottom before testing for clear identification of

crack pattern.

RCC - Conventional reinforced concrete column.

RRCC - Retrofitted reinforced concrete column.

Test setup and instrumentation:

Compression strength test of cubes:

The primary objective is to determine the compressive

strength. The cube specimens of 150 mm size were

tested at the age of 28 days. All the specimens were

tested under saturated surface dry condition. Three

identical specimens were tested in all the mixtures. This

test is carried out by using compression testing machine.

Test Procedure of RC Column:

The column is fully white washed and it is segmented

into 5 cm on each side. The steel plate is fixed on the

top of the column for applying the load. The Universal

Testing Machine (UTM) of capacity 1000 kN is used to

find the behavior of column. The column is placed in

the UTM at the depth of 600 mm. Another two sets of

steel plates are pasted on the sides of the column for

fixing the dial gauge. The dial gauges are placed both

sides of the column to note the deflection of the column.

The load is gradually applied for every 50 kN up to 300

kN. The deflection is noted both left and right side of

the column. After that, the load is gradually released

and deflection is noted for every 25 kN. The same

procedure is repeated for the load 300 kN, 350 kN, 400

kN up to the ultimate load of 800 kN. Then the

deflection is noted. The RC Column was tested as

shown in figure3. At each increment of loads, deflection

and crack pattern were recorded. The failure mode of

the specimen also observed.

Figure3. Test setup of RC column (RCC)

Test Procedure of Retrofitted RC Column:

The RC Column which is tested first in UTM

withstands the load up to 800 kN. The 80% of ultimate

load by first RC Column is given to the second RC

column of same size. The maximum load of 640 kN is

given to the second column and released with failure

pattern. The RC Column with 80% of ultimate load was

tested and due to this load cracks are formed. Cracks

formed in this column are as shown in figure 4.Then

retrofitting techniques is applied to the damaged column

to withstand a high load by arresting the crack and to

recover the damaged column. The Retrofitting

Procedure is as follows. The column is fully cleaned.

The fiber glass powder is mixed with unsaturated

polymer resin and it is applied to the full column for the

smooth surface. The accelerator is mixed with the

catalyst. The fiber glass mat is fully covered over the




column and the mixed accelerator and catalyst is applied

over the glass mat for the purpose of pasting the mat

with the column. The specimen is drying for 24 hours

before testing. After drying, the column is tested by

following the same procedure of testing RC Column.

The testing of Retrofitted RC Column is shown is

figure5.

Results and discussions:

Load carrying capacity:

The first crack was visible in the RCC at 400 kN and

ultimate failure of RCC was at 800 kN. In RRCC first

crack was not visible and the ultimate failure of the

column was witnessed at 900 kN. The ultimate load

carrying capacity of the RRCC was 15% greater than

RCC due to high elastic modulus. It prolongs the first

crack load and increases its load carrying capacity.

Hence there was a significant enhancement in ultimate

load carrying capacity of RRCC when compared to

RCC. The load sequence diagram for RCC and RRCC

are shown in figure 6 and figure 7 respectively.

Figure 4. Test setup for 80% ultimate load of RC

column (RCC)

Figure 5. Test setup of retrofitted RC Column (RRCC)

Figure 6. Load sequence diagram for RCC

Figure7. Load sequence diagram for RRCC

Load deflection behavior:

The load deflection behavior of RCC and RRCC

specimens is shown in figure 8 and figure 9

respectively. Envelope curve is obtained by joining the

peak points of each cycle. Referring to both figures it

can be seen that load carrying capacity of RRCC is

more that of RCC.

Figure 8. Load Vs Deflection diagram of RCC




Figure 9. Load Vs Deflection diagram of RRCC

Ductility characteristics:

The first yield deformation was determined as the

deflection corresponding to the beginning of horizontal

path. Ductility factor for any cycle of loading may be

calculated as the maximum deflection at that cycle to

the first yield deflection. The comparison of ductility

factor for various cycles of the RCC and RRCC on

forward and reversal loading was calculated and shown

in figure 10 and figure 11 respectively. Since the

ductility factor is less obviously ductility is more in

RRCC.

Figure 10. Ductility factor of RCC and RRCC in

forward loading

Figure 11. Ductility factor of RCC and RRCC in

reversal loading

Cumulative ductility:

Cumulative ductility means when a structure is

subjected to cyclic loading up to any load point is

defined as the sum of the ductility at maximum load

level attained in each cycles up to the cycles considered.

The variation of cumulative ductility versus load cycles

for the specimen RCC and RRCC on forward and

reversal loading was calculated and shown in figure 12

and figure 13 respectively. Since the cumulative

ductility factor is less obviously ductility is more in

RRCC.

Figure12: Cumulative ductility factor of RCC and

RRCC in forward loading

Figure13. Cumulative ductility factor of RCC and

RRCC in reversal loading

Energy absorption capacity:

Whenever a structure is subjected to loading some

energy is absorbed by the specimen. In this

investigation the area under the load-deflection curve

gives the amount of energy absorbed by the specimen

during loading. The cumulative energy absorption

capacities of RCC and RRCC for forward and reversal

loading are shown in figure figure 14 and figure 15

respectively.

Mode of Failure:

The failure pattern of RCC is shown in figure 16. As the

load increases the crack width is also increased and

extended towards the top of the column. The crushing




and spalling of concrete is in the top of the column. The

mode of failure was brittle in nature.

The failure pattern of RRCC is shown in figure 17.The

brittle behavior as in RCC was not observed in the

RRCC. As the load increases, the crack was not visible

in the column. This is due to the presence of retrofitting

the column. If the column is retrofitted, the mode of

failure completely changes. The mode of failure

observed was ductile.

Figure 14. Energy absorption of RCC and RRCC in

forward loading

Figure 15. Energy absorption of RCC and RRCC in

reversal loading

Figure 16. Failure pattern of RCC

Figure 17: Failure pattern of RRCC

Conclusion:

Tests on columns subjected to cyclic loading were

carried out. The tested specimens with Indian standard

IS: 13920-1993 detailing shows which is essential for

load sequence, ductility factor, cumulative ductility

factor, and energy absorption factor. Also the test results

necessitate that the requirements of ductility factor,

energy absorption factor which are essentially needed

for earthquake resistant structure. Based on the

investigation, the following conclusions are drawn.

They are summarized below.

The test results have demonstrated that externally

bonded FRP sheets can effectively improve both

shear strength and ductility of column.

The first crack occurs at a load of 400 kN in RC

column but in a retrofitted RC column the crack

is not visible.

The ultimate load carrying capacity of retrofitted

RC column is 1.125 times more when compare to

RC column.

The deflection in RC column and retrofitted RC

column is nearly same in both forward and reversal

loading.

The cumulative ductility value of retrofitted RC

column is 15% more than that of RC column.

The energy absorption capacity of retrofitted RC

column is 15% more than that of RC column.

The use of FRP proved to be advantages over the

conventional one.

The test results exhibits ductility factor, cumulative

ductility factor and energy absorption capacity are

almost same for forward and reversal loading of

RCC and RRCC.

Acknowledgement:

Thanks to all staff members and Technical staff of

Institute of Road and Transport Technology and Sona

College of Technology for their constant support for this

Investigation. Technical support rendered by Prof.




K.Prasad Babu, Head, Department of Civil Engineering,

SCT is also highly acknowledged.

Reference:

[1] Saadatmanesh, H., and Eshani, M.R., “RC Beams

Strengthened with FRP Plats Part I: Experimental

Study”, ASCE Jl. of Struct.Engg., Vol.117. (ii), pp

3417-3433., 1991.

[2] Naani, A., “Concrete Repaired with Externally

Bonded FRP Reinforcement”, Conc. Intl., Vol. 17,

(6), pp 22-26, 1995,

[3] Neelemegam, M., “Use of Fibre Reinforced

Polymer Composites for Repair, Rehabilitation and

retrofitting of R.C.Structural Members-Indian

Perspectives” Proce. of Sixth Asian Symposium on

Polymers in Conc., Vol. 85-94, October, 2009.

[4] C.V.R.Murty, “Learning Earthquake Design and

Construction”, Indian Institute of Technology

Kanpur, Kanpur 208 016, India, 2003.

[5] S.R Umaa and Sudhir K.Jainb, “Seismic Design Of

Beam Column Joints In RC Moment Resisting

Frames”, Department of civil Engineering, IIT

Madras and Kanpur, India.

[6] G.S.Thirugnanam and P.Govindan, “Use of HFRC

in Hinged Zones of Continuous Beams”, IIRT

Erode, India.

[7] S.Pampanin, G.M.Calvi and M.Moratti

Dipartimento di Meccanica structturale, “Seismic

Behavior of RC Beam Column Joints Designed for

Gravity Loads”, Universita di Pavia, Italy.

[8] Pankaj agarwal and Manish Shrikhande

,“Earthquake Resistant Design of Structures”,

prentice hall of India provate limited, New Delhi-

110001, 2006,

[9] Manoj K. Joshi, C.V.R. Murty and M. P.

Jaisingh,”Cyclic Behavior Of Precast RC

Connection”

[10] IS 1893 (Part-I): 2002, Indian Standard Code:

Criteria for Earthquake Resistant Design of

Structures, Fifth Reversion, 2007, New Delhi.

[11] IS 13920: 1993, Indian Standard Code: Ductile

Detailing of Reinforced Concrete Structures

Subjected to Seismic Forces – Code of Practice,

Eighth Reversion, 2005, New Delhi.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1171-1177


An Experimental Study on Hybrid Fibre Reinforced Concrete

NAZEER M1 AND GOURI MOHAN L

2

1Department of Civil Engineering, TKM College of Engineering, Kollam–5, India

2Bishop Jerome Institute, Kollam–1, India


Abstract: The low tensile strength and poor fracture toughness which are the prime drawbacks of conventional

concrete can be overcome by the addition of fibres to the matrix. The early development of fibre reinforced concrete

was using steel fibres in concrete. Later, combined use of metallic and synthetic fibres in concrete became popular

in the domain of fibre reinforced concrete. This new material is named as hybrid fibre reinforced concrete. In this

study, concrete containing hybrid fibres, combination of steel and polypropylene fibres, were investigated for their

compressive strength, splitting tensile and flexural properties, impact, toughness and residual strength. Concrete of

standard strength with steel fibres at 0.5, 1.0, 1.5, and 2.0percent of the total volume fraction and a fixed proportion

of polypropylene fibres of 0.2percent of the weight of cement is used for this study. The results are then analysed to

draw critical comments with respect to the improvements in the mechanical properties fibre reinforced concrete. The

results were then compared with test samples containing steel fibres only and conventional concrete. Even though

the compressive strength showed only a minor improvement; the flexural, impact, toughness and residual strength

was seen to be enhanced by the addition of the fibre combination rather than using mono fibre type.

Keywords: Fibre reinforced concrete, Hybrid fibre reinforced concrete, strength, Residual strength, Toughness,

Impact resistance.

1. Introduction:

Concrete is still the most widely used manufactured

construction material in the world even after the

invention of numerous building materials. This is

mainly due to its low cost and comparatively high

strength as compared to all other available construction

materials. The other advantages which promotes

concrete for wider application are water resistance, low

maintenance cost, ease in mouldability to required size

and shape, less energy input in manufacture and so on.

But the disadvantage with conventional concrete is its

low tensile strain capacity and poor fracture toughness.

This causes a major drawback in concrete which is

brittleness. Reinforcing concrete is the best way to

prevent the cracking of concrete subjected to tensile

stress.

The concept of reinforcing concrete with fibres to form

Fibre Reinforced Concrete (FRC) started since early

1940’s to impart ductility to conventional concrete. In

conventional concrete brittle failure occurs when the

structure is subjected to the peak tensile stress and

cannot withstand further load or deformation. The fibre

reinforced concrete structure also cracks at the same

peak tensile stress, but does not separate out and can

maintain the load to very large deformations.

It is well proved that the addition of short, discontinuous

fibres play an important role in the improvement of the

mechanical properties of concrete. It increases elastic

modulus, decreases brittleness; controls crack initiation

and its subsequent growth and propagation. In FRC,

fibres are very effective in arresting cracks at both

macro and micro levels. De-bonding and pulling out of

fibres require more energy absorption resulting in a

substantial increase in the toughness and fracture

resistance of the material to cyclic and dynamic loads.

But for an optimal response, different types of fibres

may be suitably combined to produce Hybrid Fibre

Reinforced Concrete (HFRC). The use of optimized

combinations of two or more types of fibres in the same

concrete mixture can produce a composite with better

engineering properties than that of individual fibres. In

well-designed hybrid composites, there is positive

interaction between the fibres and the resulting hybrid

performance exceeds the sum of individual fibre

performances and this phenomenon is often termed

“Synergy”. This includes combining fibres with

different shapes, dimensions, tensile strength and

young’s modulus to concrete matrices. It was seen that

hybrid fibre composites were not only stronger in

compression, but also depicted greater strength and

energy absorption capability in flexure.

The fibres, when used in a hybrid form, can result in

superior composite performance compared to their

individual fibre reinforced concretes [1]. A hybrid

1172 An Experimental Study on Hybrid Fibre Reinforced Concrete



combination of steel and polypropylene fibres enhances

the resistance to both initiation and the fracture

behaviour of high performance hybrid fibre composites

[2]. The synergy effect in the hybrid fibres cementitious

system is reported in the literature by Qian and Stroeven

[3, 4].

The increased fibre availability makes it more efficient

in delaying the growth of micro cracks and thereby

improving the ultimate tensile capacity [5]. Replacing a

portion of the large diameters crimped fibres with

smaller diameter crimped fibres can significantly

enhance toughness [6]. Hybrid fibre reinforced high

strength concrete enhances the flexural toughness in

addition to the splitting tensile strength and modulus of

rupture [7].

It is seen from the experimental investigations done so

far that in order to produce hybrid composite, steel

fibres of different sizes were used in concrete with

dense packing and dimensional stability. This is not a

viable option as the steel fibres of lower aspect ratio are

even more costly. This option will naturally increase the

overall cost of making of Hybrid Fibre Reinforced

Concrete (HFRC) and hence is not a practical solution.

The authors identified that the better option is to replace

the small steel fibres with synthetic fibres. Among the

synthetic fibres, the polypropylene fibre is expected to

have a better synergistic behaviour with steel fibres and

is cheaper. In this context an experimental investigation

is planned to study the mechanical properties of HFRC

containing corrugated steel and polypropylene fibre and

also to compare the results with plain concrete and

SFRC.

2. Materials and methods:

In this study, control samples with no fibre content and

crimped steel fibre reinforced concrete (SFRC)

containing single type of fibre were also prepared from

the same mix.

2.1. Materials:

Ordinary Portland Cement (OPC) confirming to IS

12269-1989 [8] was used. The physical properties of

cement used are shown in Table 1.

Crushed natural stone of size 20mm and 10mm size was

used as coarse aggregate. The specific gravity of 20mm

fraction was found to be 2.69 and the fineness modulus

was found to be 7.07. The specific gravity of 10mm

fraction was found to be 2.82 and the fineness modulus

is 6.67. The 10mm and 20mm aggregates were well

graded in suitable proportion as per IS383-1970 [9] and

the ratio of 10mm to 20mm aggregate was obtained as

3:1.

River sand (<4.75mm) having fineness modulus 2.82,

specific gravity 2.68 conforming to Zone II grade as per

IS 383:1970 [9] was used as fine aggregate.

Table 1. Properties of Cement

Grade OPC 53 Grade

Specific gravity 3.13

Fineness 7percent

Initial setting time 159 minutes

Final setting time 247 minutes

Standard consistency 27percent

Polypropylene fibre used in the present study is ‘Recron

3S’ of Reliance Industries Limited. The properties of

the fibre are given in Table 2.

Table 2. Properties of Polypropylene Fibre

Specific gravity 0.91g/cm3

Constituents Virgin Polypropylene C3H6

Fibre Thickness 18 micron

Fibre Length 12 mm

Young’s modulus 5500-7000 MPa

Tensile strength 350 MPa

Melting Point 160°C

The steel fibre used in this study was round crimped

type having 0.45mm diameter and 25mm length from

STEWOLS INDIA (P) LTD. The aspect ratio of the

fibre was 55 and has a density of 7.2 g/cc.

Tap water suitable for drinking was used for mixing and

curing of concrete specimens.

The grade of concrete prepared for the experimental

study was M 25. The proportion used in the

investigation, after necessary adjustments made on the

trial mixes, is shown in Table 3. The cement content in

concrete was 367 kg/m3.

Table 3. Concrete Mix Proportion

Grade of

concrete

Mix proportion

C FA CA W

M25 1 1.13 2.623 0.41

2.2. Methods:

In this investigation, properties of concrete containing

hybrid fibres, with the combination of steel and

polypropylene fibres, were investigated. The

compressive strength, splitting tensile and flexural

properties, impact, toughness, and residual strength of

hybrid fibre reinforced concrete are determined.

1173 NAZEER M AND GOURI MOHAN L



Concrete of standard strength containing steel fibres of

0.5, 1.0, 1.5, and 2.0percent by volume of concrete and

polypropylene fibres of 0.2percent by weight of cement

is used in this study. Depending on the fibre content and

fibre combination, the concrete mixes in the

investigation are designated as detailed in Table 4.

Table 4. Designation of concrete mixes

Mix

Designation

Steel Fibre,

percent

Polypropylene

Fibre, percent

SF0 0 0

SF5 0.5 0

SF10 1 0

SF15 1.5 0

SF20 2 0

SF5P 0.5 0.2

SF10P 1 0.2

SF15P 1.5 0.2

SF20P 2 0.2

Compressive strength of various mixes was determined

by conducting tests on 100mm cubes, 150mmΦx300mm

cylinders. Split tensile strength was determined by

testing 150mm diameter x 300mm height cylinders, and

flexural strength from 100mmx100mmx350mm prism

specimens [10, 11]. Impact resistance was determined

using 150mmdiameter x50mm thickness disc specimens

[12]. Flexural test for determining the toughness of FRC

specimens as per ASTM C 1018 [11] is used to establish

the load-deflection curve. The residual strength of FRC

prism specimens were determined by a method as

outlined in ASTM C1399-02 [13].

3. Results and discussions:

The workability of concrete is determined by measuring

the slump of concrete. The result is presented in Fig.1. It

is clear that, there exist a linear relationship between the

fibre content and slump of FRC. The inclusion of

Polypropylene fibres further decreases the workability

of concrete. However, the numerical value of this

reduction decreases with increase in steel fibre content.

It is observed that for SFRC mixes, the slump reduction

follow an equation in the form:

0 67.86 ................(1)vf fS S V

where Svf is the slump of concrete containing steel fibres

Vf, is the volume fraction of steel fibres and S0 is the

slump of corresponding control concrete.

y = -67.857x + 115

R2 = 0.9984

0

20

40

60

80

100

120

0 0.5 1 1.5 2Steel Fibre Volume, %

Slu

mp

, m

m

SFRC HFRC

Fig.1. Variation of Slump with Steel Fibre Content for

SFRC and HFRC

For each mix of control sample, SFRC and HFRC, three

cube specimens each of size 100mm were tested

according to the I.S specifications [14] at the age of 3, 7

and 28 days. The gain in compressive strength for

100mm cube was plotted on a column graph (Fig.2 and

3) to evaluate the increase in compressive strength with

age. The plotted values are the compressive strength at

age of 3 and 7days relative to the compressive strength

at the age of 28days. It was seen that for SFRC,

52percent of the 28days compressive strength was

attained at the age of 3daysand the compressive strength

attained at the age of 7day is about 63percent of the

28days strength.

0.52 0.52 0.52 0.52 0.540.62 0.63 0.63 0.63 0.64

1.00 1.00 1.00 1.00 1.00

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

SF0 SF5 SF10 SF15 SF20

Str

ength

Rela

tive t

o 2

8da

ys

stre

ngth

Mix Designation

3 7 28

Fig. 2. Compressive Strength of SFRC Mixes

A very similar trend is observed in the case of HFRC

specimens also (Fig. 3). For HFRC, about 53percent of

the 28day compressive strength was gained at the age of

3day; the compressive strength gain at the age of 7day is

about 64percent of the 28days strength. This clearly

indicates that the influence of fibre on compressive

strength is not significant for the type, content or

combination of the fibre used in this investigation.




0.52 0.53 0.54 0.53 0.550.62 0.64 0.64 0.64 0.66

1.00 1.00 1.00 1.00 1.00

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

SF0 SF5P SF10P SF15P SF20P

Str

eng

th R

elat

ive

to 2

8d

ays

stre

ng

th

Mix Designation

3 7 28

Fig. 3. Compressive Strength of HFRC Mixes

The results of compressive strength at 28 days

determined from 100 mm cubes of SFRC and HFRC are

plotted against the steel fibre volume in the mix and

shown in Fig.4. It is clear that, the addition of steel

fibres to concrete improves the compressive strength up

to a steel fibre content of 1.5percent [15]. The increases

in compressive strength may be attributed to the

confinement offered by the fibres against lateral

expansion of the specimen. The compressive strength is

found to be decreasing with fibre content when the fibre

content is greater than 1.5 percent. The reduction in

strength is attributed to the reduction in the filling

ability in presence of high fibre content which in turn

reduces the compaction. The addition of polypropylene

fibre to SFRC causes an improvement in the

compressive strength for all SFRC mixes.

33

33.5

34

34.5

35

35.5

0 0.5 1 1.5 2 2.5

Co

mp

. S

tren

gth

, MP

a

Steel Fibre Volume, %

SFRC

HFRC

Fig.4. Variation of 28 Days Compressive Strength

determined using100 mm Cubes

Compressive strength test was also determined using

150mm diameter x 300mm height cylinders. The

cylinder compressive strength was found to increase

with increase in fibre content up to 1.5percent of steel

fibre and found to reduce when the fibre content is

greater than 1.5 percent. The variation of compressive

strength of cylinders against the steel fibre content was

plotted and shown in Fig. 5. It was observed that,

modification of steel fibre reinforced concrete with

polypropylene fibre enhances the compressive strength

of cylindrical specimens.

The flexural strength tests were conducted on beam

specimens of size 100mmx100mmx350mm as per

ASTM C1018 [11]. Flexural strength obtained for

SFRC was found to increase with increase in fibre

content. The increase in flexural strength was found to

be better in the case of HFRC when compared to SFRC

mixes. The increase in flexural strength may be

attributed to the effectiveness of the steel fibre in taking

up the tension developed in the beam specimens. The

synergistic response of the steel and polypropylene

fibres was found to give a better result. The flexural

strength increased from 8percent to 18percent in the

case of SFRC with varying fibre dosages. The increase

in flexural strength is found to range from 16percent for

mix with steel fibre volume of 0.5percent to 32percent

for mix with steel fibre volume of 2percent and

containing polypropylene fibre at a dosage of 0.2percent

of weight of cement.

19.5

20.0

20.5

21.0

21.5

22.0

0.0 0.5 1.0 1.5 2.0 2.5

Cy

lin

der

Co

mp

. Str

eng

th, M

Pa


SFRC

HFRC

Fig.5. Variation of Cylinder Compressive Strength with

Steel Fibre Volume

The improvement in flexural strength of concrete by the

addition of steel fibre is almost linear with the steel fibre

content in the mix. The strength enhancement is

estimated to be about 0.52 MPa for every percentage of

steel fibre in the mix. Further, addition of polypropylene

fibre to SFRC mix causes enhancement in flexural

strength. The enhancement is proportional to the steel

fibre content in the mix (Fig. 6).

y = 0.52x + 5.24R² = 0.92

5

5.5

6

6.5

7

7.5

8

8.5

0 0.5 1 1.5 2 2.5

Mo

du

lus

of R

up

ture

, MP

a


SFRC

HFRC

Fig. 6. Modulus of Rupture vs. Steel Fibre Volume




The results of the split tensile strength test are as shown

in Fig.7. It was observed that the control samples

without fibre content exhibited a brittle failure whereas

the splitting tensile strength was seen to increase with

increase in fibre content.

0

1

2

3

4

5

6

0 0.5 1 1.5 2 2.5

Split

Tensile

Str

ength

, M

Pa


SFRC

HFRC

Fig.7. Split Tensile Strength vs Steel Fibre Volume

The uniform distribution of steel fibres in the matrix

resulted in enhanced tensile strength of concrete and the

increase was in the range of 5 to 23percent for SFRC

containing steel fibres in the range of 0.5 to 2.0percent

by volume. Further, addition of polypropylene fibre to

these mixes causes an increase in tensile strength. This

increase is the range of 7 to 23percent. In the case of

split tensile failure, the fibre across the entire splitting

cross section is expected to be equally effective in

resisting the splitting of cylinder into two pieces.

Impact resistance is one of the important attributes of

FRC. Depending upon the impacting mechanism and

parameters, there are different tests to evaluate the

impact resistance of concrete. Among various test

procedures, the drop weight test or the repeated impact

test as per the ACI 544-2R-89 is the simplest one. The

impact resistance or number of blows of hammer

corresponding to first crack and ultimate failure of both

SFRC and HFRC is determined. The results obtained

from the experimental investigation are presented in

Fig.8 (Nc and Nf indicate respectively number of blows

required for first crack and failure of the specimen).

0

100

200

300

400

500

600

0 0.5 1 1.5 2 2.5

No

. o

f B

low

s


Nc(SFRC)

Nf(SFRC)

Nc (HFRC)

Nf(HF RC)

Fig.8. Number of Blows vs. Steel Fibre Volume for

SFRC and HFRC

The test results indicate that the impact resistance of

concrete increases with addition of steel fibres. From

the results it can be clearly seen that in locations in the

structure requiring high impact resistance, SF20P mix

can be used .The hybrid fibre concrete is found to have

better first crack resistance. The ultimate strength is

found to increase by about 8 times for HFRC and about

5 times for SFRC. Thus, in locations where higher

impact resistance is required, HFRC can be used as an

alternative to SFRC.

Toughness is a measure of the energy absorption

capacity of a material and is used to characterize the

ability of the material to resist fracture when subjected

to static strains or to dynamic or impact loads. The

difficulties of conducting direct tension tests on FRC

prevent their use in evaluating toughness. Hence, the

simpler flexural test recommended by ASTM C 1018

[11] for determining the toughness of FRC specimens

are prepared and tested according to establish the load-

deflection response curve.

The total energy absorbed by the concrete prism

specimens tested for pure bending is used to assess the

toughness characteristics of fibre reinforced concrete.

The area under the load and deflection response of the

prism specimen is used to calculate the total energy

absorbed by the specimen. Trapezoidal rule is applied

for the calculation of the area.

The improvement in energy absorption capacity of

SFRC and HFRC specimens is plotted against the steel

fibre volume and is presented in Fig. 9. It is clear that

addition of 0.5percent steel fibre does not improve the

toughness of concrete. When the steel fibre volume is

greater than 1.5 percent, a tendency of decrease in the

toughness is observed. For 1.5percent steel fibre the

toughness value is improved by 65percent. Comparing

with the control mix, HFRC specimens also show a

similar trend.

0

10

20

30

40

50

60

70

80

0 0.5 1 1.5 2 2.5

Imp

rov

em

en

t in

En

erg

y A

bso

rpti

on

C

ap

ac

ity

, %


SFRC HFRC

Fig.9. Improvement in Energy Absorption Capacity of

SFRC and HFRC Prisms with respect to the Control

Specimen




As outlined in ASTM C1399 [13], the residual strength

of fibre reinforced concrete can be determined from a

prism flexure test. This load can be applied to the FRC

beam placed over a standard steel plate. Later, the load-

deflection test was performed on FRC beam without the

steel plate and the loads corresponding to 0.50, 0.75,

1.00 and 1.25mm deflections are determined. The

average of these loads multiplied with the factor L/bd2

gives the residual strength of the specimen. Table 5

shows the residual strength calculated for all SFRC and

HFRC specimens. The values reported in the Table 5

are the average from three tests.

The residual strength of control mix is not shown in the

table, as the failure was sudden and has not shown any

post-cracking behaviour. For SFRC specimens, the

residual strength increases with steel fibre volume.

Higher steel fibre content (above 1.5percent) adversely

affects the performance of concrete. Thus it may be

concluded that, the optimum steel fibre content for

improved post-crack behaviour of SFRC is between

1percent and 1.5percent. Also, the addition of

polypropylene fibre to SFRC is not found to influence

the post cracking behaviour except for SF5P.

Table 5. Residual Strength of SFRC and HFRC Prism

Specimens

Mix

Desig-

nation

Load, N

Residual

Strength,

MPa.

P0.50 P0.75 P1.00 P1.25 Pav

SF5 1300 2250 3400 4700 2912.5 0.87

SF10 1750 2800 4500 6000 3762.5 1.13

SF15 1800 3000 4200 5250 3562.5 1.07

SF20 1200 2000 2700 3750 2412.5 0.72

SF5P 1750 2750 4100 6000 3650 1.1

SF10P 1750 2700 3750 5000 3300 0.99

SF15P 1250 2250 3000 4500 2750 0.83

SF20P 1250 2000 3000 3750 2500 0.75

4. Conclusions:

Experimental investigations are carried out on SFRC

and HFRC using steel fibres and polypropylene fibres

on their physical properties of concrete. The various

physical properties examined are compressive strength,

tensile strength, impact strength, toughness and residual

strength. In light of the present investigation the

following conclusions are drawn.

[1] Addition of fibres to concrete causes reduction in

workability in terms of slump irrespective of the

type of fibre used. Addition of steel fibre causes

slump reduction at about 70mm for every

percentage of steel fibre for the concrete mix tested

in this study. Polypropylene fibre also causes slump

reduction when used in combination with steel

fibre. Hence use of a super plasticizing admixture

becomes a need for FRC.

[2] The rate of development of compressive strength

was not affected by the addition of fibres.

[3] The addition of steel fibre improves the flexural

strength of concrete depending on the steel fibre

volume. The enhancement is at the rate of 0.52

MPa for every percentage of steel fibre. The

enhancement in flexural strength due to addition of

polypropylene fibre is increasing with the steel

fibre content.

[4] Addition of fibre to concrete causes increase in split

tensile strength. The enhancement due to

polypropylene fibre addition is almost uniform.

[5] The impact resistance both up to first crack and

failure increases due to addition of polypropylene

fibre in SFRC.

[6] Addition of polypropylene fibre improves the

toughness of SFRC up to steel fibre volume of

1.5percent. Addition of 0.5percent steel fibre did

not cause improvement in toughness. In SFRC the

toughness decreases beyond 1.5percent of steel

fibre volume fraction.

Addition of steel fibre improves the residual strength of

FRC up to a dosage of 1.5percent of steel fibre.

Modification by addition of polypropylene fibre did not

improve the residual strength of SFRC.

2. Reference:

[1] W.Yao, J. Li and K. Wu. “Mechanical properties of

hybrid fiber-reinforced concrete at low fiber

volume fraction.” Cement and Concrete Research,

33(1), pp. 27-30., 2003.

[2] N. Banthia and N. Nandakumar. “Crack growth

resistance of hybrid fiber reinforced cement

composites.” Cement and Concrete Composites,

25(1), pp. 3-9., 2003.

[3] C.X. Qian and P. Stroeven. “Development of

hybrid polypropylene-steel fibre-reinforced

concrete.” Cement and Concrete Research. 30(1),

pp. 63-69., 2000.

[4] C.X. Qian and P. Stroeven. “Fracture properties of

concrete reinforced with steel-polypropylene hybrid

fibres.” Cement and Concrete Composites, 22(5),

pp. 343-351., 2000.

[5] A. Sivakumar and M. Santhanam. “Mechanical

properties of high-strength concrete reinforced with

metallic and non-metalic fibres.” Cement and

Concrete Composites. 29(8), pp. 603-608., 2007.

[6] N. Banthia and M. Sappakittipakorn. “Toughness

enhancement in steel fiber reinforced concrete

through fibre hybridization.” Cement and Concrete

Research, 37(9), pp. 1366-1372., 2007.

[7] A. Ravichandran, K. Suguna and P.N. Ragunath.

“Strength modelling of high-strength concrete with




hybrid fibre reinforcement.” American Jl. of

Applied Sciences. 6(2), pp. 219-223., 2009.

[8] IS: 12269-1987- Specification for 53 Grade

Ordinary Portland Cement, Bureau of Indian

Standards, New Delhi, 2000.

[9] IS: 383–1970 - Specification for coarse and fine

aggregate from natural sources for concrete, Bureau

of Indian Standards, New Delhi.

[10] ASTM C 78-02- Standard test Method for Flexural

Strength of Concrete (Using simple beam with third

point loading).

[11] ASTM C 1018- 97- Standard test method for

Flexural Toughness and first crack strength of Fibre

Reinforced Concrete (using beam with third point

loading).

[12] ACI 544.2R – 89 – Measurement of properties of

Fibre Reinforced Concrete.

[13] ASTM C 1399- 02- Test method for obtaining

Average Residual- Strength of Fibre Reinforced

Concrete.

[14] IS: 516–1959 – Method of tests for strength of

concrete, Bureau of Indian Standards, New Delhi.

[15] Job Thomas and Anand Ramaswamy. “Mechanical

properties of steel fibre reinforced concrete”.

Journal of Materials in Civil Engineering. 19(5),

pp. 385–392. 2007.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1178-1185


Roof Failure Mechanism of Salt-solution Goaf in bedded salt

deposits

LIU JX1, 2

, LIU YT1, CHEN J

3, SHI XL

2, HOU JJ

4 AND CHEN XL

1

1School of Civil Engineering and Architecture, Southwest University of Science and Technology, Mianyang,

Sichuan, 621010, China; 2State key Laboratory of Rock and Soil Mechanics & Engineering, Institute of Rock and Soil Mechanics, Chinese

Academy of Sciences, Wuhan Hubei, 430071, China; 3College of Resources and Environment, Chongqing University, Chongqing, 404100, China;

4School of civil engineering and architecture, Hubei University of Technology, Wuhan, 430071, China


Abstract: Serious disasters such as excessive subsidence, collapse, surface sink holes, might be induced by the roof

failure of the salt solution goafs in underground space. It is an urgent target to investigate the mechanical mechanism

of the roof failure subjected to vertical stress, longitudinal stress and brine buoyancy and injection pressure. By

implementation the theory of circular plate into study, the mechanical model of the circular roof is established. The

specific resolving solution based on Bessel function is obtained. Then, connected with the geology and mechanical

states of the roof, two main failure model and criteria are founded respectively for roof failure. In the end the roof

failure check process adopting the established models and criteria, as well as the improvement and promotion of the

method is presented. The mechanical models and failure criteria of this study provides significant references for

forecasting and controlling the roof failure and surface collapse in solution mining in salt deposits.

Keywords: Roof fracture, salt mine goaf, longitudinal stress mechanical model, failure criteria.

1. Introduction:

Salt mine, a significant source of large amount need in

industries and routine use, is generally exploited by way

of seawater evaporation (seashore regions), salt lake

exploitation and underground salt mine excavation [1].

In China, there are abundant salt mine in depth range of

10 ～ 4000 m, so approximate 90% of the salt mine is

exploited from underground salt mines. The main

contents of the salt mine are NaCl, KCl, NaSO4, and

some other soluble minerals. The rest insoluble

compositions are generally clay, anhydrite, mudstone,

organics, etc. The salt we mentioned in routine life

usually refers to sodium chloride (NaCl). Due to the

water-solution characteristics, much different with the

mental and coal exploitation, water-well-solution

method is used to fetch the salt. This method is carried

out by first drilling a well right through the destination

layer of salt rock, and then installing pipe systems.

Fresh water from internal casing is injected into the salt

mine; after solution, brine is driven out to the surface

from the annular space of the outer and internal pipes.

With the ongoing of the leaching, large cavern gradually

forms in deep underground, similar with the goafs of

coal or mental mine. The cavern is entirely filled with

compressive brine and sediment in the bottom.

All around the world, salt mine consists of two main

geological structures, dome salts and bedded rock salts.

The former is marine sedimentary and the latter is

lacustrine sedimentary [2]. In general, the dome salt is

characterized by content of NaCl, large thickness, few

nonsaline interbeds, so the salt exploitation in such

lithology is high efficient and easy roof controlling.

However, in China, mainly of the salt mines used for

salt products are bedded rock salts, which are composed

of thin salt layers and numerous insoluble interbeds

(usually, anhydrite, mudstone, argillaceous siltstone,

etc.)[2]. This complex conditions result in the insoluble

interbeds become the directly roof in or after

exploitation. Subjected to the comprehensive actions,

these roofs are easy to fracture and consequently induce

high pressure brine injects and erodes the upper strata,

even causes pipes damage and upper ground subsides.

In serious condition, the roof failure might further cause

collapse of the surface ground and brine eruption,

etc[3]. The impacts on the surface facilities will be very

threatening. Such as, as shown in Fig. 1, a sudden

surface collapse occurred in Russia above an abandoned

salt cavern [4]; and collapse occurred and formed

sinking lake on the ground surface above salt caverns, in

Dingyuan of Anhui Province, China [5]. The roof

failure has become the key factor for relative disasters,

app:ds:sodium

app:ds:chloride

1179 LIU JX, LIU YT, CHEN J, SHI XL, HOU JJ AND CHEN XL



so the accurate investigation of the mechanism of the

roof failure has become a significant issue to be

developed [6].

Many researches have been carried out on the failure

mechanism of the roof of salt goaf. Such as, Bauer et al

[7] from the Sandia National Laboratories, using the

bending theory of beam, roughly studied the mudstone

roof’s fracture properties of the salt goaf. By

implementing the circular plate model, Bekendam et al

[8], aimed at the subsidence problem of the salt cavity

after the brine moved from cavity, studied the collapse

mechanism of the roof in salt cavity, and listed the

several possible fracture modes, but he didn’t

considered the effect of the horizontal stress on the

roof’s failure. Jiang et al [10] introduced the catastrophe

theory into the study roof’s span and stability and

proposed the necessary criteria for roof instability. This

mechanical model is boundary fixed large deformable

circular plate, but the horizontal pressure of it is not

considered. Zhang et al [9], by using thin rectangular

plate, studied the longitudinal and transverse stress

distribution of the thin roof of salt cavity. Through he

indicated that failure would occur when the longitudinal

stress passed the threshold value of strength, but the

rectangular mode is much different with the real one.

Shi et al [11] introduced the circular shell into the

analysis of interlayer collapse and control, but the

model was aimed at the thin interlayer in gas storage in

salt cavern not for salt leaching goaf.

Fig. 1 Surface collapse disasters above salt cavities

In addition, numerical simulation or similar simulation

has been adopted to study the failure characteristics of

the cavity roof. By method of numerical simulation,

DeVries et al [12] researched how the roof span, roof

depth, top salt thickness, mudstone/shale thickness and

stiffness influence the stability of the gas storage in salt

deposits under operation duration, and also proposed

corresponding mechanical model. But the analysis is

purposed for energy storages, and the direct roof is salt

protection layer but not direct interbed. Yu et al [13]

conducted similar simulation tests to reveal the stability

state of the salt cavity under different mining depth, and

then obtained the limit roof span in point of simulation

in condition of different mining depth. Tao et al [14], by

using the finite differential element software FLAC,

researched the stability of the salt cavern and proposed

the model for limit span calculation. Wan et al [15]

investigated the mechanism and technologies of the

morphologies control of the salt cavity by connected

method of mechanical analysis and similar simulation

tests, but he didn’t expressed the concrete operation for

these. For the existing model of roof in salt goaf, the

horizontal stress and the constraint conditions are more

or less different with the real ones, thus the mechanical

model which can more approaching to the reality is in

urgent need.

In this paper, we propose a roof failure mechanical

model which is more suitable adoptable for the single-

well solution cavern. In this new model, the intact

circular plate model is adopted, and the vertical stress

and the radial horizontal stresses are considered when

analyzing the fracture features of the roof. For the local

failure due to the stress reaches the threshold of strength

and the entire failure, analysis has been carried out in

point of mechanism. The research can provide some

important references for the forecasting and controlling

of the surface collapse of the salt goaf.

2 Mechanical model

In the process of single-well solution, fresh water is

injected along the internal pipe to the salt mine, then

saturated brine of the goaf is driven along the annular

space of the outer and internal pipes. Along with the

continuous erosion of the water to the goaf wall and the

exploitation ongoing, the goaf volume gradually

enlarges and the salts are more and more exploited. The

progress can be illustrated as Fig.2.

In bedded salt formations, it is inevitable to avoid the

influence of the nonsaline interbeds. Different method

will be adopted to different lithotypes of interbeds. For

thin interbeds, the span of them will be enlarged by

excessive horizontal solution, then tensile concentration

will naturally forms in the interbeds and collapse with

10 m

(a) sink hole reduced by cavity collapse (b) sinking brine lake above abandoned salt cavity

30 m

1180 Roof Failure Mechanism of Salt-solution Goaf in bedded salt deposits



fragments sinking to the bottom. Then the solution can

continue to the upper salt layer. However, for thin

interlayers, under the condition of solution properties

and safety need, ideal span of the interbeds is almost

impossible. The lower and upper salt layers are hard to

be connected by way of interbeds collapse. The

insoluble interbeds will appear on the top of the goaf

and become direct roof, as seen in Fig. 3.

The roof is subjected to several parts of forces, mainly

consisting of the following three parts:

(a)The first one is the horizontal stress, its direction

perpendicular to the roof boundary and inward to the

roof center, composed with the initial in situ stress and

the increment due to the excavation;

(b)The second part is the vertical stress, which is

downward perpendicular to the horizontal plane of the

roof, mainly inducing by the overlying strata rock mass;

Fig. 2 Diagrammatic sketch of the solution mining for

salt rock with single well

Fig. 3 schematic diagram of the stresses situation of the

roof

(c) The third part is the buoyancy of the brine, whose

direction is upward perpendicular to the down surface of

the roof.

In general, the cavern formed by water solution is

symmetry shape, can be regarded as a revolution body

along with a fixed axis. In addition, the well diameter is

often 10 ～ 30 mm, much smaller than the roof diameter

of 30 ～ 50 m, thus the roof can be rationally simplified

as intact circular plate. The established plate model of

the roof is illustrated in Fig. 4.

The lithology of the roof is assumed as homogenous,

and cylindrical coordinate is adopted for solution. The

center of the coordinate is located in the center of the

roof, r directs along with the radius direction from

center to outward; z-axis is downward perpendicular to

the middle plane of the roof.

The boundary of the roof is subjected to radially

distributed stress Fn. The rock salt is famous for its

rheology characteristics, so in the long-term of tectonic-

evolution, the in situ stresses perform hydrostatic

distribution. Ma et al [18] indicated that even in bedded

salt formations, if there is no obvious tectonic-activities,

the in-situ stress was almost hydrostatic state once the

depth is more than 400 m. Thus, in the following

analysis, the radial pressure Rh along the boundary

circumferential direction can be treated as evenly

distributed and its value is defined by the following

equation.

i

n

iiin hF

1

(1)

where, i is the factor of the horizontal stress of

stratum i, which is relative with the tectonic physics; hi

is the thickness of stratum I, rate m; i is the gravity

density of stratum I, rate kN/m3; n is the total number of

the upper strata; H is the depth of the roof, defined as

n

iihH

1 .

In vertical direction, the roof is subjected to evenly

distributed pressure of q, it is the differential part of the

vertical in-situ stress qv and the brine pressure brineq ,

that is:

n

ibrinebrineiibrinev HHhqqq

1

)( (2)

where, qv is the vertical in situ stress; qb is the brine

pressure; brine is the gravity density of the brine, hi is

defined as in Eq. (1); and is the mean density of the

overlying strata, defined as：

n

i

n

iiii hh

1 1

/ (3)

Horizontal

stress

Vertical stress

Buoyant stress

Goaf

Horizontal

stress




r

¦ È

Fn

qn

Fn

qw

R

z

Fn

(a) Over-sight diagram (b) side-sight diagram

Fig. 4 Mechanical model of the circular roof of salt goaf

3 solution of the model

3.1 Deformation of the roof

For the edges simple supported sliding circular plate,

based on the stresses condition of the element in the

circular plate, Timoshenko deduced the differential

equation of the deflection of surface curvature when

subjected to horizontal longitudinal stress Fn and

vertical uniform stress q [17]:

2 2

2 2 2

d 1 d 1

d d 2

k qr

r r r R r D

(4)

Where φ is the gradient of the flexure plane; r is the

longitudinal coordinate variant; R is the radius of the

circular plate; q is the vertical stress acting on the

circular plate.

In Eq. (4), D is the flexural rigidity of the circular plate,

and its value is:

)1(12 2

3

EdD

(5)

Where E, and d are respectively the Young’s

modulus, Poisson’s ratio and thickness of the plate. k is

a parameter introduced for the convenience of solution,

its value is determined by:

D

RFk n

2

2 (6)

Eq. (4) belongs to Bessel function [18], and for the no-

hole circular plate, its general solution can be written as:

1 1 0

krC J

R

(7)

Where, J1 is a first-order Bessel function; C1 is a

parameter; and φ0 is a special solution of Eq. (4).

Generally, for the annular plate under boundary

conditions of GSS and GRS, as well as the uniform

distributed vertical stress q, the follow equation can be

used as one special solution of Eq.(4):

n

2

0 22 2

qrR qr

k D F

(8)

Thus, at this moment, the general solution of Eq.(4) can

be written as:

n

1 1

d

2 d

kr qrC J

R F r

(9)

Conducting integration on the both sides of Eq. (9), one

will obtain the expression of deflection:

n

2

1

0 24

C R kr qrJ C

k R F

(10)

Where J0 is the zero-order Bessel function; and C2 is

constant.

Boundary conditions:

When r=R, φ=0；

Subsequently, separately put the conditions r=R and

ω=0intothe expressions of φ Eqs.(7) and (10), then the

expressions of the constants C1 and C2 can be obtained

as:

3

1 2

12

qRC

k J k D

,

4 4

0

2 3 2

12 4

qR J k qRC

k J k D k D

Then put the expressions of constants C1and C2 into

Eqs. (7) and (10), one can obtain the expression of

curvature and deflection of the flexure plane:

3

1 1

2

12

krqR J R J k r

R

k J k D

(11)

Fn

w

z

q=qv-qbrine

r Fn




4

2 2 20 0

3 2

12 4

krqR J k J

qR R rR

k J k D k D

(12)

3.2 Stress distribution in the roof

In respect to the three dimensional axisymmetric

problem, the deflection ω of the roof plate is only

related with the radius r. In references [19, 20], the

moment and generalized shear stress of the circular

plate is proposed as the following equations:

d

d

d

d

0

d d

d d 2

r

r r

r

M Dr r

M Dr r

M M

qrQ D

r r r

(13)

Where M is the moment under bending; and Q is the

shear stress.

In which, in reference [19], the calculation stress

components of the bending stress of the circular plate

were given:

3

3

2

2

3

12

12

6

4

0

r

r

r

rz

r r z

Mz

h

Mz

h

Q hz

h

(14)

Where σ stands for normal stress, and τ stands for shear

stress.

Inputting Eq. (11) into Eq. (12), and Eq. (13) into Eq.

(14), one can obtain the expressions of bending stress

components; then by superposition the bending stress

components with the horizontal stress, the total stress

distribution in the entire circular plate can be obtained.

In reality, the expression of gradient of the bending

plane contains Bessel function; its unfolding term will

be very complex. So there are no resolving stress

components for Eq. (15). Even through, for real

engineering, we can look reference to the table of Bessel

function, and then put it into relative formula to

calculate the concrete value.

Set σrt and τrt as the horizontal stress component and

shear stress component of the total stress. Then the

separate maximum and minimum values of these two

stress components can be calculated by using Eq.

(15)[11]:

,max , /2

,min , /2

,max , 0

,min 0

h

rt r r r a z h

h

rt r r r a z h

rt rz r a z

rt

(15)

where σrh is the horizontal stress that the roof is

subjected to.

Based on the solution of Eq. (15), it is obviously found

that the threshold values of the stress components

appear nearby the edges and central regions of the roof.

Fig. 5 presents the stress distribution along with the

outer surface of the edges in z-direction. If the roof is

subjected to relative large horizontal stress, σrt,max might

become compression stress.

Fig. 5 Stress distribution of the roof along the vertical

direction[11]

4 Failure criteria of the roof

4.1 Local failure criteria

In general, the lithology of the roof in bedded rock salt

is anhydrite, argillaceous siltstone, glauberite mudstone

and silt mudstone. All of them are characterized by

lower strength, exerting a tensile strength much lower

than the compression strength. So in regions near the

lower location of the plate center, as well as in regions

near the upper edges, tensile fracture is high probable to

occur, shown as Fig.6. So, for this failure mode, a

criterion of the maximum tensile strength is available,

the formula of which is:

t 3 (16)

Where t is the tensile strength of the roof; and 3 is

the minimum principle stress in the roof.

Radial normal

stress

Shear stress

σrt,min

σrt,max

τrt,max

，max




According to Eq. (16), one can find that the first

principle stress in the roof can determined as

max,3 rt . When the absolute value of max,rt

exceeds the tensile strength σt, local tensile failure in

the roof will happen, and then cause fracture and

collapse of the roof.

Fig. 6 Diagrammatic sketch of the partially failure of

the roof

4.2 Criteria of entire failure

During the salt exploitation process or after the

abandonment of the salt goaf, the creep of salt will

induce the stress increasing of the roof edges. Before the

stress exceeds the roof’s strength, large deformation of

the entire roof will perform induced by the horizontal

stress, and then cause large area of collapse of the roof.

This kind of failure is not caused by insufficient

strength but caused by insufficient stability, seen as

Fig.7.

Fig. 7 Diagrammatic sketch of the integral instability of

the roof

In reality, the goaf is subjected to both solution erosion

and creep shrinkage, the horizontal stress of the roof

edges is variable under different conditions. Thus for

the entire analysis of the roof stability, both the goaf

excavation and the rheology properties should be

comprehensive considered.

Based on Eq. (11), if J1(k)=0, the value of ω will be

infinitely great, which is incorrect for real conditions.

Assuming the zero-point of function J1 is

j1、j2、j3、…, and gradually become larger and larger

one by one. Then k = j1 can determine the minimum of

the horizontal force, that is:

n

2

1

2cr

DjF

R

(17)

According to the value table of Bessel function [18], we

can find the minimum of the root j1=3.832. Putting this

root value into Eq. (17), we can then get the value

expression of Fncr:

n 2

14.682cr

DF

R

(18)

In the above Eq. (18), Fn in fact is the uniform

distributed membrane stress of the circular plate edges.

If we set σr0 as the membrane stress evenly distributed

along with thickness, we can get[16]:

0

rn dF (19)

In conjunction with Eqs. (18) and (19), the threshold

value of the membrane stress σr0 in the edges of the

roof can be calculated[11]:

2

2

0 )(1

224.1)(

R

dEcrr

(20)

Where E is the Young’s modulus of the plate; μ is the

Poisson’s ratio of the roof; d is the mean thickness of

the roof; and R is the appearing radius of the roof.

For the mechanical parameters, such as E, μ, one can

obtained by method of laboratory tests; and for the

values of d and R, they can be valued by well-log

measurements of senor measurements.

However, the value that Eq. (20) gives is the theory

value of the threshold value, which is somewhat

different with the reality. Due to the uncertain

influences of sedimentary, in real engineering, the

lithologies, thickness, as well as mechanical strength

exists some certain of random. So for real application,

the Eq. (19) should be modified as following:

20 )(3.1)(R

dEcrr (21)

Where is modifying parameter, determined by the

specific petrophysical properties of the inerbeds. Its

value is lower than 1.0 and can be obtained according to

specific engineering. In Eq.(21), the Poisson’s ratio is

valued as 0.25.

From the above equation, it is found that the exposuring

span of the roof increases along with the solution

ongoing of the cavern. When the radius increases to

reach the above relationship, entire failure of the roof

will occur.

In addition, under different conditions of solution or

abandonment, the mechanical behaviors of the roof is

variable, thus the mechanical parameters should also be

modified. We recommend checking the softening

mechanism of the brine to the interbeds, thus the

mechanical behaviors versus time can be made, and

then rational modification can be conducted.

Brine

Roof

Brine

Roof




4.3 Other failure models

In reality, the failure models might not only the two

mentioned above. The analysis above approaches more

to the mechanism point. According to the previous

analysis, it is found the mineral components are another

important factor influencing the fracture properties. The

interbeds contain a certain soluble mineral, such as

NaCl, Na2SO4, and CaCl2. In cavern, the brine

concentration is different due to its density, in which the

top region has the lowest value, thus top solution is

most fast. The roof of interbed is inevitable influenced

by the unsaturated brine. There are two factors the brine

acts on the interbeds, the one is the high brine has

infiltration effect on the roof which will increase the

fracture of the interbed; the other is the solution effect,

which will destroy the matrix of the interbeds. So, in

this way, the failure model might be more complex that

analyzed above.

Above all, for the roof stability analysis, the available

process is proposed: first, the location and size of the

roof can be detected by Sonar measurement. Then, the

mechanical parameters, such as Young’s modulus E and

Poisson’s ratio μ, can be obtained by laboratory tests; if

the interbed contains soluble mineral compositions, the

immersion test should also be conducted to investigate

the softening performances. In end, the failure models

can be used to check whether the roof will collapse or

not.

5 For connected two-well solution mining

In addition, in real practice, the two-goaf system is

widely used. That is one goaf used for fresh water

injection and the other goaf is used for brine

withdrawing out. Due to the pressure state different

between the roofs of the two connected goafs, the

failure state may be different with each other. Generally

speaking, the goaf of brine withdrawing goaf has a

lower brine supporting pressure, thus collapse is higher

probable to occur.

Fig.8 connected two-well mining solution

For this connected goaf system, it is easy to establish

relative failure criteria for this kind of roof. What we

need do is only to change the brine pressure under the

roof, and then the criteria can be made.

6. Conclusions

In this paper, elastic theory of circular plate is

implemented to analyze the failure mechanism of the

roof in single well-solution salt cavern. The mechanical

model of sliding simple supported circular plate

subjected to vertical and horizontal stresses was

established. Elastic theory of circular intact plate was

used to fetch the resolving solution. According to the

solutions, mechanical analysis has been conducted on

the local failure and entire instability, and then

respective failure criteria of the two failure models have

been proposed. The investigation of the mineral solution

and brine softening has also been suggested. The

researches of this paper provide essential reference for

the roof failure mechanism and criteria, and also supply

important guidance for the surface subsidence

prediction and control of the salt cavern.

7. Reference:

[1] Wang Qingming. Rock salt deposit and

prospecting. Beijing, Chemical industry Press,

2007. (in Chinese)

[2] Kong Junfeng. Study on Subsidence above Salt

Cavern Underground Gas Storage and Related

Engineering Application[D]. Wuhan: University of

Chinese Academy of Sciences, 2012.

[3] Qiu Zhiyong. Mechanism analysis of surface

Goafs

Injection well Withdrawing well

Injection well

Withdrawing well




collapse in the area of solution salt mining. Journal

of Safety Science and Technology, 2011, 12(7): 27-

31.

[4] Yang Chunhe, Jing Wenjun, JJK Daemen. Analysis

of major risks associated with hydrocarbon storage

caverns in bedded salt rock. Reliability Engineering

& System Safety, 2013, 113, pp94-111.

[5] Luo Xiaohui, Xi Zuping, Zhou Hong. Brief

Analysis on How to Prevent the Mined Ground

from Subsiding During Mining. China Well and

Rock Salt, 2011, 42, pp17-18.

[6] Luo Xiaohui, Xi Zuping, Zhou Hong. Brief analysis

on how to prevent the mined ground from subsiding

during mining[J]. China Well and Rock Salt,

2011,42(2), pp17-18

[7] Stephen Bauer, Brian Ehgartner, Bruce Levin,

James K. Linn. Waste Disposal in Horizontal

Solution Mined Caverns--Considerations of Site

Location, Cavern Stability, and Development

Considerations[C]// SMRI Fall Meeting, 1998.

[8] Roland Bekendam, Wim Paar. Induction of

Subsidence by Brine Removal[C] // SMRI Fall

Meeting, Austria, 2002.

[9] Jiang Deyi, Ren Song, Liu Xinrong, et al. Stability

analysis of rock salt cavern with catastrophe

theory[J]. Rock and Soil Mechanics, 2005, 26(7):

1099-1103.

[10] Zhang Peng, Lu Qingfeng, Zhang Wenguang, et al.

Stability analysis and control of rock salt cavern

roof[J]. China Well and Rock Salt, 2001, 42(2): 11-

14.

[11] Shi Xilin, Li Yinping, Yang Chunhe, Qu Dan’an,

Ma Hongling. Research on mechanical mechanism

of interlayer collapse in solution mining for salt

cavern gas storage. Rock and Soil Mechanics,

2009, 30(12), pp3615-3622.

[12] DeVries KL, Mellegard KD, Callahan GD,

Goodman WM.. Cavern roof stability for natural

gas storage in bedded salts. Final Report, National

Energy Technology Laboratory, United States

Department of Energy. Pittsburgh, Pennsylvania;

2005.

[13] Yu Hailong, Tan Xueshu, Xian Xuefu, et al. Model

test study on cavern stability of the rock salt[J].

Ground Pressure and Strata Control, 1995, 3(4):

156-159.

[14] Tao Lianjin, Jiang Deyi. Analysis on nonlinear

large deformation of saltrock cavity[J]. Journal of

Beijing Polytechnic University, 2001, 27(1): 64-67.

[15] Wan Yujin. Shape-controlling mechanism of gas

storage building in salt beds[J]. Natural Gas

Industry, 2004, 24(9): 130-132.

[16] MA Linjian, LIU Xinrong, MA Shu’na, et al.

Numerical analysis of in-situ ground stresses in

deep rock salt stratum containing mudstone

interlayers[J]. Journal of PLA University of Science

and Technology (Natural Science Edition), 2009,

10(6), pp604-609.

[17] Timoshenko S, Woinowsky KS. Theory of plates

and shells[M]. New York, USA: McQraw-Hill

Book Companty, 1959.

[18] Liang Kunmiao. Methods of Mathematical physics

[M]. Beijing: Higher Education Press, 2010. (in

Chinese)

[19] Liu Hongwen. Theory of plate and shell.

Hangzhou: Zhejiang University Press, 1987.

[20] Yu M. Double-shear strength theory. Beijing:

Science Press, 1998.

http://www.sciencedirect.com/science/journal/09518320

http://www.sciencedirect.com/science/journal/09518320

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1186-1191


Effective Utilisation of Pond Ash as Fine Aggregate in Cement

Concrete under Flexure

K ARUMUGAM1, R ILANGOVAN

1 AND A V DEEPAN CHAKRAVARTHI

2

1Department of Civil Engineering, BIT Campus, Anna University, Triuchirappalli, India

2Department of Civil Engineering, Velammal College of Engineering and Technology, Madurai, India

Email: [email protected], [email protected], [email protected]

Abstract: With an increase of power consumption due to industrial development, the generation of coal ash has

been growing tremendously. Also environmental concern over its disposal has been raised. In order to examine the

usability of coal ash as an aggregate for concrete, the mechanical properties and durability of concrete using pond-

ash was analyzed in terms of pond ash content (10%,20%,30% wt %) as a part of fine aggregate. In this

experimental investigation, a mixes M25 grade was obtained by the proper proportion of pondash which gives

higher strength than the normal concrete. For that purpose we have tried with the replacement of pondash in

different percentage and compared the behavior with that of conventional concrete.

Keywords: Pond Ash, Electrostatic Precipitator, Wet Disposal, Fine Aggregate.

Introduction:

Use of waste and by products as concrete aggregate is of

great practical significance, because about 75% of

concrete comprises aggregate. There are various types of

waste materials that can be considered for usage as

aggregates. India’s power generation has undergone a

tremendous growth since independence. The production

of ash has also increased from 17.06million tones during

1900-91 to 68.82million tones in 1996-97and has

crossed 100million tones in year 2000. The ash needs to

be managed properly (or) otherwise it will cause land,

air and water pollution. Here there is serious concern

about utilizing it to the maximum extent. With the

increasing use of low grade coal of high ash content, the

current production of ash will be more then 100million

tones per year. Generally it is seen that the bottom is

around 20% of the total ash produced. Primarily, the fly

ash is disposed of using either dry or wet disposal

scheme. In dry disposal, the flyash is transported by

trucks, chute or conveyors to the site and disposed of by

constructing dry embankments. In wet disposal, the

flyash is transported as slurry through pipe and disposed

in large ponds (or) dykes. About 1000 million tons of

such ash referred to as pond ash, is available in India,

almost free of cost. Increase in demand and decrease in

the natural resources of fine aggregate for production of

concrete has resulted in the need of identifying new

sources of fine aggregate. This paper was possibility of

utilization of thermal power plant-Neyveli by-product-

pond ash as replacement to fine aggregate in concrete.

Fly ash collected through hoppers has been widely

accepted as a pozzolona and is being used by the

construction industry. Pond ash being coarser and less

pozzolonic is not being used. Recent reports indicate

that France leads the word by utilizing 50% of fly ash

followed by West Germany and U.K, while india

utilizes hardly 1% of the annual fly ash production of

nearly 100million tones.

Properties of Materials

It is deals with the study of the properties of various

materials like cement, fine aggregates, coarse

aggregates and water, which were used for casting

concrete specimens.

The various tests conducted to study the properties of

the above said materials are also discussed in this

chapter in detail.

The various materials used in this project are listed

below:

Cement – Birla Super 53 grade cement,

Fine aggregates – locally available sand, collected

from the Amravati river bed of Karur,

Coarse aggregates – locally available stones,

Water – ordinary portable drinking water.

Properties of Pond Ash

Table1: Physical Properties

Property Values Test method

Specific gravity 2.680 IS2386 – 1963

Bulk Density

(Kg/m3)

824 IS2386 – 1963

Absorption (%) Nil IS2386 – 1963

1187 K ARUMUGAM, R ILANGOVA AND A V DEEPAN CHAKRAVARTHI



Moisture content

(%) 3% IS2386 – 1963

Fine Modulus 2.705 IS2386 – 1963

Shape Irregular Shape, Angular

Table 2: Chemical Properties

Compounds % Composition

Loss of ignition 1.38

Silica (SiO2) 81.38

Iron oxide (Fe2O3) 7.42

Titanium oxide (TiO2) Nil

Aluminium oxide (Al2O3) 3.86

Calcium oxide (CaO) 1.01

Magnesium oxide (MgO) 1.43

Sodium oxide (Na2O) 1.35

Potassium oxide (K2O) 2.17

Table 3: The mix proportion for M25 then becomes

Water

l/m3

Cement

kg/m3

Fine

Aggregate

kg/m3

Coarse

Aggregate

kg/m3

191.6 425.78 526.67 1281.7

0.45 1 1.24 3.0

Table 4: Types of mix

Mix Type of Mixes Mix Details

C1 Conventional Conventional

M1 10% Pond Ash

M2 20% Pond Ash

M3 30% Pond Ash

Table 5: Compaction factor test

Pond ash content Compaction factor

value

Conventional 0.93

10% 0.90

20% 0.87

30% 0.83

Table 6: Slump Test on Concrete

Pond ash content Slump value in mm

Conventional 25

10% 20

20% 18

30% 15

Experimental Investigation

Production of good quality concrete requires meticulous

care exercised at every stage of manufacture of

concrete. It is interesting to note that the ingredients of

good and bad concrete are the same. If meticulous care

is not exercised, and good rules are not observed, the

resultant concrete is going to be of bad quality. With the

same material if intense care is taken to exercise control

at every stage, it will result in good concrete.

The measurement of materials for making concrete is

known as batching. Here, we have adopted weigh-

batching method, and it is the correct method too. Use

of weigh system in batching, facilities accuracy,

flexibility and simplicity. Different types of weigh

batches are available; the particular type to be used

depends upon the nature of the job.

When weigh batching is adopted the measurement of

water must be done accurately. Addition of water in

terms of liter will not be accurate enough for the reasons

of spillage of water, etc.

We used cube moulds of standard size 150 mm, which

is made up of cast iron and the inside faces are

machined plane. All the faces of the mould are

assembled by using nuts and bolts and are clamped to

the base plate. It is to be noted that the entire internal

angle must 900. The faces must be thinly coated with

mould oil to prevent leakage during filling. The inside

of the mould must also be oiled to prevent the concrete

from sticking to it.

Thoroughly mixing of materials is essential for the

production of uniform course; the mixing should ensure

that the mass becomes homogeneous. In this paper we

adopted hand mixing. This uniform mixture is spread

out and water is sprinkled over the mixture and

simultaneously turned over. This operation is continued

till some time that good uniform, homogenous concrete

is obtained. It is of particular importance that the water

is not poured but it is only sprinkled.

After the materials have been mixed, the moulds are

filled immediately by pouring the concrete in to it.

Concrete is filled in three layers, and each layer is

compacted well by using a tamping rod of standard size,

so as to avoid entrapped air inside the concrete cubes

and honey combing effect on the sides. During pouring

of concrete, it is better to avoid wasting of concrete for

effective and economical usage. In order to avoid

wastage, small trowels are used to collect the concrete

that is coming out the mould while pouring, and it is

again used in the process. The mould is striped after 24

hours. The test cubes were cured for different duration

like 7 days, 28 days and 56days in a curing tank. After

the wet curing the specimens were air cured for

minimum period 7 days under laboratory conditions.

1188 Effective Utilisation of Pond Ash as Fine Aggregate in Cement Concrete under Flexure



Name of

the test Size of mould

Number

of days

Compression

Test 150 x 150 x 150 mm 7 28 56

Split tensile

Strength

Ht - 300 & Dia - 150

mm 7 28 56

Flexural

Strength 100 x 100 x 500 mm 7 28 56

Methodology

The methodology adopted for experimental study has

fully been described the hardened properties of concrete

such as compressive Strength, Split tensile Strength and

Flexural Strength parameters of M25 grade concrete as

per relevant codes of practice, replacing cement by

Pond Ash partially in refer table 2.4. This experimental

study consisted of collection of different materials to be

used for the investigation, including Pond Ash sample

collection from Neyveli, characterization of materials,

designing the mix as per IS 10262-2009, mixing of

concrete, casting the specimen for various test

considered for the study, testing the specimens,

tabulation and analysis of data and finally discussion

followed by conclusions based on test results.

Collection of Sample and its preparation - Pond Ash

sample collected (Fig.4.1) from Neyveli was used for

the investigation, which partially contributes to Cement

content of the concrete mix.

Fig1: Pond Ash

Testing of Specimens

The concrete specimens are tested using a compression

testing apparatus. The specimen is rested over the

loading frame. Care should take that the specimen

should rest within the frame for effective load

application. Then the load is applied and the value of

the load for the first visible crack and the maximum

load is noted down to calculate the compressive strength

of the concrete, which is got by dividing the maximum

load attained by the area of contact.

Result Analysis

The mix design ratio for M25Grade Concrete was

derived as 0.45:1:1.24:3.0 (weight basis) respectively.

M25 grade concrete was cast by using cement of 53

grade of its specific gravity was 3.16. Sand conforming

to zone II was used its specific gravity was 2.648. The

coarse aggregate of 20mm size was used and its specific

gravity was 2.86. Then cubes, cylinders and prisms

were cast with concrete by using the moulds of size

150mm x 150mm x 150mm, 150mm x 300 mm and

100mm x 100mm x 500mm. After 24 hours the

specimens were removed from the mould and cured for

7 days, 28 days and 56days. After curing the specimens

were tested using compression testing machine and

flexure testing machine. Then the values are noted. The

above process was repeated for replacement of sand by

Pond ash. The above experimental results were

compared with ordinary concrete results.

Table 7: Compressive Strength of concrete for Different

mixes

MIX

7 Days

Strength

(N/mm2)

28 Days

Strength

(N/mm2)

56 Days

Strength

(N/mm2)

C1 16.2 26.35 28.47

M1 16.8 27.53 28.96

M2 17.7 28.23 29.54

M3 17.8 28.38 29.86

Fig 2: Compressive Strength with Different age

Fig 3: Compressive Strength Testing sample




Table 8: Split Tensile Strength of concrete for Different

mixes

MIX

7 Days

Strength

(N/mm2)

28 Days

Strength

(N/mm2)

56 Days

Strength

(N/mm2)

C1 2.04 2.88 2.95

M1 2.12 2.97 3.12

M2 2.2 3.12 3.21

M3 2.22 3.16 3.33

Fig 4: Split Tensile Strength with Different age

Fig 5: Split Strength Testing sample

Table 9: Flexural Strength of concrete for Different

mixes

MIX

7 Days

Strength

(N/mm2)

28 Days

Strength

(N/mm2)

56 Days

Strength

(N/mm2)

C1 3.38 3.98 4.15

M1 3.41 4.1 4.21

M2 3.49 4.18 4.32

M3 3.53 4.25 4.45

Fig 6: Flexural Strength with Different age

Fig 7: Flexural Strength Testing sample

Results Discussion

With the addition of pond ash there is reduction in

slump value of fresh concrete.

The unit weight of concrete gets reduced through

the addition of pond ash as replacement of fine

aggregate since it has lesser specific gravity than

fine aggregate.

The 7 days and 28 days strength shows that the

strength increases from standard concrete up to the

addition of 30% replacement of fine aggregate with

pond ash.

The compressive strength various from 29.86 N/mm2 to

28.47 N/mm2 for 56 days and 28.38 N/mm2 to 26.35

N/mm2 for 28 days and 17.8 N/mm2 to 16.2 N/mm2

for 7 days. This is the different range of replacement

of pond ash.

1190 Effective Utilisation of Pond Ash as Fine Aggregate in Cement Concrete under Flexure



Graph 1: Compressive Strength of concrete with age

The split tensile strength various from 3.33

N/mm2 to 2.95 N/mm2 for 56 days and 3.16

N/mm2 to 2.18 N/mm2 for 28 days and 2.22

N/mm2 to 2.04 N/mm2 for 7days. This is the

different range of replacement of pond ash.

Graph 2: Split Tensile Strength concrete with age

The flexural strength various from 4.45N/mm2 to

4.15 N/mm2 for 56 days and 4.25 N/mm2 to 3.98

N/mm2 for 28 days and 3.53 N/mm2 to 3.38

N/mm2 for 7 days. This is the different range of

replacement of pond ash.

Graph 3: Flexural Strength of concrete with age

Conclusion

The following conclusions are drawn from the

observations of the compressive strength test, Split

tensile test and Flexural Strength test made by using

the pond as partly replacement for fine aggregate from

0% to 30%

The density of concrete reduces with the increase in

the percentage of pond ash.

The Split tensile strength of concrete with pond ash

increases with increased curing period.

The Flexural strength o f concrete with pond ash

increases with increased curing period.

While pond ash is being used the workability gets

reduced. For obtaining the required workability the

super plasticizers are added while preparing the

concrete.

The more pond ash we add the more super

plasticizers are required to be added for obtaining

the required workability.

With increasing replacement of fine aggregate

with pond ash, the average density of concrete

shows a linear reduction due to its lower specific

gravity.

Strength gets increased by increasing the days of

curing.

The test result shows that the replacement has

beneficial effects in improvement of properties

mainly at M3 and M5.

Acknowledgements

The authors are highly grateful to

Dr.A.M.Vasumathi, Professor & Head,

Department of Civil Engineering, K.L.N.college of

Information Technology, Madurai, Tamilnadu

(India) and Dr.L.Andal, Professor & Head,

Department of Civil Engineering, Velammal

College of Engineering and Technology, Madurai,

Tamil Nadu for constant inspiration and support.

Reference:

[1] Andrade L.B, et.al. “Influence of coal bottom ash as

fine aggregate on fresh properties of concrete”,

Construction and Building Materials,Vol.23,pp609-

614,2009

[2] .Monzo,J,et.al “A Preliminary study of flyash

granulometric influence on Mortar Strength”, Cement

and concrete research,Vol.24,No.4,791-796(1994)

[3] Aggarwal P.et al. “Effect of bottom ash as

replacement of fine aggregate in concrete”, Asian

journal of civil Engineering (building and housing),

Vol.8 No.1pages49-62, 2007.

[4] Mangaraj B.K. “Use of Ponded Fly Ash as part of

replacement of Fine Aggregate in Mortar and

Concrete”, ICJ, PP 279 – 282, May 1994.




[5] Ranganath R.V “A study on the charecterisation

and use of Ponded Fly Ash as Fine Aggregate in in

Mortar and Concrete” Ph.D Thesis, Report 1995,

IITD – Delhi,

[6] Ranganath R.V, et.al. “Influence of Size fraction of

ponded ash on its pozilanic activity” Cement

concrete Research, Vol.28, No.5, pp749-761, 1994.

[7] P. K. Kolay, D. N. Singh “Physical, chemical,

mineralogical, and thermal properties of cenospheres

from an ash lagoo” Cement and Concrete Research,

Issue 4, Volume 31, PP 539-542, April 2001,

[8] A.K.Mullick” Use of Industrial wastes for

sustainable cement and concrete constructions,”

The Indian concrete journal, Vol 81,No.12,16-

24(2007)

[9] Pranesh R.N. “Pond Ash in Cement Concrete –

Some studies on its feasibilities” Ph.D Thesis

Report 2008.VTU, Belgaum.

[10] Bharathi Ganesh “Pond Ash: An Alternative

Material as Fine Aggregate in Concrete for

Sustainable Construction” International Congress

on Advanced Materials, 2011.

[11] Leonards G. A. “Pulverized coal ash as structural

fill”, J. Geotech. Engng Div., ASCE, PP 517– 531,

1982.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1192-1198


Variable Tap Parameter (α) Techniques for Sato based Blind

Equalizer

K SUTHENDRAN1 AND T ARIVOLI

2

1Department of ECE, Kalasalingam Academy of Research and Education, Krishnankoil-626 126, India

2Department of ECE, Vickram College of Engineering, Enathi-630 561, India


Abstract: It is renowned that blind equalizers do not require training sequence to track the time varying

characteristics of the channel. But, it ends up in slow convergence to realize a selected Signal to Noise Ratio (SNR).

However, variable tap parameter (α) will speed up the convergence rate and also minimizes the mis adjustment for a

blind equalizer. In this work, two variable tap parameter techniques were used for Sato based blind equalizer

algorithm. Simulation results for Pulse Amplitude Modulated (PAM) signal show that the projected approach has a

higher convergence rate than the prevailing Sato algorithm with fastened α value.

Keywords: Blind Equalization, Convergence, Sato algorithm.

1. Introduction:

In the trendy data communication, plenty of effort has

been devoted to utilize the accessible channel

bandwidth with efficiency. Inter-Symbol Interference

(ISI) and Thermal noises are the two main factors,

which are limiting the performance of data transmission

systems. In essence, the ISI is generated by dispersion

within the transmit filter, the transmission medium, and

receive filter. In the band-limited (frequency selective)

time dispersive channel, the ISI is caused by multipath

propagation. The impact is that the modulated pulses are

spread in time into adjacent symbols, and it distorts the

transmitted signals inflicting data errors at the receiver.

Thermal noise is generated at the forepart of the

receiver. For bandwidth-limited channels, the ISI has

been recognized as the major drawback in high speed

data transmission over wireless channels. The

traditional band restricted filters fail to recover the

information once the received symbol contains ISI and

in-band noise. The Inter-Symbol Interference will be

removed by victimization equalization techniques.

Generally, the term equalization is used to describe any

signal process operation that minimizes the ISI [1]. An

equalizer within the receiver compensates for average

range of expected channel amplitude and delay

characteristics. Equalizer algorithm, equalizer structure

and the rate of modification of the multipath radio

channel are three main factors that affect the time

spread over that an equalizer converges. Two vital

problems in equalizer design and implementation are its

complexity and its training. For frequency selective

channel, the equalizer enhances the frequency

components with little amplitudes and attenuates the

sturdy frequencies within the received frequency

response and for a time-varying channel.

The mobile weakening channel could be a random and

time varying; equalizers should track the time varying

characteristics of the channel, and therefore referred to

as adaptive equalizers. Adaptive channel equalization is

an efficient tool in mitigating Inter-Symbol Interference

(ISI) caused by linear distortions in unknown channels

[2]. An adaptive filtering algorithmic program needs the

information regarding the “known” response so as to

estimate the error signal required for adaptive method.

In follow, the noted signals are often generated at the

receiver aspect in two ways. In first technique, the

transmitted training sequence is retrieved by the replica

of the known response that is stored within the receiver.

The synchronization should be done between

transmitted training sequence and better-known

response that is hold on within the receiver. With a

famed training sequence, the adaptive filtering

algorithmic program used to regulate the equalizer

coefficients, which correspond mathematically to

checking out the distinctive minimum quadratic error

performance surface. The second technique may be a

decision directed technique, in which, instead of the

known training sequence, a sequence of data symbols

are estimated from the equalizer can be used. This

estimated output could also be unreliable, so this may

not enable the tap weight coefficients to be optimized.

In general, the computation of error estimation is done

with the aid of the input vector and desired response,

and it is used to make the control over the adjustable

filter coefficients values. Depending on the filter

structure chosen, the adjustable coefficients may be in


1193 K SUTHENDRAN AND T ARIVOLI



the form of tap weight reflection coefficients or rotation

parameters[2]. However, the elemental distinction

between the various applications of adaptive filtering

arises in the manner in which the required response is

extracted.

Training and tracking are the two general operational

modes of an adaptive equalizer. First, a better-known

training sequence -pseudorandom binary signal of fixed

length is sent by the transmitter. With this, the equalizer

at the receiver facet may adapt to a correct weight for

minimum bit error rate (BER) detection. Following this

training sequence, original data is sent and adaptive

equalizer utilizes the recursive algorithm to evaluate the

channel, and thus estimates the filter coefficients to

compensate the distortion created by multipath in the

channel. Equalizers require periodic training in order to

keep up effective ISI cancellation. In data

communication systems, user data is often segmented

into short time blocks or time slots. Time division

multiple access (TDMA) wireless systems are notably

well matched for equalizers. Due to time varying nature

of wireless channels, training signals should be sent

often and this occupies additional bandwidth. In several

applications better-known training sequence is needed

to adapt the equalizers by minimizing the mean square

error [MSE], however this being impractical and

expensive when long training sequence is necessary [3].

As an example, in step with 900MHz GSM standard, 26

bits out of each 148 bit frame are used as training

signals [4].

Recently, many researchers have aimed to enhance the

convergence rate and to reduce the steady state

estimation error based on variable step size algorithms

[5], [6], [7] and [8]. They have proposed totally

different criteria for the variable step size. R.Kwong and

E.W. Johnston in [5], regulate the variable step size

based on instantaneous squared predication error. In [7],

the variable step size is updated using the changes of

successive samples gradient. In [8], the authors have

projected a variable step size to reduce the squared error

at every step. Mayyas et al. in [6] have proposed a

variable step size algorithm that has quick convergence

at the early stages of adaptation. All these algorithms

are based on Least Mean square algorithm. In this

paper, a replacement variable step size algorithm is

proposed based on Sato’s blind algorithm.

2. Blind Algorithm:

Even though trained strategies have several

disadvantages, they are often adequate. The throughput

of the system drops due to the time slots occupied by

the training signal. Another disadvantage is that the

training signal is not forever known at the receiver, e.g.,

in a non cooperative (surveillance) surroundings.

Finally, the quicker time variable channel needs training

sequence more often to train the equalizer. This leads to

further reduction in the throughput of the system[9].

The Blind algorithms are ready to exploit characteristics

of the transmitted signals and do not need training

sequences. They are called so because they provide

equalizer convergence without burdening the transmitter

with training overhead. These trendy algorithms are

ready to acquire equalization through property restoral

techniques of the transmitted signal. In general, even

when the initial error rate is large, blind equalization

methodology directs the coefficient adaptation process

towards the optimal filter parameters. A Blind Equalizer

is ready to compensate amplitude and delay distortions

of communication channel victimization only the

channel output samples and the knowledge of the basic

statistical properties of the data symbols. The foremost

advantage of blind equalizers is that there is no training

sequence to calculate the tap weight coefficients;

therefore no bandwidth is wasted by its transmission.

Blind equalization is effective for a high-speed digital

radio, GPS, digital mobile communication systems,

multi-point networks, cable TV, and digital terrestrial

TV broadcasting [10], [11], [12] and [9].

The major downside is that the equalizer can generally

take an extended time to converge as compared to a

trained equalizer. The necessity for blind equalizers in

the field of information communications is greatly

mentioned by Godard [13], in the context of multipoint

networks. Blind joint equalization and carrier recovery

may find application in digital communication system

over multipath weakening channels. Moreover, it is

applied in extremely non-stationary digital mobile

communications, wherever it is impractical to use

training sequences. These techniques include algorithms

such as the Sato algorithm and Constant modulus

algorithm (CMA).

2.1. Sato’s Blind Algorithm:

Sato was the one who first introduced the concept of

blind equalization in 1975 for multilevel pulse

amplitude modulation, where there is no reference

sequence accessible and afterward Godard combined

Sato’s idea with a decision Directed (DD) algorithm and

obtains a brand new blind equalization scheme for

Quadrature Amplitude Modulation(QAM) data

transmission. Since blind equalization has attracted vital

scientific interest due to its potentials in terms of

overhead reduction and simplification of point to

multipoint communication. Sato proposed algorithm

which was designed only for real valued signal and

PAM [14]. However, its complex valued extension is

straightforward, which was derived by Godard. The cost

function proposed by Sato is given in (1),

)})(.{()( 2

kk

sato

ysignyEAJ (1)

1194 Variable Tap Parameter (α) Techniques for Sato based Blind Equalizer



Where,

ky = output of the equalizer

0,1

0,1)( k

kk

y

yysign (2)

|)(|

)( 2

k

k

aE

aE

(3)

Sign denotes the usual signum function of a real scalar.

γ referred as scaling factor and ak denotes the input data

sequence.

Fig1: General block diagram for Blind equalizer

Fig. 1 shows the general block diagram of the Blind

Equalizer. It seems that Sato’s proposal appears to be

developed over Least Mean Square (LMS) algorithm,

which uses steepest decent criteria for convergence

method. Mathematically, differentiate any equation and

equate it to zero, then this offers the minimum; work it

to the steepest-descent criteria; it gives the tap weight

coefficient for the equalizer. So, differentiate (1) and

substitute it to the steepest-descent criteria and getting

(4) as shown below. The algorithm of Sato’s blind

equalization is based on (4), which is used for training

the output sequences[14].

)](.[.^

1

^

kkkkk ysignyrAA (4)

Where,

Ak = Weight used for training

α = Tap-adjusting coefficient

yk = Output sequence

rk = Input sequence

and

1.

k

i

ik xar (5)

Since this algorithm works under iteration basis, at

every iteration it tries to adapt its output sequence to the

self realized input sequence. Thus, it is also known as

self-learning equalizer. The convergence rate and

precision to output sequence are the two main design

considerations in Sato’s blind equalization. To get the

best result from Sato’s algorithm, the design

considerations should be optimized on the basis of its

parameters, in order that it will converge in no time with

a high precision output sequence. This can be more or

less guided by tap-adjusting coefficient ‘α’, because the

remaining parameters are not variable according to [15]

Fig2: The Sato based Blind equalizer with 5 taps

3. Proposed Algorithm:

In Sato’s blind equalization algorithm the initial value

of the tap parameter (α) is chosen between the minimum

and maximum values and these ranges of values are

finite to confirm the convergence and stability of the

algorithms. The fastened αmax is chosen with respect to

the stability condition of the algorithm, while αmin is

chosen to confirm desired steady-state performance

[16], [17] and [18]. In proposed approach the tap

parameter value starts with 0.0006 (optimum value

identified by Sato for PAM signal) to reconstruct the

very first symbol and this value is incremented by small

constant (s) for each iteration. The output distinction

between successive iterations is calculated and using

this value the iteration for reconstruction of very first

symbol is stopped. The updated tap parameter value is

chosen as beginning value for subsequent symbols. In

the reconstruction of the subsequent symbols α value is

decremented by small constant at each iteration and,

using this value the iteration is stopped. For first symbol

estimation, the specified SNR can be achieved by

changing the output difference value to stop the iteration

and for subsequent symbols estimation; by changing

αmin value the specified SNR output can be obtained.

The flowchart of the projected algorithm shown in fig.3

provides the detailed

nk

a

k

A

k

r

k Decision

Circuit

BlindEq

ualizer




Fig3: Flow Chart for Proposed algorithm

4. Simulation Results:

The performance of the improved blind algorithm has

been studied for PAM symbols as done by Sato. The

PAM symbols, having ISI with five reflections with

relative amplitude (0.7, 0.6, 0.5, 0.3 and 0.1) as shown

in fig. 4 for symbol 3and Additive White Gaussian

Noise (AWGN) with 25dB SNR, are taken as the input

to the equalizer. The equalizer has been implemented by

a linear transversal filter with a five tap complex

circuitry as shown in fig.2.

Experiment1: Constant α (αconstant)

Small α value results minimum steady state error but

results in slow convergence. High α value will speed-up

the convergence however lead to maximum

misadjustment[16].The variable tap parameter value is

restricted to the range [αmin=0.0006, αmax=0.18] to

guarantee stability of the algorithm.

Fig.4. ISI model for Symbol 3

Experiment 2:α with Linear increment and decrement

(αlinear)

For the above mentioned input, the parameter value

chosen as 0.0006 and tap weights are initialized with

center tap (only center tap has one others are near zero)

[19] and [20]. The very first symbol is reconstructed by

using linear increment in α. i.e., α is incremented by

constant factor (s=0.02) for every iteration as

αk+1 = αk + s; where s=0.02

The output distinction between current iteration and

former iteration is calculated at in three completely

different sampling points and if all the three output

values are less than 0.001 (based on experimental

analysis, output distinction value 0.001 is chosen as

better value to stop the iteration) means that the iteration

for reconstruction of very first symbol is stopped. The

updated α =0.3006 is fixed as optimum or starting value

for subsequent symbol. In the reconstruction of the

following symbols the α is decremented by a the same

factor (say 0.02) at each iteration as

αk+1 = αk – s; where s=0.02

When α reaches 0.0006, the iteration is stopped. On trial

and error basis, for constant α input the optimum α

valueis 0.18. But in proposed approach it is found

0.3006 as αmax and with updated tap weights consequent

symbolsreconstructed with 30dB SNR in few iterations

with stability. While giving α as 0.3006 with center tap

initialization for experiment 1 means, it ends up in

misadjustment and subsequent symbols can’t be

reconstructed moreover.

Experiment 3:α with Nonlinear increment and

decrement (αnonlinear)

For an equivalent input, step size parameter value

chosen as 0.0006 and tap weights are initialized with

center tap. The very first symbol reconstructed by using

nonlinear increment of step size parameter value. That

is step size value calculated with the assistance of

iteration count as




αk+1 = αk + count * 0.001

Count value is incremented at every iteration. The

output difference between current iteration and previous

iteration is calculated in three different sampling points

and if all the three output values are less than 0.001

means that the iteration for reconstruction of very first

symbol is stopped. The updated step size parameter

value (0.3006) is fixed as optimum or beginning value

for future symbol. In the reconstruction of the next

symbol the step size parameter value is decremented in

nonlinear method as

αk+1 = αk - count * 0.001

Using α (when it less than or equal to 0.0006) the

iteration is stopped. By varying the output difference

value and step size parameter value the required

SNRoutput can be obtained.

The waveforms shown in Fig. 5, Fig. 6 , Fig. 7 and Fig.

8 are the results of simulations for received symbol 3

(with Noise), self realized output symbol 3 by using

Blind approach, self realized output symbol 3 by using

variable α Blind approach and MSE comparison

between Blind and variable α Blind approaches

respectively.

Fig5: Received Symbol 3 with Noise

Fig6: Self Realized Symbol 3 by Blind Approach withα

=0.6X10-3

and SNR=30dB (22305 iterations)

Fig7: Self Realized Symbol 3 by variable α Blind

Approach with SNR=30dB (22 iterations).

Fig8: Mean Square Error comparison between Blind

and variable α Blind Approach.

Table 1, Table 2 and Table 3 shows the number of

iterations taken by Sato’s blind algorithm , proposed

variable α approach (Linear) and (Non Linear), with

different Signal to Noise ratio value for the

reconstruction of symbol 1, 2 and 3 respectively. In this

paper, the same tap adjusting coefficient value(α =

0.6x10-3

) is used as proposed by Sato to reconstruct the

PAM signal. For a variable α blind approach, better

convergence is obtained as shown in table 2. The

Simulation results show that the proposed variable α

blind approaches has increased the convergence rate

compared to that of Sato’s blind algorithm. That is, the

number of iterations to obtain the same output SNR for

identical symbol is much lesser in the variable α blind

approach.

Table1: Comparison of SNR vs. Iterations for Sato

based Blind Equalizer with Step Size α = .0006

SNR

in dB

Number of Iterations for Sato Blind

Approach with α = 0.0006

Symbol 1 Symbol 2 Symbol 3

10 154 135 278

15 604 506 1514

20 1312 1999 3936

25 2149 7331 6810

30 15503 15042 22305




Table2: Comparison of SNR vs. Iterations for Variable

α Blind approach (Linear)

SNR

in dB

Number of Iterations for Variable α

Blind approach(Linear)


10 4 1 16

15 7 4 17

20 9 20 18

25 12 21 19

30 15 22 22

Table3: Comparison of SNR vs. Iterations for Variable

α Blind approach (Non Linear)

SNR

in dB

Number of Iterations for Variable α

Blind approach(Non Linear)


10 9 1 19

15 13 21 20

20 17 21 20

25 21 21 20

30 24 23 23

5. Conclusions:

SNR and MSE are used to estimate the quality of the

reconstructed symbols, which can even be used to stop

the tap weight calculation or stop the iterations for blind

equalizers. Increase in α value gives a much better

convergence with increased mis adjustment i.e., very

high α value does not converge. Using variable α

values, the quicker convergence can be obtained with

minimum mis adjustment. In this paper, variable tap

parameter (α) techniques were proposed for Sato based

blind equalizer and also utilized two different methods

to stop the iterations in these proposed techniques. First

method uses the differences in successive outputs and

second method uses particular tap parameter (α) value.

The simulation results show that using the proposed

techniques, the desired SNR can be obtained with less

number of iterations and also with minimum steady

state error compared with Sato’s blind algorithm. From

the results it is observed that variable αlinear offers

quicker convergence than variable αnonlinear. But the

computational complexity of proposed technique is

slightly higher than Sato’s blind algorithm. The

computational time taken for reconstructing three

symbols sequentially by Sato algorithm is 5.989076

seconds and by proposed algorithm is 3.674535

seconds.

6. Future Work:

The convergence rate of proposed algorithm is very

much encouraging .The only disadvantage of Sato’s

algorithm is that it recover only single carrier, whereas

in practice the most communication system employs

dual carrier modulation systems, like quadrature

amplitude modulation. This limitation is overcome by

Godard proposal [13].By using variable tap parameter

technique convergence of Godard algorithm can also be

improved.


The authors are grateful to Editor in Chief Dr.

D.V.Reddy and Editorial board for processing our work.

Thanks are also due to reviewers for their valuable

suggestions to improve our work.

2. Reference:

[1] S.U.H.Qureshi, “Adaptive Equalization”, in proc.

IEEE, vol. 73, pp. 1349-1387, September1985.

[2] Rappaport Theodore S. Wireless communications:

principles and practice, 2nd ed., Pearson Education,

India, 2010.

[3] Ye Li and Zhi Ding, “Convergence Analysis of

Finite Length Blind Adaptive Equalizers” IEEE

Transactions on Signal Processing, vol. 43, no. 9,

pp. 2120-2129, September 1995.

[4] M. Hodges, “The GSM Radio Interface”, British

Telecom Technological Journal, vol. 8, no. 1, pp.

31-43, 1990.

[5] R.Kwong and E.W Johnston, “A variable step size

LMS algorithm,” IEEE Trans. Signal Processing,

vol 40, pp.1633-1642, July 1992.

[6] T.Aboulnasr and K.Mayyas, “aroubust variable

step-size lms-type algorithm: Analysis and

simulations,” IEEE Trans. Signal Processing, vol

45, pp.631-639, Mar.1997.

[7] R.Harris, D. Chabries, and F.A. Bishop, “ A

variable step(VS) adaptive filter algorithm,” IEEE

Trans. Acoust., Speech, Signal Procssing, vol.

ASSP-34, pp. 309-316, Apr. 1986.

[8] V.J. Mathews and Z. Xie, “ A stochastic gradient

apadtive filter with gradient adaptive step size,”

IEEE Trans. Signal Processing, vol 41, pp.2075-

2087, June 1993.

[9] Yun Zhao, Xiaonan Xue, Tingfei Zhang "Receiver-

channel based adaptive blind equalization approach

for GPS dynamic multipath mitigation", Chinese

journal of Aeronautics, vol. 26(2), pp.378-384,

March 2013.

[10] Kil Nam Oh and Jae Hong Park, “Property Restoral

Approach to Blind Equalization of Digital

Transmission Channels”, IEEE Transactions on

Consumer Electronics, vol.43, no.3, August 1997.

[11] K. Suthendran, V.R.S.Mani, V.Vijayarengan,

“Design of Blind Equalizer,” in Proc.

NATCON’06, paper 43, pp. 206-212, 2006.

[12] K.Suthendran, T.Arivoli, “Performance comparison

of adaptive and blind equalization algorithms for




wireless communication,” Bonfring International

journal of research in communication engineering,

vol 3, pp.1-6,March 2013.

[13] Dominique N. Godard, “Self-Recovering

Equalization and Carrier Tracking in Two

Dimensional Data Communication Systems”, IEEE

Transactions on Communications, vol.com-28,

No.11, pp. 1867-1875, 1980.

[14] Yoichi Sato, “A Method of Self-Recovering

Equalization for Multilevel Amplitude-Modulation

Systems”, IEEETransactions on Communications,

pp. 679-682, 1975.

[15] S. GuoYecai, He Longqing and Zhang Yanping,

“Design and Implementation of Adaptive Equalizer

Based on FPGA”, The Eighth International

Conferenceon Electronic Measurement and

Instruments (ICEMI,) pp. 790-794, 2007.

[16] Zhao Shengkui, Man Zhihong and KhooSuiyang,

“A Fast Variable Step-Size LMS Algorithm with

System Identification”, Second IEEE conference on

Industrial Electronics and Applications 2007, pp,

2340-2345, May 2007.

[17] Xue Wei, Yang Xiaoniu and Zhang Zhaoyang, “A

Variable step size Blind equalization algorithm for

QAM signals”, in Proc. IEEE ICMMT, pp.1801-

1804, 2010.

[18] Yuan Gao and XinyunQiu, “A new variable step

size CMA blind equalization algorithm”, 24th

IEEE

Chinese control and decision conference, pp.315-

317, 2012.

[19] David Smalley, “Equalization Concepts: A

Tutorial”, Application Report, Atlanta Regional

Technology Centre, SPRA140, October 1994.

[20] A. Benveniste and M. Goursat, "Blind equalizer."

IEEE Transactions on Communications, vol. COM-

32, pp.871-883, Aug. 1984.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1199-1204


Inverse Deduction of Unsaturated Soil Parameters Based on RBF

Neural Network Model

LIU J X1,2

, W LIU2, Y T LIU

1 AND NI JUNJIE

3

1School of Civil Engineering and Architecture, Southwest University of Science and Technology, Mianyang,

Sichuan, 621010, China 2State Key Laboratory of Geomechanics and Geotechnical Engineering, Institute of Rock and Soil Mechanics,

Chinese Academy of Sciences, Wuhan, 430071, China 3Administration of Coal Industry of Hechuan district of Chongqing, Chongqing, 631520, China


Abstract: In order to calculate the seepage time of model test unsaturated parameters of red layers soil fillers with

different compaction states, finite element method (FEM) model of unsaturated water-gas two phase flows was

established and analyzed. The unsaturated parameters were inversely deducted based on radial basic function (RBF)

neural network model. The numerical simulation results for input targets along with different levels of parameters

for output were considered. By varying the compaction parameters and initial water saturation corresponding to real

experimental models, numerical simulations on seepage of real progress of model test were carried out. The results

indicated that the seepage time of the numerical simulations and the inverse model tests are exactly the same by

comparison, which shows that it is high feasible to inversely deduct unsaturated parameters by use of the RBF neural

network model.

Keywords: unsaturated soil, seepage time, RBF neural network model, numerical simulation, inverse deduction of

parameters.

1. Introduction:

With the high-speed development of massive

constructions in China, more and more engineering

projects face up to unsaturated soil, such as building

foundations, soil slopes and roads embankments and so

on. The accurate investigation of the physical and

mechanical characteristics of unsaturated soil is the

basic to guarantee the stability and safety of the

engineering. The determination of the soil-water

characteristic curve and permeability coefficients of the

unsaturated soil is the key issue to put the theory of

unsaturated soil into engineering application [1, 2]. To

determine these parameters, uniformly, the direct

method is used, that is, carrying on the humidification

or dehumidification and permeability experiments of the

unsaturated soil to chase relative parameters. However,

these operations occupy a long time to accomplish the

experiments, and it is also high cost of money and high

demand of equipment, resulting in many difficulties to

apply the theory of unsaturated soils to the practical

productions. Therefore, it has become an urgent and

necessary target to look for an alternative and available

method for the determination of these parameters.On the

other side, the artificial neural network is a typical non-

linear dynamic system, which is characterized with

good adaptability, self-organization, strong self-

learning, association, fault tolerance, and anti-

interference ability [3]. It is very convenient to establish

models to analyze the unknown complex system with

multiple causes. Compared with other neural network

models, the artificial neural network has many obvious

advantages, such as approximation ability, classification

ability and learning speed. In addition, it can avoid the

phenomenon of local minimum [4], so it is widely used

in back-analysis of geotechnical engineering [5]-[8].

2. RBF network structure

The structure of the Radial basis function (RBF)

network structure is shown in Figure 1, which is

composed of three layers of neurons. The first layer is

the input layer, the second layer is the radial base (RBF

layer), and the third layer is the output layer of linear

[9].

Nx

1x

2x

3x

The input layer The output layer RBF layer

1d

Md

2d

Fig.1 RBF network structure, i

x is the input information

and di is output information

1200 Inverse deduction of unsaturated soil parameters based on RBF neural network model



It is assumed that the input neuron number is N, and

then RBF node layer neuron number is R, meanwhile,

the output node layer neuron number is M; the

connection weights between radial primary neurons j

and input layer neurons i is ji

w,

, and the weight vector

between the radial primary neurons j and the input layer

neurons is as following:

),...,(,2,1, Njjjj

wwww Rj ,...,3,2,1 (1)

Subsequently, the connection weight matrix between the

radial primary neurons and input layer neurons can be

expressed as:

1 1 2( , , )T

RW w w w (2)

1x

2x

3x

Nx

1,1 1,Nw w

w x

+

b

z y

Fig.2 RBF structure of the neurons

The radial basis, whose structure of neurons is shown in

Figure 2, and the radial basis function is used as the

activation function. The pure linear function is used as

the activation function of the linear output layer.

3 Inverse deductions of unsaturated parameters

3.1 Model production and design

Figure 3 illustrates the soil column infiltration test

equipment, which is 0.5 m in height, 0.1 m in diameter,

and with a constant head of 130.54 cm in the low part of

the specimen applied. The control of constant water

head is completed by the electronic valve and a water

level controller, and an electronic trigger is located on

the top of the specimen. When water seepage starts at

the top of the specimen, electronic trigger poles contact

water seepage, so that the electronic trigger can observe

short circuit. Subsequently, table automatically stops

and records water seepage time. The particles of all the

samples used for red clay soil were pulverized to be less

than 2 millimeter, according to the soil compaction test

in "Specification of Soil Test" (SL237-1999), the

maximum dry density is 1.855 g/cm3, and the optimum

moisture content is 12.81 %. According to "Handbook

for Design of Highway Subgrades", the permeability

test was carried out on the specimens’ compaction of 87

%, 90 %, 93 %, 95 % and 98 %, and each degree of

compaction (defined as the ratio of controlling dry

density to the maximum density) was carried out into

four groups for testing, according to different initial

water saturations. By series computation, the test results

were shown in Table 1.

Fig.3 Schematic diagram of test device

Table 1 Test results for the inverse deduction

No. name 87% 90% 93% 95% 98%

1 initial saturation 0.42 0.48 0.53 0.51 0.49

time/×104s 2.604 5.794 11.680 41.800 98.930


time/×104s 2.478 5.507 11.111 37.237 82.850


time/×104s 2.353 5.219 10.542 32.673 66.770


time/×104s 2.227 4.932 9.973 28.110 50.690

3.2 Finite element simulation

In order to obtain the trained samples, the course of soil

column test is analyzed by finite element analysis,

realizing by using the ANSYS simulation system. The

equation of unsaturated soil-water characteristic curve

and coefficient of permeability are suitable for model of

Van Genuchten et al. [10], whose expressions are shown

as follows:

1201 LIU J X., W LIU, Y T LIU AND NI JUNJIE



0.5 1/ 2

0.5 1/ 2

1/ 1

0

( ) [1 (1 ( ) ) ]

(1 ( )) [1 ( ) ]

(( ) 1)

1

w w w w a a

s e e

g w w w a aws e e

g

w a a

c e

ww w re w

r

k k S S

k k S S

p p S

S SS

S

(3)

Where w is the water dynamic viscosity (pas); g is

the gas dynamic viscosity (pas), which is a known

value, so they can be obtained from the form manual; w

sk is the saturation of "1" water permeability

coefficient, and it can be obtained from conventional

permeability test; w

rS is the residual saturation, and it

can be determined by the conventional methods (see

reference [11] for more details). Therefore, the

unknown parameters are only 0p and a .

Table 2 is test parameters obtained from conventional

tests under different degrees of compaction.

Table 2 test parameters under different degrees of

compaction

name 87 % 90 % 93 % 95 % 98 %

porosityn

0.42 0.40 0.38 0.37 0.35

w

SK /ms-

1 6.79E-7

2.81E

-7

1.33

E-7

5.09

E-8

2.75E

-8

w

rS /% 3.33 4.04 4.60 4.88 6.05

It is considered that the parameters 0p is 40 kPa, 45

kPa, 50 kPa, 55 kPa and 60 kPa, a is 0.20, 0.24, 0.28,

0.32, 0.36 and 0.40, the total of 0p and a is up to 36,

which are formed to the orthogonal mode. In order to

obtain the different parameters of water seepage, the test

process was numerically simulated. Figure 5 shows the

grid of numerical model (10×100). The pressure head

h=1.3054 m is applied to the bottom of model. The top

is the seepage boundary, which keeps the top gas

pressure constant as zero, so as to allow the gas phase to

flow.

The calculation process ran under the condition of

circulating mode, and the condition is the average value

of the top node saturation, when the value is less than

0.99, the iterative algorithm of next period (20s) will be

conducted, or that the top node is degree of saturation,

the top of the water has seeped.

X/m

Y/m

0 0.05 0.10

0.1

0.2

0.3

0.4

0.5

Frame 001 16 Jan 2007 Flac2d Mesh to Tecplot Version 10

Fig.5 grid of numerical model

Table 3 is the water seepage time under different

degrees of compaction, different initial saturation and

different combination of the parameters from the

numerical simulation results.

Table 3 numerical water seepage time of combination of parameters under initial saturations corresponding of

different degrees of compaction

Compaction

coefficient 87% 90% 93% 95% 98%

Initial saturation 0.42 0.5 0.48 0.55 0.53 0.59 0.51 0.63 0.49 0.66

p0/Pa a time/×104s time /×104s time /×104s time /×104s time /×104s

40000 0.2 7.636 5.218 13.310 9.900 21.740 17.140 60.520 37.550 119.100 71.850

40000 0.24 4.735 3.662 9.039 7.295 15.640 13.030 42.420 29.330 80.310 55.110

40000 0.28 3.588 2.941 7.146 5.993 12.690 10.870 34.020 24.790 63.190 46.070

40000 0.32 2.985 2.518 6.064 5.189 10.930 9.486 29.110 21.810 53.520 40.350

40000 0.36 2.638 2.258 5.423 4.683 9.829 8.600 26.100 19.840 47.850 36.650

40000 0.4 2.456 2.122 5.089 4.430 9.271 8.200 24.570 18.900 44.930 34.900

45000 0.2 7.868 5.300 13.590 10.000 22.060 17.300 61.570 37.700 121.700 71.900




45000 0.24 4.747 3.640 9.011 7.230 15.530 12.900 42.230 29.000 80.180 55.100

45000 0.28 3.532 2.880 7.011 5.860 12.430 10.600 33.330 24.200 62.050 46.100

45000 0.32 2.901 2.440 5.879 5.030 10.590 9.190 28.200 21.100 51.990 40.400

45000 0.36 2.586 2.210 5.302 4.580 9.614 8.410 25.550 19.400 46.840 36.700

45000 0.4 2.419 2.090 5.025 4.380 9.129 8.040 24.210 18.700 44.290 34.900

50000 0.2 8.085 5.390 13.850 10.100 22.340 17.400 62.590 37.800 124.300 72.700

50000 0.24 4.766 3.630 8.996 7.180 15.440 12.800 42.050 28.600 80.050 53.900

50000 0.28 3.482 2.830 6.890 5.740 12.190 10.400 32.750 23.700 61.030 44.000

50000 0.32 2.858 2.390 5.786 4.930 10.390 8.970 27.730 20.600 51.290 38.300

50000 0.36 2.548 2.170 5.217 4.530 9.457 8.300 25.220 19.200 46.260 35.500

50000 0.4 2.385 2.070 4.954 4.290 9.014 7.970 23.990 18.500 43.850 34.100

55000 0.2 8.289 5.470 14.110 10.200 22.630 17.500 63.570 37.900 126.600 73.100

55000 0.24 4.781 3.610 8.975 7.130 15.370 12.700 41.910 28.400 79.990 53.500

55000 0.28 3.439 2.780 6.783 5.640 11.970 10.200 32.200 23.200 60.140 43.200

55000 0.32 2.833 2.370 5.730 4.870 10.290 8.890 27.440 20.400 50.780 37.900

55000 0.36 2.533 2.160 5.196 4.480 9.429 8.260 25.080 19.100 46.140 35.300

55000 0.4 2.351 2.040 4.875 4.240 8.957 7.870 23.550 18.400 43.340 33.900

60000 0.2 8.484 5.550 14.110 10.300 22.900 17.600 64.480 38.100 128.900 73.500

60000 0.24 4.797 3.600 8.980 7.090 15.300 12.600 41.800 28.100 79.900 53.000

60000 0.28 3.418 2.750 6.780 5.560 11.800 10.000 31.900 22.800 59.800 42.600

60000 0.32 2.818 2.350 5.730 4.830 10.200 8.810 27.300 20.200 50.500 37.600

60000 0.36 2.505 2.140 5.200 4.460 9.330 8.200 24.900 19.100 45.800 35.200

60000 0.4 2.316 2.010 4.880 4.210 8.830 7.840 23.500 18.300 42.600 33.400

65000 0.2 5.620 5.620 14.600 10.400 23.200 17.700 65.300 38.200 131.000 74.000

65000 0.24 3.590 3.590 8.960 7.050 15.200 12.500 41.700 27.800 79.900 52.600

65000 0.28 2.740 2.740 6.710 5.540 11.800 9.990 31.800 22.600 59.800 42.500

65000 0.32 2.340 2.340 5.690 4.810 10.200 8.800 27.300 20.200 50.600 37.600

65000 0.36 2.130 2.130 5.140 4.420 9.310 8.140 24.800 19.000 45.300 35.100

65000 0.4 1.990 1.990 4.760 4.170 8.730 7.790 23.100 18.000 41.900 33.200

3.3 Parameters inversion

The numerical simulations of different saturated water

seepage time were used as input vectors, and then the

corresponding parameters 0p and a were used as the

output vectors. The different seepage time of initially-

saturated water by tests (its saturation corresponds to the

saturation of numerical simulation) was regarded as the

test input vectors. All data was processed by waveform

normalization into (0, 1) according to Eq. 4.

min

max min

x xx

x x

(4)

The normalized input vectors and output vectors were

presented to RBF neural network model for training. In

next, the normalized testing datasets were used to

compute the output so that the model prediction

capability can be tested. The prediction results were

conducted by anti-normalization at the same time, and

then the predicted parameters of different degrees of

compaction can be obtained, as shown in Table 5.

Table 5 predictive values of parameters under different

degrees of compaction

parameter 87% 90% 93% 95% 98%

p0/kPa 42.82 47.79 53.53 59.38 63.88

a 0.356 0.319 0.288 0.241 0.218

In order to evaluate whether the inversion parameters

are correct or not, the seepage time of the initially-

saturated water corresponds to different compaction

through the corresponding numerical simulation (the

saturation is different from the saturation of different

inversion), compared with the corresponding actual

water seepage time, as specified in Table 6. In the

Table, compared with the actual test results, the

http://dict.cnki.net/dict_result.aspx?r=1&t=degree+of+compaction&searchword=%e5%8e%8b%e5%ae%9e%e5%ba%a6

1203 LIU J X., W LIU, Y T LIU AND NI JUNJIE



maximum error of the numerical simulation results

occurs in the 98% degree of compaction. When the

initial saturation is 0.55, its value is 1.819%. According

to Table 5, one can find that the error increases along

with the increment of the degree of compaction. Due to

the increase of the degree of compaction, the decrease

of the permeability coefficient resulted in the decrease

of the top exudation water, which consequently caused

the time of electronic trigger has certain errors with the

actual time at the same time. The increase of degree of

compaction leaded to the difficulty of sample

preparation, so that the samples are different, and this is

one of the main reasons, but overall speaking, the

inversion results and the actual results are basically

consistent.

Table 6 Numerical simulation results in contrast with actual values of water seepage times under different degrees

of compaction

Compaction

coefficient Initial saturation

Actual results

/×104s

Results of numerical

simulation/×104s

Percentage

difference %

87% 0.45 2.478 2.496 0.713

0.47 2.353 2.369 0.694

90% 0.50 5.507 5.547 0.726

0.53 5.219 5.257 0.728

93% 0.55 11.111 11.010 0.909

0.57 10.542 10.450 0.873

95% 0.55 37.237 36.813 1.137

0.59 32.673 32.327 1.061

98% 0.55 82.850 84.357 1.819

0.60 66.770 67.813 1.563

4. Conclusions

The water seepage time of the numerical simulation

results are set as the input targets, and different ranges

of parameters are considered as output targets, which

are trained by RBF neural network model. The model

test results are set as the test input variables, and RBF

neural network model was used to get the parameters

inversion.

In order to evaluate whether the inversion parameters

are invalid or not, the corresponding numerical

simulations were carried out on the seepage time of

initially-saturated water corresponding to different

compactions of soil. And compared with the

corresponding actual water seepage time, the maximum

error of the water seepage time is as low as 1.819%,

which shows that it is feasible enough to inversely

deduct the unsaturated parameters by RBF neural

network model.

5. Acknowledgement

This research work was partially funded by the National

Science Foundation of China (NSFC) (No. 41272297),

technology support project and international

cooperation project (No.2013GZ0071 and No.

2014HH0007) of Sichuan, China.

6. Reference:

[1] Qi Guoqing, Huang Runqiu. An universal

mathematical model of soil-water characteristic

curve, Vol.2 No.12, Journal of Engineering

Geology, pp. 182-186.

[2] Liu Haining, Jiang Tong, Liu Handong, Indirect

determination of permeability function equation of

unsaturated soils, Vol.11 No.25, Rock and Soil

Mechanics, pp. 1795-1799.

[3] Zhang Zhiguo. Study on Artificial Neural Networks

and their Applications in Geoscience[Ph. D.

Thesis][D], Changchun, Jilin University, 2006.

[4] Chen Guoxing, Li Fangming. Probabilistic

estimation of sand liquefaction based on neural

network model of radial basis function, Vol.3

No.28, Chinese Journal of Geotechnical

Engineering, pp.301-305.

[5] Wang Xuhua, Evaluation and Prediction of Slope

Stability Based on Engineering Fuzzy Set

Theory[Ph. D. Thesis][D], Dalian: Dalian

University of Technology, 2005.

[6] V. Genuchten, M. Th, A closed form equation for

predicting the hydraulic conductivity of unsaturated

soils, Vol.44, Soil Science Society of American

Journal, pp, 892-898.

[7] C. J. Miller, N. Yesiller, K. Yaldo, S. Merayyan,

Impact of soil type and compaction conditions on

soil water characteristic, Vol.9 No.128, Journal of

Geotechnical and Geoenvironmental Engineering,

pp.733-742.

[8] Ahmat Nor N L, S. Harun, Mohd Kassim AH.

Radial basis function modeling of hourly






streamflow hydrograph. Vol.1 No.12, Journal of

Hydrologic Engineering, pp.113-123.

[9] Yang Zhengrong, A novel radial basis function

neural network for discriminant analysis 2006(03).

DOI: 10. 1109 / TNN. 2006. 873282.

[10] Idri, A. Abran, S. Mbarki, An experiment on the

design of radial basis function neural networks for

software cost estimation, Vol.1 No.1, Information

and Communication Technologies, pp. 1621-1617.

[11] Wang Z L, Li Y C, Shen R F, Correction of soil

parameters in calculation of embankment

settlement using a BP network back-analysis

model, Vol.91, No.2-4, Engineering Geology, pp.

168-177.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1205-1209


Integral Action of Slab and Beam in Grid Floor Systems

S VIMALA1 AND P GOPALSAMY

2

1Department of civil engineering, PSNA College of Engineering and Technology, Dindigul, Tamil Nadu, India

2Department of civil engineering, Anna University, Madurai, Tamil Nadu, India


Abstract: Engineers while designing the grid floor systems usually ignore the integral action of slab and grid beams

because of the complex behavior of these elements. In this paper the interaction between slab and beam is brought

into mathematical formulations. This interaction in (3m x 3m) grid system with fixed boundaries contributes to the

overall flexural strength of the systems by about 10 to 50% and it is dependent on the actions of width of floor to

beam width and depth of floor to beam heights. These aspects are clearly illustrated by means of graphical relations.

Suggestive ascriptions are ascribed to frame new code rules in the light of analytical studies presented in this paper

as Engineers usually underestimate the potential strength of grid floors by analyzing them as open grid systems.

Key words: Grid floor system, Slab and grid beam interaction, Joint force, Deflection.

1. Introduction:

Inhabitants of metropolitan cities particularly in

developing countries fostering modern culture are

invariably facing the problem of shortage in housing

because of the abundant influx of immigrants from

adjoining towns and villages. Nowadays, government

agencies like metropolitan development authorities in

developing countries are planning the mass housing

schemes to accommodate large number of people in the

limited available land- space, as ground cost might

corrode large portions of finance allocated for it. Mass

housing projects in dense inhabitations are envisaged to

have vertical expansion by providing storeyed

apartmental buildings.

While planning mass housing projects city and country

planning regulations stipulate the allocation of certain

portion of land public utilization by constructing

community halls, chapels, auditoria, marriage halls,

recreation halls etc. These buildings necessitate to have

unobstructed view and column free space inside the

building. This type of large span buildings could be

roofed by stressed skin structures, shells and grid floors.

Among these types of roofing grid floor systems are

having the unique advantages like flexibility to raise

further floors, creating aesthetically pleasing soffits,

cost reduction in construction and acquiring sites for

buildings.

An assemblage of beams spanning in different

directions with intersecting rigid joints, loads acting

perpendicular to the plane of the structure is known as

the interconnected grid system. When the beams

forming the grid lie in one and the same plane they are

called as flat grids. When the two sets of parallel beams

are lying in the one and the same plane; intersecting at

right angles, forming rigid joints at nodes, creating

rectangular meshes, then the resulting structure is called

as orthogonal grid system. Tiers of beams with rigid

intersecting joints form the layered grid systems.

In this paper a study on the effect of interaction forces

existing between slab and beam in 3 m x 3m square grid

system with fixed boundaries has been made.

2. Review of Literature:

Huber10

(1914) laid the base for the analysis of grids by

evolving plate bending differential equation to an

isotropic plate. Timoshenko25

(1940) realized the

mathematical similarity between plates and grid frames.

Guyon8 (1949) developed this concept of Timoshenko

to grids of zero torsional stiffness. Massonet15

(1950)

included the effect of torsion in his analysis. Ghosh5

(1958), Dutta and Som1 (1956) considered only the link

forces at the nodes ignoring rotations in the analysis

with the final results having appreciable variation from

those of exact methods.

Ewell-etal3 (1952), Reddy and Hendry

21 (1959) used the

moment distribution method to orthogonal grid frames,

facing the difficulty of solving too many simultaneous

equations.

Scordelis24

(1952) suggested a relaxation technique

obviating the need of solving simultaneous equations.

Reddy etal22

(1964), applied this procedure of Scordelis

to corner supported grids with the caption as

translational moment distribution. Ratnalikar and

Megre20

(1976) formulated a very slow convergent

Moment distribution process for Skew


1206 Integral Action of Slab and Beam in Grid Floor Systems



Grids.Paramasivam19

(1964) and Second author17

(1974)

have developed iteration procedures to grids similar to

Kani’s method of plane frame analysis. Reaction

distribution process originated by Fadir4 (1963) has its

own limitations of assumption of no torsion and

tediousness in approach as the number of translating

joints increases. Hendry and jager9 (1955) replaced the

transverse members of the frame by a continuous

medium, which in turn resulted in a number of linear

differential equations.

Ellington2 (1957) and Renton

23 (1964) used finite

difference calculus in their works with appropriate

assumptions to simplify the analysis. Virtual work

method developed by, Martin13

(1963) lacks with the

disadvantage of having the cumbersome analytical work

as the number of nodes increases. Under the head of

exact analysis Martin and Hernadez14

(1966) framed

equations, similar to slope- deflection equations of plane

frames, necessitating the use of computers. Light –

foot12

(1959), Kesava Rao11

(1969) and Pandit18

(1980)

evolved computer programmes for the analysis of grid

frames using slope – deflection equation only to limited

grid frames having medium degree of indeterminacy.

Ongoing through the procedures of analysis cited above,

it is found that they stressed the importance of the

analysis of open grid frames only and no attention has

been paid to study the interaction between the slab and

grid beams.

3. Importance of Slab and Grid Beam Interaction:

Even though the floor slabs do contribute to the strength

of the grid system, their presence is neglected because

of the complexity involved in analysis, behaviour at the

interface of the grid beams and floor slabs. On

reviewing the earlier works of analyses it is found that

the analysis based on the plate bending theory alone

takes into account to certain extent the presence of floor

slabs, whereas other methods ignore the contribution of

the floor slabs to overall stiffness of the grid floor.

Developing countries could not afford to lose

construction, materials because of the ignorance of the

knowledge of analyzing the interaction behaviour of

slab and grid beam. Here in this paper an attempt is

made to bring the interaction between floor / slabs and

grid frames into mathematical formulations.

Considerable amount of steel and concrete would be

conserved without foregoing the potential strength when

the proposed procedure of analysis is adopted in the

design office.

4. Effect of Floor Slab in Grid Floors:

The floor grid consists of floor slab and an open grid

frame with integral fixity all along the beam lengths.

This fixity with beam and floor would create horizontal

shear, joint moments and joint holding forces. As an

initial step in this regard the horizontal shear and joint

moments are neglected. Only the joint holding forces

are assumed to be present. These forces are otherwise

called as link forces between floor slabs and open grid

frame at the nodes. These link forces of equal and

opposite in nature would be acting in the upward

direction over the floor slab at nodal points and in the

downward direction at the nodes of open grid frame.

(Figure – 1).

As the proposed analysis is meant for grid floors with

fixed boundaries, initially a double trigonometric series

satisfying the boundary conditions for the deflection of

clamped edge plate of size ( a x b) is assumed as:

W= ∞ ∞

∑ ∑

m =1 n =1 1- Cos 2m II x

a

1- Cos 2n II y

B - (1)

1207 S VIMALA AND P GOPALSAMY



The parameters are to be determined from the

conditions of minimum potential energy of the system.

Q = (U- W) - (2)

is a minimum with respect to these parameters.

Where U = ] dx dy

W = P w (x, y) (P – external load over the plate)

Substituting U and W for Q and using the condition

, the following expression is obtained.

- (3)

Where (x1, y1) defines the position of load ‘P’

D denotes the Torsional rigidity of the plate

D = E h3

12 (1 –µ2)

In this equation E, h & µ are the Young’s modulus,

Thickness of plate and Poisson’s ratio respectively. The

co-efficient of could be estimated using the

equation (3).

After substituting the coefficient in Equation (1)

the deflection of the plate could be estimated.

5. Application to (3m x 3m) Square Grid Floor:

Maximum deflection in the floor slab of this (3mx3m)

grid with fixed boundaries could be treated as the case

of plated deflection for

Fixed edge conditions

Similarly the joint forces would cause upward

deflections of the plate due to joint

Force (Pj )

The net down ward deflection

of the plate = 6.76183x10-3 (P - Pj) a

2

Deflection of open grid in

The downward direction = 0.0763 Pj

As there is contact between the grid frame and the plate,

the resulting downward deflection of the plate and open

grid frame would be equal to each other.

Applying this Condition Pj = 0.07924 p

0 .07924+0.034(a/B)

α3

Where

α= Depth of beam / Plate thickness

B= Breadth of the grid beam.

Graphical relations (Figure-2) are obtained in

consideration of the variation in the ratios of floor width

to beam width and floor thickness to beam height. On

studying these relations, for beams with lower range of

the depth of beams (D/h ≤ 2) the interaction between

beam and slab is having its predominance to reduce the

deflection by about 50% when compared with the

corresponding open grid deflections. But very little

impact is observed in respect of interaction by about 10

percent in grids having higher values of beam depth

(D/h ≥ 7).

5.1 Illustration:

A grid floor of size 3m x3m is provided in a hall of size

15 m x 15 m with the following data. Calculate the

effect of floor slab in overall stiffness of the grid floor.

Data:

1208 Integral Action of Slab and Beam in Grid Floor Systems



Floor slab thickness (h) = .15m

Depth of beam (D) = .60m

Rib width of beam (B) = .25m

Parameter α = 0.60 / 0.15 = 4.00

Joint holding force = 0.0792p

(0.07924 + (0.034(15/.25) )

43

= 0.710 p

Engineers usually assume the nodal load as ‘P’ for

regular design work as they are ignoring the joint forces.

Now in this illustration the nodal force has assumed a

value of “.710 p” by the proposed analysis. Thus the

floor slab reduces the overall deflection of grid frame by

about 29%.

6. Conclusion:

Here in this paper a rigorous mathematical analysis has

been presented to account for the integral action of floor

slabs and grid frame. This procedure could be extended

to grid floors of larger size, but the volume if

computational work would be enlarging as the size of

grid floor increases. In the graphical relations it is seen

that the maximum deflections are reducing by about 10

to 50% to (3m x 3m) grid. As deflection surface is one

of the viable methods through which the estimation of

moments and shears in the structure is possible, the true

moments and shears to grid floor would be reduced to a

greater extent, when the interaction between slab and

beam is taken into account. Code rules may be framed

to allow for the reduction in the design stress elements

after elaborate analytical study of various types of grids

on the lines illustrated in this paper. This in turn aids to

conserve construction materials like steel and concrete

to a greater extent.


The authors are grateful to Editor in Chief

Dr.D.V.Reddy and Editorial board for processing our

work. Thanks are also due to reviewers Dr.P.Kathirvel

and Dr.V.G.Srisanthi for their valuable suggestions to

improve our work.

8. Reference:

[1] Dutta, A. and Som, P. – Analysis of Reinforced

Concrete Grid – The Indian concrete Journal –

December 1956- pp. 381 – 384.

[2] Ellington, J.P. and Mc Callian, H. – Moments and

Deflections in simply Supported Beam Grillages –

Aeronautical Quarterly – November 1957 – Vol.1–

pp. 360 – 365.

[3] Ewell, W.W. Okubo, S. & Abrams, J.I. –

Deflection in Grid Works and slabs – Trans. ASCE

–– 1952 – Vol.117 -pp. 869 – 890.

[4] Fadir, I. – Grid analysis by Reaction distributional

Method – Trans. ASCE – 1963– Vol.128 – pt.2 -

pp. 256-283.

[5] Ghosh, R.K. – Design of Simple Rectangular

Reinforced Concrete Grid – The Indian Concrete

Journal – June 1958 – pp.200 – 212.

[6] Gopalsamy, P. – Study on Interconnected Grid

systems- Thesis submitted to Madurai Kamaraj

university for Ph.D. degree – April 1985.

[7] Gopalsamy, P., Etal – Analysis of large span Grid

Floors for institutional Buildings – proceedings of

International symposium on school Buildings for

Afro Asian Countries – March 12-14 , 1986 –

CBRI- Roorkee, India - pp. 117 – 125.

[8] Guyon, Y. – Calculdes ponts Delles – Ann. Ponts

of Chausses– 1949 – Vol. 119 – pp. 553 – 612.

[9] Hendry, A.W. & Jager, L.G. – A General Method

for Analysis of Grid Frame Works – proc. Inst.

Civil Engineers (London- December (1955) – Vol.

4. pp. 937-971.

[10] Huber, M.T. – Die Ground Lagen Einer Rationellen

Berechnung Der Keru Z weise Bewhrten – Eisenbe

to r platten, zeitechrift Des 0 sterreichischen

ingenier U. Architekten – Vereiues– 1914 – vol.66

– no. 30.

[11] Kesava Rao – M.N. – Computer Analysis of grids –

Grillages and interconnected Girder systems – The

Indian concrete journal– January 1969- vol. 43 –

pp25 – 32.

[12] Light Foot, E. & Sanko, F – frame Works Resolved

by Generalized slope – Deflection – engineering –

January2 – 1959 – pp. 18- 20.

[13] Martin, J.B. – The Elastic Analysis of simple

Rectangular Grids by Virtual work method – The

structural engineer– June 1963 –– vol. 46 pp. 199 –

201.

[14] Martin, I. and Hernandex, J. – Orthogonal Grids

loaded Normal to their plane – proceedings of

ASCE. ST- January 1960 – pp.1- 12.

[15] Massonet, C. – Methods of calculations of Bridges

with Several Longitudinal Beams Taking into

Consideration Their Torsional resistance –

publication – International Association of bride and

structural Engineering– 1950 – vol. 10 – pp. 147 –

182.

[16] Megre, A.S. – Iteration method of Grids –

Institution of engineers (India) – July 1976 – Vol.

57– pp. 72 – 78.

[17] Meiyappan, Pl. and Gopalsamy , p. – A method of

Iteration to Grid Frames – National symposium on

Economy in construction- Madurai – March 1974 –

pp. A. 12-A. 19.

[18] Pandit, G.S. and Guptha, S.P. – Displacement

Approach to Irregular Grids – The Indian concrete

journal – September 1980 – pp. 238 – 244.

1209 S VIMALA AND P GOPALSAMY



[19] Paramasivan, V. Dravid, P.S. and Ramesh, C.K. -

Grid Analysis by Rotation contribution Method-

The Indian concrete Journal – July 1967 – pp. 263 –

292.

[20] Retnalikar, M. and Megre, A.S. – Analysis of Skew

Grids by Moment distribution – proc. Institution of

Engineers (India)– September 1976 – vol. 57 – pp.

102 – 120.

[21] Reddy, D.V. and Hendry, A.W. – A Rapid moment

Torque Distribution method for grid Frame work

Analysis – Civil Engineering and Public works

Review – July / August 1959– vol. 54- no. 637 –

pp. 867 – 872.

[22] Reddy, D.V., Krithivasan and Vengatraman, T.S.-

Analysis of Corner Supported Grid Frame work by

translational Moment distributional – The Indian

Concrete journal – October 1964 – pp. 383 – 388.

[23] Renton, J.D. – A Finite Difference Analysis of

Flexural – Torsional Behaviour of Grillages – inst.

Journal of Mechanical Science– 1964-– vol. 6 - pp.

209 – 220.

[24] Scordelis, A.C. – discussion on deflection in Grid

frame works and slabs by Ewell – Etal – Trans.

ASCE– 1952 - vol.117 – pp. 892 – 897.

[25] Timoshenko, S. – Theory of plates and shells – Mc

– Graw hill book company – NewYork-1940.

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1210-1215


Strength, Swelling and Durability Characteristics of Fly-Lime

Stabilized Expansive Soil-Ceramic Dust Mixes

AKSHAYA KUMAR SABAT AND BIDULA BOSE Department of Civil Engineering, Institute of Technical Education and Research, Siksha ‘O’ Anusandhan

University, Khandagiri Square, Bhubaneswar, OR, India


Abstract: Strength, swelling and durability are the three important characteristics to judge the efficiency of a

stabilized soil. Construction on expansive soil creates problems because of its alternate swell-shrink behavior and

low strength. Stabilization, using industrial wastes with binder, is one of the different methods to increase the

strength and reduce the swelling characteristics of the soil. However, seasonal climate change exposes the stabilized

soils to alternate wet-dry cycles, which affect these properties severely. Unconfined compressive strength, California

bearing ratio, swelling pressure and durability of expansive soil-ceramic dust mixes stabilized with an optimum

percentage of fly ash and lime have been discussed in this paper.

Keywords: strength, swelling, durability, expansive soil, fly ash, lime, ceramic dust.

1. Introduction:

Expansive soil also called as black cotton soil is a good

soil for farmers for growth of cotton crop but a bad soil

for civil engineers for construction. It expands

significantly when comes in contact with water and

shrinks when water squeezes out. This alternate swell-

shrink behavior of the soil damages lightly loaded

structures severely. Cracks of different shapes and

varying depths are found in pavements, buildings, canal

linings, retaining walls etc. These soils occupy more

than 20% of total soils of India and mostly found in the

states like, Karnataka, Uttarpradesh, Maharastra,

Andhrapradesh etc. Some prominent techniques to

counteract the damaging effect of this soil are,

prewetting, application of surcharge, sand cushion,

cohesive non swelling soil(CNS) cushion, under reamed

pile foundation,granular pile anchor, mechanical

stabilization, chemical stabilization, stabilization using

industrial wastes with or without a binder etc. Soil

stabilization may be defined as any process by which a

soil material is improved and made more stable,

resulting in improved bearing capacity, increase in soil

strength, and durability under adverse moisture and

stress conditions [1]. Three important characteristics to

assess the effectiveness of a stabilized soil are its

strength, swelling behaviour and durability. The

durability of a stabilized soil is assessed in the

laboratory by subjecting it to alternate wet-dry (W-D)

cycles simulating the worst field conditions.

Fly ash is an industrial by-product, produced by the

combustion of pulverized coal in thermal power plants

during production of electricity. Ceramic dust is another

industrial waste produced in ceramic industry, the

disposal of both the wastes creates a lot of

geoenvironmental problems. These two wastes can be

utilized in the stabilization of expansive soil along with

a binder like lime which, not only will save the

environment from their polluting effects, but also will

save a lot of valuable land, which could have been

utilized for their disposal. The positive effects of

stabilization of expansive soil using industrial wastes

with binders have been documented, very well in

literature. Some of the prominent wastes with binders

which have been successfully utilized for stabilization

of expansive soil are, rice husk ash-marble dust[2], rice

husk ash - lime [3], rice husk ash - cement[4,5], quarry

dust-lime[6],brick dust-lime[7],fly ash-lime[8],fly ash-

cement[9],granulated blast furnace slag-cement[10],red

mud -cement[11], bagasse ash - lime sludge[12],etc. To

the knowledge of the authors no investigation has been

reported so far, regarding the stabilization of expansive

soil using fly ash-lime and ceramic dust.

The objective of the present investigation is to study the

unconfined compressive strength at 0, 7 and 28 days of

curing, California bearing ratio, swelling pressure and

durability of fly ash -lime stabilized expansive soil-

ceramic dust mixes.

2. Materials and Methods:

The materials used and the methods adopted in the

experimental programme have been discussed below.

2.1 Materials


1211 AKSHAYA KUMAR SABAT AND BIDULA BOSE



The materials used in the experimental programme are

mainly, expansive soil, fly ash, lime and ceramic dust.

Soil:

Expansive soil from Bhubaneswar city is used in the

experimental programme. The geotechnical properties

of the soil are:

i) Grain size distribution:

Sand size -14%, Silt size -24 %, Clay size – 62%

ii) Specific gravity: - 2.67

iii) Atterberg’s Limit:

Liquid limit - 61%, Plastic limit -30%, Plasticity index-

31%

iv) Compaction characteristics:

OMC – 20.6%, MDD – 15.4 kN /m³

v) UCS: – 55kN/m²

vi) Soaked CBR: - 1.82%

vii) Shear strength parameters:

Cohesion -17 kN/m2, Angle of internal friction -13

0

viii) Swelling pressure:-132 kN/m2

Fly ash:

The fly ash used in the experimental programme is, a

class -F Fly ash (having Cao -0.89%,Al2O3-28.1% and

SiO2-61.39%) collected from the hopper of the

Nalco,Anugul, Odisha. The geotechnical properties of

the fly ash are

i) Grain size distribution:

Clay size-1.92%, Silt size-81.52%, Sand size-16.56%

ii) Specific Gravity:- 2.21

iii) Compaction properties:

OMC=24.1%,MDD=12.85kN/m³

iv) Soaked CBR=0.98%

Ceramic Dust:

Crushed ceramic dust having specific gravity 2.82

passing 75µ IS sieve is used in the experimental

programme.

2.2 Methods:

To find the optimum percentage of fly ash and lime in

stabilization of expansive soil, soil –fly ash-lime mixes

were prepared by addition of fly ash from 0 to 20% at

an increment of 5% and lime from 2 to 6 % at an

increment of 1%.The guidance on the preparation of

soil-fly ash-lime mixes were taken from the

literature.Standard Proctor compaction tests were

conducted on soil-fly ash –lime mixes/samples without

curing to find OMC and MDD. The UCS tests were

conducted on soil-fly ash –lime mixes preparing the

samples at their OMC and MDD at 28 days of curing in

desiccators. After getting optimum percentage of fly ash

and lime, UCS tests were conducted on fly ash –lime

stabilized soil mixes cured at 0, 7 and 14 days. Then

samples were prepared by adding ceramic dusts from 5

to 55% at an increment of 10% to soil stabilized with

optimum percentage of fly ash and lime. Compaction

tests were conducted on these mixes to find OMC and

MDD of the different mixes to prepare samples for

UCS, CBR and swelling pressure tests. Two sets of soil-

fly ash-lime –ceramic dust UCS samples were prepared

by compacting them in their OMC and MDD, cured for

0, 7, 14 and 28 days and then UCS tests were conducted

on one set of these mixes without subjecting them to

any W-D cycles and another set subjecting them to 3, 6,

9 and 12 W-D cycles. The ASTM standard test method

for evaluation of durability of soil-cement mixture

(ASTMD 559-2003) was adopted to study the effect of

W-D cycles on fly ash-lime stabilized soil-ceramic dust

mixes with a minor modification of the process. Each

W-D cycle consisted of submerging the samples in

water for 5 hours and then placing them in oven at 710C

for 42 hours. The volumetric changes were measured

after each W-D cycle. Samples having volumetric strain

more than 10% were rejected. The test continued for 12

W-D cycles. After 3, 6, 9 and 12 cycles the UCS tests

were conducted again on these samples. The application

of scratch brush (as prescribed by ASTMD 559-2003)

was not done in these tests. CBR tests were conducted

by preparing the samples like UCS samples, but without

curing and soaked in water for 96 hours under a

surcharge of 5kg. Samples for swelling pressure tests

were also prepared by compacting them in their OMC

and MDD, but tests were conducted without curing. For

each reported results, three samples were prepared and

average value of the test results are taken. All tests were

conducted according to the procedures given in the

relevant Indian standard codes.

3. Analysis of test results:

Fig.1 shows the variation of MDD of soil with the

addition of fly ash and lime.MDD goes on decreasing

with an increase in percentage of both the fly ash and

lime.

Fig.2 shows the variations of OMC of soil with the

addition of fly ash and lime. OMC goes on increasing

with increase in percentage of both the fly ash and lime.

0 5 10 15 20

11.2

11.6

12.0

12.4

12.8

13.2

13.6

14.0

14.4

14.8

15.2

MD

D(k

N/m

3 )

Fly Ash(%)

2 % Lime

3 % Lime

4 % Lime

5 % Lime

6 % Lime

Figure 1: Variation of MDD with Fly ash and Lime

1212 Strength, Swelling and Durability Characteristics of Fly-Lime Stabilized Expansive




0 5 10 15 20

21

22

23

24

25

26

27

28

29

OM

C(%

)

Fly Ash (%)

2 % Lime

3 % Lime

4 % Lime

5 % Lime

6 % Lime

Figure 2: Variation of OMC with Fly ash and Lime

Fig. 3 shows the variations of UCS of soil with the

addition of fly ash and lime at 28 days of curing. UCS

increases with increase in percentage of both the fly ash

and lime up to a certain percentage of fly ash and lime,

thereafter it decreases. The UCS of soil increases to 352

kN/m2 from 55 kN/m

2 when the percentage of fly ash is

10% and lime is 5%. The strength increases due to

pozzolanic reaction between alumina and silica present

in fly ash and soil with calcium present in lime, the

strength decreases due to a carbonation reaction which

occurs due to availability of excess lime to react with

insufficient alumina and silica present in soil and fly

ash.

Fig.4 shows the variations of MDD of expansive soil

stabilized with an optimum percentage of fly ash and

lime with different percentage of ceramic dust. From the

figure it is observed that with increase in percentage

addition of ceramic dust the MDD goes on increasing.

Ceramic dust fills up the voids formed on soil-fly ash-

lime mix due to flocculation, when ceramic dust is

added to any soil-fly ash-lime mix it results in higher

density as the void space is filled with ceramic dust,

compared to that of soil –fly ash-lime mix where the

void space is empty.

Fig.5 shows the variation OMC of expansive soil

stabilized with an optimum percentage of fly ash and

lime, with different percentage of ceramic dust. From

the figure it is observed that with increase in the

percentage of addition of ceramic dust the OMC goes

on decreasing. Due to reduction in clay content of soil

the attraction of water molecules towards soil –fly ash-

lime –ceramic dust mixes decreases, hence OMC

decreases.

0 5 10 15 20

220

230

240

250

260

270

280

290

300

310

320

330

340

350

360

UC

S(k

N/m

2 )

Fly Ash (%)

2 % Lime

3 % Lime

4 % Lime

5 % Lime

6 % Lime

Figure 3: Variation of UCS (28 days Curing) with

Fly ash and Lime

0 5 10 15 20 25 30 35 40 45 50 55

0

3

6

9

12

15

18

MD

D(k

N/m

3 )

Ceramic dust(%)

Figure 4: Variation of MDD of Fly ash - Lime

stabilized soil with Ceramic dust(%)

0 5 10 15 20 25 30 35 40 45 50 55

0

5

10

15

20

25

30

OM

C(%

)

Ceramic dust(%)

Figure 5: Variation of OMC of Fly ash - Lime stabilized

soil with Ceramic dust(%).

Fig.6 shows the variations of UCS of soil stabilized with

optimum percentage of fly ash and lime, with different

percentage of ceramic dust, at different curing periods.

UCS of fly ash-lime stabilized soil increases with

increase in percentage addition of ceramic dust up to

35%, thereafter it decreases irrespective of the curing




periods. The UCS of soil increases to 497kN/m2 from

86 kN/m2

(0% ceramic dust and 0 day curing period)

when the percentage of ceramic dust is 35% and the

curing period is 28 days. The optimum percentage of

ceramic dust in stabilization of the soil may be taken as

35% .There is a 4.8 times increase in UCS as compared

to fly ash-lime stabilized soil(0% ceramic dust) and 8

times increase in UCS as compared to virgin soil at this

proportion of ceramic dust and at this curing period. The

reason of such type of behavior is that due to pozzolanic

reaction of calcium present in lime with alumina and

silica present in fly ash and soil cementing gels(calcium

silicate hydrates and calcium aluminate hydrates) are

formed which bind the ceramic dust particles with soil

particles, resulting increase in strength of the soil. As

the pozzolanic reaction is time dependant its strength

increases with increase in curing period. The maximum

strength is obtained when the ceramic dust is 35%,

further addition of ceramic dust decreases the UCS

values irrespective of curing period. The reason of such

behavior may be due to decrease in clay content

resulting, carbonation reaction due to non availability of

sufficient alumina and silica to react with excess lime.

0 5 10 15 20 25 30 35 40 45 50 55

0

100

200

300

400

500

600

UC

S(k

N/m

2 )

Ceramic dust(%)

0 Day

7 Days

14 Days

28 Days

Figure 6: Variation of UCS of Fly ash - Lime stabilized

soil with Ceramic dust(%)

0 4 8 12 16 20 24 28

0

50

100

150

200

250

300

350

400

450

In

crea

se i

n U

CS

(%)

Curing period(days)

Figure 7: Variation of increase in UCS (%) of Fly ash-

Lime stabilized soil-Ceramic dust mixes (at optimum

values) with Curing period

Fig.7 shows the variation of percentage increase in UCS

of fly ash-lime stabilized expansive soil-ceramic dust

mixes (at optimum proportion) with curing period. The

percentage increase in UCS increases with increase in

curing period. There is 174%, 281% and 382% increase

in UCS when the curing period increases to 7,14 and 28

days respectively from 0 day.

0 5 10 15 20 25 30 35 40 45 50 55

0

1

2

3

4

5

6

7

8

9

10

Soa

ked

CB

R (

%)

Ceramic dust (%)

Figure 8: Variation of Soaked CBR of Fly ash - Lime

stabilized soil with Ceramic dust(%)

Fig.8 shows the variations of soaked CBR of soil

stabilized with optimum percentage of fly ash and lime,

with different percentage of ceramic dust, without

curing. With the addition of optimum percentage of fly

ash-lime the soaked CBR of the soil increases to 6.82%.

From the fig. it can be observed that the soaked CBR of

the fly ash –lime stabilized soil goes in increasing with

increase in percentage addition of ceramic dust up to

35%, there after it decreases. The soaked CBR of soil

increases to 9.46% from 6.82% when the percentage of

ceramic dust is 35%. The optimum percentage of

ceramic dust in stabilization of the soil was found to be

35%. There is 38% increase in soaked CBR as

compared to fly ash-lime stabilized soil and 420%

increase in soaked CBR as compared to virgin soil at

this proportion of ceramic dust. The reason of such type

of behavior is similar, as discussed in UCS test results.

1214 Strength, Swelling and Durability Characteristics of Fly-Lime Stabilized Expansive




0 2 4 6 8 10 12

0

100

200

300

400

500

600

700

UC

S(k

N/m

2 )

Wet- Dry- Cycles

7 Days

14 Days

28 Days

Figure 9: Effect of W-D cycles on UCS of Fly ash -

Lime stabilized soil- Ceramic dust mixes

Fig. 9 shows the effects of 12 W-D cycles on UCS of

soil stabilized at optimum percentage of stabilizers (fly

ash-10%, lime -5% and ceramic dust -35%) at 7, 14 and

28 days of curing. The UCS samples without curing

were unable to survive 3 W-D cycles hence, these

values have not been shown in the figure. From the

figure it is found that increase in W-D cycles decreases

the UCS of the stabilized soil, curing decreases the loss

in UCS values. Increase in curing period makes a strong

bond between stabilizers and soil hence, decreases the

losses in UCS values.

0 2 4 6 8 10 12

0

5

10

15

20

25

Red

uct

ion

in U

CS

(%)

Wet-Dry-Cycles

7 Days

14 Days

28 Days

Figure 10: Reduction in UCS (%) of Fly ash - Lime

stabilized soil- Ceramic dust mixes by the effect of W-D

cycles

Fig.10 shows the variation in percentage reduction in

UCS of fly ash – lime stabilized soil-ceramic dust mixes

(at optimum proportion) by the effect of W-D cycles.

From the figure it can be seen that the percentage

reduction decreases to 8.35% from 19.1% when the

curing period increases from 7days to 28 days when the

samples were subjected to 12 W-D cycles.

Fig.11 shows the variation of volumetric strain of fly

ash - lime stabilized soil- ceramic dust mixes (at

optimum proportion) with W-D cycles. The volumetric

strain goes on increasing with increase in W-D cycles

and goes on decreasing with curing period. As the

change in volume of stabilized soil takes place with

increase in W-D cycles the volumetric strain goes on

increasing. As curing increases the bonding between

stabilizers and soil, the change in volume of the samples

decreases which results, decrease in volumetric strain of

the samples.

0 2 4 6 8 10 12

0

2

4

6

8

10

12

Vol

um

etri

c st

rain

(%)

Wet-Dry-Cycles

7 Days

14 Days

28 Days

Figure 11: Variation of Volumetric strain of Fly ash -

Lime stabilized soil- Ceramic dust mixes with W-D

cycles

0 5 10 15 20 25 30 35 40

0

10

20

30

40

50

60

Sw

elli

ng

Pre

ssu

re

Ceramic dust(%)

Figure 12: Variation of Swelling pressure of Fly ash-

Lime stabilized soil with Ceramic dust (%)

Fig.12 shows the variation of swelling pressure of fly

ash-lime stabilized expansive soil with different

percentage of ceramic dust. With the addition of

optimum percentage of fly ash and lime, the swelling

pressure reduces to 56 kN/m2

from 132 kN/m2 . It can be

seen from the figure that with the increase in addition of

different percentage of ceramic dust the swelling

pressure of soil goes on decreasing, decreases to zero,

when the percentage of ceramic dust is 35%, making the

stabilized soil a non-swelling material. When fly ash-

lime (at optimum proportion) is added to soil, because

of the development of pozzolanic reaction between

calcium present in lime with alumina and silica present




in fly ash and soil a strong cementing bond develops

between soil particles which prevents the swelling of

soil resulting, reduction in swelling pressure of

expansive soil. The swelling pressure of soil reduces

further by addition of ceramic dust. As the clay particles

are replaced by ceramic dust particles, the attraction of

the stabilized soil for water molecules decreases, hence

swelling pressure decreases. Another reason of such

phenomenon is, the development of strong inter-particle

bond between soil, fly ash-lime and ceramic dust

particles, which prevents swelling of the soil, hence

swelling pressure decreases.

4. Conclusion:

The following conclusions are drawn from this study.

The optimum percentages of fly ash, lime and

ceramic dust for stabilization of expansive soil are,

fly ash -10%, lime -5% and ceramic dust- 35%.

There is an eight times increase in UCS values of

the soil, stabilized at optimum percentages of fly

ash, lime and ceramic dust and cured for 28 days,

over virgin soil.

There is a 4.2 times increase in soaked CBR values

of the soil, stabilized at optimum percentages of fly

ash, lime and ceramic dust, over virgin soil.

At the optimum percentages of fly ash, lime and

ceramic dust the stabilized soil is a non-swelling

material having no swelling pressure.

An increase in the number of W-D cycles decreased

the UCS values and increased the volumetric strain

of soil stabilized at optimum percentages of fly ash,

lime and ceramic dust. Increase in curing period

decreases, the percentage loss of UCS and

volumetric strain of the stabilized soil.

The soil stabilized at optimum percentages of

stabilizers, with a minimum curing period of 7 days

was found to be durable, as the percentage loss in

strength is found to be below 20% after 12 W-D

cycles.

Acknowledgement:

The Authors gratefully acknowledge the reviewers and

the Editor in Chief of IJEE for their useful comments

and suggestions for improving the quality of this

manuscript.

2. Reference:

[1] M. Joel, and I.O. Agbedi, “Mechanical-Cement

stabilization of laterite for use as flexible pavement

material”, Journal of material in Civil Engineering,

Vol.23 (2), PP.146-152., 2011.

[2] A. K. Sabat, and R.P. Nanda, “Effect of marble dust

on strength and durability of rice husk ash

stabilised expansive soil”, International Journal of

Civil and Structural Engineering, Vol.1 (4) PP.939-

948., 2011.

[3] A. S. Muntohar and G. Hantoro, “Influence of rice

husk ash and lime on engineering properties of a

clayey subgrade”, Electronic Journal of

Geotechnical engineering, Vol.5., 2000.

[4] E.A.Basha, R.Hashim, and A.S.Muntohar, “Effect

of the cement-rice husk ash on the plasticity and

compaction of soil”, Electronic Journal of

Geotechnical engineering, Vol.8, Bundle A., 2003.

[5] A. N. Ramakrishna and A.V. Pradeep Kumar,

“Stabilisation of black cotton soil using rice husk

ash and cement’’, National conference on Civil

Engineering meeting the challenges of tomorrow -

2006, PP.215-220.,2006.

[6] A. K. Sabat, “A study on some geotechnical

properties of lime stabilised expansive soil –quarry

dust mixes”, International Journal of Emerging

trends in Engineering and Development,

Vol.1(2),PP.42-49.,2012.

[7] M. Abd EI-Aziz and M.A. Abo-Hashema,

“Measured effects on engineering properties of

clayey subgrade using lime-homra stabiliser”,

International Journal of Pavement Engineering,

Vol.14 (4)., PP.321-332.,2013.

[8] F. Zha, S. Liu, Y. Du, and K. Cui, “Behavior of

expansive soils stabilized with fly ash”, Natural

hazards, Vol.47 (3), PP.509-523., 2008.

[9] O.O.Amu, A.B.Fajobi and S.O. Afekhuai,

“Stabilizing potential of cement and fly ash mixture

on expansive clay soil”, Journal of Applied

Sciences, Vol.5 (9), PP.1669-1673., 2005.

[10] E. Cokca, V. Yazici, and V. Ozaydin, “Stabilisation

of expansive clays using granulated blast furnace

slag (GBFS) and GBFS-cement”, Geotechnical and

Geological Engineering, Vol.27 (4), PP.489-499.,

2009.

[11] E. Kalkan, “Utilization of red mud as a stabilization

material for the preparation of clay liners”,

Engineering Geology, Vol. 87(3-4), PP. 220-229.,

2006.

[12] A. K. Sabat, “Utilization of bagasse ash and lime

sludge for construction of flexible pavements in

expansive soil areas”, Electronic journal of

Geotechnical Engineering, Vol.17, Bund.

H.PP.1037-1046., 2012.

http://scialert.com/asci/author.php?author=S.O.%20Afekhuai

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1216-1222


Workability and Mechanical Performance of Concrete with Copper

Slag

M VELUMANI1 AND K NIRMALKUMAR

2

1K.S. Rangasamy College of Technology, Tiruchengode, India

2Kongu Engineering College, Perundurai, Erode, India


Abstract: Copper slag (CS) is a by-product obtained during the metal smelting and refining of copper. It is one of

the materials being considered as a waste and many researchers has been investigated the possibility of using copper

slag as a replacement of fine aggregate in concrete matrix. This paper reports the results of feasibility studies of

copper slag as fine aggregate replacement in concrete. The physical and chemical properties of copper slag were

determined experimentally. The concrete mixtures were prepared with the proportions of copper slag as fine

aggregate replacement ranging from 0% (for the control concrete) to 100% (copper slag concrete.). The specimens

were carried out for testing to determine the compressive strength, tensile and flexural strength for age 7, 28 and 60

days of curing and also to determine the workability by slump test. Adding up to 40% of copper slag as sand

replacement yielded more strength with that of conventional M35 grade concrete. However, further additions of

copper slag caused reduction in strength with that of control concrete specimens due to the increase of free water

content in the mix. Hence it is suggested that 40% weight of copper slag can be used as replacement of fine

aggregate for the nominal standard M35 grade concrete for the Construction Industries.

Keywords: Copper Slag, Workability, Compressive Strength, Tensile Strength and Flexural Strength.

1. Introduction:

In this modern era, Concrete is one of the major

construction materials which is used in Construction

Industries in all over the world. Aggregate which makes

up 70% of the concrete volume, is one of the main

constituent materials in concrete production. However,

due to the higher cost of natural sand which is used as a

fine aggregate and the rising emphasis on sustainable

construction, there is a need for the construction

industry to search for alternative materials as fine

aggregates in concrete production.

Large quantities of industrial by-products are produced

every year by various industries. Utilizing the industrial

by-products has the advantages of decreasing the

depletion of natural resources, eliminating the costs of

dumping and minimizing the environmental pollution

problems. Copper slag is one of the industrial by-

products produced from the process of manufacturing of

copper. It is suitable for a variety of applications such as

cement manufacturing, in aggregates, land fill, ballast,

abrasives, roofing granules, glass, tiles, bituminous

pavements etc. Its application as a fine aggregate in

concrete production reaps many environmental benefits

such as waste recycling and solves disposal problems.

There have been various experiences in using copper

slag as a replacement for fine aggregate in concrete. Al-

Jabri et al. (1) reported that 40% weight of copper slag

is permissible as replacement of sand in High

Performance Concrete with good strength and durability

properties. Beyond that level of replacement, copper

slag caused reduction in strength with that of control

mix.

Wei Wu et al. (2) observed that less than 40% copper

slag as sand substitution in High Strength Concrete

achieved good strength which was comparable and

better to control mix, beyond which however its

behaviours decreased significantly. It was also

recommended that up to 40-50% (by weight of sand) of

copper slag can be used as a replacement for fine

aggregate for nominal M25 grade concrete to obtain

good strength and durability requirements (3). The

results showed that 50% by weight of copper slag

substitution for Portland cement gave similar strength

performance as the control mix especially at low water -

to-binder ratios of 0.5 and 0.6. Higher copper slag

(13.5%) replacement yielded lower strength values (4).

Even though many researchers reported the replacement

effect of copper slag as aggregates on the performance

of high performance and high strength concretes, there

has been little research concerning the replacement of

copper slag as fine aggregates to produce standard M35

grade concrete for medium concrete structures. Hence

this research has been performed for the feasibility of


1217 M.VELUMANI AND K.NIRMALKUMAR



using copper slag as fine aggregate in M35 grade

concrete.

1.1 Literature Review:

There have been various experiences in using copper

slag as a replacement for fine aggregate in concrete. (5)

reported that Addition of up to 50% of copper slag as

sand replacement yielded comparable strength with that

of the control mix. However, further additions of copper

slag caused reduction in the strength due to an increase

of the free water content in the mix. Mixes with 80%

and 100% copper slag replacement gave the lowest

compressive strength value of approximately 80 MPa,

which is almost 16% lower than the strength of the

control mix. The results also demonstrated that the

surface water absorption decreased as copper slag

quantity increases up to 40% replacement; beyond that

level of replacement, the absorption rate increases

rapidly. Therefore, it is recommended that 40 wt% of

copper slag can used as replacement of sand in order to

obtain HPC with good strength and durability

properties.(2.Wei Wu, et.al) This study investigated the

mechanical properties of high strength concrete

incorporating copper slag as a fine aggregate and

concluded that less than 40% copper slag as sand

substitution can achieve a high strength concrete that

comparable or better to the control mix, beyond which

however its behaviors decreased significantly. The

workability and strength characteristics were assessed

through a series of tests on six different mixing

proportions at 20% incremental copper slag by weight

replacement of sand from 0% to 100%. (6) The results

demonstrated that surface water absorption decreased as

copper slag content increases up to 50% replacement.

Beyond that, the absorption rate increased rapidly and

the percentage volume of the permeable voids was

comparable to the control mixture. Therefore, it is

recommended that up to 40–50% (by weight of sand) of

copper slag can be used as a replacement for fine

aggregates in order to obtain a concrete with good

strength and durability requirements.(7.D.Brindha) For

this research work, M20 grade concrete was used and

tests were conducted for various proportions of copper

slag replacement with sand of 0 to 60%, cement of 0 to

20% in concrete. The obtained results were compared

with those of control concrete made with ordinary

Portland cement and sand. (1.S. Al-Jabri, et. al)This

research study was conducted to investigate the

performance of high strength concrete (HSC) made with

copper slag as a fine aggregate at constant workability

and to study the effect of superplasticizer addition on

the properties of HSC made with copper slag. Two

series of concrete mixtures were prepared with different

proportions of copper slag. The first series consisted of

six concrete mixtures prepared with different

proportions of copper slag at constant workability. The

water content was adjusted in each mixture in order to

achieve the same workability as that for the control

mixture. Even though many researchers reported the

replacement effect of copper slag as aggregates on the

performance of high performance and high strength

concretes, there has been little research concerning the

replacement of copper slag as fine aggregates to

produce standard M35 grade concrete for medium

concrete structures. Hence this research has been

performed for the feasibility of using copper slag as fine

aggregate in M35 grade concrete.

1.2 Research objectives:

The main objective of this research is to study the use of

copper slag as fine aggregate replacement in M35 grade

concrete. The following are the main objectives:

1. Determination of physical properties and chemical

compositions of copper slag.

2. Comparison of the results of the particle size

distribution of copper slag and sand (fine

aggregate).

3. Evaluating the compressive, tensile and flexural

strength of concrete specimens at different

proportions of copper slag admixed concrete and

M35 grade control concrete.

4. Also to determine the workability of copper slag

replacement concrete

1.3 Materials:

1.3.1 Cement:

The cement used in this study was Ordinary Portland

Cement (OPC) of 53 Grade with specific gravity of

3.15. This cement is one of the most widely used

cements in construction industry in India

1.3.2 Coarse and fine aggregates:

The coarse aggregate of 20 mm size with a specific

gravity of 2.65 was used in this study. The fine

aggregate is natural sand with a specific gravity of 2.64

and the fineness modulus of 2.95. The particle size

distribution was well-graded with over 60-80% of the

sand over the size range of 0.6 to 1.18mm. It is well

shown that the coarse and fine aggregates met the

specifications requirements as per the Indian Standard

codes.

1.3.3 Copper Slag (CS):

Copper slag is a by-product material obtained from the

process of manufacturing of copper. Copper slag used in

this work was brought from Sterlite Industries India

Limited (SIIL) Tuticorin, Tamil Nadu which produces a

huge amount of copper slag.

1.3.4 Water:

It is strictly adhered that the potable water is used for

making the concrete specimens because it plays a major

1218 Workability and Mechanical Performance of Concrete with copper slag



role in water to cement ratio, consistency and slump

values of fresh concrete. Also the durability properties

get affected due to the use of the nature of the water

used. Hence most care had been taken for the water as

constituent in making the concrete specimens.

1.4 Physical properties and mechanical properties of

copper slag:

The physical and mechanical properties of the copper

slag are tabulated in Table1. Air cooled copper slag

was a black in colour and have glassy appearance. The

specific gravity of the copper slag was found as 4.00.

The bulk density of copper slag is 2.20 g/cm3. The

density varies from 3.16 to 3.87 g/cm3 which are higher

than that of conventional aggregate. Since the copper

slag used in this study is a granulated one and hence the

water absorption is somewhat higher (0.41%) than that

of conventional aggregate. The copper slag is made up

of regularly shaped, angular particles mostly between

4.75 mm and 2.36 mm in size.

The copper slag used in this study has a number of

favourable mechanical properties for aggregate use,

including good soundness characteristics, abrasion

resistance and good stability. It has high friction angle

due to sharp angular shape. However the slag tends to

be vitreous or glassy which adversely affects their

frictional properties, a problem arises if used in

pavement surfaces.

1.5 Chemical compositions of Copper Slag (CS):

Table 2 shows the chemical compositions present in the

waste copper slag and Ordinary Portland Cement.

In comparison with the chemical composition of natural

pozzolanas, the summation of three oxides (silica,

alumina and iron oxide) in copper slag is 93% which

exceeds nearly 28% of OPC. Hence it is recommended

that the copper slag is expected to have good potential

to produce high quality pozzolanas.

1.6 Sieve analysis of Copper slag and sand:

Gradation test was conducted on the samples of copper

slag and sand, and showed that both copper slag and

sand have comparable particle size distribution as

shown in Fig.1.

1.7 Experimental testing program:

1.7.1 Mix Design and preparation of copper slag

admixed concrete specimens:

Nine concrete mixtures with different proportions of

copper slag ranging from 0% (for the control mix) to

100% were considered as shown in Table 3.

1.7.2 Laboratory tests conducted for copper slag

admixed concrete:

The ingredients of concrete making materials were

weighed separately and mixed manually. The mixes

were compacted using vibrating table. The slump of the

fresh concrete was determined to ensure that it would be

within the design value and to study the effect of copper

slag replacement on the workability of concrete. The

specimens were remoulded after 24 hours, cured in

water and tested at room temperature for 7, 28 and 60

days.

To determine the unconfined compressive strength, nine

cubes (150mm x 150mm x 150mm) were casted for

each mix and water cement ratio, and three number of

samples were tested after 7, 28 and 60 days of curing.

Nine cylinders of 150mm diameter x 300mm long were

prepared for each mix to determine the split tensile

strength of concrete for 7, 28 and 60 days of curing.

Flexural strength (modulus of rupture) of the specimen

is determined for each mix by casting nine numbers of

prisms of 100mm x 100mm x 500mm tested after 7, 28

and 60 days of curing.

1.8 Results and Discussions:

1.8.1 Effect of copper slag replacement on the

workability of concrete:

The slump value of the concrete incorporating the

copper slag as fine aggregate increased when more

percentage of natural sand was replaced by the copper

slag as shown in Fig.2.The maximum slump value

obtained was 120mm at 40% copper slag replacement

compared to the slump value of 60 mm of control mix.

In general it can be concluded that the addition of

copper slag improves the workability of concrete mix. It

appears a much smoother surface than sand, which

reduces shear resistance in the concrete mix, thus

increasing its flowability. Water not absorbed by the

copper slag acts as lubricant between the solid particles,

reducing the inter-particle frictional force.

The effect of copper slag replacement as fine aggregates

on the workability of M35 grade concrete is presented

in Table 4.

1.8.2 Effect of copper slag replacement on the

compressive strength of concrete

The measured compressive strength value for 0 to 100%

a mix proportion was shown in Table 4. The

compressive strength values for concrete mixtures with

0 to 100% replacement of copper slag cured at 7, 28 and

60 days are plotted as shown in Fig.3.The test results

indicate that for the mixtures prepared up to 40% copper

slag replacement, the compressive strength of concrete

is comparable to the strength of the control mix with

100% sand. But for mixtures with 80% and 100%

copper slag the compressive strength decreased rapidly

below the strength of the control specimens. Hence the

40% copper slag content yielded the highest 60 days




compressive strength of 62.32 N/mm2 compared with

42.13 N/mm2 for the control mix, when tested after 60

days where as the lowest compressive of 18.04 N/mm2

was obtained for Mix No.9 with 100% replacement.

This reduction in strength for concrete mixtures with

high copper slag contents is due to the increase in free

water content that results from the low water absorption

characteristics of copper slag in comparison with sand.

1.8.3 Effect of copper slag replacement on the split-

tensile strength of concrete:

The 7, 28 and 60 days tensile strength of M35grade

copper slag concrete is given in Table 5. Direct tensile

strength of concrete specimens cannot be determined

owing to difficulty in preparation of test specimen and

applying truly axial tensile load. Hence the specimen

was placed horizontally with its axis perpendicular to

the loading direction. The load is then applied and it is

calculated from the formula as given below:

Split tensile strength = 2P

LD where P=maximum load

at failure, L= length of cylindrical specimen in mm,

D=diameter of cylindrical specimen in mm.

The results show that the maximum tensile strength lies

at 40% replacement of copper slag in M35 grade

concrete at 7th

day (4.27 N/mm2), 28

th day (5.15

N/mm2) and 60day (5.83 N/mm2) Where the lowest

tensile strength of 2.67 N/mm2 was obtained for Mix

No.9 with 100% replacement. This reduction in strength

for concrete mixtures with high copper slag contents is

due to the increase in free water content that results

from the low water absorption characteristics of copper

slag in comparison with sand.

Fig.4 clearly represents that the copper slag admixed

concrete of M35 grade achieves 40% replacement of

copper slag is maximum compared with the other

proportions.

1.8.4 Effect of copper slag replacement on flexural

strength of concrete:

The flexural strength study is useful in the design of

pavement slabs and airfield runway as flexural tension

is critical in these cases. The results are affected by the

size of the specimens casting, curing and moisture

conditions, type manner of loading (three point or

central point loading) rate of loading etc,

The test was conducted for three points loading and the

strength is obtained according to the prescribed

standards. The 7, 28 and 60 days flexural strength

(modulus of rupture) values for copper slag concrete are

presented in Table 6.

The results showed that the 40% replacement of copper

slag yields optimum value than that of conventional

concrete at the age of 7th

day (3.81N/mm2), 28

thday

(7.89 N/mm

2) and 60 day (8.88 N/mm

2).Where the

lowest flexural strength of 1.12 N/mm2 was obtained for

Mix No.9 with 100% replacement. This reduction in

strength for concrete mixtures with high copper slag

contents is due to the increase in free water content that

results from the low water absorption characteristics of

copper slag in comparison with sand.

Fig.5 shows the typical flexural strength values of

copper slag concrete and conventional concrete. It is

clear that the presence of up to 40% copper slag as a

replacement for fine aggregate has resulted in increase

in the flexural strength.

1.9 Conclusions:

The following conclusions may be drawn from the

present study:

1. The workability of M35 grade copper slag concrete

increased rapidly with increase in copper slag

percentage of even up to 40% replacement for fine

aggregate.

2. Up to 40% increase of copper slag as sand

replacement yielded higher compressive strength

with that of control mix. Also up to

60%replacement of copper slag will be within the

desirable limit, beyond that there will be decrease

in compressive strength of concrete.

3. Similarly for 40% replacement of copper slag as

sand replacement, the tensile and flexural strengths

were also yielded higher strength than that of

control specimen and reduction in strength started

from 50% replacement. It was due to the lowest

surface water absorption beyond the addition of

40% replacement.

4. It is suggested that 40% weight of copper slag can

be used as replacement of sand in medium level

constructions where nominal concrete grades are

used.

5. Further research work is needed to explore the

effect of copper slag as fine aggregates combined

with other materials such as fly ash, crushed sand

stone powder, and Super plasticizers in other

superior concretes

2. Acknowledgement:

The author express their sincere thanks to the reviewer

Dr.S.Gopalakrishnan, Professor, EBETGroup of

Institution and Dr.N.Ramesh, Professor,

K.S.Rangasamy College of Technology, Tiruchengode

and editorial in-charge Dr.Reddy for their valuable

guidance in improving the paper.

3. Reference:

[1] Khalifa S.Al-Jabri, et.al. Copper slag as sand

replacement for high performance concrete. Cement

and Concrete composites 2009; 31:483-488.




[2] Wei Wu, Weide Zhang, Guowei Ma. Optimum

content of copper slag as fine aggregate in high

strength concrete. Materials and Design 2010;

31:2878-2883.

[3] Khalifa S.Al-Jabri, et.al. Effect of copper slag as a

fine aggregate on the properties of coment mortars

and concrete. Construction and Building Materials

2011; 25:933-938.

[4] Meysam Najimi, Ali Reza Pourkhorshidi.

Properties of concrete containing copper slag waste.

Magazine of Concrete Research, 2011; 63(8):605-

615.

[5] Meenakshi S., Illangovan R. Performance of

Copper slag and Ferrous slag as partial replacement

of sand in Concrete. International Journal of Civil

and Structural Engineering. 2011; 1(4): ISSN 0976-

4399.

[6] A.S.Alnuaimi. Use of copper slag as a replacement

for fine aggregate in reinforced concrete slender

columns. Computational Methods and Experiments

in Materials Characterisation 2099; 64: ISSN 1743-

3533.

[7] D.Brindha and Nagan S. Durability studies on

copper slag admixed concrete. Asian Journal of

Civil Engineering (Building and Housing) 2011;

12(5):563-578.

Table 1: Typical physical properties and mechanical properties of copper slag (CS)

Physical Properties Copper slag Mechanical Properties Copper slag

Appearance Black, glassy Aggregate crushing value (%) 10–21

Particle shape Irregular Aggregate impact value (%) 8.2–16

Specific Gravity 4.00 Abrasion loss (%) 24.1

Bulk density gm/cc 2.20 Conductivity (_s/cm) 500

Water absorption % 0.41 Type Air cooled

Fineness Modulus 3.4 Soundness (%) 0.8–0.9

Density (g/cm3) 3.16–3.87 Angle of internal friction 51° 20’

Table 2: Chemical compositions of Ordinary Portland cement and Copper slag (CS)

Chemical Compositions Ordinary Portland Cement (%) Copper slag (%)

Iron Oxide- Fe2O3 3.51 55. 00

Silica – SiO2 20.85 35. 00

Aluminium Oxide Al2O3 4.78 3. 01

Calcium Oxide CaO 63.06 0. 20

Magnesium Oxide MgO 2.32 0. 90

Mn2O3 0.05 0.06

SO3 3.00 0.32

K2O 0.50 0.70

Table 3: Copper slag admixed concrete with nine proportions in terms of Kg/m3

Mix ID CC CS10 CS20 CS30 CS40 CS50 CS60 CS80 CS100

Cement 448 448 448 448 448 448 448 448 448

F.A 656 590.4 524.8 459.2 393.6 328 262.4 131.2 0

C.A 1254 1254 1254 1254 1254 1254 1254 1254 1254

C.S 0 65.6 131.2 196.8 262.4 328 393.6 524.8 656

Water 197 197 197 197 197 197 197 197 197

Table 4: Slump value and Cube compressive strengths of M35 grade concrete

Sl. No. Mix ID

(%) Mix type Slump (mm)

Strength(N/mm2)

CS7 CS28 CS60

1 0 Control (100%S) 32 26.33 36.12 42.13

2 10 S90% + CS10% 38 31.56 44.23 53.13

3 20 S80% + CS20% 52 34.36 47.65 56.51

4 30 S70% + CS30% 61 36.02 50.12 57.79




5 40 S60% + CS40% 89 36.5 52.66 62.32

6 50 S50% + CS50% 112 31.6 45.31 54.86

7 60 S40% + CS60% 122 26.25 38.76 47.69

8 80 S20% + CS80% 139 20.15 29.81 32.65

9 100 S0% + CS100% 147 11.71 17.13 18.04

CS7=Compressive strength at 7days, CS28=Compressive strength at 28days, CS60=Compressive strength at 60days

Table 5: Split tensile strength of copper slag concrete at 7, 28 and 60day

Sl. No. Mix ID


Strength(N/mm2)

TS7 TS28 TS60

1 0 Control (100%S) 32 4.12 4.31 4.72

2 10 S90% + CS10% 38 4.05 4.72 5.31

3 20 S80% + CS20% 52 4.20 4.91 5.53

4 30 S70% + CS30% 61 4.23 5.05 5.59

5 40 S60% + CS40% 89 4.27 5.15 5.83

6 50 S50% + CS50% 112 4.05 4.87 5.51

7 60 S40% + CS60% 122 3.65 4.69 4.95

8 80 S20% + CS80% 139 3.58 3.75 4.13

9 100 S0% + CS100% 147 2.67 2.84 2.81

TS7= tensile strength at 7 days, TS28= tensile strength at 28 days, TS60= tensile strength at 60 days

Table 6: Flexural strength of copper slag concrete at 7, 28 and 60days

Sl. No. Mix ID


Strength(N/mm2)

FS7 FS28 FS60

1 0 Control (100%S) 32 2.76 4.77 5.94

2 10 S90% + CS10% 38 3.44 5.98 6.35

3 20 S80% + CS20% 52 3.58 6.99 7.54

4 30 S70% + CS30% 61 3.60 7.61 8.01

5 40 S60% + CS40% 89 3.81 7.89 8.88

6 50 S50% + CS50% 112 3.13 5.02 6.85

7 60 S40% + CS60% 122 2.49 4.69 5.63

8 80 S20% + CS80% 139 1.93 3.19 3.66

9 100 S0% + CS100% 147 1.12 1.76 2.04

FS7= flexural strength at 7th

days, FS28= flexural strength at 28th days, FS60= flexural strength at 60days

Fig.1. Particle size distribution curves on copper slag and sand




Fig.2. Slump value for replacement of copper slag in

M35 grade concrete

Fig.3. Cube compressive strength of copper slag

concrete at 7, 28 and 60 days

Fig.4. Split-tensile strength of copper slag concrete at

7,28 and 60 days

Fig.5. Flexural strength of copperslag concrete at 7,28

and 60 days

www.cafetinnova.org

Indexed in





ISSN 0974-5904, Volume 07, No. 03

June 2014, P.P.1223-1230


Economic Comparison of Semi Continuous Mining System with

Intermittent System in Special Cycle Time

ZHAODONG WANG College of Mining Engineering,Liaoning Technical University ,Fuxin,123000,China

China Coal Technology & Engineering Group Corp ,Beijing 100013,China


Abstract: Today, the mining process is dominated by large-scale mechanized mining with the coexistence of

various mining processes. With the continuity of mineral flow, open pit mining technology has three kinds of

processes: continuous, semi-continuous, and intermittent technology. Due to various factors involved in the mining

process system, optimization work is quite complex, and it is always an important research field in open pit mining.

Semi continuous mining technology is a comprehensive process based on the other two process. In the paper, we

give the transport distance under different mining process according to the recent level of open-pit mining

characteristics, and take minimum total cost in calculation period as the goal to establish the open pit mining process

optimization model. Calculation method of economic indicators of intermittent mining technology and semi

continuous process has been presented. And energy use of truck and belt conveyor has been compared. The semi

continuous technology combines the advantages of single bucket excavator, automobile and belt conveyor and can

be used in open pit mine to get high total benefit with low cost.

Keywords: open pit mining, semi continuous system, intermittent system, economic comparison.

1. Introduction:

With rapid and continuous development of the mining

industry, two or more production processes used in

open-pit mines are no longer uncommon. Integrated

production processes ensure the respective advantages

of each single production process under appropriate

circumstances with its wide range of adaptability. Its

high reliability and economic benefits are favored by

numerous open-pit mines [1]. Development of the open

mining process has been experienced from the initial

open-pit mining technology of simple human

exploitation to today's mechanized mass production. In

modern times, the mining process is dominated by

large-scale mechanized mining with the coexistence of

various mining processes. Open pit mining technology

has three kinds of processes: continuous, semi-

continuous, and intermittent with the continuity of

mineral flow.

(1) Intermittent continuous process. Intermittent

continuous technology is mainly in the manners of

shovel-truck technology. This process appears at an

early time and has wide application. Various research

focuses on the truck or other machinery dispatching [2-

6].

(2) Continuous mining process. Continuous mining

technology [7-11] originates from Germany. Typical

system is composed of bucket shovel, wheel conveyor

and dumping machine. It is originally from Germany

and greatly improves the production efficiency.

(3) Semi continuous process. Semi continuous process

[12-14] is based on the continuous production

technology and intermittent continuous mining

technology. It combines the advantages of continuous

and intermittent continuous processes. Some studies on

special application, such as crusher [15-16], reliability

[17-18].

Objects of the three kinds of process are processes from

mining to transport. In actual open pit mining, mining

sequence of ores or blocks requires responsible

attention. Perfect mining sequence will lead to higher

economic benefit. And the exploitation sequence with

the consideration of time and space [19-22] and random

distribution of mineral [23-24] is much easier to reduce

the cost of fixed investment and boost their total benefit.

In the paper, we mainly discussed the characteristics

and development status of semi continuous process.

Economic problems of the two processes have been

studied. With a certain open pit mine, economic

indicator of intermittent technology and semi

continuous process has been analyzed. Energy

efficiency of belt transportation and truck transportation

has been compared.

The main contributions of this paper is an investigation

on transport distance under different mining process

according to the recent level of open-pit mining


1224 Economic Comparison of Semi Continuous Mining System with Intermittent

System in Special Cycle Time



characteristics, and take minimum total cost in the

calculation period as the goal to establish the open pit

mining process optimization model. The remainder of

the paper can be organized as the following: Section 2 is

the introduction of semi continuous technology in open

pit mining, including characteristics, application

conditions, and development. Calculation method of

transport distance of overburden is given in section 3.

Optimization model is described in section 4, including

objective and calculation method. Section 5 gives the

comparison results of semi continuous and intermittent

processes with a certain open pit mine. The conclusion

is given in section 6.

2. SEMI CONTINUOUS IN OPEN-PIT MINING

2.1. Characteristics of semi continuous in open-pit

mining

Semi continuous mining technology refers two kinds of

forms. One form includes three parts: the mining

process is intermittent; material transportation is by

truck with an intermittent manner from working face to

crush station; material transports continually after it

moving out of the crushing station. The other form is all

continuous transportation. The former one is combined

with the characteristics of high adaptability of shovel

truck technology, flexibility, and belt conveyor with

advantages of large climbing ability, high efficiency,

and low transportation cost. It is rapidly developed in

recent years, both technology and equipment. Two

kinds of semi continuous mining technology are widely

applied, and the technology is relatively mature. Both of

the two kinds are with less or without trucks and partly

or completely with the belt conveyor instead. This can

reduce the production cost and environmental pollution.

The advantages and disadvantages of semi continuous

mining technology are shown in table 1.

2.2. Conditions for application of semi continuous

mining technology

1) Geological conditions. Geological condition is the

fundamental condition to choose the semi continuous

mining technology, including coal, rock, soil thickness,

number, angle, lithology, geological structure and

mining depth complexity.

Mining system can use semi continuous mining

technology with single bucket-truck and semi portable

crusher station. Some nearly horizontal or gently

inclined seam, which has the advantages of simple

structure, small depth of pit, can use semi continuous

mining technology with single bucket and mobile

crusher.

The majority of open-pit stripping system is suitable for

semi continuous mining technology with single bucket

and mobile crusher in geological terms. But open pit

mine with complex mining condition is more suitable

for the semi continuous mining technology with single

bucket-truck + semi portable crusher station.

If the inner dump condition is poor, open pit mining

depth is larger, and stripping haul distance is far and

with the heavy car uphill, most stripped quantity should

be discharged. This would make the height of outer

dump area increasing. The truck would pass a long way

to dump. When uphill distance of heavy car is up to 2.8

km or above, semi continuous mining technology

should be used to get a higher benefit.

2) Climatic conditions. Freezing adhesive in equipment

is a not well resolved technical problem.

3) Equipment manufacturing conditions for semi

continuous technology. The key equipment of the semi

continuous technology is crushing machine, including

the semi mobile station and self-mobile crusher. This is

decive factors of mature of semi continuous technology.

Other equipment in semi continuous process system,

such as single bucket excavators, truck, belt conveyor,

dumping machine is mature and reliable.

Single bucket-truck + semi mobile crusher station

technology is mature. Semi mobile crushing station is

mature and various series of models are completed.

Single bucket and self-mobile crusher technology

crusher have existed but with no series of products.

Especially greater than 2500 ht / , mobile crusher has

just few application servers, and equipment in each link

matches unreasonable. Some key technologies are still

in the stage of research and development. For example:

overall structural design of self-mobile crusher itself,

operation stability, and single bucket excavator in the

working face with high efficiency, all these technologies

have not been fully tested in practice. So, single bucket

and self-mobile crusher have a low technology maturity,

and it is the main obstacle to effective decision-making.

Table1 Advantages and disadvantages of semi

continuous process

Content remarks

Adva

ntage

s

Low operating costs,

obvious advantages when

the material transport path

is uphill slope and

distance is long;

When

transportatio

n distance is

more than

2.8km, semi

continuous

technology

has obvious

economic

advantages

compared

with the

single

Less unit energy

consumption;

Low cost of spare parts

and repair maintenance;

Little constrain by fuel

shortage and less

environmental pollution;

1225 ZHAODONG WANG



Less persons and low

labor intensity;

bucket truck

mining

technology.

disad

vanta

ges

Series production process

system, great fluctuations

in short-term production

capacity, buffer link;

Semi

continuous

mining

technology

is combined

process and

delivery

transportatio

n. It’s

mainly

manifested

in the

system

reliability

and initial

investment

Different levels of

freezing adhesive crusher

bin with the temperature

is below -10℃, clean

work is great, and low

production efficiency;

Higher initial investment,

but due to the long service

life, depreciation expense

is not high;

Cycle time of

manufacturing and

installation is long;

2.3. Development of semi continuous process

Semi continuous mining technology is developed based

on continuous mining technology and intermittent

mining technology. It has advantages of continuous

process and intermittent technology. It has strong

adaptability and can mine hard rock and freezing soft

rock mining in winter. It can give the continuous

operation on ore transportation, expand the scale of

production and reduce the cost of production.

At present, semi continuous mining technology mainly

has 2 forms: single bucket truck and semi mobile

crusher station and belt conveyor; single bucket and

self-mobile crusher and belt conveyor.

Semi continuous mining technology with single bucket

truck and semi mobile crusher station and belt conveyor

has been widely used. So far, hundreds of mines have

applied the mining technology. Maximum production

capacity of semi mobile crushing station has reached

14000 ht / . It feeds with the bucket capacity of 60 3m

of single bucket excavator and load of 365 t dump

truck. The semi continuous mining technology is fit for

mining field with irregular shape, thickness and

inclination of useful minerals, especially the concave

pit.

Semi continuous mining with single bucket and self-

mobile crusher is mining technology without fuel truck.

Its production capacity is generally 300 ~ 5000 ht / .

Due to the rapid development of large tonnage dump

truck, semi continuous mining technology with single

bucket truck and semi self-mobile crusher station and

belt conveyor. Production capacity of semi mobile

station improves especially faster. It greatly increases

production efficiency and reduces production cost soon.

In this stage, semi continuous mining technology with

single bucket and mobile crusher is gradually eliminated

due to the little improvement efficiency.

3. TRANSPORT DISTANCE OF OVERBURDEN

Under ideal conditions, material transport distance can

be calculated according to the loading position shown as

the following:

||,)()(max 22

i

zzyyxxS zx

zxzx

(1)

S is the transport distance, m ;

zx xx , are the X coordinates of unloading and loading

points;

zx yy , are Y coordinates of unloading and loading

points;

zx zz , are z coordinates of unloading and loading

points;

I is the climbing ability of transportation equipment,

% .

Compared to the calculation of transport distance with

equation (1) based on the loading and unloading

position, transport distance of overburden in open pit

mining is mainly due to the layout decisions of transport

system. As shown in Figure 1.

In order to shorten the haul distance of overburden,

shovel-truck process often adopts the manners of a

double row and inner dumping. At a specified mining

level, average transporting distance can be expressed as

the following:

)cotcot(cot2

4

cot2)cot(cot

4

cot2

c

cc

ck

HWL

HLHW

HLS (2)

Where, kS is the average haul distance, m ;

L is the bottom line length in the open pit mining, m ;

cH is the elevation difference between analysis level

height and bottom level height, m ;

is the slope angle of open-pit mine side, (°);

W is the bottom width of the open pit mining, m ;

is the slope angle of working side in open pit

mining, (°);

is the slope angle of dumping side in open pit

mining, (°)





Minelayer

Dumparea

Truck transportation

Truck transportation

Work slopeStep size SyStep size Sy

crushercrusher

Belt conveyerWorking platform

Width of bottom

Inner dump

ll

HH

h

φ H

α

W

β

Figure1. Transport system of open pit mine

When the open pit mine determines to use the integrated

technology in vertical direction, some specific elevation

can be used as assumed bottom elevation. At the

distance between working slope and dump would be

taken as the bottom width. Under specific conditions of

single dump manners in open pit mine, average

transport distance for truck is:

)cot2cot(cot ck HWLS (3)

Truck haul distance of semi continuous process of

overburden is composed of truck transporting distance

and belt conveyor distance. Due to the need of crushing

station service for multiple steps and take end slope

crushing station arrangement as an example, material

hauling distance can be described as the following:

cot*2

c

y

q HsL

S

(4)

)cotcot(cot2

c

y

j HWsL

S (5)

Where, qS is the average truck transporting distance,

m ;

yS is the step distance of crusher, m ;

jS is the average transporting distance of belt conveyor

under the condition of semi continuous process, m ;

)cot2cot(cot cd HWLS (6)

Where, dS is laying length of belt conveyor, m ;

4. OPTIMIZATION MODEL

4.1. Optimization objective

Economic indicators in open pit mining include

investment and production cost. The objective of the

optimization is the minimum total cost in a period. The

objective function with the consideration of investment

and cost can be expressed as the following:

min * * y FC M n Q c C (7)

C is the total cost of the mining system in calculation

period;

M is the system investment; n is the length of

calculation period, year ;

Q is the production capacity of system, am /3 ;

yC is the system operation cost; FC is the additional

costs for the production system.

We would give some following assumptions for the

optimization model:

1) According to the articles and experience of open-pit

stripping with continuous and semi continuous

technology, application experience, transportation link

standby of belt conveyor and buffer hopper would not

be taken into consideration;

2) Crushing station is a key link of semi continuous

process system. Production capacity of the belt

conveyor transportation and disposal system should be

equal to or greater than the capacity of crushing station.

The capacity of crushing station would be taken as the

evaluation index of semi continuous technology system;

3) Take differences of mining equipment depreciation

into consideration, and refer to the experience of open

pit mine technology selection, depreciable life of dump

truck in mine would adapt 8a as a calculation period for

program comparison;

4) Due to the selection of a short period, the production

capacity, cost, and consumable price fluctuations would

not to take into account. That is to say that the static

index would be taken as an evaluation criterion.

4.2. Shovel –truck technology

(1) Investment of the shovel-truck technology

According to the model assumptions, shovel-truck

technology would just consider investment of transport

equipment. According to the system capacity, number

of trucks should be matched and the responsible

investment is:

k

k

kh

c

k

k

k m

Tt

V

Qm

Q

QM *

*

* (8)

Where,

10003600k

k

z

k

xckh

v

S

v

S

ttt

(9)

Where, kM is the truck investment of shovel-truck

system;

kQ is the truck transportation capacity, am /3 ; km is

the price of truck;

cV is the effective loading amount, 3m ; kht is the truck

transport cycle time, h ;

1227 ZHAODONG WANG



kT is the design operating time of truck in each year,

ah / ;

cT is the truck loading time, including actual loading

time, waiting time at the working platform, unloading

time, and so on, s ;

zv is the average running speed when truck is

overloading, hkm / ;

kv is the average running speed when truck is

unloading, hkm / .

(2) system operating expenses

According to the production experience of open-pit

mine, the system operating costs usually include

material costs, energy costs and personnel costs. Of the

three kinds of costs, personnel costs are relatively fixed.

Salary and welfare expenses of operators’ are especially

taken into consideration.

k

rtkkmkckrkkmkcky

Q

knScccSccc

**)(*)( (10)

Where, kc is unit cost of truck transportation;

kcc is the

material costs of truck transportation; knc is the energy

cost of truck transport; krc is the personnel expenses of

shovel-truck system; rn is the number of operators

matching for single equipment; rk is the annual salary

and welfare costs of operating staffs.

4.3. Semi continuous process

(1) System investment

According to the assumption of the model, semi

continuous technology investment including trucks,

crushing equipment, belt conveyor, conveying

equipment, dumping equipment etc. The truck

investment can be calculated by equation (8). The cycle

time of the operation can be corrected with the transport

distance as:

10003600k

q

z

q

xcsh

v

S

v

S

ttt

(11)

pddsqb MSmMMM * (12)

k

k

sh

c

k

q

q m

Tt

V

Qm

Q

QM *

*

* (13)

Where, sht is the transportation time of truck in semi

continuous process, h ;

bM is the investment of semi continuous system;

bm is the truck investment under semi continuous

process;

sM is the crushing equipment investment; dM is the

investment of belt conveyor;

pM is the investment of dumping and transfer

equipment in working surface;

qQ is the truck transportation capacity under semi

continuous technology, 3m ;

(2) System operating expenses

From Figure1 described before, fixed equipment in semi

continuous process system includes crushing equipment,

conveying equipment, belt conveyor, dumping face

conveyor, dumping machine and working platform.

Then, we can calculate the operation cost of semi

continuous process system as the following:

tbrjjnjcqknkcby ccSccSccc *)(*)( (14)

Where,

q

rrrrbr

Q

kn

Q

knc

**6 (15)

byc is the unit cost of operation under semi continuous

process; jcc is the material transport costs of belt

conveyor; jnc is the energy cost of belt conveyer;

brc is

the personnel costs in semi continuous process; tc is

other costs in semi continuous process, including other

link broken, dumping reproduced, etc.

(3) Additional costs

The additional cost considered mainly consists of three

parts:

1) Loosing blasting cost. Loosing blast cost index of

particle size of ore and rock and muck pile shape under

in semi continuous process conditions should consider

mining equipment efficiency, material crushing, and

belt conveyor transportation requirements, and the loose

blasting cost is generally higher;

2) The cost of crusher station equipment moving in

production process. This part of cost can be calculated

according to the moving distance of working slope and

step moving parameters of crusher station in one

calculation period.

3) Opportunity cost of semi continuous technology

investment. This is due to the increased investment

costs. In the model, we would use the investment loan

interest of higher part of semi continuous process to

intermittent process.

eMM

Q

C

S

vQcCCCc kb

y

y

spFJFYFPF *)(***8 (16)





In the equation, FPc is the loosing blasting cost; FYc is

the additional costs of crusher moving; FJc is the

additional costs by additional investment; spcis the

loose blasting cost differences between semi continuous

process and shovel-truck process; v is the push

schedule of open pit mine, am / , Yc is the step cost of

crusher moving; e is the loan interest rate.

4.4. Solution

Based on the analysis above, if using semi continuous

process, cost of it must be less than that of an

intermittent process. Then,

kykkfbybb cQMCCcQMc *8*8 (17)

sp

y

Ykbbyky c

SQ

CvMM

Q

ecc

*

*)(*

8

812

(18)

Then, we can get the boundary conditions of application

of semi continuous technology.

5. APPLICATION

An open pit mining in northern part of China has been

used to verify the theory. The original approved

production capacity of 30 aMt / . The program would

use semi continuous technology with shovel, truck,

ground crushing station and belt semi continuous

process. The stripping is mainly using single bucket and

truck technology. The rock thickness is 32 m on

average and the approved capacity is 45 aMm /3 . Due

to the parameters of tested and experience, operation

forecast of the two process plan can be calculated. As

shown from table 2 to table 5.

The results show that, the upper step of open-pit with

semi continuous process for stripping is reasonable, and

the application can be further expanded.

Payload of truck and belt conveyor can be compared as

shown in table 6.

Table.2 Comparison of mining system on transporting

device

Item Intermittent

technology

Semi

continuous

technology

Truck transport

distance/ m 4323 1253

Truck speed with

load/ 1* hkm 30 25

Truck speed

without

load/ 1* hkm

40 32

Cycle time/ h 278 278

Transport

capacity/13 * aMm

0.45 0.21

Number of trucks 49 24

Length of

conveyer 5498

Transport distance

of conveyer 4245

Table.3 Comparison of mining system on investment

Item Intermittent

technology

Semi

continuous

technology

Truck

investment/ yuan 720 000 000 350 000 000

Investment of

conveyer/ yuan 0.00 124 350 000

Investment of other

equipment/ yuan 0.00 340 000 000

Total investment/ yuan 720 000 000 814 350 000

Table.4 Comparison of mining system on operation

Item Intermittent

technology

Semi

continuous

technology

Material

expenses/ yuan 1 027 500 000 881 320 000

Energy

expenses/ yuan 1 680 630 000 1 703 240 000

Personnel

expenses/ yuan 240 000 000 124 400 000

Total 2 948 130 000 2 708 960 000

Table.5 Comparison of mining system on additional

expenses

Item Intermittent

technology

Semi

continuous

technology

Step moving

expenses/yuan

211 320 000

Interest of

investment/yuan

63 240 000

Total 274 560 000

From table 6 we can see automobile transportation

energy consumption is about 60% for self-weight, and

only about 40% is used for the material carrier. The belt

conveyor transportation energy consumes 80% of the

1229 ZHAODONG WANG



total to transport materials, and just about 20% for the

self-weight. Thus, in suitable conditions, belt conveyor

can be used instead of automobile transportation to save

a lot of energy consumption cost and production cost. It

is especially more obvious of the advantage in the long

distance transportation situation.

Table.6 Payload of the truck and belt conveyor

Item Truck

(154t)

Truck

(290t)

belt

conveyor

Load /t 154 290 605

Weight /t 101 211 140

Total weight of

load and unload /t

101*2+154

=356

211*2+290

=712

605+140=

645

Ratio of load and

total weight 1:2.31 1:2.46 1:1.2

Energy use rate of

load /% 43.2 40.7 81

According to the production experience of large open

pit mine, large truck fuel consumption can get

0.1~0.2 kmL/m3 , fuel costs is accounted for more than

50% of truck operating costs [25]. Therefore, sharp

fluctuations in oil prices will constitute a significant

impact on the shovel truck operation cost. Semi

continuous production process use belt conveyer instead

of trucks transport to increase the transportation

quantity and energy utilization rate. Due to the small

fluctuations in the power price, belt transportation cost

fluctuation is also small. This is good for the cost

accounting of mining and maintaining price stability.

When the semi continuous system is into production,

the cost of the process is low and the benefit would be

improved.

6. CONCLUSION

Due to various factors involved in the mining process

system, optimization work is quite complex, and it is

always an important research field in open pit mining.

Automobile transportation has the characteristics of

high flexibility and strong adaptability. But high self-

weight lead to high energy consumption, high tire

consumption, high repair costs, and high transportation

costs. Especially in the present stage, oil shortages and

rising prices, tire supply shortages and rising prices,

labor force rising prices, environmental protection

requirements to reduce carbon emissions and social

responsibility of corporate, the development of long

distance of truck transport are not conducive to the

mining with economic benefits and social benefits. So,

reduction of truck number is necessary.

In the paper, we give the transport distance under

different mining process according to the recent level of

open-pit mining characteristics, and take minimum total

cost in calculation period as the goal to establish the

open pit mining process optimization model.

Calculation method of economic indicators of

intermittent mining technology and semi continuous

process has been presented. And energy use of truck and

belt conveyor has been compared.

The semi continuous technology combines the

advantages of single bucket excavator, automobile and

belt conveyor. It has rapid development all over the

world, especially used in large scale open pit mine.

Semi continuous process used in open pit mine can get

high total benefit at low cost.

7. Reference:

[1] Wang Xifu, Zhang Daxian, Peng Shij. Engineering

Reliability Study on Combined Mining

Technology. Journal of Fuxin Mining Institute

(Natural Science), 16(1):28-31. (1997).

[2] Ma Ming, Li Hong-Jie, Lei He. 3D fracture

network and block analysis of the south end-slope

of Antaibao Coal Mine. Advances in Civil and

Industrial Engineering, 353-354: 1165-1170.

(2013).

[3] Blackwell G.H. Estimation of large open pit

haulage truck requirements, CIM Bulletin,

92(1028): 143-69. (1999).

[4] Hardy R.H. Selection criteria for loading and

hauling equipment – open pit mining

applications”, PhD thesis, Curtin University of

Technology, Sydney, 2: 558. (2007).

[5] Bastos Guilherme S., Souza Luiz E., Ramos Fabio

T., Ribeiro Carlos H. C. A single-dependent agent

approach for stochastic time-dependent truck

dispatching in open-pit mining. IEEE Conference

on Intelligent Transportation Systems,

Proceedings, ITSC, 1057-1062. (2011).

[6] Bastos Guilherme S. Decision making applied to

shift change in stochastic open-pit mining truck

dispatching?. IFAC Proceedings Volumes (IFAC-

Papers Online), 15(PART 1): 34-39. (2013).

[7] Ringleff Horst. MODERN OPEN-PIT MINING

TECHNOLOGY. Bulk Solids Handling, 5(1): 165-

166. (1985).

[8] Raaz Viktor, Mentges Bergbau Ulrich. Emissions

reduction by means of continuous open pit mining

technology. ThyssenKrupp techforum, 2: 54-59.

(2007).

[9] Benndorf Jörg. Application of efficient methods of

conditional simulation for optimising coal

blending strategies in large continuous open pit

mining operations. International Journal of Coal

Geology, 112:141-153. (2013).

[10] Schroeder D. Technical and economical aspects of

continuous mining with bucket wheel excavators.





ZKG International, Edition B, 47(9): 513-518.

(1994).

[11] Glanz W., Weissflog J. CONTINUOUS

CONVEYING IN OPEN PIT MINES. Bulk Solids

Handling, 2(1): 81-89. (1982).

[12] Wang Hai-Jun, Li Ke-Min, Chen Shu-Zhao. Study

on application scope of stripping semi-continuous

system in internal dumping open pit mine. Journal

of Mining and Safety Engineering, 30(5): 761-765.

(2013).

[13] Jin-Long Sun, Chang-Sheng Ji, Shu-Zhao Chen. A

research on the optimization of layout depth of

semi-continuous technology crusher station in flat

dipping open pit mine. Journal of Convergence

Information Technology, 7(21): 44-52. (2012).

[14] Chen Shu-Zhao, Optimization and application

research on semi-continuous technology system

with mobile crusher in large open pit coal mines.

Journal of the China Coal Society, 37(3): 533-534.

(2012).

A. Kumar, H. Blumberg. A probabilistic model for a

repairable mining equipment inventory system and

procurement policy determination [J]. International

Journal of Surface Mining and Reclamation, (5):

117~121. (1991).

[15] K. Mukhopadhyay. Open pit system reliability [J].

Journal of Mines, Metals & Fuels, (8): 389~392.

(1988).

[16] CHE Zhao-xue, CAI Qing-xiang. Study on Semi-

Contiuous System for Loess Stripping in Surface

Mines of Heidaigou. Journal of China University

of Mining & Technology. 33(6): 673-677. (2004).

[17] M Cardu.C.Clcrici. An experimental research on

the commination law and work index concepts in

jaw crushers. 28th International Mineral

Processing Congress. (1993).

[18] Karparov K.N., Handley, M.F.. Engineering

application of thrust block analysis in slope

stability problems in open pit mines. Journal of the

Southern African Institute of Mining and

Metallurgy, 109(5): 309-327. (2009).

[19] Boland Natashia, Dumitrescu Irina, Froyland Gary,

Gleixner Ambros M.. LP-based disaggregation

approaches to solving the open pit mining

production scheduling problem with block

processing selectivity. Computers and Operations

Research, 36( ): 1064-1089. (2009).

[20] Chicoisne Renaud, Espinoza Daniel, Goycoolea

Marcos, Moreno Eduardo, Rubio Enrique. A new

algorithm for the open-pit mine production

scheduling problem. Operations Research, 60(3):

517-528. (2012).

[21] Yang Biao, Luo Zhou-Quan, Lu Guang, Lu Hao,

Wang Yi-Wei. Open pit mining limit dynamic

optimization based on economic time series

forecasting. Journal of the China Coal Society,

36(1): 29-33. (2011).

[22] Li Shouguo, Yang Chang, Lu Caiwu. Cut-off grade

optimization using stochastic programming in

open-pit mining. 6th International Conference on

Internet Computing for Science and Engineering,

ICICSE 2012, 66-69. (2012).

[23] Asad Mohammad Waqar Ali, Dimitrakopoulos

Roussos. A heuristic approach to stochastic cutoff

grade optimization for open pit mining complexes

with multiple processing streams. Resources

Policy, 38(4): 591-597. (2013).

[24] DENG Ming. The measures to reduce production

cost of Heidaigou open-pit coal mine. Opencast

Mining Technology, (3): 119-121. (2012).

Documents

INTERNATIONAL JOURNAL OF EARTH SCIENCES AND ENGINEERING CAFET-INNOVA Technical Society Hyderabad, INDIA