Performance of geopolymer concrete in fire · Performance of Geopolymer Concrete in Fire . AHMAD ZURISMAN MOHD ALI . This thesis is presented as part of the requirement for the

Performance of Geopolymer Concrete in Fire

AHMAD ZURISMAN MOHD ALI

This thesis is presented as part of the requirement for the

Award of the Degree of

Doctor of Philosophy

Department of Civil and Construction Engineering

Faculty of Science, Engineering and Technology

Swinburne University of Technology

August 2015

ii

Abstract

Ordinary Portland Cement (OPC) concrete can be classified as a fire resistant

construction material. However, spalling may occur in fire especially in high strength

concretes. The need to address this spalling problem in fire has intensified since the

use of high rise buildings. Geopolymer is a cementless concrete binder which is based

on aluminosilicate reaction of fly ash, a power plant by-product and activated by

alkaline solution. Recent research has shown that geopolymer concrete has great

potential in resisting fire. In addition, without the presence of Portland cement in the

concrete has led to more sustainable construction material since cement is the second

highest CO2 emitter after fossil fuels.

In this research, the severity of large scale high strength concrete spalling under

hydrocarbon fire was investigated. The results show that high strength OPC concrete

severely spalled during the first 30 minutes of fire exposure in explosive manner. There

is a thermal diffusivity drop when the temperature is between 110oC and 155oC due to

water in concrete changing phase to steam. Maximum steam pressure corresponds to

155oC temperature is 0.44 MPa which shows that steam pressure alone is not a critical

factor for concrete spalling.

Further investigations on factors affecting concrete spalling show that larger specimen

spalled more than smaller scale specimen based on 3 m x 3 m walls, 1 m x 1 m walls,

columns and cylinders. Aggregates size effect was also observed with increasing

spalling with decreasing aggregate size. No obvious trend observed for different

aggregate types (granite and basalt).

For making geopolymer, fly ashes from Gladstone, Tarong and Microash were

examined for physical properties. Gladstone fly ash demonstrated the best workability

compared with Tarong and Microash. The reasons for workability differences were

investigated and the conclusion was agglomerated particle size distribution was the

cause

Effect of dry and wet condition on strength of geopolymer was investigated. Strength

reduction of 17% was observed.

The high strength geopolymer 1 m x 1 m wall panels and two sizes of cylinders (150,

100 mm) were tested. From hydrocarbon fire test, high strength geopolymer concrete

exhibited good fire resistance performance due to no explosive spalling observed. Two

iii

panels and all cylinders remained un-spalled. Less than 1% spalling (excluding

moisture loss) observed on 2 spalled panels. There are 11% - 65% of residual

strengths observed on fire tested specimens as compared to OPC which has no

remaining strength.

To investigate aggregate and geopolymer binder thermal incompatibility factor on

spalling, aerated geopolymer panels were tested. From hydrocarbon fire test, no

spalling observed was further proof of thermal incompatibility was the cause of spalling.

Under room temperature, bearing load, axial load, corner bearing and flexural tests on

aerated geopolymer panel were conducted. Results show that aerated geopolymer

panel has the capacity and strength to be used as structural wall in fire application.

iv

Declaration

This thesis contains no material that has been accepted for the award of any Degree or

Diploma in any University. To the best of my knowledge and belief, the thesis contains

no material previously published or written by another person except where due

reference is made in the text.

………………………………

Ahmad Zurisman Mohd Ali

v

Acknowledgement

I wish to express my sincere gratitude and appreciation to many people for their

support and encouragement throughout this research. First and foremost, I would like

to express my sincere thanks to my research supervisor, Professor Jay Sanjayan for

his wonderful assistance, invaluable guidance, encouragement and support during the

course of this research. This thesis would not have been possible to complete without

his help. Special thanks to my co-supervisor, Professor Kwesi Sagoe-Crentsil for his

advice, assistance and invaluable guidance.

I gratefully acknowledge Ministry of Education Malaysia and my employer, Universiti

Tun Hussein Onn Malaysia for providing me the financial support and opportunity to

carry out this research.

I would like to extend my gratitude to Smart Structure Lab staff, Michael Culton, Kia

Rasekhi and Sanjeet for their technical assistance throughout my laboratory works.

I would like to thank Dr Maurice Guerrieri from Centre of Risk and Safety Engineering

(CESARE), Victoria University, Werribee for his assistance on fire tests.

I am also indebted to Mr David Devenish from Commonwealth Scientific and Industrial

Research Organisation (CSIRO) Highett for helping me casting the geopolymer

concrete panels and Dr Rackel San Nicolas from Melbourne University for assisting me

on chemical laboratory works.

To my fellow research colleague, Hasnat, Chandani, Behzad, Gias, Myintzu, Fahad,

Nainesh and others, thanks for the moral support, listening time and stimulating

discussions.

Last but not least, my sincerest gratitude to wife, Siti Nazahiyah and my wonderful kids,

Ilham Amani and Irdina Amani for the unconditional love, understanding, sacrifice and

support they provide during all these years.

vi

Table of Content Abstract ....................................................................................................................... ii

Declaration ................................................................................................................. iv

Acknowledgement ...................................................................................................... v

Table of Content ........................................................................................................ vi

List of Figures ........................................................................................................... xii

List of Tables............................................................................................................ xvi

List of Abbreviations ............................................................................................. xviii

CHAPTER 1 ................................................................................................................. 1

Introduction ............................................................................................................... 1

1.1 Background ................................................................................................. 1

1.2 Aim of research ........................................................................................... 2

1.3 Research objectives .................................................................................... 2

1.4 Scope of the thesis ...................................................................................... 2

1.5 Thesis Organisation .................................................................................... 4

References ............................................................................................................ 6

CHAPTER 2 ................................................................................................................. 7

Literature Review ...................................................................................................... 7

2.1 Background ................................................................................................. 7

2.2 Geopolymer Binder in Concrete for Spalling Solution ................................ 11

2.3 Geopolymer Binder in Concrete for Environmental Benefits ...................... 12

2.4 Geopolymer History and Reaction Mechanisms ........................................ 13

2.5 Fly Ash ...................................................................................................... 17

2.6 Alkaline activator ....................................................................................... 20

2.7 Geopolymer workability ............................................................................. 21

2.8 Geopolymer compressive strength ............................................................ 22

2.9 Geopolymer under Environmental Exposure ............................................. 25

2.10 Geopolymer fire test .................................................................................. 26

References .......................................................................................................... 28

vii

CHAPTER 3 ............................................................................................................... 37

Hydrocarbon Fire testing of full size Portland Cement High Strength Concrete Wall

Panels ..................................................................................................................... 37

3.1 Introduction ............................................................................................... 37

3.2 Experimental Programme .......................................................................... 38

3.2.1 Material .............................................................................................. 38

3.2.2 Concrete Properties ........................................................................... 39

3.2.3 Test Setup ......................................................................................... 40

3.2.4 Compressive strength and moisture content ...................................... 41

3.2.5 Thermal Data Collection ..................................................................... 42

3.2.6 Spalling analysis ................................................................................ 44

3.2.7 Thermal diffusivity analysis ................................................................ 45

3.3 Result and Discussion ............................................................................... 46

3.3.1 Compressive Strength and Moisture Content ..................................... 46

3.3.2 Spalling .............................................................................................. 47

3.3.3 Temperature Result ........................................................................... 48

3.3.4 Thermal Diffusivity.............................................................................. 53

3.4 Conclusion ................................................................................................ 55

References .......................................................................................................... 56

CHAPTER 4 ............................................................................................................... 59

Specimen’s Size, Aggregate Size and Aggregate Type Effect on Spalling of Concrete

in Fire ...................................................................................................................... 59

4.1 Introduction ............................................................................................... 59

4.2 Experimental Programme .......................................................................... 60

4.2.1 Concrete Test Specimens .................................................................. 61

4.2.2 Hydrocarbon Fire Test ....................................................................... 65

4.2.3 Spalling Measurements ...................................................................... 67

4.2.4 Nominal Spalling Depth Analysis ........................................................ 67

4.3 Result and Discussion ............................................................................... 69

viii

4.3.1 Spalling .............................................................................................. 69

4.3.2 Nominal Depth ................................................................................... 71

4.3.3 Specimen’s size effect ....................................................................... 71

4.3.4 Aggregate Size Effect ........................................................................ 74

4.3.5 Aggregate Type Effect ...................................................................... 75

4.4 Conclusion ................................................................................................ 76

References .......................................................................................................... 78

CHAPTER 5 ............................................................................................................... 80

Investigation of the Effects of Fly Ash Types and Properties on the Workability of

Fresh Geopolymer................................................................................................... 80

5.1 Introduction ............................................................................................... 80

5.2 Experimental Program............................................................................... 82

5.2.1 Raw Material ...................................................................................... 82

5.2.2 Workability Test ................................................................................. 82

5.2.3 Investigation of the Relevant Physical Properties of Fly Ash .............. 83

5.2.4 Investigation of the Relevant Chemical Properties of Fly Ash ............. 85

5.3 Results and Discussions ........................................................................... 86

5.3.1 Workability Results............................................................................. 86

5.3.2 Analyses of the Measured Physical Properties of Fly Ash .................. 88

5.3.2.1 Particle Size Analyses .................................................................... 88

5.3.2.2 Particle Shape Analyses ................................................................. 94

5.3.3 Chemical Analyses............................................................................. 95

5.4 Conclusion ................................................................................................ 97

References ........................................................................................................ 100

CHAPTER 6 ............................................................................................................. 103

Strength of Geopolymers in Saturated and Dry Conditions.................................... 103

6.1 Introduction ............................................................................................. 103

6.2 Experimental Programme ........................................................................ 104

6.2.1 Material ............................................................................................ 104

ix

6.2.2 Specimens Details ........................................................................... 105

6.2.3 Drying and Saturation Method .......................................................... 105

6.2.4 Compressive Test ........................................................................... 106

6.2.5 Inductively Couple Plasma Test ...................................................... 106

6.3. Result and Discussion ............................................................................ 106

6.3.1 Compressive Strength ..................................................................... 106

6.3.2 Effect of Strength on Reduction of Strength in Saturated Geopolymer

109

6.3.3 Excess Sodium Silicate Factor ........................................................ 110

6.4 Conclusion .............................................................................................. 111

References ........................................................................................................ 112

CHAPTER 7 ............................................................................................................. 114

Hydrocarbon Fire testing of Geopolymer High Strength Concrete Wall Panels and

Cylinders ............................................................................................................... 114

7.1 Introduction ............................................................................................. 114


7.2.1 Material ............................................................................................ 115

7.2.2 Geopolymer Concrete Mix ................................................................ 117

7.2.3 Geopolymer concrete casting ........................................................... 118

7.2.4 Geopolymer concrete properties ...................................................... 123

7.2.5 Hydrocarbon fire test setup .............................................................. 125

7.2.6 Thermal diffusivity ............................................................................ 127

7.2.7 Compressive strength ...................................................................... 128

7.3 Results and Discussion ........................................................................... 129

7.3.1 Spalling ............................................................................................ 129

7.3.2 Temperature Results ........................................................................ 131

7.3.3 Thermal Diffusivity............................................................................ 135

7.3.4 Residual Compressive Strength After Fire Exposure ........................ 136

7.4 Conclusions ............................................................................................ 138

x

References ........................................................................................................ 140

CHAPTER 8 ............................................................................................................. 142

Investigation on Aerated Geopolymer Wall Panels for Fire Applications ................ 142

8.1 Introduction ............................................................................................. 142


8.2.1 Material ............................................................................................ 143

8.2.2 Aerated Geopolymer Panel Properties ............................................. 145

8.2.3 Room Temperature Test .................................................................. 146

8.2.3.1 Bearing Load Test ........................................................................ 147

8.2.3.2 Axial Load Test ............................................................................. 148

8.2.3.3 Corner Bearing Test ..................................................................... 148

8.2.3.4 Flexural Load Test ........................................................................ 149

8.2.4 Hydrocarbon Fire Test ..................................................................... 150

8.2.4.1 Test setup..................................................................................... 150

8.3 Results and Discussion ........................................................................... 152

8.3.1 Room Temperature Test .................................................................. 152

8.3.1.1 Bearing Load Test ........................................................................ 152

8.3.1.2 Axial Load Test ............................................................................. 154

8.3.1.3 Corner Bearing Test ..................................................................... 157

8.3.1.4 Flexural Load Test ........................................................................ 158

8.3.2 Hydrocarbon Fire Test ..................................................................... 161

8.3.2.1 Spalling ........................................................................................ 161

8.3.2.2 Temperature Results .................................................................... 162

8.4 Conclusions ............................................................................................ 164

References ........................................................................................................ 165

CHAPTER 9 ............................................................................................................. 167

Summary, Conclusions and Recommendations .................................................... 167

9.1 Summary................................................................................................. 167

9.2 Conclusions ............................................................................................ 168

xi

9.3 Recommendations .................................................................................. 170

xii

List of Figures Figure 2.1: Types of polysialates (Davidovits, 2002) ................................................... 15

Figure 2.2: Conceptual model for geopolymerisation (Duxson et al., 2007) ................. 16

Figure 2.3: SEM pictures (a) original fly ash, (b) fly ash activated with NaOH

(Fernández-Jiménez et al., 2005) ........................................................................ 16

Figure 2.4: Fly ash production (Flyash Australia, 2010) .............................................. 18

Figure 3.1: Reinforcement details of concrete wall panel (top view) ............................ 38

Figure 3.2: Concrete wall panel dimension and fire exposed area .............................. 39

Figure 3.3: Casting of Ordinary Portland Cement High Strength Concrete .................. 40

Figure 3.4: Casted Full-scale (3.36 m x 3.38 m) Portland Cement High Strength ........ 40

Concrete ..................................................................................................................... 40

Figure 3.5: TramexTM CME 4 moisture meter .............................................................. 42

Figure 3.6: Hydrocarbon and Standard Fire Temperature versus Time Profile ............ 42

Figure 3.7: Wall Panel and Furnace Setup .................................................................. 43

Figure 3.8: Hydrocarbon Fire Furnace ........................................................................ 43

Figure 3.9: Illustration of Thermal Couple’s Location................................................... 44

Figure 3.10: Thermal Couple Set Up ........................................................................... 44

Figure 3.11: Specimen After Fire Test ........................................................................ 47

Figure 3.12: Spalling Concrete Percentage ................................................................. 48

Figure 3.13: Temperature for specimens with 7 mm basalt aggregates ...................... 48

Figure 3.14: Temperature for specimens with 14 mm basalt aggregates .................... 49

Figure 3.15: Temperature for specimens with 20 mm basalt aggregates .................... 49

Figure 3.16: Temperature for specimens with 7 mm granite aggregates ..................... 50

Figure 3.17: Temperature for specimens with 14 mm granite aggregates ................... 50

Figure 3.18: Temperature across the thickness of the specimens with 7 mm basalt

aggregates........................................................................................................... 51


aggregates........................................................................................................... 51


aggregates........................................................................................................... 52

Figure 3.21: Temperature across the thickness of the specimens with 7 mm granite

aggregates........................................................................................................... 52


aggregates........................................................................................................... 53

xiii

Figure 3.23: Thermal diffusivity vs Temperature ......................................................... 54

Figure 4.1: Specimens’ dimension .............................................................................. 63

Figure 4.2: Casting of Ordinary Portland Cement High Strength Concrete .................. 63

Figure 4.3: Reinforcement bar details (a) large panel (b) medium panel (c) column .... 65

Figure 4.4: Specimen Fire Test Setup (a) Large panels (b) Medium panels (c) Columns

(d) Cylinders ........................................................................................................ 67

Figure 4.5: Fire Exposed Area (a) Large panels (b) Medium panels (c) Columns (d)

Cylinders ............................................................................................................. 69

Figure 4.6: Images of specimens after fire test (a) Large Panel (b) Medium Panel

(c) Column (d) Cylinder ........................................................................................ 71

Figure 4.7: Nominal Spalling Depth. ............................................................................ 73

Figure 4.8: Nominal Spalling Depth for Maximum Aggregate Size .............................. 75

Figure 4.9: Nominal Spalling Depth for Maximum Aggregate Type ............................. 76

Figure 5.1: Measurement of d1 and d2. ........................................................................ 82

Figure 5.2: CILAS Particle Size Analyzer .................................................................... 84

Figure 5.3: ZEISS Supra 40VP Scanning Electron Microscope ................................... 84

Figure 5.4: Belsorp Max Adsorption Measurement ..................................................... 85

Figure 5.5: Fly ash density measurement equipment .................................................. 85

Figure 5.6: Varian 720-ES inductively coupled plasma equipment .............................. 86

Figure 5.7: Geopolymers with alkaline solution to fly ash ratio of 0.4 for (a) Gladstone.

............................................................................................................................ 87

(b) Tarong. (c) Microash ............................................................................................. 87

Figure 5.8: Particle size distribution graphs for (a) Gladstone fly ash. (b) Tarong fly ash.

(c) Microash ......................................................................................................... 90

Figure 5.9: Adsorption and desorption curves for (a) Gladstone fly ash (b) Tarong fly

ash (c) Microash .................................................................................................. 92

Figure 5.10: Images of Gladstone fly ash in loose grainy form (left) and solid form

(right). .................................................................................................................. 93

Figure 5.11: Scanning Electron Microscope (SEM) Images comparison for Gladstone

Power Station fly ash and Tarong Power Station fly ash. ..................................... 95

Figure 5.12: Mixtures of demineralized water and fly ash from (a) Gladstone Power

Station. (b) Tarong Power Station. (c) Microash .................................................. 97

Figure 6.1: Compressive Strength of Dry and Saturated Geopolymer ....................... 110

Figure 7.1: Aggregate Particle Size Distribution Graph ............................................. 116

Figure 7.2: Sand Particle Size Distribution Graph ..................................................... 116

Figure 7.3: Alkaline activator’s mixer ......................................................................... 119

Figure 7.4: High shear concrete mixer ...................................................................... 119

xiv

Figure 7.5: Dry components mixing process ............................................................. 120

Figure 7.6: Alkaline activator added into the mix ....................................................... 120

Figure 7.7: Geopolymer concrete in high shear concrete mixer ................................ 121

Figure 7.8: Geopolymer fresh concrete ..................................................................... 121

Figure 7.9: Cast geopolymer concrete panel ............................................................. 122

Figure 7.10: Geopolymer concrete panel being steam cured .................................... 122

Figure 7.11: Demoulded hardened geopolymer concrete panels .............................. 123

Figure 7.12: Geopolymer concrete dimensions (a) panel (b) cylinders ...................... 124

Figure 7.13: TramexTM CME 4 moisture meter .......................................................... 124

Figure 7.14: Reinforcement details for geopolymer concrete panel ........................... 125

Figure 7.15: Geopolymer concrete panels and cylinders furnace setup illustration ... 126

Figure 7.16: Images of panels (left) and cylinders (right) hydrocarbon fire test setup 126

Figure 7.17: ProceqTM SilverSchmidt rebound hammer ............................................ 128

Figure 7.18: Geopolymer concrete panels after 2 hours hydrocarbon fire exposure .. 129

Figure 7.19: Geopolymer concrete cylinders after 2 hours hydrocarbon fire exposure

.......................................................................................................................... 130

Figure 7.20: Temperature for panel 1 (P1) ................................................................ 131



Figure 7.23: Temperature across thickness for panel 1 ............................................. 133



Figure 7.26: Temperature across thickness comparison between geopolymer concrete

and Portland cement HSC ................................................................................. 135

Figure 7.27: Thermal diffusivity vs Temperature ....................................................... 136

Figure 7.28: Fire tested geopolymer concrete cylinders 150 mm diameter (left) and 100

mm diameter (right) after compressive strength test .......................................... 138

Figure 8.1: Sand Particle Size Distribution Graph ..................................................... 144

Figure 8.2: Aerated geopolymer panel cross section ................................................ 146

Figure 8.3: Reinforcement details for aerated geopolymer panel .............................. 146

Figure 8.4: TramexTM CME 4 moisture meter ............................................................ 147

Figure 8.5: Bearing load test setup ........................................................................... 147

Figure 8.6: Axial load test setup ................................................................................ 148

Figure 8.7: Corner bearing test setup ........................................................................ 149

Figure 8.8: Flexural load test setup ........................................................................... 149

Figure 8.9: Hydrocarbon temperature versus time curves ......................................... 150

Figure 8.10: Aerated geopolymer concrete panels furnace setup illustration ............. 151

xv

Figure 8.11: Image of aerated geopolymer panels fire test setup .............................. 151

Figure 8.12: Bearing load vs displacement graph for panel 1 .................................... 153



Figure 8.15: Photos of panels’ shear failure after bearing load test ........................... 154

Figure 8.16: Axial load vs displacement for panel 1 .................................................. 155



Figure 8.19: Buckling failure of panel 1 ..................................................................... 156

Figure 8.20: Localised failure at bottom of panels after axial load test ...................... 157

Figure 8.21: Photos of panels after corner bearing test ............................................. 158

Figure 8.22: Flexural load versus displacement for panels with Crust up .................. 159

Figure 8.23: Flexural load versus displacement for panels with Crust down .............. 159

Figure 8.24: Photos of panels’ cracks after flexural load test..................................... 161

Figure 8.25: Aerated geopolymer panels after 30 minutes hydrocarbon fire exposure

.......................................................................................................................... 162

Figure 8.26: Temperature for aerated geopolymer panel .......................................... 163


xvi

List of Tables Table 2.1: Bibliographic history of some important events/articles about alkali-activated

cements (Li et al., 2010) ...................................................................................... 13

Table 2.2: Chemical requirement for Class F and Class C fly ashes according to ASTM

618 (1994) ........................................................................................................... 18

Table 2.3: Average chemical content for fly ashes in Australia (Heidrich, 2003) .......... 19

Table 3.1: Concrete (per cubic meter) mix summary ................................................... 39

Table 3.2: Concrete Casting and Fire Test Details ...................................................... 41

Table 3.3: Compressive strength and moisture content summary ............................... 46

Table 4.1: Concrete (per cubic meter) mix summary ................................................... 61

Table 4.2: Specimens’ fire test details ......................................................................... 64

Table 4.3: Spalling Percentage Summary for All Specimens....................................... 72

Table 5.1: Relative slump summary ............................................................................ 88

Table 5.2: Average Diameter Summary ...................................................................... 89

Table 5.3: Adsorption test (BET Theory) results summary .......................................... 92

Table 5.4: Bulk density of fly ash ................................................................................ 93

Table 5.5: Water Absorption Test Results ................................................................... 93

Table 5.6: Chemical Component and slump-flow of Fly Ash ....................................... 96

Table 5.7: Chemical concentration for water filtered from fly ash and demineralized

water mixture ....................................................................................................... 97

Table 6.1: Chemical composition of fly ash (Kong and Sanjayan, 2010) ................... 105

Table 6.2: Result summary of geopolymers with 0.4 alkaline solution to fly ash ratio and

2.5 Na2SiO3 to NaOH ratio ................................................................................. 107


1.75 Na2SiO3 to NaOH ratio ............................................................................... 107


1.0 Na2SiO3 to NaOH ratio ................................................................................. 108

Table 6.5: Result summary of geopolymers with 0.57 alkaline solution to fly ash ratio

and 2.5 Na2SiO3 to NaOH ratio .......................................................................... 109

Table 6.6: ICP Test Results ...................................................................................... 110

Table 7.1: Chemical composition of Gladstone fly ash .............................................. 115

Table 7.2: Sodium silicate specification .................................................................... 117

Table 7.3: Geopolymer concrete mixture proportions ................................................ 118

Table 7.4: Moisture content and testing date details ................................................. 124

Table 7.5: Geopolymer concrete spalling summary .................................................. 130

Table 7.6: Compressive Strength and Density Summary .......................................... 136

xvii

Table 7.7: Residual compressive strength summary ................................................. 137

Table 8.1: Chemical component of Gladstone fly ash ............................................... 144

Table 8.2: Sodium silicate specification .................................................................... 145

Table 8.3: Panels weight and density summary ........................................................ 152

Table 8.4: Results of Bearing Load Tests ................................................................. 152

Table 8.5: Results of Axial Load Tests ...................................................................... 156

Table 8.6: Results of Corner Bearing Tests .............................................................. 157

Table 8.7: Results of Flexural Load Tests ................................................................. 159

Table 8.8: Calculated Maximum Loads for various d, f’c and fy ................................. 160

Table 8.9: Aerated geopolymer panel spalling summary ........................................... 162

xviii

List of Abbreviations

ft - tensile strength of concrete

fy - compressive strength of concrete

HSC - high strength concrete

ICP - inductively couple plasma

K - thermal diffusivity

LOI - loss on ignition

PS - polysialate

PSS - polysialate-siloxo

PSDS - polysialate-diloxo

PC - pulvarized coal

rp - relative slump

SEM - scanning electron microscopic

SF - slump-flow

t - time (minutes)

T - temperature (oC)

XRF - X-Ray fluorescence

1

CHAPTER 1

Introduction

1.1 Background

Concrete structures were designed to withstand various types of environmental

conditions categorised as from mild to very severe conditions. Fire represents one of

the most severe environmental conditions to which concrete structures may be

subjected to such as close conduit structure like tunnel (Kim et al., 2010; Khaliq and

Kodur, 2011). There are many reasons that can trigger fire event in a tunnel or high

rise building. Fire event in Tunnel Mont Blanc in 1999 was initiated from a breakdown

of a truck carrying margarine and flour. In 2010, a 28-story high-rise apartment caught

on fire ignited from welding spark killing at least 50 people. A recent fire event occurred

in 2012, when two buses collided in the 12.9 km Hsuehshan Tunnel, Taiwan causing a

passenger bus to catch fire after it was rear-ended by another bus. Generally,

concrete is regarded as a fire resistant construction material, especially when

compared to the alternatives such as steel and timber. However, concrete is

susceptible to a less known phenomenon termed spalling in fire. Spalling normally

occurs on concrete exposed to fire especially high strength concrete structure in which

widely used nowadays in high rise buildings and tunnels. (Sanjayan and Stocks, 1993).

The new breed of concrete binder considered to be the binder of the future due to less

CO2 emission, geopolymer is a cementless concrete binder. The name geopolymer

2

was coined by Davidovits in 1975 (Davidovits, 1991; Van Jaarsveld et al., 2002).

Geopolymer concrete is known as very high fire resistance material due to its ceramic-

like properties (Davidovits and Davidovics, 1991). Geopolymer concrete strength after

elevated temperature exposure can increase due to further geopolymerisation (Pan et

al., 2009). However, there is decrease in strength of geopolymer concrete after

elevated temperature exposure especially geopolymer concrete with coarse aggregate

due to thermal incompatibility between aggregates and binder which can contribute to

the decline in geopolymer concrete strength (Kong and Sanjayan, 2008).

To-date, the chemistry of geopolymer has been extensively focused by many studies.

Some investigations have been undertaken to study the material properties of

geopolymer concrete and structural behaviour of geopolymer concrete structural

elements such as beams and columns, high performance geopolymer concrete and

steel fibre reinforced geopolymer concrete. Only a few have investigated the

performance of geopolymer concrete in fire, let alone large scale size geopolymer

concrete and hydrocarbon fire exposure.

1.2 Aim of research

The aim of the research is to develop a fire resistant geopolymer concrete when

exposed to hydrocarbon fire.

1.3 Research objectives

The objectives of the research are listed below:

1) To investigate the performance of large scale high strength geopolymer

concrete panels and aerated geopolymer concrete in hydrocarbon fire.

2) To investigate the severity of large scale high strength Portland cement

concrete spalling when exposed to hydrocarbon fire.

3) To investigate several issues affecting high strength geopolymer concrete

casting in mass production.

1.4 Scope of the thesis

This research was undertaken in two main parts; Portland cement high strength

concrete part and geopolymer concrete part. In Portland cement high strength part, the

severity of high strength Portland cement concrete spalling was investigated by

exposing large scale 3 m square wall panel to hydrocarbon fire. The weight loss of

panels after fire test was measured and the spalling percentage was calculated based

3

on the weight loss. The panel’s temperatures were measured at several depths of wall

for further thermal properties analysis.

In addition, specimen size effect on spalling was investigated by comparing the spalling

of 3 m square panels with 1 m square panels, columns and cylinders. All panels were

cast with different maximum aggregate size (7 mm, 14 mm and 20 mm) and type

(basalt and granite) for investigation on aggregate size and type effect on spalling.

In geopolymer concrete part, 3 types of fly ash were initially selected as aluminosilicate

source of geopolymer concrete namely Gladstone fly ash (from Gladstone power

station), Tarong fly ash (from Tarong power station) and Microash. Since the

geopolymer concrete were cast in large scale production, workability of fresh

geopolymer was used as main criteria in selecting the fly ash. Therefore, Gladstone fly

ash was selected because fresh geopolymer exhibited the best workability compared to

Tarong and Microash. Factors affecting the workability of these fly ashes were

investigated out by examining the physical properties of the fly ashes such as mean

particle size, particle size distribution, pore volume, water absorption and density. The

effect of water to geopolymer mortar was investigated by examining the strength of

geopolymer paste in saturate and dry conditions. The high strength geopolymer

concrete mix was determined based on the outcome of fly ash workability and water

effect investigation.

The investigation on the performance of high strength geopolymer concrete was

carried out by exposing 1 m square wall panels and 2 sizes of cylinders (100 mm and

150 mm diameter) to hydrocarbon fire. Concrete spalling, temperature profile and

residual strength were measured. Thermal incompatibility between aggregate and

geopolymer effect on spalling was also investigated. Aerated geopolymer concrete

(without coarse aggregate) wall panels were tested for hydrocarbon fire exposure. In

addition, investigation on the suitability of aerated geopolymer concrete to be used as

structural member was carried out by testing the panels for axial load capacity, bearing

capacity, flexural strength and corner bearing capacity. The overview of the research is

illustrated in Figure 1.1

4

Performance of Geopolymer Concrete In

Fire

Portland Cement High

Strength Concrete

Hydrocarbon Fire testing of full size Portland Cement

High Strength Concrete Wall

Panels

Geopolymer Concrete

Specimen’s Size, Aggregate

Size and Aggregate Type

Effect on Spalling of

Concrete in Fire

Investigation of the Effects of Fly Ash Types and Properties

on the Workability of

Fresh Geopolymer

Strength of Geopolymers in Saturated

and Dry Conditions

Hydrocarbon Fire testing of Geopolymer

High Strength Concrete Wall

Panels and Cylinders

Investigation on Aerated

Geopolymer Wall Panels for

Fire Applications

Figure 1.1: Research Overview

1.5 Thesis Organisation

This thesis consists of nine chapters. Chapter 1 describes the motivation of

developing fire resistant geopolymer concrete, the aim of the research, objectives of

the research and brief description on the research content. Chapter 2 of the thesis

covers literature review of the general scenario of concrete performance in under high

temperature. Several fire events which cause devastating effect on concrete

structurally, the potential of geopolymer concrete as fire resisting material and the

benefits of geopolymer concrete environmentally are mentioned in this chapter. In

addition, literature on geopolymer concrete including fundamental chemistry of

geopolymer, investigation on geopolymer compressive strength, investigation on

geopolymer workability and fire exposure on geopolymer concrete investigation are

also presented.

Chapter 3 of the thesis presents the results of the severity of full size 3 m square

Portland cement high strength concrete spalling when exposed to hydrocarbon fire. In

addition, thermal diffusivity values for the concrete are also determined.

Chapter 4 of the thesis presents the investigation on factors affecting high strength

concrete spalling. Factors that have been investigated are specimen’s size, aggregate

size and aggregate type.

5

Aluminosilicate source (fly ash) for geopolymer was investigated and reported in

Chapter 5. This chapter presents the effect of physical properties of fly ash on

workability of geopolymer. The source of fly ash used in this research was decided

based on the outcome of this chapter.

Chapter 6 of the thesis presents the strength of geopolymer paste in both saturated

and dry condition. Geopolymer is basically an aluminosilicate reaction to produce gel

and hardening process through polymerization. Unlike hydration process in Portland

cement binder, geopolymer’s reaction does not involve free water during mixing stage.

Therefore, strength reduction when geopolymer immersed in water for saturation

process was investigated.

Chapter 7 of the thesis presents the performance of high strength geopolymer concrete

exposed to hydrocarbon fire. High strength geopolymer mix calculation and casting

procedure are also presented. High strength geopolymer concrete spalling,

temperature profile, thermal diffusivity graphs and residual strength are discussed and

reported in this chapter.

Chapter 8 of the thesis describes the investigation of aggregate and geopolymer binder

thermal incompatibility effect on spalling and the suitability of aerated geopolymer

concrete wall to be used as structural member in fire application. Aerated geopolymer

concretes wall without coarse aggregate were tested under room temperature (flexural,

bending, axial load and corner bearing) and in fire conditions.

Finally in Chapter 9 of the thesis, overall conclusions made in this research are

provided. In addition, recommendations of future work and discussions on enhancing

geopolymer concrete research are also presented.

6

References

Davidovits, J. (1991). "Geopolymers - Inorganic polymeric new materials." Journal of

Thermal Analysis 37(8): 1633-1656.

Davidovits, J. and Davidovics, M. (1991). Geopolymer. Ultra-high temperature tooling

material for the manufacture of advanced composites. International SAMPE

Symposium and Exhibition (Proceedings).

Khaliq, W. and Kodur, V. K. R. (2011). "Effect of High Temperature on Tensile Strength

of Different Types of High-Strength Concrete." ACI Materials Journal 108(4):

394-402.

Kim, J. H. J., Mook Lim, Y., Won, J. P. and Park, H. G. (2010). "Fire resistant behavior

of newly developed bottom-ash-based cementitious coating applied concrete

tunnel lining under RABT fire loading." Construction and Building Materials

24(10): 1984-1994.

Kong, D. L. Y. and Sanjayan, J. G. (2008). "Damage behavior of geopolymer

composites exposed to elevated temperatures." Cement and Concrete

Composites 30(10): 986-991.

Pan, Z., Sanjayan, J. G. and Rangan, B. V. (2009). "An investigation of the

mechanisms for strength gain or loss of geopolymer mortar after exposure to

elevated temperature." Journal of Materials Science 44(7): 1873-1880.

Sanjayan, G. and Stocks, L. J. (1993). "Spalling of high-strength silica fume concrete in

fire." ACI Materials Journal 90(2): 170-173.

Van Jaarsveld, J. G. S., Van Deventer, J. S. J. and Lukey, G. C. (2002). "The effect of

composition and temperature on the properties of fly ash- and kaolinite-based

geopolymers." Chemical Engineering Journal 89(1-3): 63-73.

7

CHAPTER 2

Literature Review

2.1 Background

Generally, concrete is regarded as a fire resistant construction material,

especially when compared to the alternatives such as steel and timber. However,

concrete is susceptible to a well-known phenomenon termed spalling in fire. Spalling is

a physical process of the breakdown of surface layers of concrete which crumble into

small pieces in response to high temperatures and/or mechanical pressure. Spalling of

concrete in fire is dislodgement of small pieces of concrete (chips up to 50 mm)

popping out from the surface of concrete, often explosive in nature. Explosive spalling

may have a very severe impact on the surrounding environment. Pieces of smashed

concrete can fly with high speed and explosive energy causing severe casualty (Ali et

al., 2001). For example, a fire in Channel tunnel (35 km railroad tunnel connecting

England and France) in 1996 caused severe damage to concrete tunnel rings owing to

the spalling of concrete and resulted in six-month closure for repairs costing US$1.5

million per day (Ulm et al., 1999). On 24 March 1999, a cargo truck carrying margarine

and flour in the Mont Blanc tunnel connecting France and Italy caught fire and stopped

at Kilometre 6.7 causing 39 human deaths and severely damaging 900 m long tunnel

roof due to spalling (Roh et al., 2008). A fire occurred in the Great Belt tunnel in

Denmark in 1995 also causing severe spalling of concrete tunnel rings (Hertz, 2003).

Steam pressure build-up in the pores of concrete in fire is believed to cause moisture

clog spalling, first proposed by Shorter and Harmathy (1961). Spalling was not

8

considered a major problem until the advent of high strength concrete (HSC) and its

widespread use since 1990’s. HSCs are significantly more vulnerable to spalling in fire.

Sanjayan and Stocks (1993) is the first published research work in international journal

identified this problem. Since HSC has become the dominant construction material,

there has been a renewed interest in spalling research. Highly reputed researchers in

the field of concrete have now cast doubts on the moisture clog spalling hypothesis.

Bazant (1997) hypothesized that spalling results from restrained thermal dilation close

to the heated surface, which leads to compressive stresses parallel to the heated

surface, leading to brittle fractures of concrete. This hypothesis is further developed by

Ulm et al. (1999) - chemoplastic softening model, Stabler and Baker (2000a and

2000b) - coupled thermo-mechanical damage model and Nechnech et al. (2002) -

elasto-plastic damage model.

National Research Council of Canada (NRC), as well as a number of organizations

world-wide, reported that factors affecting fire performance of high strength concrete

are concrete strength, concrete density, load intensity, moisture content, fire intensity,

aggregate type and specimen’s dimension (Kodur, 2000)

Spalling may also result from the thermal incompatibility between the aggregates and

the cement paste, in particular in concrete with silicious aggregates (Kong and

Sanjayan, 2010; Pan et al., 2012). The exact mechanism of spalling is still hotly

disputed.

Experimental results published by Crozier and Sanjayan (2000) showed that areas of

concrete surfaces under compressive stresses are more prone to spalling than the

ones under tensile stresses. This evidence supports the hypothesis of brittle fracture of

concrete under compressive stresses due to restrained thermal dilation. The fact that

HSC is highly brittle also supports this theory. HSC is predominantly used in structural

elements subjected high level of compressive stresses, e.g., the tunnel’s rings and

columns in high-rise buildings.

9

Due to the highly publicized events such as Channel and Great Belt fires, many

infrastructure owners are demanding the spalling issue to be addressed at the design

stage. Currently, it is addressed by providing some sort of fireproofing to concrete

tunnel rings and columns in high-rise buildings. These fireproofing add significant

expense to the construction costs due to materials cost and high manpower needed for

complicated installation procedures. In addition, sprayed type fire proofing on exposed

structural member does not enhance the architectural design of a building.

Spalling occurs in the initial stages of the fire, i.e., within 15 to 30 minutes (Sanjayan

and Stocks, 1993; Crozier and Sanjayan, 2000) – a critical period for fire control and

escape. Hydrocarbon fire in a tunnel can have catastrophic consequences due to the

tunnel collapse caused by concrete spalling, while fire fighters are assisting people to

escape.

Further, the large columns in lower storeys of high-rise buildings constructed during the

last decade are almost always constructed with high strength concrete, which are

susceptible to spalling of concrete in a fire (Sanjayan and Stocks, 1993). Lower storey

columns virtually carry the entire building and destruction of these columns in a fire can

have catastrophic consequences. Spalling results in rapid loss of the surface layers of

the concrete columns exposing the steel reinforcement, which quickly loses strength

when exposed to fire. Ali et al. (2001) reported that 17 out of 18 columns tested for fire

test spalled explosively.

Chan et al. (1999) conducted fire tests using standard’s fire curve on 100 mm cubic

concrete cube made with silica fume. All specimens spalled explosively and Chan et al.

(1999) concluded that moisture content and strength are the two main factors

governing explosive thermal spalling of concrete. Moisture content has a dominant

influence on spalling.

Noumowe et al. (2006) reported that 160 mm diameter cylinders and 100 x 100 x 400

mm prism self-compacted high strength concrete spalled during low heating rate

(0.5oC/min) with maximum 400oC temperature while normal high strength (vibrator

10

compacted) concrete remained un-spalled. Normal high strength concrete recorded

55% residual strength after the heating cycle. However, both self-compacted high

strength and normal high strength concrete spalled explosively when exposed to

standard fire curve.

Hernández-Olivares and Barluenga (2004) also reported similar outcome when 200 x

300 x 50 prism high strength concrete made with silica fume spalled when exposed to

standard fire temperature curve.

Noumowe et al. (2009) reported that 160 × 320 mm specimens of lightweight

aggregate concrete spalled during the heating phase. The explosion took place when

the temperature at the surface of the specimens was between 290 and 430°C and

concluded that combination of high thermal gradient (which induces high thermal

stresses) and low permeability (which induces high vapour pressure) is main concern

in the concrete spalling at high temperature.

Arioz (2007) reported that surface cracks of 70×70×70 mm concrete became visible

when the temperature reached 600°C. The cracks were very pronounced at 800°C and

increased extremely when the temperature increased to 1000°C. The weight of the

concrete specimens reduced significantly as the temperature increased. This reduction

was gradual up to 800°C. A sharp reduction in weight was observed beyond 800°C.

The increments of crack and weight reduction are due to chemical degradation of OPC

cement binder at 800°C temperature.

Most of researches investigated the damage of fire exposure to concrete on small lab

scale specimens and fire exposure are based on Standard Fire (or cellulosic fire)

specified by ISO 834 (International Standard, 1999). The size of these specimens are

not representing the level of concrete deterioration in terms of spalling and other

thermal properties when exposed to fire in actual condition. In addition, risk of spalling

is significantly raised when the rate of temperature rise is rapid (Copier, 1983). In a

hydrocarbon fire according to EN 1991, rate of heating is twice as the standard fire

(room temperature to 1000oC within 10 minutes) (BSI, 2005). Hydrocarbon fire

11

simulates fire event in a tunnel, petrochemical refineries plant, combustible chemical

storage warehouse, etc. Therefore, in this research, the performance of high strength

concrete exposed to fire was conducted in large scale high strength concrete wall

panels and results from large scale specimens were compared with small scale ones.

Fire temperature curve for fire exposure was based on hydrocarbon fire to maximise

the spalling risk. This research also investigates the size effect as discussed above

where there is no consensus as to how the size effects the spalling of concrete.

2.2 Geopolymer Binder in Concrete for Spalling Solution

During the last decade, remarkable achievements have been made through

geosynthesis and geopolymerisation. It is an excellent alternative to Portland cement

binder concrete due to the elastic properties of hardened geopolymer concrete and the

behaviour and strength of reinforced geopolymer concrete structural members are

similar to those observed in the case of Portland cement concrete (Rangan, 2008)

Geopolymers are very-low viscosity inorganic resins, hardened like thermosetting

resins, but have very high strength and fire resistance, and are ceramic-like in their

properties (Davidovits, 1991). An ultra-high temperature tooling material for the

manufacture of advanced composites was made using geopolymers, which performed

better than the ceramic tooling materials (Davidovits and Davidovics, 1991). During the

Grand Prix season 1994 and 1995, Benetton-Renault Formula 1 Sport Car designed a

unique thermal shield made out of carbon/geopolymer composite. It helped Michael

Schumacher to win twice the world championship and offered to his technical team to

become World Champion of car builders during these two years. Since then, most

Formula 1 teams are using geopolymer composite materials (www.geopolymer.org).

Lyon et al. (1997) reported that geopolymer composites is ideally suited for

construction, transportation and infrastructure where fire endurance is part of needed

requirement .

Low cost geopolymer resins can be produced by activation of fly ash. Geopolymer

concrete is produced by combining these resins with coarse and fine aggregates using

the conventional concrete technology methods. Since fly ash is an industrial by-product

from coal power stations (largely wasted by dumping in landfills), the cost of

manufacturing this concrete can be potentially lower than the conventional Portland

cement based concrete.

http://www.geopolymer.org/

12

Research works reported so far in the literature indicate that geopolymers have

superior fire resistance when compared to conventional concretes (Kong et al., 2008;

Kong and Sanjayan, 2008; Pan et al., 2009). The superior fire resistance properties are

attributed to the ceramic-like properties of geopolymers, including the way it looks:

smooth, glassy and shiny (Palomo et al., 1999).

2.3 Geopolymer Binder in Concrete for Environmental Benefits

Production of 1 ton of Portland cement consumes 1½ tons of raw materials and

is responsible for the release of about 0.75 ton of CO2 into the atmosphere. Portland

cement production releases 6.5 million tons of CO2 in Australia. Worldwide,

greenhouse gas emission from the Portland cement production is about 3 billion tons

annually or about 7% of the total greenhouse gas emissions to the earth’s atmosphere.

Further, the production of Portland cement worldwide is increasing 3% annually

(Collins and Sanjayan, 2002; Hardjito et al., 2004). The amount of CO2 emissions due

to concrete using conventional Portland cement is the fourth largest contributor to

global carbon emissions after oil, coal and natural gas. Portland cement was found to

be the primary source of CO2 emissions generated by typical commercially produced

concrete mixes, being responsible for 74% to 81% of total Portland cement concrete

CO2 emissions (Flower and Sanjayan, 2007).

Utilisation of unused industrial by-products such as fly-ash, which otherwise are

dumped in landfills and contribute to land pollution, to make concrete without the use of

any Portland cement will not only reduce the cost of construction materials but also

reduce the greenhouse gas emissions arising from Portland cement manufacture.

Power stations around the world are currently still using coal as fuel to generate

electricity. These power stations produce enormous quantities of fly-ash residues every

year. In 2011, about 13 million tons per annum of fly ash is produced in Australia

(Cement Australia, 2011), in which only small quantity is used for cementitious

applications (e.g. Portland cement/fly ash blended cements, road stabilisation, low

strength fills, asphaltic fillers etc) (Heidrich, 2003). Worldwide, the production of fly ash

is 390 million tons per annum, but its utilisation was less than 14%. In the future, fly ash

production will increase, especially in countries such as China and India (Hardjito et al.,

2005). From China alone, 2.1Gt produced in 2011 which accounting for 58% of world

total cement production (Oss, 2011). Therefore, utilizing the fly ash in geopolymer

13

concrete as fire resistance construction materials in this research will not only have a

huge impact on concrete in fire development, but also provided a better sustainable

environment.

2.4 Geopolymer History and Reaction Mechanisms

Alkali-activated cement researches begin in 1939 with the work of Feret (slags

used for cement) and alkali-slag combinations work by Purdon in 1940 (Roy, 1999).

Glukhovsky (1959) used “soil cement” term for concrete binder which using “soil

silicates” In 1979, Davidovits introduced “geopolymer” terms for alkaline activated

material and the term is generally accepted to-date. Table 2.1 summarises the history

of alkali-activated cement (Roy, 1999; Li et al., 2010).

Table 2.1: Bibliographic history of some important events/articles about alkali-activated cements (Li et al., 2010) Author Year Significance

Feret 1939 Slags used for cement

Purdon 1940 Alkali-slag combination

Glukhovsky 1959 First called “alkaline cements”

Davidovits 1979 “Geopolymer” term

Forss 1983 F-cement (slag-alkali-superplastizer)

Davidovits and Sawyer 1985 Patent of “Pyrament’ cement

Krivenko 1986 DSc thesis, R20-RO-SiO2-H20

Deja and Maloplepsy 1989 Resistance to chlorides shown

Roy and Langton 1989 Ancient concretes analogs

Talling and Brandstetr 1989 Alkali-activated slag

Wu et al. 1990 Activation of slag cement

Roy et al. 1991 Rapid setting alkali-activated cements

Wang and Scivener 1995 Slag and alkali-activated microstructure

Fernandez-Jimenez and Puertas 1997 Kinetic studies of alkali-activated slag cements

Davidovits 1999 Chemistry of geopolymeric system

Palomo 1999 Alkali-activated fly ash – a cement for the future

Palomo and Palacios 2003 Immobilization of hazardous waste

Duxson 2007 Geopolymer technology: the current state of art

Provis and Deventer 2009 Geopolymers: structure, processing, properties

and industrial application

14

Table 2.1 only lists the major works in the chemical aspects of the alkali activated

cements and have missed some contributions by our group. Alkali activated research

commenced in Melbourne in 1995 with the investigation on microstructure and

durability of alkali activated cementitious paste (Bakharev and Patnaikuni, 1997). In

1997, Collins and Sanjayan (1998) used alkali activated cement primarily to find a way

to increase early age strength characteristics of slag concrete . The slag blended

cement with Portland cements are notoriously low in early age strength and therefore

have difficulty being used in precast applications. The group has also demonstrated

that alkali activated cements have superior sulphate resistance (Bakharev et al., 2002)

and acid resistance (Bakharev et al., 2003). Fire resistance properties of geopolymer

using fly ash showed excellent results. Some of the major publications in this area are:

Kong et al. (2008), Kong and Sanjayan (2010) and Pan and Sanjayan (2010). The

mechanical properties of geopolymer show that the properties are comparable to

Portland cement concretes (Pan et al., 2011).

The polymerisation process involves a chemical reaction under highly alkaline

conditions on Al-Si minerals (e.g. fly ash, kaolin, metakaolin), yielding polymeric Si-O-

Al-O bonds, as described by Davidovits (1991): Mn [ – ( Si – O2 ) z – Al – O ] n . wH2O,

where M is the alkaline element, the symbol – indicates the presence of a bond, z is 1,

2 or 3, and n is the degree of polymerisation.

Sialate is an abbreviation for silicon-oxo-aluminate in which the alkali is sodium (Na+),

potassium (K+), lithium (Li+) or calcium (Ca2+). Polysialates are chain and ring polymers

with Si4+ and Al3+ in IV-fold coordination with oxygen and range from amorphous to

semi-crystalline (Davidovits, 1991). Types of polysialates identified in geopolymer are

polysialate (PS), polysialate-siloxo (PSS), polysialate-disiloxo (PSDS) as shown in

Figure 2.1

15

Figure 2.1: Types of polysialates (Davidovits, 2002)

The hardening processes of geopolymer normally occur through polymerisation

process. General mechanism of alkali activation for primarily comprising silica and

reactive alumina as proposed by Glukhovsky in 1950’s consist three main processes

namely destruction-coagulation stage, coagulation-condensation stage and

condensation-crystallisation stage. The first step (destruction-coagulation stage)

consists of a breakdown of the covalent bonds Si–O–Si and Al–O–Si, which happens

when the pH of the alkaline solution rises, so those groups are transformed into a

colloid phase. Then in coagulation-condensation stage, destroyed products accumulate

and interact among them to form a coagulated structure. Finally, condensed structure

generated and crystallized (Li et al., 2010).

More comprehensive review of the current state of knowledge on inorganic polymers

was compiled by Duxson (2007) in which a further enhancement on general

mechanism proposed by Glukhovsky in 1950’s. A conceptual model for inorganic

polymerisation illustration with high level of simplification (excluding raw material

processing such as fine grinding, heat treatment etc.) as shown in Figure 2.2.

16

Figure 2.2: Conceptual model for geopolymerisation (Duxson et al., 2007)

Dissolution of the solid aluminosilicate source by alkaline hydrolysis (consuming water)

produces aluminate and silicate species started by the alkaline solution attack on the

fly ash (Fernández-Jiménez et al., 2005). Figure 2.3 shows the scanning electron

microscopic (SEM) images of fly ash before and after dissolution.

Figure 2.3: SEM pictures (a) original fly ash, (b) fly ash activated with NaOH

(Fernández-Jiménez et al., 2005)

17

Once in solution the species released by dissolution are incorporated into the aqueous

phase, which may already contain silicate present in the activating solution. A complex

mixture of silicate, aluminate and aluminosilicate species is thereby formed.

Supersaturated aluminosilicate solution is created at high rate due to rapid dissolution

of amorphous aluminosilicate at high pH. This supersaturated aluminosilicate solution

resulted in gel formation. The system continues to rearrange and reorganize after the

gel formation resulting in the three-dimensional aluminosilicate network commonly

attributed to geopolymers (Duxson et al., 2007).

2.5 Fly Ash

Fly ash also known as pulverised fuel ash, is a fine grey powder consisting

mostly of spherical glassy particles. It is residue product from burning coal in coal fired

power station. Fly ash is collected from the exhaust gases from the combustion

chambers with electrostatic precipitators or bag houses before the gases are released

to atmosphere (Flyash Australia, 2010; Cement Australia, 2011). Explanation of fly ash

production at a power station is shown in Figure 2.4.

Coal is injected into the furnace and ignited while in suspension. During combustion,

minerals in coal become fluid at high temperature and are then cooled. In a pulverized

coal (PC) fired boiler, the furnace operating temperatures are typically in excess of

1400°C. As the particles are heated, volatile matter is vaporized and combustion

occurs. Minerals undergo thermal decomposition, fusion, disintegration and

agglomeration. The final products of combustion are usually spherical ash particles,

which may subsequently undergo other processes such as coalescence with other

particles or expansion due to internal gas release. The formation of molten ash droplets

marks the highest temperature, and a significant fraction of the volatile forms of

elements will exist in the gas phase. The main formation mechanism for coarse ash

particles (>2 micrometer) is carryover of a proportion of the mineral matter in the feed

coal. A portion of the incombustible material is retained in the furnace as bottom ash.

The rest of the inorganic material leaves in the flue gases as fly ash. Rapid cooling in

the post-combustion zone results in the formation of spherical, amorphous (non-

crystalline) particles (Clarke, 1993; Kutchko and Kim, 2006).

18

Figure 2.4: Fly ash production (Flyash Australia, 2010)

The fly ashes produced in Australian power stations are light to mid-grey in colour and

have the appearance of cement powder. The particle sizes range from less than 1 to

200 micrometer (Heidrich, 2003). According to the ASTM C 618, there are two classes

of fly ash; Class C and Class F. Class C fly ash, contains more than 20% of calcium

oxide (CaO) in its composition, is produced from burning lignite or sub-bituminous coal

and is sometimes called high-calcium fly ash. Meanwhile, class F fly ash is also known

as low calcium fly ash. The Loss on Ignition (LOI) is also an important characteristic for

fly ash, especially when it is used in concrete production. LOI indicates the amount of

unburned carbon present in the fly ash. Table 2.2 summarizes the chemical

requirement of Class F and Class C fly ashes according to ASTM C618 .

Table 2.2: Chemical requirement for Class F and Class C fly ashes according to ASTM 618 (1994) Chemical Properties Class F Class C

Silicon dioxide (SiO2) plus aluminum oxide (Al2O3) plus iron oxide (Fe2O3)

Minimum 70% Minimum 50%

Sulfur trioxide (SO3) Maximum 5% Maximum 5%

Moisture content Maximum 3% Maximum 3%

Loss on ignition (LOI) Maximum 6% Maximum 6%

19

The use of Class F pozzolan containing up to 12.0% loss on ignition may be approved

if either acceptable performance records or laboratory test results are made available.

The Loss on Ignition (LOI) is also an important characteristic for fly ash, especially

when it is used in concrete production. LOI indicates the amount of unburned carbon

present in the fly ash.

In terms of mineral composition, fly ash consists of a glassy matrix with noticeable

crystalline phases of quartz (SiO), magnesioferrite (MgFe2O4), hematite (Fe2O32),

anhydrite (CaSO), lime (CaO), and portlandite (Ca(OH)) (Hanjitsuwan et al., 2011).

In Australia, majority of fly ash produced is categorised as Class F mainly due to the

amount of silica and alumina range between 80 – 85%. Class F fly ash is pozzolanic

and reacts with various cementitous materials. Most of the fly ash being used to

enhance the properties of concrete (as admixture), OPC replacement as concrete

binder, road base binders and asphalt filler (Heidrich, 2003). Average chemical content

of 10 fly ashes in Australia is summarised in Table 2.3.

Most researches in Australia used fly ash from power plants in Western Australia and

Gladstone power plant in Queensland (Hardjito et al., 2004; Hardjito et al., 2005; Kong

and Sanjayan, 2008; Kong and Sanjayan, 2010; Pan et al., 2012; Giasuddin et al.,

2013; Deb et al., 2014; Nematollahi and Sanjayan, 2014)

Table 2.3: Average chemical content for fly ashes in Australia (Heidrich, 2003) Chemical Component Weight (%)

SiO2 61.1

Al2O3 26.2

Fe2O3 4.2

CaO 1.7

MgO 0.9

Na2O 0.8

K2O 1.1

SO3 0.2

Loss on Ignition (LOI) 2.6

In this research, there are fly ashes from Gladstone power plant and Tarong power

plant were initially used as geopolymer aluminosilicate source. Based on workability

20

test and further investigation on physical properties effect on workability, fly ash from

Gladstone power plant was used.

2.6 Alkaline activator

Basically, geopolymer can be produced by any strong alkali solution as the

alkaline activators (Rowles, 2004). Glukhovsky et al. (1980) classified alkaline activator

into six groups according to their chemical composition : (1) Caustic alkalis: MOH; (2)

Non-silicate weak acid salts: M2CO3, M2SO4, M3PO4, MF, etc; (3) Silicates: M2O·nSiO2

(4) Aluminates: M2O.nAl2O3; (5) Aluminosilicates M2O·Al2O3·(2-6)SiO2; and (6) Non-

silicate strong acid salts: M2SO4.

The type of alkaline solution plays an important role in the geopolymerisation process.

An investigation on three alkaline solutions are sodium hydroxide (NaOH), sodium

carbonate (Na2CO3) and sodium silicate solutions concludes that NaOH has the

highest compressive strength followed by sodium silicate solution and higher

concentration of Na2O corresponds to higher compressive strength (Fernández-

Jiménez et al., 2005).

Xu and Van Deventer (2000) reported that the ionic sizes of alkali-metal cation also

affect the dissolution rate when NaOH solution exhibits a higher dissolution rate with

aluminosilicate source materials as the smaller Na+ cation even though Na+ and K+

have the same electric charge. This is due to cation-anion pair interaction becoming

less significant as the cation size increases (Xu and Van Deventer, 2000).

Alkaline activator can be a single material such as sodium hydroxide, sodium silicate,

calcium hydroxide or combination of any silicate and hydroxide (Brough and Atkinson,

2002; Buchwald and Schulz, 2005).

NaOH is used because it is cheap, has low viscosity and is the most widely available

alkaline hydroxide. Also, the hydroxyl ion - in NaOH is an important element in starting

the geopolymerisation process (Provis and Van Deventer, 2009).

21

Sodium silicate (Na2SiO3) is a high viscosity chemical in liquid or powder form. It

influences the geopolymer mixture workability when a highly concentrated one is used

in liquid and a powder form. Na2SiO3 in the geopolymer system increase the final

strength of the paste and bind the material together to produce a dense paste (Jo et al.,

2007).

NaOH plus sodium silicate have been found to produce the fastest setting time and

promotes the collapse of both micropores and mesopores, thus increasing density and

strength (Jiang, 1997). The majority of research has found that activation with sodium

silicate or sodium silicate blended with NaOH has given the best strength (Pimraksa et

al., 2008; Adam et al., 2010; Guo et al., 2010; Nazari et al., 2011; Zhao and Sanjayan,

2011).

Therefore, this research used the combination of sodium silicate and sodium hydroxide

as alkaline activator.

2.7 Geopolymer workability

Workability is often referred as the ease with which a concrete can be

transported, placed and consolidated without any loss of stability or homogeneity. It is

greatly affected by the characteristics of the constituent materials of concrete (Leite et

al., 2013). The workability properties are measured for fresh concrete before it has set

and hardened.

High viscosity of sodium silicate and sodium hydroxide made geopolymer concrete less

workable compared to Portland cement concrete. Since geopolymer is based on

aluminosilicate reaction for producing gel and hardening, no free water were added into

fresh concrete and reduced the workability. Sathonsaowaphak et al. (2009) reported

that main factors affecting fresh geopolymer mortar workability were sodium

silicate/hydroxide ratio and sodium hydroxide concentration. The increase of both

sodium silicate/hydroxide ratio and sodium hydroxide concentration resulted in less

workable mortar mixes due to higher viscosity of sodium silicate and sodium hydroxide.

The setting time of geopolymer paste was found out to increase proportionally with the

increasing of sodium hydroxide concentration because of hardening and setting of the

paste is governed by geopolymerisation process. The geopolymerisation process

22

which usually occurred in a slower rate as compared with C-S-H and C-A-H dependent

cementitious system (Hanjitsuwan et al., 2014).

Even though there are a lot of commercial superplasticizers in the market that are

suitable that can be used in Portland cement concrete to improve the workability, only

polycarboxylate (PC) based superplasticizers were effective in increasing the

workability. This is due to its ether chain in its structure which resulted in steric

repulsion (Nematollahi and Sanjayan, 2014).

Most of researches on geopolymer workability focus on chemical reaction of alumina

and silica in which the basic of geopolymerisation process. However, there is still no

firm conclusion on the ratios of silicate/hydroxide, liquid/fly ash, the amount of water or

the concentration on hydroxide needed to obtained high workability. In this research,

the focus on factors affecting fly ash workability based on physical properties of fly ash

particle such as particle sizes, particle distribution, pore volume etc.

2.8 Geopolymer compressive strength

Compressive strength is the one of most important concrete property from

structural engineering point of view. In geopolymer concrete, there are several factors

affecting its compressive strength. Sukmak et al. (2013) studied the effect of sodium

silicate/sodium hydroxide and alkaline activator solution/fly ash ratios on compressive

strength of geopolymer bricks. With curing regime of 75oC for 48 hours, results showed

that optimum values for sodium silicate/sodium hydroxide and alkaline activator

solution/fly ash ratios are 0.7 and 0.6 respectively.

Ryu et al. (2013) studied on hydroxide concentration, hydroxide/silicate and curing

method and curing period influences on compressive strength of geopolymer mortar.

The results show that the combination of highest concentration of NaOH which is 9 M,

hydroxide/silicate ratio of 1 and 60oC for 24 hours curing regime provides the highest

compressive strength of geopolymer mortar. Further investigation on pore size of

geopolymer paste at 28 days indicates that the number of small pores increases with

higher molarity of the alkaline activator. The reduction of the porosity accompanying

the lowering strength observed in fly ash-based geopolymer can be explained by the

geopolymerisation process, which is the hardening mechanism of geopolymer based

on coal ash rich in Al and Si components.

23

Pimraksa et al. (2011) reported that concentration on NaOH and KOH as alkaline

activator directly influence the strength of geopolymer. Higher concentration on alkaline

activator molarity in which represents higher Na2O/Al2O3 and Na2O/SiO2 ratios

promoted the strength enhancement. The strengths increased with the increases in

curing temperature and time. Optimum curing regime suggested is 75oC for 5 days.

There was a significant increase in strength between Na2O/SiO2 ratio 0.75 and 1.0 but

strength increment was reduced when the ratio reached 1.25. The increase in strength

observed is attributed to an increase in the dissolution process, thus resulting in a

higher reaction rate and in fewer unreacted FA particles (Law et al., 2014).

Gorhan and Kurklu (2014) investigated the influence of sodium hydroxide (NaOH)

solution concentration, curing time and temperature on geopolymer mortars. NaOH

concentrations of 3, 6 and 9 M were used throughout the laboratory work. Based on

compressive strength results, NaOH concentration of 6 M with 85oC curing temperature

and 2 – 5 hours curing time produced the best compressive strength for geopolymer

mortars.

Ahmari and Zhang (2012) studied the feasibility of utilizing copper mine tailings for

production of eco-friendly bricks based on the geopolymerisation technology found out

that geopolymer bricks made from NaOH concentration of 15 M compressive strength

is higher than geopolymer bricks made from 10 M NaOH. The authors also reported

compressive strength increase for brick with higher NaOH/mining tailings ratio. This is

due to higher Na/Al and Na/Si ratio for those bricks which produced much thicker

geopolymer binder gels and bind the unreacted particles.

While many research papers reported on compressive strength increment with higher

hydroxide concentration particularly NaOH, research done by He et al. (2013) shows a

total contrast in compressive strength development. Several reasons such as Si and Al

ions leaching disruption due to high viscosity of NaOH and premature precipitation of

geopolymeric gels because of excessive OH- concentration are postulated for the

24

decrease of geopolymer composites compressive strength in higher NaOH

concentration.

Apart from hydroxide concentration effect on compressive strength of geopolymers,

ratio of silicate to hydroxide influence was investigated by some researches. Ridtirud et

al. (2011) reported that geopolymer mortars with sodium silicate/sodium hydroxide

ratio of 1.5 provide the highest compressive strength (45 MPa) compared to mortars

made with lower sodium silicate/sodium hydroxide ratios. The increasing trend is

mainly attributed to the increasing Na content in the mixture where Na+ ion plays a

critical role in the formation of geopolymer by acting as a charge balancing ions.

However, the authors also reported that the compressive strength of mortars with ratio

of sodium silicate/sodium hydroxide of 3 decreased as excess sodium silicate

hampered the evaporation of water and also disrupt the formation of three dimensional

networks of aluminosilicate geopolymers.

Some other researches highlighted the need of proper adjustment of silicate/hydroxide

ratio to improve the compressive strength of geopolymer concrete/mortar/brick

(Sathonsaowaphak et al., 2009; Guo et al., 2010; Nazari et al., 2011).

Kromjlenovic et al., (2010) concluded that the physical characteristic of fly ash, the

fineness, or particle size distribution was the key parameter in compressive strength of

geopolymer mortar. The compressive strength of geopolymers predominantly

depended on the content of FA fine particles (smaller than 43 um). In all cases, the

highest mortar compressive strength showed the FA which had the highest amount of

fine particles.

In this research, further investigation on various silicate/hydroxide ratios, liquid/fly ash

ratio, hydroxide concentration and fly ash source was carried out in order to achieve

high strength geopolymer concrete more than 50 MPa. This is because most of

literatures reported compressive strength of geopolymer concrete is less than 50 MPa.

The high strength classification of geopolymer concrete is based on high strength of

Portland cement concrete classification in accordance to IS EN 206 (European

Standard, 2013).

25

2.9 Geopolymer under Environmental Exposure

Concrete structures were designed to be able to withstand aggressive

environmental exposure. Fly ash-based geopolymer concrete is better than Portland

cement concrete in many aspects such as compressive strength, exposure to

aggressive environment, workability and exposure to high temperature (Al Bakri et al.,

2011). Giasuddin et al. (2013) discovered that geopolymer binder has the potential to

replace traditional oil well cements in CO2 geo-sequestration in saline aquifier. Results

show that the saline water cured geopolymer exhibited higher compressive strength

compared to normal water cured geopolymer.

Main phases of blended ash geopolymer concrete such as sodalite, gmelite and

natrolite were still intact even after 18 months of acid sulphuric exposure. This is due to

no ettringite detected as the Al ions participated in the formation of N-A-S-H gels, thus

making the available Al ions insufficient to form ettringite as in Portland cement binder

system (Ariffin et al., 2013).

There are compressive strength losses ranging from 53.3% to 78.4% when geopolymer

bricks are exposed to nitric acid (pH 4). The losses are caused by incomplete

geopolymerisation, modification of chemical composition of geopolymer gels and high

degree of unreacted alkali specimen (Ahmari and Zhang, 2013).

Other than acidic exposure, water (wetting and drying cycles) can introduce

deterioration problems in geopolymer. Unlike OPC, geopolymer reaction in producing

aluminosilicate gel and geopolymerisation hardening process do not involve any

hydration process. Therefore, water play no roles in hardened geopolymer structure

(Davidovits, 1989). Further investigation on strength reduction due to saturation

process was carried out in this research.

26

2.10 Geopolymer fire test

As mentioned in Section 2.2, geopolymer concrete has better fire resistance

compared with Portland cement concrete due to its ceramic-like properties (Davidovits,

1991). Cheng and Chiu (2003) investigated the fire resistance of geopolymer panel

with potassium hydroxide as alkaline activator and granulated blast furnace slag as

filler. Fire resistance test were conducted by exposing 10 mm thick geopolymer panel

to 1100oC flame, and measured the reverse-side temperatures. From the test,

geopolymer panel reverse side temperature is less than 350oC after 35 minutes fire

exposure.

Bakharev (2006) observed the thermal stability of the geopolymer materials prepared

with sodium containing activators was rather low and significant changes in the

microstructure occurred during curing process. However, geopolymer concrete

prepared using Gladstone fly ash and silicate-containing activator had better strength

than geopolymers prepared using potassium-containing activators. However, after fire

test (1000oC), geopolymer with sodium-silicate activator recorded high strength loss

while potassium-silicate activator had improved strength. This is mainly due to

presence of significant amount of crystalline phases after firing at 800°C.

Kong and Sanjayan (2008) investigated damage behaviour of geopolymer composite

exposed to elevated temperature. After 800oC temperature exposure, geopolymer gain

53% strength while geopolymer/aggregate composite strength decrease by 63%. This

is because the thermal incompatibility between geoppolymer binder and aggregate and

it is proved by dilametry measurement of geopolymer and aggregate (Kong and

Sanjayan, 2010).

The strength losses of geopolymer mortar decrease with increasing ductility, even

there are gain of strength in geopolymer with high levels of ductility. This correlation is

attributed to the fact that mortars with high ductility have high capacity to accommodate

thermal incompatibilities. It is believed that the two opposing processes occur in

mortars: (1) further geopolymerisation and/or sintering at elevated temperatures

leading to strength gain; (2) the damage to the mortar because of thermal

27

incompatibility arising from non-uniform temperature distribution. The strength gain or

loss occurs depending on the dominant process (Pan et al., 2009) .

Different pore structures of geopolymer concrete and Portland cement concrete

attributes to situation where no spalling was found in geopolymer concretes, whereas

the companion Portland cement concrete exhibited spalling. The sorptivity test found

that geopolymer concrete had a significantly higher sorption, therefore more connected

pores, than Portland cement concrete when compared at the same strength level

(Zhao and Sanjayan, 2011).

As mentioned in Section 2.1, testing large specimen in fire is essential in order to

enhance the knowledge of geopolymer concrete behaviour and thermal properties. It

provides a better representation of the actual condition. In addition, high strength of

geopolymer concrete and hydrocarbon fire exposure used in this research resulted

discovering the maximum potential of geopolymer concrete as fire resisting

construction material.

As reported by Kong and Sanjayan (2010) one of the main deterioration mechanisms in

fire is the thermal incompatibility between the geopolymer paste and aggregate. To

overcome this issue, concrete without aggregates was tested in this research. Aerated

geopolymer wall panels without coarse aggregates were fire tested and observations

were made.

28

References

Adam, A., Molyneaux, T., Patnaikuni, I. and Law, D. (2010). "Strength, sorptivity and

carbonation of geopolymer concrete." Challenges, Opportunities and Solutions

in Structural Engineering and Construction: 563-568.

Ahmari, S. and Zhang, L. (2012). "Production of eco-friendly bricks from copper mine

tailings through geopolymerization." Construction and building materials 29:

323-331.

Ahmari, S. and Zhang, L. (2013). "Durability and leaching behavior of mine tailings-

based geopolymer bricks." Construction and Building Materials 44: 743-750.

Al Bakri, M. M., Mohammed, H., Kamarudin, H., Niza, I. K. and Zarina, Y. (2011).

"Review on fly ash-based geopolymer concrete without Portland Cement."

Journal of engineering and technology research 3(1): 1-4.

Ali, F., O'Connor, D. and Abu-Tair, A. (2001). "Explosive spalling of high-strength

concrete columns in fire." Magazine of Concrete Research 53(3): 197-204.

Ariffin, M., Bhutta, M., Hussin, M., Mohd Tahir, M. and Aziah, N. (2013). "Sulfuric acid

resistance of blended ash geopolymer concrete." Construction and Building

Materials 43: 80-86.

Arioz, O. (2007). "Effects of elevated temperatures on properties of concrete." Fire

Safety Journal 42(8): 516-522.

ASTM (1994). ASTM C618: Standard Specification for Fly Ash and Raw or Calcined

Natural Pozzolan For Use as Mineral Admixture in Portland Cement Concrete.

West Conshohocken, Pennsylvania, American Society for Testing and

Materials. ASTM C618-92a.

Bakharev, T. (2006). "Thermal behaviour of geopolymers prepared using class F fly

ash and elevated temperature curing." Cement and Concrete Research 36(6):

1134-1147.

Bakharev, T. and Patnaikuni, I. (1997). Microstructure and durability of alkali activated

cementitious pastes. Proceedings of the fifth international conference on

structural failure, durability and retrofitting. Singapore Concrete Institute,

Singapore.

29

Bakharev, T., Sanjayan, J. G. and Cheng, Y. B. (2003). "Resistance of alkali activated

slag concrete to acid attack." Cement and Concrete Research 33(10): 1607-

1611.

Bakharev, T., Sanjayan, J. G. and Y.B.Cheng (2002). "Sulphate Attack on Alkali

Activated Slag Concrete”, Cement and Concrete Research." Cement and

Concrete Research 32(2): 211-216.

Bazant, Z. P. (1997). Analysis of pore pressure, thermal stress and fracture in rapidly

heated concrete. Proc., Int. Workshop on Fire Performance of High-Strength

Concrete.

Brough, A. R. and Atkinson, A. (2002). "Sodium silicate-based, alkali-activated slag

mortars: Part I. Strength, hydration and microstructure." Cement and Concrete

Research 32(6): 865-879.

BSI (2005). BS EN 1991-1-2: Actions on Structures. Part 1 -2: General Actions -

Actions on structures exposed to fire. Brussels, European Committee for

Standardization.

Buchwald, A. and Schulz, M. (2005). "Alkali-activated binders by use of industrial by-

products." Cement and Concrete Research 35(5): 968-973.

Cement Australia (2011). Sustainable Development – Fly ash.

Chan, S. Y. N., Peng, G.-F. and Anson, M. (1999). "Fire behavior of high-performance

concrete made with silica fume at various moisture contents." ACI Materials

Journal 96(3).

Cheng, T. W. and Chiu, J. P. (2003). "Fire-resistant geopolymer produced by

granulated blast furnace slag." Minerals Engineering 16(3): 205-210.

Clarke, L. B. (1993). "The fate of trace elements during coal combustion and

gasification: an overview." Fuel 72(6): 731-736.

Collins, F. and Sanjayan, J. G. (1998). "Early Age Strength and Workability of Slag

Pastes activated by NaOH and Na2CO3." Cement and Concrete Research

28(5): 655-664.

Collins, F. and Sanjayan, J. G. (2002). The challenge of the cement industry towards

the reduction of greenhouse emissions. IABSE Symposium Report.

30

Copier, W. (1983). "The spalling of normal weight and lightweight concrete exposed to

fire." ACI Special Publication 80.

Crozier, D. A. and Sanjayan, J. G. (2000). "Tests of load-bearing slender reinforced

concrete walls in fire." ACI Structural Journal 97(2): 243-251.

Davidovits, J. (1989). "Geopolymers and geopolymeric materials." Journal of Thermal

Analysis 35(2): 429-441.



Davidovits, J. (2002). 30 years of successes and failures in geopolymer applications.

Market trends and potential breakthroughs. Keynote Conference on

Geopolymer Conference.




Deb, P. S., Nath, P. and Sarker, P. K. (2014). "The effects of ground granulated blast-

furnace slag blending with fly ash and activator content on the workability and

strength properties of geopolymer concrete cured at ambient temperature."

Materials & Design 62(0): 32-39.

Duxson, P., Fernández-Jiménez, A., Provis, J. L., Lukey, G. C., Palomo, A. and Van

Deventer, J. S. J. (2007). "Geopolymer technology: The current state of the art."

Journal of Materials Science 42(9): 2917-2933.

European Standard (2013). IS EN 206:2013 Concrete - Specification, performance,

production, and conformity. Brussel, Belgium, European Committee for

Standardization

Fernández-Jiménez, A., Palomo, A. and Criado, M. (2005). "Microstructure

development of alkali-activated fly ash cement: a descriptive model." Cement

and Concrete Research 35(6): 1204-1209.

Flower, D. J. M. and Sanjayan, J. G. (2007). "Green house gas emissions due to

concrete manufacture." International Journal of Life Cycle Assessment 12(5):

282-288.

31

Flyash Australia (2010). "http://www.flyashaustralia.com.au/whatisflyash.aspx.

http://www.flyashaustralia.com.au/whatisflyash.aspx (22/10/2014).

Giasuddin, H. M., Sanjayan, J. G. and Ranjith, P. (2013). "Strength of geopolymer

cured in saline water in ambient conditions." Fuel 107: 34-39.

Glukhovsky, V. (1959). "Soil silicates." Gostroiizdat Publish. Kiev, USSR.

Glukhovsky, V., Rostovskaja, G. and Rumyna, G. (1980). High strength slag-alkaline

cements. Proceedings of the 7th international congress on the chemistry of

cement, Paris.

Görhan, G. and Kürklü, G. (2014). "The influence of the NaOH solution on the

properties of the fly ash-based geopolymer mortar cured at different

temperatures." Composites Part B: Engineering 58: 371-377.

Guo, X., Shi, H. and Dick, W. A. (2010). "Compressive strength and microstructural

characteristics of class C fly ash geopolymer." Cement and Concrete

Composites 32(2): 142-147.

Hanjitsuwan, S., Chindaprasirt, P. and Pimraksa, K. (2011). "Electrical conductivity and

dielectric property of fly ash geopolymer pastes." International Journal of

Minerals, Metallurgy, and Materials 18(1): 94-99.

Hanjitsuwan, S., Hunpratub, S., Thongbai, P., Maensiri, S., Sata, V. and Chindaprasirt,

P. (2014). "Effects of NaOH concentrations on physical and electrical properties

of high calcium fly ash geopolymer paste." Cement and Concrete Composites

45: 9-14.

Hardjito, D., Sumajouw, D. M., Wallah, S. and Rangan, B. (2005). "Fly ash-based

geopolymer concrete." Australian Journal of Structural Engineering 6(1): 77.

Hardjito, D., Wallah, S. E., Sumajouw, D. M. J. and Rangan, B. V. (2004). "On the

development of fly ash-based geopolymer concrete." ACI Materials Journal

101(6): 467-472.

He, J., Jie, Y., Zhang, J., Yu, Y. and Zhang, G. (2013). "Synthesis and characterization

of red mud and rice husk ash-based geopolymer composites." Cement and

Concrete Composites 37: 108-118.

http://www.flyashaustralia.com.au/whatisflyash.aspx

http://www.flyashaustralia.com.au/whatisflyash.aspx

32

Heidrich, C. (2003). Ash Utilisation-An Australian Perspective. Proceedings of the 2003

international ash utilization symposium, Centre for Applied Energy Research,

University of Kentucky.

Hernández-Olivares, F. and Barluenga, G. (2004). "Fire performance of recycled

rubber-filled high-strength concrete." Cement and Concrete Research 34(1):

109-117.

Hertz, K. D. (2003). "Limits of spalling of fire-exposed concrete." Fire Safety Journal

38(2): 103-116.

International Standard, O. (1999). ISO 834-1999: Fire Resistance Tests: Elements of

building construction. Part 1: General Requirements. Geneva, Switzerland,

International Organization for Standardization. ISO 834-1999.

Jiang, W. (1997). Alkali-activated cementitious materials: Mechanisms, microstructure

and properties. 9817499. Ann Arbor.

Jo, B.-W., Park, S.-K. and Park, M.-S. (2007) "Strength and hardening characteristics

of activated fly ash mortars." Magazine of Concrete Research 59.

Kodur, V. (2000). Spalling in high strength concrete exposed to fire—concerns, causes,

critical parameters and cures. Proceedings, ASCE Structures Congress,

Philadelphia, PA.

Komljenović, M., Baščarević, Z. and Bradić, V. (2010). "Mechanical and microstructural

properties of alkali-activated fly ash geopolymers." Journal of hazardous

materials 181(1): 35-42.

Kong, D. L., Sanjayan, J. G. and Sagoe-Crentsil, K. (2008). "Factors affecting the

performance of metakaolin geopolymers exposed to elevated temperatures."




Composites 30(10): 986-991.

Kong, D. L. Y. and Sanjayan, J. G. (2010). "Effect of elevated temperatures on

geopolymer paste, mortar and concrete." Cement and Concrete Research

40(2): 334-339.

33

Kutchko, B. G. and Kim, A. G. (2006). "Fly ash characterization by SEM–EDS." Fuel

85(17–18): 2537-2544.

Law, D., Adam, A., Molyneaux, T., Patnaikuni, I. and Wardhono, A. (2014). "Long term

durability properties of class F fly ash geopolymer concrete." Materials and

Structures: 1-11.

Leite, M., Figueire do Filho, J. and Lima, P. (2013). "Workability study of concretes

made with recycled mortar aggregate." Materials and Structures 46(10): 1765-

1778.

Li, C., Sun, H. and Li, L. (2010). "A review: The comparison between alkali-activated

slag (Si+ Ca) and metakaolin (Si+ Al) cements." Cement and Concrete

Research 40(9): 1341-1349.

Lyon, R. E., Balaguru, P., Foden, A., Sorathia, U., Davidovits, J. and Davidovics, M.

(1997). "Fire-resistant aluminosilicate composites." Fire and Materials 21(2): 67-

73.

Nazari, A., Bagheri, A. and Riahi, S. (2011). "Properties of geopolymer with seeded fly

ash and rice husk bark ash." Materials Science and Engineering: A 528(24):

7395-7401.

Nechnech, W., Meftah, F. and Reynouard, J. M. (2002). "An elasto-plastic damage

model for plain concrete subjected to high temperatures." Engineering

Structures 24(5): 597-611.

Nematollahi, B. and Sanjayan, J. (2014). "Effect of different superplasticizers and

activator combinations on workability and strength of fly ash based

geopolymer." Materials & Design 57: 667-672.

Noumowe, A., Carré, H., Daoud, A. and Toutanji, H. (2006). "High-strength self-

compacting concrete exposed to fire test." Journal of materials in civil

engineering 18(6): 754-758.

Noumowe, A. N., Siddique, R. and Debicki, G. (2009). "Permeability of high-

performance concrete subjected to elevated temperature (600 °C)."

Construction and Building Materials 23(5): 1855-1861.

34

Oss, H. G. v. (2011). "Cement Statistics and Information.

http://minerals.usgs.gov/minerals/pubs/commodity/cement/ (21 October 2014).

Palomo, A., Grutzeck, M. W. and Blanco, M. T. (1999). "Alkali-activated fly ashes: A

cement for the future." Cement and Concrete Research 29(8): 1323-1329.

Pan, Z. and Sanjayan, J. G. (2010). "Stress-strain behaviour and abrupt loss of

stiffness of geopolymer at elevated temperatures." Cement and Concrete

Composites 32(9): 657-664.

Pan, Z., Sanjayan, J. G. and Kong, D. L. Y. (2012). "Effect of aggregate size on

spalling of geopolymer and Portland cement concretes subjected to elevated

temperatures." Construction & Building Materials 36: 365-372.




Pan, Z., Sanjayan, J. G. and Rangan, B. V. (2011). Fracture Properties of geopolymer

paste and concrete. Magazine of Concrete Research 63:(10): 763 – 771.

Pimraksa, K., Chareerat, T., Chindaprasirt, P., Mishima, N. and Hatanaka, S. (2008).

Composition and microstructure of fly ash geopolymer containing metakaolin.

Excellence in Concrete Construction through Innovation: Proceedings of the

conference held at the Kingston University, United Kingdom, 9-10 September

2008.

Pimraksa, K., Chindaprasirt, P., Rungchet, A., Sagoe-Crentsil, K. and Sato, T. (2011).

"Lightweight geopolymer made of highly porous siliceous materials with various

Na2O/Al2O3 and SiO2/Al2O3 ratios." Materials Science and Engineering: A

528(21): 6616-6623.

Provis, J. L. and Van Deventer, J. S. J. (2009). Geopolymers : structure, processing,

properties and industrial applications. J. L. Provis and J. S. J. Van Deventer.

Boca Raton, Fla. : CRC

Oxford : Woodhead, Boca Raton, Fla. : Oxford.

Rangan, B. V. (2008). "Fly ash-based geopolymer concrete." Your Building

Administrator 2.

http://minerals.usgs.gov/minerals/pubs/commodity/cement/

35

Ridtirud, C., Chindaprasirt, P. and Pimraksa, K. (2011). "Factors affecting the shrinkage

of fly ash geopolymers." International Journal of Minerals, Metallurgy, and

Materials 18(1): 100-104.

Roh, J. S., Yang, S. S., Ryou, H. S., Yoon, M. O. and Jeong, Y. T. (2008). "An

experimental study on the effect of ventilation velocity on burning rate in tunnel

fires—heptane pool fire case." Building and Environment 43(7): 1225-1231.

Rowles, M. R. (2004). The structural nature of aluminosilicate inorganic polymers: a

macro to nanoscale study. Curtin University of Technology, Department of

Applied Physics.: 248.

Roy, D. M. (1999). "Alkali-activated cements Opportunities and challenges." Cement

and Concrete Research 29(2): 249-254.

Ryu, G. S., Lee, Y. B., Koh, K. T. and Chung, Y. S. (2013). "The mechanical properties

of fly ash-based geopolymer concrete with alkaline activators." Construction

and Building Materials 47(0): 409-418.



Sathonsaowaphak, A., Chindaprasirt, P. and Pimraksa, K. (2009). "Workability and

strength of lignite bottom ash geopolymer mortar." Journal of Hazardous

Materials 168(1): 44-50.

Shorter, G. and Harmathy, T. (1961). "Discussion on the fire resistance of prestressed

concrete beams." Proceedings of the Institution of Civil Engineers 20(2): 313-

315.

Stabler, J. and Baker, G. (2000). "Fractional step methods for thermo-mechanical

damage analyses at transient elevated temperatures." International Journal for

Numerical Methods in Engineering 48(5): 761-785.

Stabler, J. and Baker, G. (2000). "On the form of free energy and specific heat in

coupled thermo-elasticity with isotropic damage." International Journal of Solids

and Structures 37(34): 4691-4713.

Sukmak, P., Horpibulsuk, S. and Shen, S.-L. (2013). "Strength development in clay–fly

ash geopolymer." Construction and Building Materials 40: 566-574.

36

Ulm, F. J., Acker, P. and Lévy, M. (1999). "The "Chunnel" fire. II: Analysis of concrete

damage." Journal of Engineering Mechanics 125(3): 283-289.

Ulm, F. J., Coussy, O. and Bažant, Z. P. (1999). "The "chunnel" fire. I: Chemoplastic

softening in rapidly heated concrete." Journal of Engineering Mechanics 125(3):

272-281.

Xu, H. and Van Deventer, J. S. J. (2000). "The geopolymerisation of alumino-silicate

minerals." International Journal of Mineral Processing 59(3): 247-266.

Zhao, R. and Sanjayan, J. (2011). "Geopolymer and Portland cement concretes in

simulated fire." Magazine of concrete research 63(3): 163-173.

37

CHAPTER 3

Hydrocarbon Fire testing of full size Portland Cement High

Strength Concrete Wall Panels

3.1 Introduction

Concrete structures were designed to withstand various types of environmental

conditions from mild conditions to very severe conditions. Fire represents one of the

most severe environmental conditions to which concrete structures may be subjected

to especially in close conduct structure like tunnel (Kim et al., 2010; Khaliq and Kodur,

2011).

Spalling is one of the major risk in concrete when exposed to fire especially high

strength concrete structures (Sanjayan and Stocks, 1993). However, spalling of the

concrete at the soffit of the slab exposed the bottom bar and the flexural capacity of the

slab was drastically reduced (Hertz, 2003; Heo et al., 2010). Several researchers have

indicated that the explosive spalling of concrete especially in high strength concrete

should be considered as a failure due to its high velocity of spalling debris (Phan and

Carino, 1998; Yang and Peng, 2011) as well as loss of concrete section. There are

number of reasons why concrete spall in fire. Vapour pressure is one of the reasons

underlined by number of researches. Vapour pressure in highly impermeable concrete

38

with low diffusivity such as in high strength concrete may exceed the tensile stress and

cause explosive spalling (Anderberg, 1997; Phan and Carino, 1998; Yang and Peng,

2011). Thermal stress at high temperature due to high thermal gradient and low

thermal diffusivity is an issue closely related to concrete spalling as well (Harmathy,

1993). Khalifa (2000) described thermo-mechanical is one of a contributing factor in

concrete spalling as well. Thermal dilation gradient due to difference thermal expansion

between cement paste and aggregate can also result in initial fracture and energy

release from the fracture subsequently causing spalled concrete in fire (Bazant and

Kazemi, 1990; Pan et al., 2010; Kanéma et al., 2011). The results in this chapter

represent the performance of high strength concrete exposed to hydrocarbon fire

exposure. The amount of spalling, temperature of wall panel at various depth, thermal

diffusivity and pore pressure corresponding to thermal diffusivity were reported in this

chapter.

3.2 Experimental Programme

3.2.1 Material

Granite and basalt aggregates were used as aggregates for the wall

specimens. The coarse aggregate sizes varied and were 7 mm, 14 mm and 20 mm. In

all concrete mixes, the water to binder ratio was fixed at 0.3. The binder included 25%

to 30% of slag by weight. Superplasticizers were introduced to increase the workability

with targeted slump of 150 mm. Design strength for all concrete mixes was 80 MPa. All

concrete were tested for compressive strength at the age of 28 days. Table 3.1 shows

concrete mixes details. All specimens were cast at a precasting yard (Hollowcore Pty

Ltd). Surface moisture content was measured before the fire test. All specimens were

reinforced with mild strength steel 6 mm diameter bar and 100 mm spacing mesh.

Reinforcement bars were located at the middle of the concrete wall panel thickness.

Figure 3.1 shows the reinforcement details for the panels.

R6-100 cc

100 mm200 mm

3.38 m Figure 3.1: Reinforcement details of concrete wall panel (top view)

39

Table 3.1: Concrete (per cubic meter) mix summary Aggregate

Type

(Max Size)

Aggregate Cement

(kg)

Water

(kg)

Super-plasticizer

(Litre) Coarse

(kg)

Fine

(kg)

Basalt

(7 mm) 1100 577

550

(25% Slag) 150

7.5

(Viscoscreet)

Basalt

(14 mm) 1100 577

550

(25% Slag) 150

7.5

(Viscoscreet)

Basalt

(20 mm) 1200 620

600

(33% Slag) 180

9.6

(Super Viscoscreet)

Granite

(7 mm) 870 759

633

(25% Slag) 190

7 – 17

(ADVA 142)

Granite

(14 mm) 901 784

600

(25% Slag) 180

7 – 17

(Super Viscoscreet)

Granite

(20 mm) 956 785

567

(25% Slag) 170

7 – 17

(Super Viscoscreet)

3.2.2 Concrete Properties

All specimens’ sizes for this research are 3.38 m x 3.36 m x 0.2 m (thickness).

Figure 3.2 illustrates the dimension of concrete wall panel and area which the panel

exposed to fire.

3.36 m

3.0 m

3.3

9 m

3.0

m

Fire Exposed Area

0.2 m

Figure 3.2: Concrete wall panel dimension and fire exposed area

40

3.2.3 Test Setup

All specimens were cast at a precast factory and aged for at least at 28 days

before tests. Figure 3.3 shows the image of concrete casting at precasting yard

(Hollowcore Pty. Ltd.). Image of casted concrete wall panel is shown in Figure 3.4.

Compressive test were performed at the age of 28 days for all specimens. Fire tests

were carried out during the age between 44 days and 115 days for the investigation of

concrete age effect on spalling. Details of concrete casting and fire tests are

summarised in Table 3.2.

Figure 3.3: Casting of Ordinary Portland Cement High Strength Concrete

Figure 3.4: Casted Full-scale (3.36 m x 3.38 m) Portland Cement High Strength

Concrete

41

Table 3.2: Concrete Casting and Fire Test Details Aggregate Type

(Max. Size)

Casting Date

(dd/mm/yyyy)

Fire Test

Date

Age During Fire Test

(days)

Basalt

(7 mm) 3/8/2010 16/9/2010 44

Basalt

(14 mm) 5/8/2010 11/10/2010 67

Basalt

(20 mm) 20/7/2010 3/9/2010 67

Granite

(7 mm) 13/8/2010 8/11/2010 87

Granite

(14 mm) 12/8/2010 3/11/2010 83

Granite

(20 mm) 17/8/2010 10/12/10 115

3.2.4 Compressive strength and moisture content

Compressive strength test was carried out on 100 mm diameter x 200 mm

height concrete cylinder. The concrete cylinders were tested at the age of 28 days in

accordance to AS 1012.9-1999 (Standard Australia, 1999). Moisture content of

concrete wall panel was measured 1 hour before the fire test commenced. Moisture

content was measured by using TramexTM CME 4 moisture meter. Figure 3.5 shows

the moisture meter used to measure moisture content of concrete wall panel.

42

Figure 3.5: TramexTM CME 4 moisture meter

3.2.5 Thermal Data Collection

Although the standard fire in accordance to ISO 834:1999 (International

Standard, 1999) curve has been in use for many years, it soon became apparent that

the burning rates for certain materials e.g. fire in a tunnel, petrol gas and chemicals

were well in excess of the burning rate suggested in standard fire. As such, there was a

need for an alternative exposure for the purpose of carrying out tests on structures and

materials, thus the hydrocarbon curve was developed. The temperature for

hydrocarbon fire reaches up to 1000oC within 10 minutes as compared to 90 minutes

for standard fire. Figure 3.6 shows comparison of the hydrocarbon fire temperature

versus time curve and standard fire temperature versus time.

Figure 3.6: Hydrocarbon and Standard Fire Temperature versus Time Profile

Hydrocarbon fire temperature versus time in accordance to EN 1991-1-2 was the

temperature versus time curve used for exposing the specimens. Equation 3.1 is the

hydrocarbon fire temperature equation (BSI, 2005).

𝑇 = 1080(1 − 0.325𝑒−0.167𝑡 − 0.675𝑒−2.5𝑡) + 20 Equation 3.1

Where T = Temperature (oC)

t = Time (minutes)

0

400

800

1200

0 30 60 90 120 150 180

Tem

pera

ture

(o C)

Time (min)

Hydrocarbon and Standard Fire Temperature vs Time Curves

Hydrocarbon Fire (EN 1991-1-2)

Standard Fire (ISO 834-1999)

43

The fire tests conducted for minimum of 120 minutes. The temperature of the furnace

(near exposed surface) was measured by the furnace’s temperature sensor. Figure 3.7

illustrates the position of wall panel attachment to the furnace. Picture of attached wall

panel ready for the fire test is shown in Figure 3.8.

Furnace Wall Panel

Figure 3.7: Wall Panel and Furnace Setup

Figure 3.8: Hydrocarbon Fire Furnace

The temperature of wall panels specimens at the depth of 25 mm, 50 mm, 75 mm, 100

mm, 150 mm and unexposed surface were also measured and recorded. The

44

illustration of thermal couple’s locations and image of thermal couple set up are shown

in Figure 3.9 and Figure 3.10.

50 mm

25 mm

25 mm

25 mm

25 mm

200 mm

Thermal Couples

50 mmFire Exposed Surface

Figure 3.9: Illustration of Thermal Couple’s Location

Figure 3.10: Thermal Couple Set Up

3.2.6 Spalling analysis

The weight of the specimens was measured before and after the specimens

was exposed to fire. The weight difference is considered as the weight of spalled

45

concrete. The loss of concrete wall panel moisture content was also considered as part

of spalled concrete weight.

3.2.7 Thermal diffusivity analysis

Thermal diffusivity is a thermal conductivity divided by the volumetric heat

capacity measured in mm2/s. It is a material specific thermal property which describes

how quickly a material reacts to a change in temperature. It is also a measurement of

thermal inertia of solids. In order to predict cooling processes or to simulate

temperature fields, the thermal diffusivity values are essential. The higher thermal

diffusivity value will increase the heat propagation rate into the material and can be

classified as good heat conductor material. The ideal insulator materials with large heat

capacity will have a very low thermal diffusivity (Kaviany, 2011).

Thermal diffusivity for each specimen under fire test was calculated by using the Finite

Difference back calculation method at the depth of 50 mm measured from exposed

surface. Based on the temperature data from the fire test, thermal diffusivity value were

calculated on trial and error basis based on the general equation (Welty, 1974):

Equation 3.2

where K is thermal diffusivity

Central difference equation was used for second spatial derivative,

𝜕2𝑇

𝜕𝑥2 =𝑇𝑖+1

𝑛 −2𝑇𝑖𝑛+ 𝑇𝑖−1

𝑛

ℎ2 Equation 3.3

Where n = temperature at calculated time step

i = temperature at calculated thermal couple location (∆x = 50 mm)

h = total thickness for wall specimen

Forward equation was used for first time derivative,

tT

xT

12

2

46

𝜕𝑇

𝜕𝑡=

𝑇𝑖𝑛+1− 𝑇𝑖

𝑛

∆𝑡 Equation 3.4

where, ∆𝑡 = time interval

Therefore, the final equation for thermal diffusivity calculation was:

𝑇𝑖𝑛+1 = 𝐾∆𝑡 ⌊

𝑇𝑖+1𝑛 −2𝑇𝑖

𝑛+ 𝑇𝑖−1𝑛

ℎ2⌋ + 𝑇𝑖

𝑛 Equation 3.5

Trial value for thermal diffusivity, K substituted in Equation 3.5 and the correct value of

thermal diffusivity is determined when the temperature, 𝑇𝑖𝑛+1 calculated from thermal

diffusivity equation equates to the temperature from experimental data.

3.3 Result and Discussion

3.3.1 Compressive Strength and Moisture Content

Compressive strength of the concrete at the age of 28 days ranged between

69.4 MPa and 75.1 MPa. Therefore, the concrete can be classified as high strength

concrete. Average moisture content for concrete wall panels was 5%. Table 3.3

summarises the compressive strength of concrete and moisture content of wall panels.

Table 3.3: Compressive strength and moisture content summary

Aggregate Type

(Max Size)

Moisture Content

(%)

Compressive Strength at

28 days (MPa)

Basalt

(7 mm) 5.2 70.0

Basalt

(14 mm) 5.2 69.4

Basalt

(20 mm) 5.1 70.0

Granite

(7 mm) 4.8 70.7

Granite

(14 mm) 5.1 73.8

Granite

(20 mm) 4.8 75.1

47

3.3.2 Spalling

From the observation during fire exposure, concrete spalling occurred within the

first 30 minutes of fire exposure. The spalling happened in explosive manner. The

spalling of concrete weight was inclusive with the loss of moisture content from the fire

test. Specimen with 7 mm basalt aggregate exhibited the most severe spalling with

32.5% weight loss including 5.2% loss of moisture content. Figure 3.12 shows the

percentage of concrete spalling after the specimens were exposed to fire. Wall panels

with 14 mm and 20 mm basalt aggregates spalled 7.5% and 12.7% respectively.

Granite aggregate wall panels’ spalling percentages are 22.9%, 23.5 and 13.5% for 7

mm, 14 mm and 20 mm aggregate size respectively. From these results, it is an

indication that aggregate size is not exhibiting a clear trend to concrete spalling. It is

also observed that there is no clear trend between the amounts of spalling and the age

of concrete.

Figure 3.11: Specimen After Fire Test

48

Figure 3.12: Spalling Concrete Percentage

3.3.3 Temperature Result

The thermal analyses for this investigation were done by using the first 30

minutes thermal data obtained from the data logger due to the observation that no

concrete spalling was observed after 30 minutes of the test. All thermal data was

successfully recorded into the data logger except the thermal data for 20 mm granite

aggregates specimen. Unfortunately, the data gauges for the specimen with 20 mm

granite aggregates were damaged during the fire exposure and data logger could not

retrieve most of the data. Figures 3.13, 3.14, 3.15, 3.16 and 3.17 show that the

temperatures vs. time for the specimens at each depth during fire exposure.

Figure 3.13: Temperature for specimens with 7 mm basalt aggregates

0

10

20

30

40

7 mm 14 mm 20 mmSpal

ling

Con

cret

e (%

)

Aggregate Size

Basalt Aggregate

Granite Aggregate

0

200

400

600

800

1000

1200

0.00 10.00 20.00 30.00

Tem

par

atu

re (

oC

)

Time (Minutes)

Hydro Carbon FireTemp.Temp. at 0 mm

Temp. at 25 mm

Temp. at 50 mm

Temp. at 75 mm

Temp. at 100 mm

Temp. at 150 mm

Temp. at 200 mm

* Including 4.8% -

5.2% moisture

content loss.

49



0

200

400

600

800

1000

1200

0.00 10.00 20.00 30.00

Tem

par

atu

re (

oC

)

Time (Minutes)


Temp. at 25 mm

Temp. at 50 mm

Temp. at 75 mm

Temp. at 100 mm

Temp. at 150 mm

0

200

400

600

800

1000

1200

0.00 10.00 20.00 30.00

Tem

par

atu

re (

oC

)

Time (Minutes)


Temp. at 25 mm

Temp. at 50 mm

Temp. at 75 mm

Temp. at 100 mm

Temp. at 150 mm

Temp. at 200 mm

50

Figure 3.16: Temperature for specimens with 7 mm granite aggregates

Figure 3.17: Temperature for specimens with 14 mm granite aggregates

Figure 3.13 shows that temperature at 25 mm depth significantly increased after 10

minutes of fire exposure and identical with temperature at exposed surface. At this

stage, the concrete surrounding the thermal couple at 25 mm depth spalled and the

thermal couple is exposed directly to the fire.

Figures 3.18 – 3.22 show the temperature at every depth of the specimens for 10

minutes, 20 minutes and 30 minutes fire exposure. The graphs represent the

0

200

400

600

800

1000

1200

0.00 10.00 20.00 30.00

Tem

par

atu

re (

oC

)

Time (Minutes)


Temp. at 25 mm

Temp. at 50 mm

Temp. at 75 mm

Temp. at 100 mm

Temp. at 150 mm

Temp. at 200 mm

0

200

400

600

800

1000

1200

0.00 10.00 20.00 30.00

Tem

par

atu

re (

oC

)

Time (Minutes)


Temp. at 25 mm

Temp. at 50 mm

Temp. at 75 mm

Temp. at 100 mm

Temp. at 150 mm

Temp. at 200 mm

51

temperature gradient across the thickness of the specimen. Temperature gradient can

be related with thermal diffusivity of the specimen. Higher temperature gradient reflects

lower thermal diffusivity of the specimen. It proved that the spalling of high strength

concrete with low thermal diffusivity is due to high density of high strength concrete

itself (Sanjayan and Stocks, 1993; Phan and Carino, 1998; Khaliq and Kodur, 2011).


aggregates


aggregates

0

400

800

1200

0 50 100 150 200

Tem

per

atu

re (C

)

Distance From Exposed Surface (mm)

10 minute

20 minute

30 minute

0

400

800

1200

0 50 100 150 200

Tem

par

atu

re (o

C)

Distance from exposed surface (mm)

10 minute

20 minute

30 minute

52


aggregates


aggregates

0

400

800

1200

0 50 100 150 200

Tem

per

atu

re (C

)


10 minute

20 minute

30 minute

0

400

800

1200

0 50 100 150 200

Tem

pe

ratu

re (C

)


10 minute

20 minute

30 minute

53


aggregates

3.3.4 Thermal Diffusivity

Using the temperature data presented in Section 3.3.2, thermal diffusivity of

concrete was calculated at various temperatures using the method described in

Section 3.2.7. The calculation was based on temperature at 50 mm from fire exposed

surface. Common value of thermal diffusivity for concrete at room temperature is 0.75 x

10-6 m2/s (Çengel, 2003).

Thermal diffusivity for all specimens is shown in Figure 3.23. All specimens exhibit

sharp drop in thermal diffusivity at a temperature. This temperature ranged between

110oC to 155oC for different specimens. This phenomenon is due to water changing

phase and the thermal capacity drop at this point. At this temperature, water in the

concrete absorbs the heat to transform from liquid phase to steam phase. This

temperature can be defined as saturated stream temperature. The saturated steam

temperature corresponds to a steam pressure, which is a fixed value. This value can

be obtained from standard steam tables such as (SpiraxSarco, 2011). Values from

saturated steam pressure and saturated steam temperature are used to estimate

superheated steam pressure based on fire temperature. However, the difference

between saturated steam pressure and superheated steam pressure is very small and

can be considered as equal.

0

400

800

1200

0 50 100 150 200

Tem

pe

ratu

re (C

)


10 minute

20 minute

30 minute

54

Specimen with 14 mm granite aggregates saturated steam temperature is 110oC,

which is lowest saturated steam temperature. The saturated steam pressure

correspond to the temperature is 0.042 MPa. The saturated steam temperatures for

specimens with 14 mm granite aggregates, 7 mm basalt aggregates, 14 mm basalt

aggregates and 20 mm basalt aggregates are 113oC, 120oC, 130oC and 155oC

respectively. The highest saturated steam pressure that corresponds to the highest

saturated steam temperature (20 mm basalt aggregate) is 0.442 MPa. This is

significantly less than tensile strength (ft) of concrete, which is estimated to be about 3

MPa.

The tensile strength of concrete can be estimated using the Australian Standard,

AS3600 (Clause 3.113) formula (Standard Australia, 2009).

𝑓𝑡 = 0.6√𝑓𝑐 Equation 3.6

For 70 MPa compressive strength (fc) concrete, the tensile strength may be estimated

as 3.3 MPa. At temperatures ranging from 110 to 155 degrees Celcius the tensile

strength may decrease by no more than 10% which will bring down the tensile strength

to 3 MPa (Behnood and Ghandehari, 2009; Khaliq and Kodur, 2011; Kodur, 2014). This

strength is significantly larger than the steam pressure created by the saturated steam

pressure.

Figure 3.23: Thermal diffusivity vs Temperature

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1.40

1.60

1.80

2.00

50 100 150 200 250 300

Dif

fusi

vity

(m

2/s

) x

10-6

Temperature (C)

7mm (Basalt)

14mm (Basalt)

20 mm (Basalt)

7mm (Granite)

14mm (Granite)

55

3.4 Conclusion

1 Full-scale high strength concrete wall panels spalled between 7.5% and 32.5%

when exposed to hydrocarbon fires. Concrete wall panels spalled explosively in

the first 30 minutes of fire exposure. There are no further concrete spalling

observed after this period.

2 Thermal diffusivity dropped sharply at temperatures which ranged from the

temperatures of 110oC and 155oC. This temperature can be described as phase

changing steam temperature. Heat was absorbed by water in the concrete to

change phases from liquid to steam.

3 Maximum saturated steam pressure induced by the steam corresponds to

maximum saturated steam temperature of 155°C is 0.442 MPa. This is

significantly less that tensile strength of concrete which is estimated to be 3

MPa. Hence, steam pressure itself is not a critical factor for concrete spalling.

However, it is suggested that combining steam pressure together with thermal-

mechanical mechanism such as expansion of aggregates and contraction of

cement paste may explain the spalling in fire.

56

References

Anderberg, D. Y. (1997). Spalling Phenomena of HPC and OC.L. T. Phan, N. J. Carino,

D. Duthinh and Garboczi. NIST Workshop on Fire Performance of High

Strength Concrete, Gaithersburg. 69-73.

Bazant, Z. P. and Kazemi, M. T. (1990). "Determination of fracture energy, process

zone longth and brittleness number from size effect, with application to rock and

conerete." International Journal of Fracture 44(2): 111-131.

Behnood, A. and Ghandehari, M. (2009). "Comparison of compressive and splitting

tensile strength of high-strength concrete with and without polypropylene fibers

heated to high temperatures." Fire Safety Journal 44(8): 1015-1022.



Standardization.

Çengel, Y. A. (2003). Heat Transfer: A Practical Approach. McGraw-Hill Companies

Inc., New York, USA.

Harmathy, T. Z. (1993). Fire Safety Design and Concrete. F.K.Kong and R. H. E. CBE.

Concrete Design and Construction. Longman Scientific and Technical, Essex,

England.

Heo, Y. S., Sanjayan, J. G., Han, C. G. and Han, M. C. (2010). "Synergistic effect of

combined fibers for spalling protection of concrete in fire." Cement and

Concrete Research 40(10): 1547-1554.


38(2): 103-116.

International Standard, O. (1999). ISO 834-1999: Fire Resistance Tests: Elements of

building construction. Part 1: General Requirements. Geneva, Switzerland,

International Organization for Standardization. ISO 834-1999.

Kalifa, P., Menneteau, F.-D. and Quenard, D. (2000). "Spalling and pore pressure in

HPC at high temperatures." Cement and Concrete Research 30(12): 1915-

1927.

57

Kanéma, M., Pliya, P., Noumowé, A. and Gallias, J. L. (2011). "Spalling, Thermal, and

Hydrous Behavior of Ordinary and High-Strength Concrete Subjected to

Elevated Temperature." Journal of Materials in Civil Engineering 23(7): 921-

930.

Kaviany, M. (2011). Essential of Heat Transfer: Principals, Materials and Application.

Cambridge University Press, New York, USA.



394-402.

Kim, J. H. J., Mook Lim, Y., Won, J. P. and Park, H. G. (2010). "Fire resistant behavior

of newly developed bottom-ash-based cementitious coating applied concrete

tunnel lining under RABT fire loading." Construction and Building Materials

24(10): 1984-1994.

Kodur, V. (2014). "Properties of Concrete at Elevated Temperatures." ISRN Civil

Engineering 2014: 15.


paste and concrete. Magazine of Concrete Research 63:(1): 1-9.

Phan, L. T. and Carino, N. J. (1998). "Review of mechanical properties of HSC at

elevated temperature." Journal of Materials in Civil Engineering 10(1): 58-64.



SpiraxSarco (2011). "Steam Tables. http://www.spiraxsarco.com/resources/steam-

tables.asp (19 September 2011).

Standard Australia, L. (1999). AS 1012.9: Methods of testing concrete. Method 9:

Determination of the compressive strength of concrete specimens. Sydney,

Australia, Sai Global Limited: 3.

Standard Australia, L. (2009). AS 3600-2009: Concrete Structures. Design Properties

of Materials. Sydney, Australia, SAI Global Limited. AS 3600-2009.

Welty, J. R. (1974). Engineering heat transfer. New York : Wiley, New York.

http://www.spiraxsarco.com/resources/steam-tables.asp

http://www.spiraxsarco.com/resources/steam-tables.asp

58

Yang, J. and Peng, G. F. (2011). The Mechanism of Explosive Spalling and Measures

to Resistant Spalling of Concrete Exposed to High Temperature by

Incorporating Fibers: a Review. L. J. Li. Advances in Building Materials, Pts 1-3.

Trans Tech Publications Ltd, Stafa-Zurich 168-170: 773-777.

59

CHAPTER 4

Specimen’s Size, Aggregate Size and Aggregate Type

Effect on Spalling of Concrete in Fire

4.1 Introduction

Spalling is one of the defects of concrete exposed to fire especially high

strength concrete structure and high strength concrete rather than normal strength

concrete is more prone to (Sanjayan and Stocks, 1993; Phan and Carino, 1998). Bailey

in 2002 categorized spalling into three categories; aggregate spalling, explosive

spalling and corner spalling in which he indicated that explosive spalling is the most

extremely violent and dangerous (Bailey, 2002). Number of factors affecting spalling of

high strength concrete have been reported which include moisture clog, vapour

pressure, pore pressure and thermal properties such as thermal stress mechanism,

thermal dilation and incompatibility between cement paste and aggregate (Anderberg,

1997; Kalifa et al., 2000; Yang and Peng, 2011; Pan et al., 2012).

Specimen size effect is one of the factors discussed in this chapter. Erdem (2014)

reported that the specimen size did not play an important role on the residual

mechanical properties of ECC when the specimens were heated up to 800oC although

60

the choice of specimen’s maximum size was limited by the size of the furnace . A slight

different focus on specimen size effect on mechanical properties did prove that

specimen’s size has significant effect on shear capacity and bearing capacity. The

tests results are in good agreement with Bazant’s size effect law (SEL) theory which

stated that strength decreases as the specimen size increase (Bazant and Sun, 1987;

Bažant, 2000; Ince and Arici, 2004).

Other than specimen size effect, aggregate size effect was also investigated in this

chapter. Similar study on aggregate size effect on spalling was carried out by Pan et.

al. (2012). This study shows aggregate size has a significant effect on spalling of

concrete especially when the maximum aggregate size is below 10 mm explosive

spalling was observed. Pan et al. (2012) explained the aggregate size effect on

spalling as due to higher fracture zone process for bigger aggregate size. On broader

scope, fracture zone process was reported as the main reason in many of the

aggregate size effects studies on mechanical performances such as compressive

strength, bending strength and shear strength investigations (Chen and Liu, 2004;

Meddah et al., 2010; Pan et al., 2010).

Apart from aggregate size, the effect of aggregate type was also investigated in this

chapter. Limited research has been done to investigate aggregate type effect on

spalling as a sole purpose but Chen et al. (2014) in their study on bonding behaviour

between basic oxygen furnace slag and asphalt binder reported that basalt and

aggregate and granite aggregate have similar properties and bonding behaviour. Kong

and Sanjayan (2008) reported thermal expansion for basalt aggregate and granite

aggregate does not differ much with linear expansion percentage in the range of 1.6%

to 1.8% at 700oC to 800oC temperature. This chapter mainly focused on the effect of

specimen’s size, aggregate size and aggregate type on spalling. Various sizes from

small cylinder (100 mm x 200 mm height) to large full scale wall panel (3.36 m x 3.38 m

x 0.2 m thickness) with different aggregate size and type were tested under

hydrocarbon fire to investigate the effect of specimen’s size, aggregate size and

aggregate type on spalling.


In all concrete mixes, the water to binder ratio was fixed at 0.3. The binder

included 25% to 30% of slag by weight. Superplasticizers were introduced to increase

the workability with targeted slump of 150 mm. Design strength for all concrete mixes

61

was 80 MPa. All concretes were tested for compressive strength at the age of 28 days

in accordance with AS 1012.9-1999 (Standard Australia, 1999). Table 4.1 shows

concrete mixes details.

Table 4.1: Concrete (per cubic meter) mix summary

Aggregate

Type

(Max Size)

Aggregate

Cement

(kg)

Water

(kg)

Super-

Plasticizer

(Liter)

Compressive

Strength at

28 days

(MPa)

Coarse

(kg)

Fine

(kg)

Basalt

(7 mm) 1100 577

550 (25% Slag)

150 7.5

(Viscoscreet) 70.0

Basalt

(14 mm) 1100 577

550 (25% Slag)

150 7.5

(Viscoscreet) 69.4

Basalt

(20 mm) 1200 620

600 (33% Slag)

180

9.6

(Super

Viscoscreet)

70.0

Granite

(7 mm) 870 759

633 (25% Slag)

190 7-17

(ADVA 142) 70.7

Granite

(14 mm) 901 784

600 (25% Slag)

180

7-17

(Super

Viscoscreet)

73.8

Granite

(20 mm) 956 785

567 (25% Slag)

170

7-17

(Super

Viscoscreet)

75.1

All specimens were cast at a precasting yard (Hollowcore Pty Ltd). Surface moisture

content was measured before the fire test.

4.2.1 Concrete Test Specimens

Specimen’s sizes were varied for specimen’s size effect. Basalt and granite

aggregate types were used for aggregate type effect investigation purposes. Maximum

aggregates sizes were varied from 7 mm, 14 mm and 20 mm to investigate the effect of

aggregate size on spalling. Specimen’s dimensions illustration and fire test detail were

summarized in Figure 4.1 and Table 4.2 respectively. Figure 4.2 shows image of

concrete casting at precasting yard.

62

3.38 m

3.36 m

0.2 m

(a) Large panel

1.075 m

1.075 m

0.2 m

(b) Medium Panel

63

400 mm

400 mm

1000 mm

(c) Column

100 mm

200 mm

(d) Cylinder

Figure 4.1: Specimens’ dimension

Figure 4.2: Casting of Ordinary Portland Cement High Strength Concrete

64

Table 4.2: Specimens’ fire test details

Specimen

Type/Dimension

Aggregate

Type

(Max. Size)

Ave.

Surface

Moisture

Content (%)

Casting Date

(dd/mm/yyyy)

Fire Test

Date

Age

During Fire

Test

(days)

Large Panel

Basalt

(7 mm) 5.2 3/8/2010 16/9/2010 44

Basalt

(14 mm) 5.2 5/8/2010 11/10/2010 67

Basalt

(20 mm) 5.1 20/7/2010 3/9/2010 67

Granite

(7 mm) 4.8 13/8/2010 8/11/2010 87

Granite

(14 mm) 5.1 12/8/2010 3/11/2010 83

Granite

(20 mm) 4.8 17/8/2010 10/12/10 115

Medium Panel

Basalt

(7 mm) 4.9 3/8/2010 1/5/2012 640

Basalt

(14 mm) 4.8 5/8/2010 1/5/2012 638

Basalt

(20 mm) 4.8 20/7/2010 10/4/2012 656

Granite

(7 mm) 4.9 13/8/2010 24/4/2012 630

Granite

(14 mm) 4.9 12/8/2010 10/4/2012 631

Granite

(20 mm) 4.9 17/8/2010 24/4/2012 626

Column All specimens

NA

Between

20/7/2010 to

17/8/2010

7/12/2010

Between

112 - 140

days

Cylinder All specimens

NA

Between

20/7/2010 to

17/8/2010

7/12/2010

Between

112 - 140

days

65

Large panels and medium panels were reinforced with 6 mm mild yield strength

reinforcement bar mesh with 100 mm spacing between bars in both directions. The

mesh was located at the middle of panel depth. Columns were reinforced by using four

8 mm mild yield strength bars with 6 mm bar for stirrup. Reinforcement arrangement for

large panel, medium panel and column are shown in Figure 4.3 (a-c).

R6-125 cc

100 mm200 mm

3380 mm (a)

R6-100 cc

100 mm200 mm

1075 mm (b)

400 mm

400 mm

50 mm

4R8

R6-325 cc

(c)

Figure 4.3: Reinforcement bar details (a) large panel (b) medium panel (c) column

4.2.2 Hydrocarbon Fire Test

Hydrocarbon fire temperature versus time in accordance to EN 1991-1-2 was

used for exposing the specimens. Equation 4.1 is the hydrocarbon fire temperature

equation (BSI, 2005).

𝑇 = 1080(1 − 0.325𝑒−0.167𝑡 − 0.675𝑒−2.5𝑡) + 20 (Equation 4.1)

where T = Temperature (oC)

t = time (minute)

66

The fire tests conducted for minimum of 120 minutes. The temperature of the furnace

(near exposed surface) were measured and recorded by the furnace’s temperature

sensor to the data logger.

(a)

(b)

67

(c)

(d)

Figure 4.4: Specimen Fire Test Setup (a) Large panels (b) Medium panels (c) Columns

(d) Cylinders

4.2.3 Spalling Measurements

The weight of the specimens were measured before and after the specimens

were exposed to fire. The weight difference is considered as the weight of spalled

concrete.

4.2.4 Nominal Spalling Depth Analysis

Nominal spalling depth is a method of presenting the severity of spalling.

Nominal spalling depth is an effective method to evaluate the severity of concrete

spalling which is proposed in this study. In nominal spalling depth analysis, spalling

depth was calculated based on the area in which the specimen was exposed to fire.

Nominal spalling depth is calculated as average depth of spalled concrete after fire

68

test. Nominal depth is calculated by dividing the volume of spalled concrete with fire

exposure area. Fire exposed areas for large panels, medium panels, columns and

cylinders are illustrated in Figure 4.5(a-d).

The nominal spalling depth concept is developed in this study so that

specimens of different shape and size can be directly compared without the results

being distorted by the weight of the large specimens in other previous methods such as

% weight loss.

3.36 m

3.0 m

3.3

9 m

3.0

m

Fire Exposed Area

0.2 m

(a) One side exposed to fire

1.075 m

1.0 m

1.0

75

m

1.0

m

Fire Exposed Area

0.2 m

(b) One side exposed to fire

69

0.4 m

1.0

m

0.4 m

Fire Exposed Area

(c) All four sides exposed to fire

100 mm

20

0 m

m

Fire Exposed Area

(d) the cylindrical surface exposed to fire

Figure 4.5: Fire Exposed Area (a) Large panels (b) Medium panels (c) Columns (d)

Cylinders

4.3 Result and Discussion

4.3.1 Spalling

Large panels and medium panels were observed to spall in explosive manner.

All of the spalling occurred in the first 10 minutes of fire test. No spalling observed after

70

10 minutes of fire test. The same phenomena were also reported by (Hertz, 2003).

Images of specimens after fire test are shown in Figure 4.6. Large panels have the

largest amount of spalling with percentage of concrete spalled ranging between 8%

and 33%. Average spalling for all large panels are 19%. Medium panels average

spalling are 10% ranging from 6% and 13%. Columns and cylinders average spalling

were 14% and 9% respectively. There is no concrete age effect on spalling observed.

Spalling percentage for all specimens is summarized in Table 4.3.

(a) (b)

(c) (d)

71

Figure 4.6: Images of specimens after fire test (a) Large Panel (b) Medium Panel

(c) Column (d) Cylinder

4.3.2 Nominal Depth

Figure 4.7 shows the summary of nominal spalling depth for all specimens.

Based on visual inspection on specimen after the fire tests, nominal spalling depth

analysis is the better way to represent the severity of spalling compared to weight loss

percentage due to spalling. This is illustrated by comparing the weight loss

percentages in Table 4.3 which has no observable trend with specimen size. This point

can be further illustrated by an example where if the thickness of the panel is doubled

the spalling amount would not double to maintain the spalling loss weight percentage.

In fact, beyond a certain thickness of a panel, the thickness has very little effect on the

spalling of exposed surface.

4.3.3 Specimen’s size effect

Figure 4.6 shows that there is a trend of spalling severity based on specimen’s size

when exposed to fire. Bigger specimen size caused bigger nominal spalling depth.

Large panel specimens have the biggest nominal spalling depth average of 47.3 mm.

The variability of spalling at this size was high with nominal spalling depth ranging from

18.9 mm to 82.0 mm. The maximum and minimum nominal spalling depths for medium

panels were 29.0 mm and 12.9 mm respectively. The average nominal spalling depth is

22.8 mm, a 51.8% reduction compared to large panel’s average nominal spalling

depth. As the size of specimen reduces to 400 x 400 mm x 1 m height, the average

nominal spalling depths were also reduced to 12.7 mm and when the specimen’s size

reduced to small scale 100 mm diameter cylinder, average nominal spalling depth

reduced to 2.1 mm. The consistencies of nominal spalling depth for cylinder specimens

were good with the range of nominal spalling depth being between 1.8 mm and 2.6

mm. This result shows that specimen’s size did have effect to severity of spalling and

good agreement with Bazant (1984) in his research on size effect to fracture process in

concrete (Bazant, 1984). The spalling of concrete is a fracture process therefore, the

size effect is important as evidenced by these experimental results.

72

Table 4.3: Spalling Percentage Summary for All Specimens. Specimen Details Results

Type Maximum Aggregate

Size Aggregate Type

Spalling Weight

(kg)

Spalling (%)

Nominal Depth (mm)

Large Panel

7 mm Basalt 1839.0 32.5 82.0

Granite 1295.1 22.9 57.7

14 mm Basalt 423.4 7.5 18.9

Granite 1324.5 23.4 59.1

20 mm Basalt 720.6 12.7 32.1

Granite 764.7 13.5 34.1

Medium Panel

7 mm Basalt 63.4 11.9 27.4

Granite 52.4 9.1 21.0

14 mm Basalt 64.0 11.1 25.7

Granite 72.3 12.6 29.0

20 mm Basalt 32.1 5.6 12.9

Granite 51.4 9.1 20.6

Column

7 mm Basalt 80.3 20.2 18.3

Granite 41.9 10.5 9.6

14 mm Basalt 94.1 23.6 21.5

Granite 53.0 13.3 12.1

20 mm Basalt 32.1 8.1 7.3

Granite 33.5 8.4 7.6

Cylinder

7 mm Basalt 0.32 8.1 1.8

Granite 0.38 9.8 2.2

14 mm Basalt 0.45 11.7 2.6

Granite 0.36 9.3 2.1

20 mm Basalt 0.36 9.1 2.0

Granite 0.32 8.1 1.8

73

Figure 4.7: Nominal Spalling Depth.

Figu

re 4

.7: N

omin

al S

pallin

g D

epth

. For

spe

cim

ens

ID, n

umbe

rs 7

, 14,

20

repr

esen

t max

imum

agg

rega

te s

ize

of 7

mm

, 14

mm

and

20 m

m re

spec

tivel

y. L

ette

rs G

and

B re

pres

ent b

asal

t agg

rega

te a

nd g

rani

te a

ggre

gate

resp

ectiv

ely.

4.3.

3 S

peci

men

Siz

e

Effe

ct

74

4.3.4 Aggregate Size Effect

Figure 4.8 shows the nominal spalling depth for every specimen with different

maximum aggregate size. Most of specimens with 7 mm maximum size aggregate and

14 mm maximum aggregate size have spalled more than 20 mm maximum aggregate

size.

Specimens with 14 mm maximum aggregate size does not exhibit consistent aggregate

size effect when compared with 7 mm maximum aggregate size specimens. Large

panel (granite), medium panel (granite), column (basalt) and cylinder (granite) for 14

mm maximum aggregates size spalled more than 7 mm maximum aggregate size.

However, the differences of spalling depth between 14 mm and 7 mm maximum

aggregates size are relatively small except for large panel (basalt). The differences are

1.4 mm for large panel (granite), 3.1 mm for medium panel (basalt), 8 mm for medium

panel granite, 3.2 mm for column (basalt), 2.5 mm for column (granite), 0.1 mm for

cylinder (basalt) and 0.8 mm for cylinder (granite).

The effect of aggregate size can be observed when the maximum aggregate size was

increased from 7 mm to 20 mm for all specimens. Unlike specimens with 14 mm

maximum aggregate size, spalling depth for all specimens with 20 mm maximum

aggregate size were less than specimens with 7 mm maximum aggregate size. Large

panel specimens show biggest effect for aggregate size in terms of spalling depth.

Nominal spalling depth for large panel (basalt) decreased from 82 mm to 32.1 mm

when maximum aggregate size increased from 7mm to 20 mm. Large panel (granite)

spalling depth were also reduced by 23.6 mm from 57.7 mm with the same maximum

aggregate size increment. Spalling depth reduction for medium panel (basalt), medium

panel (granite), column (basalt) and column (granite) were 6.8 mm, 0.4 mm, 11 mm

and 2 mm respectively. Cylinder specimens spalling depth were consistent between

1.8 mm to 2.6 mm. Pan et al. (2012) demonstrated with small cylinders similar effect of

aggregate size. The same effect was observed with various specimen sizes confirms

the aggregate size effect hypothesis. Pan et al. (2012) postulated that the main cause

for better resistance to spalling by large aggregate samples is the longer fracture

process zone which needs larger kinetic energy to produce crack. In slightly different

focus on aggregate size effect, some researches highlighted the performance of

mechanical properties such as bending strength and compressive strength increase as

the aggregates size increases (Chen and Liu, 2004; Meddah et al., 2010).

75

Figure 4.8: Nominal Spalling Depth for Maximum Aggregate Size

4.3.5 Aggregate Type Effect

Figure 4.9 shows nominal spalling depth comparison between specimens with

basalt aggregate and granite aggregate. Both basalt and granite are most common

igneous rock used as coarse aggregate in Australia. Basalt is a dark colored aggregate

with fine grain widely used as aggregate in concrete matrix, road base, road pavement

aggregate, railway ballast and many more. Polished basalt were also used for flooring

tiles purposes. Meanwhile granite is best known igneous rock. It is light in color with

large grains. The grains are large enough to be visible with unaided eye. Granite is also

widely used as aggregate in concrete matrix and other construction purposes. Chen et

al. (2014) found out that main properties for aggregate such as bulk specific gravity and

water absorption values are similar between basalt and granite aggregates .

From Figure 4.9, it is clear that there is not a clear trend for nominal spalling depth

between basalt aggregate specimens and granite aggregate specimens. Basalt

aggregate large panel with 14 mm and 20 mm maximum aggregate size spalled more

than granite aggregate large panel but different scenario with large panel with 7 mm

maximum aggregate size where granite aggregate panel spalled less than basalt

aggregate. The same phenomena happens to nominal spalling depth for medium

panels, columns and cylinders where no decisive conclusion can be made when basalt

76

aggregate were compared to granite aggregate. This is likely due to the fact that both

basalt and granite have similar thermal expansion coefficients. Thermal expansion

coefficient for concrete with basalt and granite as coarse aggregate are 9.5 and 9.3

microstrain/oC respectively (Naik et al., 2011).

Figure 4.9: Nominal Spalling Depth for Maximum Aggregate Type

4.4 Conclusion

1 Specimen size did have effect on the spalling of concrete under hydrocarbon

fire exposure. Nominal spalling depths of concrete increases as the specimen

size increases.

2 The aggregate size effect on spalling was prominent when the maximum

aggregate size increased from 7 mm to 20 mm. For specimens with 20 mm

maximum aggregate size spalling depth is less compared to specimens with 7

mm and 14 mm maximum aggregate size.

3 Aggregate type (basalt and granite) does not exhibit any clear trend on spalling

depth. This is likely to be due to similar thermal expansion properties of basalt

and granite.

77

4 Nominal spalling depth concept proposed in this study is an effective method of

representing the degree of spalling and is better than the percentage weight

loss method commonly used.

78

References




Bailey, C. (2002). "Holistic behaviour of concrete buildings in fire." Proceedings of the

Institution of Civil Engineers: Structures and Buildings 152(3): 199-212.

Bazant, Z. P. (1984). "SIZE EFFECT IN BLUNT FRACTURE: CONCRETE, ROCK,

METAL." Journal of Engineering Mechanics 110(4): 518-535.

Bažant, Z. P. (2000). "Size effect." International Journal of Solids and Structures 37(1–

2): 69-80.

Bazant, Z. P. and Sun, H.-H. (1987). "SIZE EFFECT IN DIAGONAL SHEAR FAILURE:

INFLUENCE OF AGGREGATE SIZE AND STIRRUPS." ACI Materials Journal

84(4): 259-272.



Standardization.

Chen, B. and Liu, J. (2004). "Effect of aggregate on the fracture behavior of high

strength concrete." Construction and Building Materials 18(8): 585-590.

Chen, Z., Xie, J., Xiao, Y., Chen, J. and Wu, S. (2014). "Characteristics of bonding

behavior between basic oxygen furnace slag and asphalt binder." Construction

and Building Materials 64(0): 60-66.

Erdem, T. K. (2014). "Specimen size effect on the residual properties of engineered

cementitious composites subjected to high temperatures." Cement and

Concrete Composites 45(0): 1-8.


38(2): 103-116.

Ince, R. and Arici, E. (2004). "Size effect in bearing strength of concrete cubes."


79

Kalifa, P., Menneteau, F.-D. and Quenard, D. (2000). "Spalling and pore pressure in

HPC at high temperatures." Cement and Concrete Research 30(12): 1915-

1927.



Composites 30(10): 986-991.

Meddah, M. S., Zitouni, S. and Belâabes, S. (2010). "Effect of content and particle size

distribution of coarse aggregate on the compressive strength of concrete."


Naik, T. R., Kraus, R. N. and Kumar, R. (2011). "Influence of types of coarse

aggregates on the coefficient of thermal expansion of concrete." Journal of

Materials in Civil Engineering 23(4): 467-472.





paste and concrete. Magazine of Concrete Research 63:(1): 1-9.

Phan, L. T. and Carino, N. J. (1998). "Review of mechanical properties of HSC at

elevated temperature." Journal of Materials in Civil Engineering 10(1): 58-64.






Yang, J. and Peng, G. F. (2011). The Mechanism of Explosive Spalling and Measures

to Resistant Spalling of Concrete Exposed to High Temperature by

Incorporating Fibers: a Review. L. J. Li. Advances in Building Materials, Pts 1-3.

Trans Tech Publications Ltd, Stafa-Zurich 168-170: 773-777.

80

CHAPTER 5

Investigation of the Effects of Fly Ash Types and Properties

on the Workability of Fresh Geopolymer

5.1 Introduction

Workability is often referred as the ease with which a concrete can be

transported, placed and consolidated without any loss of stability or homogeneity. It is

greatly affect by the characteristics of the constituent materials of concrete (Leite et al.,

2013). The workability properties are measured for fresh concrete before it has set and

hardened. Adequate workability of a fresh concrete is essential in order to reduce the

amount of air voids in hardened concrete which are also affected by bad compacting

and bad consistencies (Neville, 2010 ). Strength of concrete is adversely and

significantly affected by the presence of voids in the compacted mass. In Portland

cement based concrete, amount of water in the mix is the main parameter that is used

to control workability. Higher water content causes greater workability.

81

The new breed of concrete binder considered to be the binder of the future due to less

CO2 emission, geopolymer is a cementless concrete binder which was coined by

Davidovits in 1975 (Davidovits, 1991; Van Jaarsveld et al., 2002).

Geopolymers are a new class of materials, whose potential remains to be fully

unlocked (Provis and van Deventer, 2009). Geopolymer cements provide an alternative

to the Portland cement used to manufacture concrete. The main raw material

commonly used to produce geopolymers is fly ash, which is found in the waste stream

from power generation facilities (Hardjito et al., 2004; Bakharev, 2005).

There have been many research works focused on the aluminate and silicate reaction

factors on geopolymer compressive strength (Davidovits, 1989; Palomo et al., 1999;

Hardjito et al., 2004; Duxson et al., 2007; Bakri et al., 2012). However, casting

geopolymer concrete in real life construction needs a specific attention to the

workability of the concrete itself. Among the other parameters that affect workability of

fresh concrete mix are shape of aggregates and binder’s particle size and distribution

(Leite et al., 2013; Deb et al., 2014). Alkaline solution properties such as viscosity of

sodium silicate, ratio of silicate to hydroxide and ratio of solution to fly ash were also

found out to influence workability of geopolymer (Chindaprasirt et al., 2007; Nath and

Sarker, 2014). Since fly ash is a by-product of coal generated power plant and different

power stations use different coal sources, the chemical and physical properties of one

fly ash to another are different (Hardjito et al., 2004). The workability differences

between fly ash from different power plants can be due to chemical property

differences or physical property differences. Tarong fly ash’s poor performances as

compared to other fly ash available in Australia was concluded to be due to high Si/Al

ratio (Keyte, 2008). This chapter mainly focused on fly ash physical properties effect on

geopolymer workability. Physical properties tests such as particle average diameter,

particle size distribution, particle shape, density, pore volume, specific surface, bulk

density and absorption tests for Tarong fly ash, Gladstone fly ash and Microash were

carried out to investigate the effects of fly ash types and properties on the workability of

fresh geopolymer.

82

5.2 Experimental Program

5.2.1 Raw Material

Fly ashes used were from Gladstone Power Station, Tarong Power Station and

Bayswater Power Station (Microash). These are power stations located in various parts

of Australia and use different sources of coal for their power generations. Gladstone

and Tarong fly ash were supplied by Cement Australia. Microash was supplied by Fly

Ash Australia. Alkaline activator used in the geopolymer mix was combination of

sodium silicate, Na2SiO3 and sodium hydroxide, NaOH (molarity of 8.0M). Sodium

silicate liquid and sodium hydroxide solids were supplied by PQ Australia and Sigma

Aldrich. Sodium silicate used in this experiment has a ratio of SiO2 to Na2O of 2.

Sodium hydroxide of 8.0M solution was made by mixing 26.2% of NaOH solids and

73.8% of water (Rangan, 2006). The NaOH solution was mixed at least 24 hours prior

to mixing with fly ash. Sodium silicate and sodium hydroxide were mixed to form a

single activator solution just a few minutes before the activator solution get mixed with

fly ash to form geopolymer mix. In all the geopolymer mixes investigated in this study

presented in this chapter, the alkaline solution to fly ash ratio was fixed at 0.4.

5.2.2 Workability Test

The workability of the geopolymer mix was measured by using slump test in

accordance to the ASTM C230/230M-08 (ASTM, 2010). Geopolymer mixes were

mixed, poured and compacted in three layers into a brass cone mould with an inner

diameter base (d0) of 100 mm. Each layer was tamped by 25 times by a brass tamper.

The mould was lifted and the mix was allowed to flow. The diameters of d1 and d2 of

the flowed mix was measured after one minute. The slump-flow (SF) of the final

slumped mortar was measured on the average of two perpendicular diameters (d1, d2)

after cone lifting. Figure 5.1 illustrates the measurements d1 and d2. Relative slump rp

was calculated by using Equation 5.1 (Topçu and Uygunoǧlu, 2010).

d1

d2

Figure 5.1: Measurement of d1 and d2.

83

𝑟𝑝 = (𝑆𝐹

𝑑0)

2

− 1 (Equation 5.1)

Where,

𝑆𝐹 =𝑑1+𝑑2

2 mm

d0 = 100 mm

5.2.3 Investigation of the Relevant Physical Properties of Fly Ash

Various fly ash physical properties such as particle size distribution, mean

particle diameter, shape images, specific surface area, average pore diameter and

pore volume were measured and analysed. Particle size distribution and average

diameter were measured by using CILAS Particle Size Analyzer 1190. Figure 5.2

shows the image of CILAS Particle Size Analyzer 1190 equipment. Particle shape

analyses were carried out by using Zeiss Supra 40VP Scanning Electron Microscope

(SEM) analyses. Particle shape images were captured from SEM analyses. Samples

were thinly gold coated with a DYNAVAC (CS 300) deposition system prior to the SEM

analysis due to fly ash has non-conductive surface and without electrically

conductive surface, charging effect may arise during scanning. Picture of Zeiss

Supra 40VP equipment is shown in Figure 5.3. Fly ash specific surface area and pore

properties were measured based on nitrogen adsorption method by using Belsorp Max

equipment (Figure 5.4). Specific surface area and pore volume at atmospheric

pressure were calculated based on Brunauer-Emmett-Teller (BET) theory using

BELMasterTM Version 6.3.1.0 software which was developed by BEL Japan, Inc. BET

theory explained that specific surface area of fly ash is determined by measuring

physical adsorption of nitrogen gas on the surface of the solid (Brunauer et al., 1938).

The amount of adsorbed gas to a monomolecular layer on the surface corresponds to

specific surface area and pore volume values. Density of fly ash measurement was

carried out in accordance to Indian Standard IS 4031 Part 11:1988. Weighed quantity

of fly ash was scooped in La chatelier flask filled with kerosene (Figure 5.5). Displaced

volume of kerosene in the la chatelier flask was recorded. Density of fly ash is

calculated based on the mass of the fly ash and displaced volume (Indian Standard,

1988).

84

Figure 5.2: CILAS Particle Size Analyzer

Figure 5.3: ZEISS Supra 40VP Scanning Electron Microscope

85

Figure 5.4: Belsorp Max Adsorption Measurement

Figure 5.5: Fly ash density measurement equipment

5.2.4 Investigation of the Relevant Chemical Properties of Fly Ash

Fly ash from Gladstone, Tarong and Microash were tested using X-Ray

Fluorescence (XRF) analysis to determine the chemical component. They were tested

for main oxide components such as Al2O3, SiO2, CaO, MgO, MnO, Fe2O3 and others.

Geopolymer formation is due to alluminate and silicate reaction. Unlike cement which

used water for hydration process to produce Calcium-Silicate-Hydrate CSH gel,

86

geopolymer is based on reaction from activator solution with aluminate and silicate

source in fly ash for polymerization and hardening. It is postulated that any kind of

aluminate and silicate reaction do not take place since there are no heat introduced

during the initial mixing procedure. All the fly ashes were mixed with demineralized

water with the ratio of water to fly ash being set constant at 0.4. The workability of

these mixtures were measured similar with the procedure explained in Section 5.2.2.

Water from the geopolymer mixture was filtered by using filter paper. Inductively couple

plasma (ICP) analyses were conducted on the water to identify chemical component

which have already dissolved at initial stage of geopolymer mixing even without

aluminate and silicate reaction. ICP analyses were carried out using Varian 720-ES

equipment (Figure 5.6). The oxides selected as standard references were based on

chemical components obtained from XRF analyses.

Figure 5.6: Varian 720-ES inductively coupled plasma equipment

5.3 Results and Discussions

5.3.1 Workability Results

Geopolymer with Gladstone fly ash has the best workability compared to

geopolymer using Tarong fly ash and Microash. Figure 5.7 (a-c) shows the images of

geopolymer made of fly ash from Gladsone, Tarong and Microash. Gladstone fly ash

immediately changes its form into liquid paste when mixed with alkaline solution.

87

Relative slump for Gladstone geopolymer was 7.7. Microash changed its form into very

stiff paste when mixed alkaline solution. There is no relative slump measurement for

Microash admixture due to the solid form of this Geopolymer. Tarong fly ash has the

worst workability compared to Gladstone fly ash and Microash. Tarong fly ash was

observed to absorb all the alkaline solution when the alkaline solution was poured into

the mixing bowl. The mixture maintained its powder form and not much changes were

observed. The ratio of alkaline solution to fly ash shown in these images was 0.4.

Additional alkaline solution was added into Tarong fly ash and Microash geopolymers

until relative slump values were equivalent to Gladstone geopolymer. Tarong fly ash

and Microash exhibit similar workability with Gladstone fly ash when the ratios of

alkaline solution to fly ash were increase to 0.8 and 0.7 respectively. The workability

results of geopolymers are summarised in Table 5.1

(a) (b)

(c)

Figure 5.7: Geopolymers with alkaline solution to fly ash ratio of 0.4 for (a) Gladstone.

(b) Tarong. (c) Microash

88

Table 5.1: Relative slump summary

Fly Ash Alkaline Solution

to Fly Ash Ratio

Diameter 1 (d1)

mm

Diameter 2 (d2)

mm

Relative Slump

(rp)

Gladstone 0.4 290 300 7.7

Tarong 0.4 N/A N/A N/A

Tarong 0.8 290 292 7.5

Microash 0.4 N/A N/A N/A

Microash 0.7 310 320 8.9

5.3.2 Analyses of the Measured Physical Properties of Fly Ash

5.3.2.1 Particle Size Analyses

Tarong fly ash has the largest particle average diameter of 18.63 µm followed

by Gladstone fly ash and Microash. Gladstone fly ash and Microash average diameters

were 10.11 µm and 4.20 µm respectively. Largest average diameter of Tarong fly ash

resulted in lowest density of 2.09 g/cm3. Low density can be due to more voids

between particles which provide more volume and low in mass for Tarong fly ash. The

density for Gladstone fly ash and Microash were 2.29 and 2.43 g/cm3. Table 5.2

summarises the average diameter and density of all fly ash. Based on particle size

distribution graphs, Gladstone fly ash has the most evenly distributed particle size.

Biggest particle diameter for Gladstone fly ash is 100 µm and 50% of Gladstone fly ash

particle size ranging from 1 µm to 100 µm. Tarong fly ash particle size was not evenly

distributed as compared to Gladstone fly ash. Particles diameters for Tarong fly ash

ranged from 0.04 µm to 100 µm. More than 50% of Tarong fly ash particles diameters

were more than 10 µm. However, there was a gap for fly ash particles with diameter

between 1 µm and 5 µm. Less than 5% of particle fly ash with the diameter within this

range. There were 30% fly ash particles with diameter less than 1 µm. Microash

particle size exhibit more even distribution compared to Tarong fly ash. More than 85%

of its particles size ranged between 0.8 µm and 30 µm. The distribution of particle size

within this range was equivalent to the S-Curve Cumulative distribution. There were

less than 15% of Microash particles of diameter between 0.04 µm and 0.3 µm. There

was also a gap in which less than 5% of particles of fly ash are with particle diameter

within 0.3 µm and 0.6 µm. Figure 5.8 (a-c) show particle size distribution graphs for (a)

Gladstone fly ash (b) Tarong fly ash and (c) Microash.

89

Table 5.2: Average Diameter Summary

Fly Ash Fly Ash Average Diameter (µm)

Fly Ash Density (g/cm3)

Gladstone 10.11 2.29 Tarong 18.63 2.03

Microash 4.20 2.43

(a)

(b)

90

(c)

Figure 5.8: Particle size distribution graphs for (a) Gladstone fly ash. (b) Tarong fly ash.

(c) Microash

Figure 5.9 (a-c) shows the adsorption and desorption graphs for Gladstone fly ash,

Tarong fly ash and Microash. Pressure for adsorption and desorption of nitrogen gas

measurement ranged from atmospheric pressure (relative pressure = 0.999) to very

low pressure (relative pressure = 0.01). Relative pressure is the ratio of pressure in

sample cell to atmospheric pressure (/0). From Figure 5.9, absorption and desorption

curves for all fly ashes were similar. This indicates that the adsorption of nitrogen gas

on particle surface was accurate and acceptable. Based on the results obtained from

adsorption test, specific surface area was calculated using BET theory. Gladstone fly

ash, Tarong fly ash and Microash specific surface areas were 1.15 m2/g, 1.09 m2/g and

1.16 m2/g respectively. These results are in a good agreement with findings from

Pathan (2003).

BET theory analyses also provide values such as mean pore diameter and pore

volume at atmospheric pressure. These pores can be due to pores from the particle

itself or agglomeration of particles. From the pore volume and mean pore diameter

values analyses, there was a clear trend between pore volume and workability

observed. Gladstone fly ash has best workability and the lowest volume of pores and

mean pore diameter. It has pore volume of 3.39 mm3/g and 11.8 nm mean pore

diameter. Microash has larger pore volume compared to Gladstone fly ash. The pore

volume and mean diameter were 3.8 mm3/g and 12.9 nm respectively. Tarong fly ash

has the largest pore volume amongst other fly ash with 4.05 mm3/g and 14.7 nm mean

pore diameter. Its pore volume was 20% more than Gladstone fly ash and 8% higher

91

than Microash. It is clearly indicated that high workability geopolymer made from fly ash

with highest pore volume consumed more alkaline solution. The pore volume is

postulated to be the main factor for high porosity of the fly ash which resulted in high

solution demand and low workability. Table 5.3 summarises the results calculated from

the adsorption test.

(a)

(b)

92

(c)

Figure 5.9: Adsorption and desorption curves for (a) Gladstone fly ash (b) Tarong fly

ash (c) Microash

Table 5.3: Adsorption test (BET Theory) results summary

Fly Ash Specific Surface

(m2/g)

Mean Pore Diameter

(nm)

Pore Volume

(mm3/g)

Gladstone 1.15 11.8 3.39

Tarong 1.09 14.7 4.05

Microash 1.16 12.9 3.75

Further absorption and bulk density tests were carried out to consolidate the pore

volumes and porosity effect on workability. For bulk density, fly ash weight was

measured in a known volume container. Bulk density was calculated by dividing the

weight of fly ash with the volume. For absorption test, demineralised water was added

to fly ash in steps of very small amount until fly ash changed its form from loose grainy

particle into cohesive mass. Figure 5.10 shows the images of fly ash in loose grainy

form and cohesive mass. This is the point where it is considered that all the water was

absorbed by the fly ash. Demineralised water used in this test due to the insolubility of

fly ash in water (Cement Australia, 2011). The amount of water added was recorded

and the percentage of water mass to the fly ash mass was calculated. Results from

bulk density and absorption test were summarised in Table 5.4 and Table 5.5

respectively.

From Table 5.4, bulk density for Gladstone fly ash, Tarong fly ash and Microash were

1083 kg/m3, 714 kg/m3 and 905 kg/m3 respectively. The differences in the bulk density

93

could play major roles in workability of geopolymers. This is due to the fact that fly ash

mass was used to measure the amount of fly ash needed in a geopolymer mix. For

example, Tarong fly ash will have 34% more volume compared to the same mass of

Gladstone fly ash. This extra volume makes the demand for alkaline solution higher

and reduced the workability of the fly ash.

Table 5.4: Bulk density of fly ash

Fly Ash Container Volume

(mL)

Fly Ash Weight

(g)

Bulk Density

(kg/m3)

Gladstone 487.1 527.5 1083

Tarong 487.1 348 714

Microash 487.1 440.8 905

Figure 5.10: Images of Gladstone fly ash in loose grainy form (left) and solid form

(right).

(This is representative pictures for all fly ashes)

Table 5.5: Water Absorption Test Results

Fly Ash Sample Weight (g)

Water Absorbed (g)

Water Percentage (%)

Gladstone 400 81.4 20 Tarong 400 167.9 42

Microash 400 112 28

From Table 5.5, 400 g of Gladstone fly ash needs only 81.4 g of water to change its

form from loose grainy form into solid. 81.4 g of water absorbed is 20% of total weight

94

of fly ash. The same amount of Microash absorbed 112 g of water for the ash changed

its form to solid. 112 g of water weight is 28% from fly ash weight. However, Tarong fly

ash absorbed 167.9 g of water and it is 42% from the fly ash weight. Results from

absorption test is in a good agreement with pore volume results obtained from

adsorption test and provide further evidence to conclude that porosity of fly ash plays

an important factor influencing the workability of fly ash.

5.3.2.2 Particle Shape Analyses

Figure 5.11 shows the comparison of images between Gladstone fly ash and

Tarong fly ash particle shapes on various magnifying scale. There was not much

difference or abnormality observed from SEM analyses. Both fly ash exhibited

spherical shape. However, it can be observed that Tarong fly ash particles were more

agglomerated compared to Gladstone fly ash. These images proved that

agglomeration of particles was the main cause of high pore volume in fly ash

(adsorption test) which leads to high porosity (absorption test) and low bulk density.

Gladstone Power Station Fly Ash Tarong Power Station Fly Ash

95

Figure 5.11: Scanning Electron Microscope (SEM) Images comparison for Gladstone

Power Station fly ash and Tarong Power Station fly ash.

5.3.3 Chemical Analyses

Table 5.6 summarises the chemical component obtained from XRF test. The

results were similar to XRF tests conducted by Keyte (2008) and Kong and Sanjayan

(2010) . Workability test was conducted to fly ash and demineralised water with water

to fly ash ratio of 0.4. From workability test, Gladstone fly ash still produced the best

workability followed by Microash. Gladstone fly ash immediately changes into liquid

paste form when mixed with demineralised water. Microash turns into a stiff paste

when mixed with the same amount of water. However, the Tarong fly ash still has the

worst workability compared to other fly ashes and the mixture maintain with its grainy

forms when mixed with water. The workability of fly ash and demineralised water

mixture were identical to fly ash and alkaline activator mixtures (geopolymer). Main

reaction in geopolymer was aluminosilicate source from fly ash such as Al2O3 and SiO2

and aluminate and silicate such as sodium silicate and sodium hydroxide from alkaline

solution to form gelation, polymerization and hardening (Duxson et al., 2007). However,

without the presence of any aluminate and silicate in these water mixtures, the reaction

was not taking place which rules out the dissolution effect of Al2O3 and SiO2 in

geopolymer workability. Figure 5.12 (a-c) shows the images of mixtures of fly ash and

water mixtures with 0.4 ratio of water to fly ash.

Further analyses were carried out to investigate chemical effects on workability.

Additional demineralised water was then mixed into fly ash until all of fly ash and water

mixture became more like liquid mixture. The ratio of 0.75 water to fly ash was

observed to be the optimum ratio for all fly ashes especially Tarong to change its form

into liquid like mixture. The water from the mixture was then filtered using paper filter 5

minutes after the water and fly ash were mixed to determine chemical content through

96

Inductively Couple Plasma (ICP) analysis. Results from ICP analysis are presented in

Table 5.7.

Table 5.6: Chemical Component and slump-flow of Fly Ash

Component (wt%)

Gladstone Tarong Microash Al2O3 25.56 23.22 17.66 BaO 0.09 0.03 0.05 CaO 4.3 0.07 0.49 Cr2O3 - - - Fe2O3 12.48 0.89 3.58 K2O 0.7 0.52 0.94 MgO 1.45 0.14 0.53 MnO 0.15 0.02 0.07 Na2O 0.77 0.06 0.16 P2O5 0.885 0.046 0.137 SO3 0.24 0.06 0.14 SiO2 51.11 73.15 75.16 TiO2 1.32 1.31 0.66

Loss on Ignition 0.57 0.75 0.62 Total 99.62 100.36 100.23

Si/Al Ratio 1.99 3.15 4.26 Relative Slump 7.5 N/A* N/A*

*Fly ash still in powder form. Therefore, slump measurement is unavailable.

From Table 5.7, no trace of aluminium oxide and sodium oxide was detected on filtered

water for all fly ashes. 5.7 to 5.85 part per million (ppm) of Ferrous oxide and

Magnesium oxide were detected in all fly ashes as well. Silicon oxide amount (ppm) for

Gladstone fly ash, Tarong fly ash and Microash were 15, 21 and 30 respectively. The

differences of the oxides amount for all of the fly ashes are too small to be considered

as an important factor. However, Calcium oxide amount traced in Gladstone fly ash

filtered water was high compared to Tarong fly ash and Microash. Concentration of 660

ppm calcium oxide for Gladstone fly ash filtered water followed by 260 ppm for Tarong

fly ash and 22 ppm for Microash. This is due to high amount of calcium oxide in

Gladstone fly ash referring to XRF analyses result. To further investigate whether

calcium oxide plays a major role in workability, additional calcium oxide was added into

Tarong fly ash and demineralised water mixture. In order to stimulate the amount of

calcium oxide in Gladstone fly ash, 5% of calcium oxide (by fly ash weight) was added

to the mix. However, there was no difference in workability observed. Therefore, it can

be concluded that neither calcium oxide nor any aluminosilicate component plays any

role in geopolymer mix workability.

97

(a) (b)

(c)

Figure 5.12: Mixtures of demineralized water and fly ash from (a) Gladstone Power

Station. (b) Tarong Power Station. (c) Microash

Table 5.7: Chemical concentration for water filtered from fly ash and demineralized water mixture

Fly Ash

Source

Chemical Concentration in ppm

SiO2 CaO FeO MgO Al2O3 Na2O

Gladstone 15 660 5.84 5.8 0 0

Tarong 21 260 5.71 5.7 0 0

Microash 30 22 5.85 5.8 0 0

5.4 Conclusion

1 Geopolymer from Gladstone fly ash has the best workability compared to

Tarong fly ash and Microash geopolymer. Gladstone Power Station fly ash

geopolymer mixture’s relative slump value was 7.7 with alkaline solution to fly

ash ratio of 0.4. Microash and Tarong fly need 0.7 and 0.8 ratio of alkaline

98

solution to fly ash respectively to achieve similar level of workability to

Gladstone geopolymer with 0.4 ratio.

2 High pore volume caused high porosity of fly ash and from absorption test,

Gladstone, Tarong and Microash absorbed 20%, 42% and 28% of water

respectively. The differential percentages of water absorbed between the fly ash

were similar with pore volume differential percentages. Therefore, it can be

concluded that high pore volume is the main factor affecting the workability of fly

ash

3 Bulk density values for all the fly ash indicate that Tarong fly ash volume was

34% higher than the Gladstone on the same mass. Low bulk density of Tarong

fly ash is postulated by high pore volume of the fly ash. This resulted in high

alkaline solution demand and low workability for Tarong fly ash.

4 Particle size distribution analyses show that Microash was the finest fly ash

followed by Gladstone Power Station fly ash and Tarong Power Station Fly ash

with average diameter of 4.20 µm, 10.11 µm, and 18.63 µm respectively.

According to this, Tarong fly ash should provide the highest workability since

large particles provide higher workability. Since the opposite effect was

observed, the particle size distribution can be discounted as an influencing

factor.

5 Gladstone, Tarong and Microash have pore volumes of 3.4, 4.05 and 3.75

mm3/g respectively. Pore volume can be due to pore from agglomeration and

pore from the particle surface itself.

6 SEM images provide further proofs that Tarong fly ash particles are

agglomerated compared to Gladstone even though both particle types exhibit

spherical shape. Agglomeration of fly ash particle reflected in high pore volume

of fly ash which caused high porosity of fly ash. Therefore, it can be concluded

that the porosity of fly ash is the main factor effecting workability.

7 Aluminosilicate reaction was not an influencing factor in geopolymer workability.

All fly ash mixed with demineralised water exhibit similar workability levels

compared with with fly ash mixed with alkaline solution.

8 Even though higher calcium oxide was found to have dissolved in Gladstone fly

ash mixture, there was not any improvement observed to the workability of

Tarong fly ash when additional calcium oxide was added to Tarong fly ash

geopolymer mixture. Therefore, it can be concluded that chemical properties do

not affect the workability of fly ash geopolymer since there was no major

differences between various chemicals found in the dissolution water except

calcium oxide.

99

100

References

ASTM (2010). C230/C230M – 08: Standard Specification for Flow Table for Use in

Tests of Hydraulic Cement. West Conshohocken, Pennsylvania 19428-2959,

United States, ASTM International.

Bakharev, T. (2005). "Geopolymeric materials prepared using Class F fly ash and

elevated temperature curing." Cement & Concrete Research 35(6): 1224-1232.

Bakri, A. M. M. A., Heah, C. Y., Kamarudin, H., Liew, Y. M., Luqman, M., Nizar, I. K.

and Ruzaidi, C. M. (2012). Processing and characterization of calcined kaolin

cement powder. Construction and Building Materials 30:(1): 794+.

Brunauer, S., Emmett, P. H. and Teller, E. (1938). "Adsorption of Gases in

Multimolecular Layers." Journal of the American Chemical Society 60(2): 309-

319.

Cement Australia (2011). "Fly Ash - Tarong.

http://www.cementaustralia.com.au/wps/wcm/connect/website/packaged-

products/resources/7490140045f9401681178f6685a73cbc/CA008+Tarong+Fly

+Ash.pdf (11 August 2014).

Chindaprasirt, P., Chareerat, T. and Sirivivatnanon, V. (2007). "Workability and

strength of coarse high calcium fly ash geopolymer." Cement and Concrete

Composites 29(3): 224-229.


Analysis 35(2): 429-441.



Deb, P. S., Nath, P. and Sarker, P. K. (2014). "The effects of ground granulated blast-

furnace slag blending with fly ash and activator content on the workability and

strength properties of geopolymer concrete cured at ambient temperature."

Materials & Design 62(0): 32-39.

Duxson, P., Fernández-Jiménez, A., Provis, J. L., Lukey, G. C., Palomo, A. and Van

Deventer, J. S. J. (2007). "Geopolymer technology: The current state of the art."


http://www.cementaustralia.com.au/wps/wcm/connect/website/packaged-products/resources/7490140045f9401681178f6685a73cbc/CA008+Tarong+Fly+Ash.pdf



101



101(6): 467-472.

Indian Standard (1988). IS 4031: Methods of Physical Tests for Hydraulic Cement. Part

11: Determination of Density. New Delhi 110002 Bureau of Indian Standard.

Keyte, L. (2008). What's wrong with Tarong?: the importance of coal fly ash glass

chemistry in inorganic polymer synthesis.



40(2): 334-339.

Leite, M., Figueire do Filho, J. and Lima, P. (2013). "Workability study of concretes

made with recycled mortar aggregate." Materials and Structures 46(10): 1765-

1778.

Nath, P. and Sarker, P. K. (2014). "Effect of GGBFS on setting, workability and early

strength properties of fly ash geopolymer concrete cured in ambient condition."

Construction and Building Materials 66: 163-171.

Neville, A. M. (2010 ). Concrete technology J. J. Brooks. Prentice Hall Harlow, England

; New York : .

Palomo, A., Grutzeck, M. W. and Blanco, M. T. (1999). "Alkali-activated fly ashes: A

cement for the future." Cement and Concrete Research 29(8): 1323-1329.

Pathan, S. M., Aylmore, L. A. G. and Colmer, T. D. (2003). "Properties of Several Fly

Ash Materials in Relation to Use as Soil Amendments." Journal of

Environmental Quality 32(2): 687.

Provis, J. L. and van Deventer, J. S. J. (2009). Geopolymers - Structure, Processing,

Properties and Industrial Applications, Woodhead Publishing.

Rangan, B. V. (2006). "Fly ash-based geopolymer concrete [2]." Indian Concrete

Journal 80(2): 35.

Topçu, I. B. and Uygunoǧlu, T. (2010). "Influence of mineral additive type on slump-

flow and yield stress of self-consolidating mortar." Scientific Research and

Essays 5(12): 1492-1500.

102




103

CHAPTER 6

Strength of Geopolymers in Saturated and Dry Conditions

6.1 Introduction

Concrete is the most widely used construction material (Raijiwala and Patil,

2010). Concrete structures were designed to withstand various types of environmental

conditions from mild conditions to very severe conditions (Kim et al., 2010). Resistance

to environment exposure inclusive of water exposure for reinforced concrete was dealt

with adequate concrete cover (Standard Australia, 2009). However, alternatives to

Ordinary Portland Cement (OPC) are gaining popularity due to the threat of global

warming. Carbon emissions due to concrete manufacture range between 0.3 to 0.4 ton

of CO2 per cubic meter of concrete is falling behind to only fossil fuels (oil, coal and

natural gas) in terms of man-made global carbon emissions (Sanjayan, 2010). Hence,

the new breed of concrete binder using geopolymer, a name coined by Davidovits in

1975, is the binder for the concrete of the future due to less CO2 emission (Davidovits,

1991; Van Jaarsveld et al., 2002).

In normal OPC concrete, water is not an issue due to the nature that OPC

hardening is a hydration process. Pore pressure effects in OPC concrete have been

reported to cause a difference between saturated and dry conditions up to about 5%

(Wang et al., 2009). This effect is normally ignored since it is in the same order of

magnitude of variation of test results normally observed in concrete testing. Further,

this difference between strengths in saturated and dry conditions has not been

104

consistently observed. Some early researchers reported that the saturated concrete is

slightly weaker than the dry concrete (Wittmann, 1973; Haynes, 1976) and attributed

this to the pore pressure in concrete. However, later researchers challenged this pore

pressure concept and found that there were no distinct differences between saturated

and dry conditions (Morley, 1979; Zielinski et al., 1981). In the design of concrete

structures the effect of any strength differences that may possibly exist between

saturated and dry concrete is not a consideration.

According to Davidovits (1989), water plays no roles in geopolymerization due to

no presence of any water or OH- in geopolymer structure even though there is

presence of water due to polymeric bond reaction of Al-O and Si-O in geopolymer

(Davidovits, 1991). There are very few researches works published on water effect on

geopolymer concrete. Steffens et al. (2003) investigated the effect of both water and

alkaline silica reaction (ASR) on reactive retaining wall found out that geopolymer

concrete exposed to water is susceptible to delamination degradation. Resistance to

water is an important durability parameter that needs investigation in construction

materials. The results presented in this chapter aimed to investigate the drying

(vacuum drying) and saturated (saturation in vacuum conditions) cycle in geopolymer

and its effect on strength. The results show substantial difference between saturated

and dry samples that cannot be simply explained by pore pressure concept.


6.2.1 Material

Manufacture of geopolymer requires an aluminosilicate source (such as

metakaolin or fly ash) and activators. The aluminosilicate source for the geopolymer

used in this study was fly ash sourced from Pozzolanic Gladstone in Queensland,

Australia. Chemical composition for this fly ash is shown is Table 6.1. Alkaline activator

used was the combinations of sodium silicate, Na2SiO3 and sodium hydroxide, NaOH

(molarity of 8.0M). Sodium silicate used in this experiment has a ratio of SiO to Na2O of

2. Sodium hydroxide of 8.0M solution comprises 26.2% of NaOH solids and 73.8% of

water (Rangan, 2006). The NaOH solution was mixed at least 24 hours prior to mixing

with fly ash. Sodium silicate and sodium hydroxide were mixed into form one activator

solution just a few minutes before the activator solution get mixed with fly ash to form

geopolymer mix.

105

Table 6.1: Chemical composition of fly ash (Kong and Sanjayan, 2010) Chemical Component (wt%)

Al2O3 27.0 SiO2 48.8 CaO 6.2

Fe2O3 10.2 K2O 0.85 MgO 1.4 Na2O 0.37 P2O5 1.2 TiO 1.3 BaO 0.19 MnO 0.15 SrO 0.16 SO3 0.22 ZrO2 -

Loss On Ignition (LOI) 1.7

6.2.2 Specimens Details

All specimens had dimension of 38 mm x 73 mm (diameter x height) and cast in

PVC moulds. The ratios of Na2SiO3 to NaOH were 2.5, 1.75 and 1.0. Alkaline solution

to fly ash ratio was kept constant at 0.4. Further specimens with 0.57 alkaline solution

to fly ash ratio and 2.5 Na2SiO3 to NaOH ratio were made to investigate the effect of

saturation and dry conditions of high strength geopolymer. The geopolymers were

prepared by using a mortar mixer for 3 minutes. After mixing, the geopolymer mix was

cast in moulds and vibrated on vibrating table for at least 5 minutes to ensure that there

are no voids in the specimens. The specimens were then cured in moulds in oven for

24 hours at a temperature of 60oC and 100% relative humidity.

6.2.3 Drying and Saturation Method

Cured specimens were removed from their moulds and top of the specimens

were ground flat by using grinding machine because saturated specimens need be to

test for compressive strength whilst it is fully saturated. The specimens were ground

before being saturated in order to measure the amount of water absorbed by the

specimens. The heights of the specimens were marginally reduced by not more than

0.4 mm when the specimens were ground.

Preparation of Saturated Specimens: The ground specimens were then placed in a

vacuum machine at -0.08 MPa pressure for 24 hours to dry. Dried specimens weights

were measured and then they were immersed in water. Immersed specimens with

106

water were then placed in a different vacuum machine with -0.08 MPa pressure for

another 24 hours to ensure that the saturation process is complete. The fully saturated

specimens’ weight was measured to calculate the amount of water absorbed by the

specimen.

Preparation of Dry Specimens: The dry specimens were prepared by drying in vacuum

machine at -0.08 MPa for 24 hours at room temperature.

6.2.4 Compressive Test

All specimens were tested by using a 250 kN capacity loading machine in

displacement control at a rate of 1 mm/min until failure. Maximum compressive axial

force was recorded. Dry specimens and saturated specimens were tested at the same

time to eliminate the effect of age on specimen strength.

6.2.5 Inductively Couple Plasma Test

In order to test that hypothesis that the effect of saturation in water is due to

leaching of chemicals in the immersing water, the water that was used to immerse the

geopolymer specimens were tested for chemicals that may have leached during the

geopolymer immersion. Inductively couple plasma (ICP) test were conducted on the

water used to immerse the specimens. The water was tested for chemical

concentration in parts per million (ppm) unit.

6.3. Result and Discussion

6.3.1 Compressive Strength

In dry condition, specimens with 0.4 alkaline solution to fly ash ratio produced

average compressive strength of 92.8 MPa, 90 MPa and 74.5 MPa for 2.5, 1.75 and

1.0 sodium silicate to sodium hydroxide ratios respectively. These results

demonstrated the trend that higher strength can be achieved by using higher sodium

silicate to sodium hydroxide ratio.

In saturated condition, the strengths of the specimens were less than the dry

condition. The saturated strengths were 77 MPa, 76 MPa and 64.4 MPa for 2.5, 1.75

and 1.0 sodium silicate to sodium hydroxide ratios respectively. These are 17%, 15.2

and 13.6% less than the dry strengths. Higher strength losses were recorded in higher

107

sodium silicate to sodium hydroxide ratio. Table 2-4 show compressive strength of

geopolymers in dry and saturated conditions.

Table 6.2: Result summary of geopolymers with 0.4 alkaline solution to fly ash ratio and 2.5 Na2SiO3 to NaOH ratio

Condition Specimen

No.

Dry Weight

(g)

Saturated

Weight (g)

Water

Absorption

(%)

Compressive

Strength (MPa)

Dry

1 156.3 - - 94.99

92.76

2 156.5 - - 90.74

3 156 - - 93.63

4 155.9 - - 89.20

5 156.3 - - 95.22

Saturated

1 156.5 162.7 4.0 87.44

76.98

2 156.1 163.3 4.6 77.00

3 157.2 163.8 4.2 84.74

4 155.2 162.4 4.6 66.85

5 150.2 156.7 4.3 68.84

Strength Reduction Percentage (%) : 17.0


Condition Specimen

No.

Dry Weight

(g)

Saturated

Weight (g)

Water

Absorption

(%)

Compressive

Strength (MPa)

Dry

1 154.6 - - 90.85

89.62

2 153.9 - - 85.14

3 154.6 - - 95.23

4 154.5 - - 90.10

5 155.1 - - 86.79

Saturated

1 155.9 162.6 4.3 56.07

75.99

2 155.3 162 4.3 76.30

3 155.9 162.5 4.2 85.54

4 154.7 161.6 4.5 81.58

5 155.2 161.6 4.1 80.48


108


Condition Specimen

No.

Dry Weight

(g)

Saturated

Weight (g)

Water

Absorption

(%)

Compressive

Strength (MPa)

Dry

1 155.5 - - 74.26

74.54

2 156.5 - - 75.52

3 156.8 - - 75.57

4 156.8 - - 70.70

5 155.2 - - 76.64

Saturated

1 155.8 161.5 3.7 68.06

64.40 2 156 161.7 3.7 63.36

3 155.6 161.4 3.7 67.29

4 155.8 161.7 3.8 58.88


The results in Tables 6.2 to 6.4 show that increasing Na2SiO3/NaOH ratio

increases the strength loss due to saturation. Figure 6.1 summarized the strength of

dry and saturated geopolymers. However, it should be noted that increasing Na2SiO3 /

NaOH ratio also coincides with increasing strength. The higher strength geopolymer

are likely to be more brittle which may have been responsible for damage when the

specimens were dried and saturated which forces the specimens to undergo shrinkage

and expansion during this process.

To verify that the strength is not the causal factor, same strength specimens as

Na2SiO3/NaOH ratio =1.0 (74.5 MPa) were made with Na2SiO3 / NaOH ratio = 2.5. This

was achieved by trial and error of alkaline solution to fly ash ratio and it was found a

ratio of 0.57 provided the closest strength of 76 MPa. The results of these tests are

provided in Table 6.5. The strength losses due to saturation in these specimens were

closest to the results in Table 6.2, hence the effect of strength can be eliminated as an

influencing factor.

109


Condition Specimen

No.

Dry Weight

(g)

Saturated

Weight (g)

Water

Absorption

(%)

Compressive

Strength (MPa)

Dry

1 152.6 - - 84.39

76.53

2 148.9 - - 73.97

3 152.3 - - 80.21

4 152.8 - - 64.58

5 152.8 - - 79.48

Saturate

d

1 152.1 159.1 4.6 51.99

64.57

2 151.4 158.6 4.8 77.39

3 152.6 159.5 4.5 56.51

4 151.6 159.1 4.9 62.46

5 151.2 158.6 4.9 74.49


6.3.2 Effect of Strength on Reduction of Strength in Saturated Geopolymer

It is postulated that the strength reduction in saturated samples may be due to

the high brittleness of the higher strength geopolymer. This is possible since drying and

saturation requires expansion and shrinkage which in high brittle materials can cause

damage to the specimen. Higher strength geopolymers demonstrate higher brittleness

as in normal OPC concrete. To investigate the brittleness effect, same strength

specimens were prepared by using the same activator solution with a Na2SiO3/NaOH

ratio of 2.5. Ratio of alkaline solution to fly ash was increased to 0.57 to lower the

specimens’ strength. Sodium silicate to sodium hydroxide ratio was set to 2.5. The

average compressive strength for these specimens was 74.2 MPa. The results from

these specimens show that even with lower strength, the reduction in strength was

similar which was 15.6% as opposed to 17.0% strength reduction for high strength

specimens with 0.4 alkaline solutions to fly ash ratio and 2.5 sodium silicate to sodium

110

hydroxide ratio. This clearly indicates that high strength was not a factor in strength

reduction of geopolymer exposed with saturated condition.

Figure 6.1: Compressive Strength of Dry and Saturated Geopolymer

6.3.3 Excess Sodium Silicate Factor

With the elimination of high strength factor, the only one factor effecting

strength of geopolymer in saturated condition is the amount of sodium silicate in the

mixture. Excess silicate leached out from the specimens when immersed in water

resulting in loss of strength. The leaching out of sodium silicate in the water used to

immerse specimens was observed when the water is more viscous and physically

condition like sodium silicate liquid. Eighty six ppm of sodium silicate was measured

from Inductively Couple Plasma test. Table 6.6 shows the chemical concentrations

measured in immersed water.

Table 6.6: ICP Test Results Chemical Components Concentration in ppm

Na2SiO3 86

CaO 0

FeO 5.78

MgO 5.9

Al2O3 0

Na2O 0

111

6.4 Conclusion

1. There is compressive strength reduction ranging from 13.6% to 17% for

geopolymer exposed to drying and saturation condition. Geopolymer

specimens with 2.5 ratio of Na2SiO3/NaOH and 0.4 alkaline solution to fly ash

ratio provides the highest average compressive strength of 92.8 MPa and the

highest average strength reduction of 17%.

2. Geopolymer specimens with 1.0 ratio of Na2SiO3/NaOH and 0.4 alkaline

solution to fly ash ratio provides the lowest average compressive strength of

74.5 MPa and the lowest average strength reduction of 13.6%.

3. Dry strength of geopolymer is not the influencing factor in the strength

reduction. This was proven by same strength but different Na2SiO3/NaOH ratio

specimens. The trend of increasing Na2SiO3/NaOH with increasing strength

reductions remained for the same strength specimens.

4. The strength reduction is caused by sodium silicate leaching out from

specimens when immersed in water. This conclusion was reached by

eliminating strength as the cause of this and by measuring leaching chemicals

in the immersing water.

112

References


Analysis 35(2): 429-441.



Haynes, H. H. (1976). Seawater absorption and compressive strength of concrete at

Ocean depths. R. S. Highberg and B. A. Nordby. Technical note (Naval Civil

Engineering Laboratory (Port Hueneme, Calif.)). Civil Engineering Laboratory,

Port Hueneme, Calif.

Kim, J. T., Seo, D. S., Kim, G. J. and Lee, J. K. (2010). Influence of water glass content

on the compressive strength of aluminosilicate-based geopolymer. Korean

Journal of Materials Research. 20: 488-493.



40(2): 334-339.

Morley, C. T. (1979). "Theory of pore pressure in concrete cylinders." ACI Materials

Journal(3-4): 7-45.

Raijiwala, O. B. and Patil, H. S. (2010). Geopolymer concrete: A green concrete. 2nd

International Conference on Chemical, Biological and Environmental

Engineering, ICBEE 2010, Cairo.


Journal 80(2): 35.

Sanjayan, J. G. (2010). "Concretes without Cement: Concrete of the Future?" Indian

Concrete Journal 11(3): 29-33.

Standard Australia, L. (2009). AS 3600-2009: Concrete Structure. Section 4: Design for

Durability. Sydney, Australia, SAI Global Limited.

Steffens, A., Li, K. and Coussy, O. (2003). "Aging approach to water effect on alkali-

silica reaction degradation of structures." Journal of Engineering Mechanics

129(1): 50-59.

113




Wang, H., Jin, W. and Li, Q. (2009). "Saturation effect on dynamic tensile and

compressive strength of concrete." Advances in Structural Engineering 12(2):

279-286.

Wittmann, F. H. (1973). "Interaction of Hardened Cement Paste and Water." Journal of

the American Ceramic Society 56(8): 409-415.

Zielinski, A. J., Reinhardt, H. W. and Körmeling, H. A. (1981). "Experiments on

concrete under uniaxial impact tensile loading." Matériaux et Constructions

14(2): 103-112.

114

CHAPTER 7

Hydrocarbon Fire testing of Geopolymer High Strength

Concrete Wall Panels and Cylinders

7.1 Introduction

The needs for concrete structures to withstand fire condition are far more

demanding nowadays with numbers of concrete tunnels, petro-chemical’s plant and

other mega structures are being built all around the world. Due to high usage of

concrete as construction material, it is important to develop high fire resistance

concrete that can sustain structural integrity when exposed to fire event. Spalling

phenomena occurs during fire exposure in Portland cement high strength concrete and

factors influencing the amount of spalling have been extensively discussed in the

previous Chapters 2, 3 and 4.

Geopolymer concrete known as very high fire resistance material due to its ceramic-

like properties (Davidovits and Davidovics, 1991). Geopolymer concrete strength after

elevated temperature exposure can increase due to further geopolymerisation (Pan et

al., 2009). However, there are strength decrease of geopolymer concrete after elevated

temperature exposure especially geopolymer concrete with coarse aggregate. Thermal

incompatibilities due to difference thermal expansions between aggregates and binder

115

contributes to the decline in geopolymer concrete strength (Kong and Sanjayan, 2008).

The results in this chapter represent the performance of geopolymer high strength

concrete panel and cylinder exposed to hydrocarbon fire exposure. The amount of

spalling, temperature of wall panel at various depths, thermal diffusivity and residual

compressive strength after fire exposures are reported in this chapter.


7.2.1 Material

Aluminosilicate source for making geopolymer concrete in this study is fly ash.

The fly ash was sourced from Gladstone power station in Queensland. Gladstone fly

ash was selected because it has the best workability (Section 5.3.1) and most widely

used. Chemical composition of Gladstone fly ash is shown in Table 7.1.

Table 7.1: Chemical composition of Gladstone fly ash Chemical Component Weight (%)

Al2O3 25.56 BaO 0.09 CaO 4.3

Fe2O3 12.48 K2O 0.7 MgO 1.45 MnO 0.15 Na2O 0.77 P2O5 0.885 SO3 0.24 SiO2 51.11 TiO2 1.32 LOI* 0.57 Total 99.62

Sand for the geopolymer high strength concrete was sourced from Langwarrin quarry,

Melbourne. The sand absorbed 0.3% of water. Sand and coarse aggregate need to be

in saturated surface dry condition before mixing to prevent alkaline solution being

absorbed by sand and aggregate. Coarse aggregate maximum diameter was 14 mm

sourced from Oakland Junction, Hornsfel and supplied by ReadymixTM. Coarse

aggregate water absorption was 0.8%. Particle size distribution graphs for coarse

aggregate and sand are shown in Figure 7.1 and 7.2 respectively.

116

Figure 7.1: Aggregate Particle Size Distribution Graph

Figure 7.2: Sand Particle Size Distribution Graph

Sodium silicate and sodium hydroxide were used as alkaline activator for the

geopolymer high strength concrete. D grade sodium silicate was supplied by PQ

Australia. Table 7.2 shows the D grade sodium silicate specification. Sodium hydroxide

used has molarity of 8M. Sodium hydroxide was supplied by Orica Chemical with 99%

purity. Sodium hydroxide of 8.0M solution was made by mixing 26.2% of NaOH solids

and 73.8% of water (Rangan, 2006).

0

20

40

60

80

100

0.075 mm 2.36 mm 4.75 mm 6.7 mm 9.5 mm 13.2 mm

% P

assi

ng

Aggregate Particle Size Distribution Graph

Upper Limit

Aggregate Grading

Lower Limit

0

20

40

60

80

100

Pan 75 um 150 um 300 um 600 um 1.18 mm 2.36 mm 4.75 mm

% P

assi

ng

Sand Particle Size Distribution Graph

Upper Limit

Sand Grading

Lower Limit

117

Table 7.2: Sodium silicate specification Properties Specification

Na2O (% wt) 14.5 – 14.9

SiO2 (% wt) 29.1 – 29.7

Solids (% wt) 43.6 – 44.6

SiO2/ Na2O ratio 1.95 – 2.05

Density @ 20oC (g/cc) 1.50 – 1.53

Viscosity @ 20oC (cps) 250 – 450

7.2.2 Geopolymer Concrete Mix

Targeted compressive strength for geopolymer high strength concrete is 50

MPa. The geopolymer mix was determined by casting geopolymer paste in order to

obtain the optimum ratios of alkaline activator to fly ash and sodium silicate to sodium

hydroxide (Na2SiO3/NaOH). Geopolymer paste was made with alkaline activator to fly

ash ratio of 0.4. The Na2SiO3/NaOH ratio for alkaline solution was 2.5. The workability

of geopolymer paste was measured by using slump test in accordance to ASTM

C230/230M-08 (ASTM, 2010). Relative slump for geopolymer paste was 7.7 (Section

5.3.1).The geopolymer paste was moulded in a 35 mm (diameter) x 70 mm (height)

cylinder and cured for 24 hours at 60oC temperature in an oven. Average compressive

strength of geopolymer cylinders was 92.8 MPa. Sand was introduced in geopolymer

paste mix to form mortar. Ratio of sand to fly ash was 1.8. Similar curing regime was

carried out with geopolymer mortar. Average compressive strength for geopolymer

mortar was 74.5 MPa.

Coarse aggregates were then introduced in mortar mix to form geopolymer concrete.

The amount for coarse aggregate was calculated by using specific surface area

calculation. According to Neville (2010 ), specific surface areas for sand and coarse

aggregate with 14 mm diameter maximum size are 6.5 and 0.4 m2/kg respectively.

Based on these values, total sand surface area was calculated based on total sand

weight used in geopolymer mortar trial mix. The area was considered as total

aggregates surface area for geopolymer concrete. The area and weight ratios of

coarse aggregate to sand were 9 and 1.6 respectively. Sand and course aggregates

weights were calculated based on their respective specific surface area values.

Geopolymer concrete was cast in 100 mm (diameter) x 200 mm moulds and cured with

118

similar curing regime as geopolymer paste and mortar. However, the alkaline activator

to fly ash ratio was increased to 0.5 for workability purposes. Average compressive

strength of geopolymer concrete was 64.1 MPa and it can be classified as geopolymer

high strength concrete according to EN 206:2013 (European Standard, 2013).

Geopolymer concrete mixture proportions are presented in Table 7.3.

Table 7.3: Geopolymer concrete mixture proportions Materials Weight (kg/m3)

Coarse Aggregates 1214

Sand 747

Fly ash 415

Sodium Hydroxide 148

Sodium silicate 59

7.2.3 Geopolymer concrete casting

Geopolymer concrete was casted at Commonwealth Scientific and Industrial

Research Organization (CSIRO) laboratory in Highett, Australia. The geopolymer

concrete mixing procedure was according to Hardjito et al., (2004). Sodium silicate and

sodium hydroxide were mixed together for 10 minutes prior to concrete mixing to form

an alkaline activator (Figure 7.3). Geopolymer concrete was mixed in high shear

concrete mixer as shown in Figure 7.4. Dry components of concrete (coarse

aggregates + sand) were mixed for 3 minutes (Figure 7.5). One percent of water was

added into the dry mixture to make the coarse aggregates and sand in saturated

surface dry condition. Fly ash was added and mixed with aggregates for further 3

minutes. Alkaline activator was then added into the mixer for further 5 minutes (Figures

7.6 and 7.7). The fresh geopolymer concrete had a stiff consistency and was glossy in

appearance. The geopolymer concrete was then cast and compacted in 1.075 m x

1.075 m x 0.2 m (thickness), 150 mm (diameter) x 300 mm and 100 mm (diameter) x

200 mm moulds. The cast geopolymer concrete panels and cylinders were cured using

steam curing at 60oC temperature for 24 hours before demoulded. Figures 7.8 – 7.11

show the images of cast geopolymer concrete, geopolymer concrete panel being

steam cured and demoulded. Compressive strength test for geopolymer concrete was

carried out at the age of 1 day, 7 days and 28 days in accordance to AS 1012-2.

119

Figure 7.3: Alkaline activator’s mixer

Figure 7.4: High shear concrete mixer

120

Figure 7.5: Dry components mixing process

Figure 7.6: Alkaline activator added into the mix

121

Figure 7.7: Geopolymer concrete in high shear concrete mixer

Figure 7.8: Geopolymer fresh concrete

122

Figure 7.9: Cast geopolymer concrete panel

Figure 7.10: Geopolymer concrete panel being steam cured

123

Figure 7.11: Demoulded hardened geopolymer concrete panels

7.2.4 Geopolymer concrete properties

Based on the finding in Section 4.3.3, specimen’s size has an effect on

concrete spalling in which larger specimen spalled more than smaller ones. On this

ground, geopolymer concrete was cast in 3 difference sizes; 1.075 m x 10.75 m panel,

150 mm diameter cylinder and 100 mm diameter cylinder. Total of 4 panels, 10

cylinders (150 mm diameter) and 20 cylinders (100mm diameter) were cast for fire test

and compressive test purposes. Figure 7.12 (a-b) shows the illustrations of geopolymer

concrete dimensions. Geopolymer concrete panels and cylinder were tested at the age

between 83 and 86 days. Moisture content of geopolymer concrete panel was

measured 1 hour before fire test. Moisture content was measured by using TramexTM

CME 4 moisture meter. Figure 7.13 shows the moisture meter used to measure

moisture content of concrete wall panel. Table 7.4 summarises the moisture content of,

casting and fire testing date detail for geopolymer concrete.

1.075 m

1.075 m

0.2 m

(a)

124

150 mm

300 mm

100 mm

200 mm

(b) Cylinders

Figure 7.12: Geopolymer concrete dimensions (a) panel (b) cylinders


Table 7.4: Moisture content and testing date details

Moisture

Content

(%)

Casting Date

(dd/mm/yyyy)

Fire Testing Date

(dd/mm/yyyy)

Age During

Fire Test

(days)

Panel 1 (P1) 4.1 02/09/2013 26/11/2013 86

Panel 2 (P2) 4.1 03/09/2013 26/11/2013 85

Panel 3 (P3) 4.0 04/09/2013 26/11/2013 84

Panel 4 (P4) 4.1 05/09/2013 26/11/2013 83

Cylinders N/A 02/09/2013 26/11/2013 86

125

Geopolymer concrete panels were reinforced with mild strength 6 mm diameter steel

bar and 100 mm spacing mesh. Reinforcement bars were located at the middle of the

concrete wall panel thickness. Figure 7.14 shows the reinforcement details for the

panels.

R6-100 cc

100 mm200 mm

1075 mm Figure 7.14: Reinforcement details for geopolymer concrete panel

7.2.5 Hydrocarbon fire test setup


used for exposing the specimens. Standard fire curve is commonly used for testing for

fire exposures in buildings whereas the Hydrocarbon fire test is more suitable for

infrastructure applications where the concrete is likely to be exposed to fire originated

from hydrocarbons (eg. tunnels). More details for the reasons of using hydrocarbon fire

instead of standard fire in accordance to ISO 834:1999 was discussed in Section 3.2.5.

Equation 7.1 is the hydrocarbon fire temperature (BSI, 2005).

𝑇 = 1080(1 − 0.325𝑒−0.167𝑡 − 0.675𝑒−2.5𝑡) + 20 (Equation 7.1)


t = time (minute)

The fire tests were conducted for minimum of 120 minutes. The temperature of the

furnace (near exposed surface) were measured and recorded by the furnace’s

temperature sensor to the data logger.

Geopolymer concrete panels and cylinders were tested simultaneously. Panels and

cylinders furnace setup illustration and image are shown in Figure 7.15 and 7.16

respectively.

Thermocouples were installed in geopolymer concrete panel at the depth 25 mm, 50

mm, 75 mm, 100 mm, 125 mm, 150 mm, 175 mm and 200 mm measured from fire

126

exposed surface. Temperatures at exposed (0 mm) and unexposed surface (200 mm)

were also recorded.

The weight of the specimens were measured before and after the specimens were

exposed to fire. The weight difference is considered as the weight of spalled concrete.

The loss of concrete wall panel moisture content was also considered as part of spalled

concrete weight.

Furnace

Panel

1.075 m1.0 m

Specimen’s rig

Cylinders

Figure 7.15: Geopolymer concrete panels and cylinders furnace setup illustration

Figure 7.16: Images of panels (left) and cylinders (right) hydrocarbon fire test setup

127

7.2.6 Thermal diffusivity

Thermal diffusivity is a thermal conductivity divided by the volumetric heat

capacity measured in mm2/s. It is a material specific thermal property which describes

how quickly a material reacts to a change in temperature. It is also a measurement of

thermal inertia of solids. The high thermal diffusivity in a material means the material

has high heat propagation rate and it can be classified as good heat conductor

material. Low thermal diffusivity value represents good fire resistant material due to

large heat storage capacity and/or low conductivity (Kaviany, 2011).

Thermal diffusivity for each specimen under fire test was calculated by using the Finite

Difference back calculation method at a depth of 25 mm measured from exposed

surface. Based on the temperature data from the fire test, thermal diffusivity values

were calculated on trial and error basis based on the general equation (Welty, 1974):

Equation 7.2

where K is thermal diffusivity

Central difference equation was used for second spatial derivative,

𝜕2𝑇

𝜕𝑥2 =𝑇𝑖+1

𝑛 −2𝑇𝑖𝑛+ 𝑇𝑖−1

𝑛

ℎ2 Equation 7.3

Where n = temperature at calculated time step

i = temperature at calculated thermal couple location (∆x = 50 mm)

h = total thickness for wall specimen

Forward equation was used for first time derivative,

𝜕𝑇

𝜕𝑡=

𝑇𝑖𝑛+1− 𝑇𝑖

𝑛

∆𝑡 Equation 7.4

where, ∆𝑡 = time interval

Therefore, the final equation for thermal diffusivity calculation was:

tT

xT

12

2

128

𝑇𝑖𝑛+1 = 𝐾∆𝑡 ⌊

𝑇𝑖+1𝑛 −2𝑇𝑖

𝑛+ 𝑇𝑖−1𝑛

ℎ2 ⌋ + 𝑇𝑖𝑛 Equation 7.5

Trial value for thermal diffusivity, K substituted in Equation 7.5 and the correct value of

thermal diffusivity is determined when the temperature, 𝑇𝑖𝑛+1 is calculated from thermal

diffusivity equation equates to the temperature from experimental data.

7.2.7 Compressive strength

Geopolymer concrete cylinders were tested at the age of 1 day, 7 days and 28

days. Geopolymer concrete almost reached full strength after 24 hours of 60oC heat

curing regime. Compressive strength test on cylinder was carried out in accordance to

AS 1012.9-1999.

Unspalled geopolymer concrete panels and cylinders after the fire test were also tested

for residual compressive strength. Compressive strength for fire tested geopolymer

concrete panels were carried out by using ProceqTM SilverSchmidt rebound hammer

(Figure 7.17). Fire tested 150 mm diameter and 100 mm diameter cylinders

compressive strength were tested as according to AS 1012.9-1999 (Standard Australia,

1999).

Figure 7.17: ProceqTM SilverSchmidt rebound hammer

129

7.3 Results and Discussion

7.3.1 Spalling

Figure 7.18 and 7.19 show the geopolymer concrete condition after exposed to

hydrocarbon fire for 2 hours for panels and cylinders respectively. No explosive spalling

noises were heard during the whole fire test period. After the fire test, there were some

minor surface spalling was observed on 2 out of 4 geopolymer concrete panels.

Geopolymer concrete cylinder did not have any spalling at all and all cylinders

remained in their original shape. However, there was a loss in weight for all specimens

due to loss of moisture content. Based on weight loss on unspalled panels, 3.57% -

3.75% of moisture escaped the concrete panel during the fire test. Therefore, it can be

estimated that less than 1% of geopolymer concrete panel spalled when exposed to

hydrocarbon fire. For geopolymer concrete cylinders, since there is not any spalling

observed, average of 2.86% of weight loss was purely the loss of moisture content

inside the concrete. This can be concluded that geopolymer concrete is the better fire

resistant concrete than Portland cement high strength concrete (Section 4.3.1).

Spalling results for all geopolymer concrete panels and cylinders are summarised in

Table 7.5.

Figure 7.18: Geopolymer concrete panels after 2 hours hydrocarbon fire exposure

P

1

P

2 P

3v

P

4

130

Figure 7.19: Geopolymer concrete cylinders after 2 hours hydrocarbon fire exposure

Table 7.5: Geopolymer concrete spalling summary Concrete

Type. No Weight (kg)

Spalling (%) Remark Before Test After Test

Panels

1 538 518 3.72

3.92

No spalling 2 559 533 4.65 3 560 539 3.75 4 561 541 3.57 No spalling

Cylinder (150 mm

dia.)

1 12.48 12.1 3.04

2.84

No spalling 2 12.52 12.18 2.72 No spalling 3 13.08 12.7 2.91 No spalling 4 13.06 12.7 2.76 No spalling 5 12.44 12.1 2.73 No spalling 6 12.48 12.12 2.88 No spalling

Cylinder (100 mm

dia.)

1 3.7 3.6 2.70

2.89

No spalling 2 3.7 3.6 2.70 No spalling 3 3.7 3.58 3.24 No spalling 4 3.68 3.58 2.72 No spalling 5 3.68 3.58 2.72 No spalling 6 3.68 3.56 3.26 No spalling

131

7.3.2 Temperature Results

Figures 7.20 – 7.22 show the geopolymer concrete temperature for every 25

mm depth of panels P1, P3 and P4. Unfortunately, several thermocouples for panel P2

malfunctioned and the data was not usable for any thermal analysis. From these

graphs, it is clear indication that geopolymer concrete has good heat resistant

properties. This can be observed from the temperature at depth more than 100 mm

from exposed surface ranging between 39oC to 45oC after 30 minutes of fire exposure

even though the temperature at exposed surface exceeds 1000oC.

Figure 7.20: Temperature for panel 1 (P1)

0

200

400

600

800

1000

1200

0 20 40 60 80 100 120

Tem

pera

ture

(C)

Time (minutes)

Temp at 0mmTemp at25 mmTemp at50 mmTemp at75 mmTemp at100 mmTemp at125 mmTemp at150 mmTemp at175 mmTemp at200 mm

132



The temperature gradient across panel’s depth was also high which represent the high

volume of heat capacity of geopolymer concrete. Figures 7.23 – 7.25 show the thermal

gradient of geopolymer concrete panel at first 10, 20 and 30 minutes. Figure 7.26

shows the temperature gradient across panel’s depth comparison between geopolymer

0

200

400

600

800

1000

1200

0 20 40 60 80 100 120

Tem

pera

ture

(C)

Time (minutes)


0

200

400

600

800

1000

1200

0 20 40 60 80 100 120

Tem

pera

ture

(C)

Time (minutes)


133

concrete and Portland cement concrete. It can be observed that temperature for

Portland cement HSC panels were higher compared to geopolymer concrete. It is

postulated that heat capacity of geopolymer concrete is higher than Portland cement

HSC. Furthermore, due to non-spalling nature of geopolymer concrete when exposed

to fire, the thickness of geopolymer concrete panels remain and thus provide more heat

resistance compared to spalled Portland cement HSC.

Figure 7.23: Temperature across thickness for panel 1

0

200

400

600

800

1000

1200

0 50 100 150 200

Tem

pera

ture

(C)


10 minute

20 minute

30 minute

134



0

200

400

600

800

1000

1200

0 50 100 150 200

Tem

pera

ture

(C)


10 minute

20 minute

30 minute

0

200

400

600

800

1000

1200

0 50 100 150 200

Tem

pera

ture

(C)


10 minute

20 minute

30 minute

135

Figure 7.26: Temperature across thickness comparison between geopolymer concrete

and Portland cement HSC

7.3.3 Thermal Diffusivity

Using the temperature data presented in Section 7.3.2, thermal diffusivity of

concrete was calculated at various temperatures using the method described in

Section 7.2.6. The calculation was based on temperature at the depth of 25 mm from

fire exposed surface. From the diffusivity graphs (Figure 7.23), the phenomenon as

described in Section 3.3.3 was observed where a drop in diffusivity values when the

temperature reaches between 105oC and 120oC. This phenomenon is due to water

changing phase and the thermal capacity drop at this point. At this temperature, water

in the concrete absorbs the heat to transform from liquid phase to steam.

Figure 7.27 also compares the thermal diffusivity of geopolymer concrete

panels and thermal diffusivity of same specimen’s size Portland cement high strength

concrete (HSC) with 14 mm maximum aggregate diameter. From the comparison, it is

clear that diffusivity of geopolymer concrete panel is lower than Portland HSC. Lower

diffusivity value is part of indication that geopolymer concrete has high heat storage

capacity. It makes geopolymer concrete more heat resistant than Portland HSC which

contributes to less spalling when exposed to hydrocarbon fire. The other factor that

may contribute to less spalling of geopolymer concrete is geopolymer binder itself.

Unlike hydration reaction in Portland cement which forms calcium-silicate-hydrate gel,

0

200

400

600

800

1000

1200

0 50 100 150 200

Tem

pera

ture

(C)


Geopolymer HSC (10 minutes)



Portland Cement HSC (10 minutes)



136

geopolymer is an aluminate and silicate reaction. It is postulated that geopolymer as a

binder does not shrink when exposed to fire and thus the reduced spalling amount.

Figure 7.27: Thermal diffusivity vs Temperature

7.3.4 Residual Compressive Strength After Fire Exposure

Geopolymer concrete achieved its strength at very early age. Under curing

regime of 60oC steam curing for 24 hours, the compressive strength of geopolymer

concrete was 54.1 MPa. Table 7.6 summarises density the compressive strength of

geopolymer concrete at the age of 1 day, 7 days and 28 days.

Table 7.6: Compressive Strength and Density Summary

Age No. Weight (g)

Density (kg/m3)

Compressive Strength (MPa)

1 day 1 3804 2422

2419 51.5

54.09 2 3808 2424 55.0 3 3789 2412 55.8

7 days 1 3808 2424

2430 56.2

56.06 2 3820 2432 56.0 3 3823 2434 56.0

28 days

1 3828 2437 2439

56.3 58.94 2 3840 2445 60.3

3 3827 2436 60.2 All geopolymer concrete panels and cylinders remained intact after 2 hours of

hydrocarbon fire exposure. Therefore, residual compressive strength of geopolymer

concrete could be measured. Residual compressive strength of a structure after fire

0.000.200.400.600.801.001.201.401.601.80

40 60 80 100 120 140 160 180 200 220

Diff

usiv

ity (m

2s-

1 ) x

10-

6

Temparature (C)

Diffusivity vs TemparaturePortland Cement HSC

Geopolymer HSC (1)

Geopolymer HSC (3)

Geopolymer HSC (4)

137

event is an important factor in assessing the safety and integrity of whole structure. The

decision for appropriate repair works on structure after fire is based on a reliable

assessment of residual strength (Annerel and Taerwe, 2009). Table 7.6 summarises

the average residual compressive strength of geopolymer concrete panel, 150 mm

diameter cylinder and 100 mm diameter cylinder. From Table 7.7, it can be seen that

geopolymer concrete panel has average residual compressive strength which was the

highest (35.1 MPa). Cylinders with 150 mm diameter and 100 mm diameter have

residual compressive strength of 13.3 and 6.7 MPa respectively. These results show

that geopolymer concrete is structurally viable even after fire exposure. The specimen

size effect was also exhibited with larger specimen’s size showing lesser strength loss

due to fire exposure (Li et al., 2004). Figure 7.28 shows the images of fire tested

geopolymer concrete cylinders after compressive strength test.

Table 7.7: Residual compressive strength summary

Conrete Type Residual Compreesive Strength (MPa)

Panel (exposed surface)

P1 38.9

38.5 P2 35.6 P3 34.3 P4 45.2

Cylinder (150 mm diameter)

12.1 13.3 14.3

13.5

Cylinder (100 mm diameter)

7.9 6.7 8.2

4.1

138

Figure 7.28: Fire tested geopolymer concrete cylinders 150 mm diameter (left) and 100

mm diameter (right) after compressive strength test

7.4 Conclusions

1 There were no explosive spalling observed during the whole length of

hydrocarbon fire test on geopolymer specimens. Panel 2 (P2) recorded the

highest geopolymer concrete spalling with 4.65% spalling percentage.

However, it is estimated that 3.57% - 3.75% of the weight loss was due to the

loss of moisture content during fire test. The estimation was based on loss of

weight for unspalled panels (P1 and P4). Thus, the spalling percentage of

geopolymer concrete was actually less than 1%. There was no spalling

observed on geopolymer concrete cylinders even though there were 2.84% and

2.86% weight loss for 150 mm diameter and 100 mm diameter respectively.

The weight loss was due to loss of moisture during fire test.

2 The diffusivity values for geopolymer concrete were lower than diffusivity values

for Portland cement high strength concrete (HSC). This is due to high

volumetric heat capacity of geopolymer concrete which can be related to better

heat resistant than Portland cement HSC.

3 There was residual compressive strength on geopolymer concrete after

hydrocarbon fire test. The fire tested geopolymer concrete panel, 150 mm

139

diameter cylinder and 100 mm diameter cylinder have 38.5, 13.3 and 6.7 MPa

compressive strength respectively. It proved that geopolymer concrete

structurally can withstand load even after being exposed to 2 hours of

hydrocarbon fire exposure. It can be concluded that geopolymer concrete is

good fire resistant construction material.

140

References

Annerel, E. and Taerwe, L. (2009). Approaches for the assessment of the residual

strength of concrete exposed to fire. Concrete Repair, Rehabilitation and

Retrofitting II - Proceedings of the 2nd International Conference on Concrete

Repair, Rehabilitation and Retrofitting, ICCRRR.

ASTM (2010). C230/C230M – 08: Standard Specification for Flow Table for Use in

Tests of Hydraulic Cement. West Conshohocken, Pennsylvania 19428-2959,

United States, ASTM International.



Standardization.




European Standard (2013). IS EN 206:2013 Concrete - Specification, performance,

production, and conformity. Brussel, Belgium, European Committee for

Standardization



101(6): 467-472.

Kaviany, M. (2011). Essential of Heat Transfer: Principals, Materials and Application.

Cambridge University Press, New York, USA.



Composites 30(10): 986-991.

Li, M., Qian, C. and Sun, W. (2004). "Mechanical properties of high-strength concrete

after fire." Cement and Concrete Research 34(6): 1001-1005.

Neville, A. M. (2010 ). Concrete technology J. J. Brooks. Prentice Hall Harlow, England

; New York : .

141





Journal 80(2): 35.




Welty, J. R. (1974). Engineering heat transfer. New York : Wiley, New York.

142

CHAPTER 8

Investigation on Aerated Geopolymer Wall Panels for Fire

Applications

8.1 Introduction

Concrete is susceptible to a phenomenon termed spalling in fire. Spalling of

concrete in fire is dislodgement of small pieces of concrete popping out from the

surface of concrete, often explosive in nature. Spalling occurs in the initial stages of the

fire, i.e., within 15 to 30 minutes (Sanjayan and Stocks, 1993; Crozier and Sanjayan,

2000). Spalling results in rapid loss of the surface layers of the concrete columns

exposing the steel reinforcement, which quickly loses strength when exposed to fire.

While many aspects of spalling behaviour of concrete are understood, some aspects of

the influence of various parameters in fire remain not fully understood. Further,

consensus has not been reached on the basic mechanism that causes the spalling.

Some research works indicate free water, vapour pressure and moisture gradients in

the concrete as main reason for explosive spalling (Anderberg, 1997; Hertz, 2003;

Kanema et al., 2011; Khaliq and Kodur, 2011). However, findings from Chapter 3 prove

that maximum stress that corresponds to maximum saturated steam temperature is low

and hence cannot to be considered as a main factor causing spalling.

The thermal incompatibility between the concrete binder and aggregates is the most

likely cause of strength loss in concrete specimens at elevated temperatures. The rate

143

of expansion of the aggregate with temperature is an influential factor in the

performance of concrete under elevated temperatures (Pan et al., 2009; Kong and

Sanjayan, 2010). The aggregate size is shown to have a significant effect on spalling of

concrete at high temperatures based on fracture process zone. Microcracks

accumulate around aggregates crack tip and a fracture process zone (lp) is created

along the main crack face (Pan et al., 2012).

Although geopolymer is chemically stable in high temperatures, the geopolymer

concrete deteriorates in fire due to the incompatibility between aggregates and the

binder. Therefore, in these trials, the aggregates were removed from the concrete mix

so that the aggregate effect can be eliminated. The results in this chapter represent the

performance of aerated geopolymer panels in room temperature and high temperature.

The room temperature properties such as bearing load capacity, axial load capacity,

corner bearing strength and flexural strength are reported in this chapter. Under high

temperature tests, spalling percentage, temperature in the panel and temperature

gradient when panels exposed to hydrocarbon fire are also presented and the

suitability of the panel to be used as fire resisting construction material was concluded.


8.2.1 Material

Aluminosilicate source for aerated geopolymer panel was fly ash. The fly ash

was sourced from Gladstone power station in Queensland. Gladstone fly ash was

selected because it has the best workability (Section 5.3.1). Chemical component of

Gladstone fly ash is shown in Table 8.1. Sand for the panels was sourced from

Langwarrin quarry, Melbourne. The sand was measured to have an absorption rate of

0.3% of water. Sand and coarse aggregate need to be in saturated surface dry

condition before mixing to prevent alkaline solution being absorbed by sand and

aggregate. Particle size distribution graph for sand is shown in Figure 8.1.

144

Table 8.1: Chemical component of Gladstone fly ash Chemical Component Weight (%)

Al2O3 25.56 BaO 0.09 CaO 4.3

Fe2O3 12.48 K2O 0.7 MgO 1.45 MnO 0.15 Na2O 0.77 P2O5 0.885 SO3 0.24 SiO2 51.11 TiO2 1.32 LOI* 0.57 Total 99.62

Figure 8.1: Sand Particle Size Distribution Graph

Sodium silicate and sodium hydroxide were used as alkaline activator for the aerated

geopolymer panel. Grade D sodium silicate was supplied by PQ Australia. Table 8.2

shows the D grade sodium silicate specification. Sodium hydroxide used has molarity

of 8M. Sodium hydroxide was supplied by Orica Chemical with 99% purity. Sodium

hydroxide of 8.0M solution was made by mixing 26.2% of NaOH solids and 73.8% of

water (Rangan, 2006). Aluminium powder was used to make the concrete aerated.

0

20

40

60

80

100

Pan 75 um 150 um 300 um 600 um 1.18 mm 2.36 mm 4.75 mm

% P

assi

ng

Sand Particle Size Distribution Graph

Upper Limit

Sand Grading

Lower Limit

145

Aluminium powder reacts with water to form hydrogen gas. The hydrogen gas foams

and doubles the volume of the raw mix creating gas bubbles up to 3 mm in diameter. At

the end of the foaming process, the hydrogen escapes into the atmosphere and is

replaced by the surrounding air. This is the process of producing aerated geopolymer

concrete (Wongkeo et al., 2012).

Table 8.2: Sodium silicate specification Properties Specification

Na2O (% wt) 14.5 – 14.9

SiO2 (% wt) 29.1 – 29.7

Solids (% wt) 43.6 – 44.6

SiO2/ Na2O ratio 1.95 – 2.05

Density @ 20oC (g/cc) 1.50 – 1.53

Viscosity @ 20oC (cps) 250 – 450

8.2.2 Aerated Geopolymer Panel Properties

There were 3 different concrete specimens cast for various purposes. Panel

specimens for bearing load test, axial load test and flexural load test were 2.7 m

(height) x 0.925 m (wide) x 0.1 m (thick). Panel specimens for corner bearing test was

1.35 m (height) x 0.925 m (wide) x 0.1 m (thick) in size. For hydrocarbon fire test,

panel’s dimension was 0.925 m (height) x 1.075 m (wide) x 0.1 m (thick). There was 5

mm thick of crust on one surface for all panels. The crust is geopolymer without the air

bubbles so it is denser than the rest of the aerated panel. Cross section of panel

illustration and image are shown in Figure 8.2.

Aerated geopolymer panels were reinforced with mild strength 6 mm diameter steel bar

and 100 mm spacing mesh. Reinforcement bars were located at the middle of the

panel thickness. Figure 8.3 shows the reinforcement details for the panels.

146

Smoothed Surface

Aerated geopolymer

panel

Crust

0.1 m

5 mm

Figure 8.2: Aerated geopolymer panel cross section

R6-100 cc

50 mm100 mm

0.925 m Figure 8.3: Reinforcement details for aerated geopolymer panel

8.2.3 Room Temperature Test

Overall density of panel was measured based on total weight of the panel and

its volume. Moisture content of aerated geopolymer panel was measured 1 hour before

fire test. Moisture content was measured by using TramexTM CME 4 moisture meter.

Figure 8.4 shows the moisture meter used to measure moisture content of concrete

wall panel.

147


8.2.3.1 Bearing Load Test

This test was conducted to estimate the eccentric axial load capacity of the

panels when the panel is loaded vertically as shown in Figure 8.5. The load was

applied on the top with a 30 mm wide bearing strip with a load eccentricity of 25 mm

from the centre of the panel away from the crust. The bearing strip is a steel plate with

a thickness of 20 mm. The bottom of the panel was supported by a Granor rubber of

100 mm wide and 20 mm thick for the full width of the test panel. The reaction at the

bottom is expected to be distributed with the resultant being in line with the load at the

top. The load was applied under displacement control at the rate of 1 mm/min. The

load was applied until peak load is achieved.

30 mm

P

10 mm

Crust

Figure 8.5: Bearing load test setup

148

8.2.3.2 Axial Load Test

The axial load test was conducted to estimate the concentric axial load capacity

of the panels when the panel is loaded vertically as shown in Figure 8.6. The load was

applied on the top with a 100 mm wide bearing strip with a load eccentricity of 5 mm

from the centre of the panel away from the crust. The 5 mm (0.05D) is the minimum

eccentricity for loads recommended by AS3600 (Standard Australia, 2009). The

bearing strip is a steel plate with a thickness of 20 mm, i.e. two plates of 10 mm each.

The bottom of the panel was supported by a Granor rubber of 100 mm wide and 20

mm thick. The reaction at the bottom is expected to be distributed with the resultant

being in line with the load at the top. The load was applied under displacement control

at a rate of 1 mm/min. The load was applied until peak load is achieved.

55 mmP

Crust

Figure 8.6: Axial load test setup

8.2.3.3 Corner Bearing Test

Corner bearing tests were performed with a bearing pad at the top of 180 mm x

100 mm as shown in Figure 8.7. The panels tested were 1.35 m high and 0.925 m

wide. After the test was completed, the bearing pad was moved to the other corner and

the test was repeated. Therefore, there were four tests performed on 2 panels.

149

925 mm

1350 mm

180 mm

P

Figure 8.7: Corner bearing test setup

8.2.3.4 Flexural Load Test

Six flexural strength tests were conducted in a four point bending arrangement

as shown in Figure 8.8 with a span of 2.4 m. Three of the tests were conducted with

the crust on top and other three with the crust at the bottom. The load was applied with

a 100 mm Granor rubber and a steel square hollow section (SHS). The SHS wall

thickness is 7 mm. The load was applied under displacement control with a loading rate

of 1 mm/min.

2700 mm

800 mm 800 mm 800 mm

P/2 P/2

Figure 8.8: Flexural load test setup

150


8.2.4.1 Test setup


used for exposing the specimens. The reason of using hydrocarbon fire instead of

standard fire in accordance to ISO 834:1999 was discussed in Section 3.2.5. Equation

8.1 is the hydrocarbon fire temperature (BSI, 2005) and temperature versus time curve

for hydrocarbon fire is shown in Figure 8.9.

𝑇 = 1080(1 − 0.325𝑒−0.167𝑡 − 0.675𝑒−2.5𝑡) + 20 (Equation 8.1)


t = time (minute)

Figure 8.9: Hydrocarbon temperature versus time curves

The fire tests were conducted for minimum of 120 minutes. The temperature of the

furnace (near exposed surface) were measured and recorded by the furnace’s

temperature sensor to the data logger.

Thermal couples were installed in geopolymer concrete panel at the depth 25 mm, 50

mm and 75 mm measured from fire exposed surface. Temperature at exposed (0 mm)

and unexposed surface (100 mm) were also recorded.

The weight of the specimens were measured before and after the specimens were

exposed to fire. The weight difference is considered as the weight spalled concrete.

0

200

400

600

800

1000

1200

0 30 60 90 120 150 180

Tem

pera

ture

(o C)

Time (min)

Hydrocarbon Temperature vs Time Curves

151

The loss of concrete wall panel moisture content was also considered as part of spalled

concrete weight.

Furnace

Panel

1.075 m1.0 m

Specimen’s rig

Figure 8.10: Aerated geopolymer concrete panels furnace setup illustration

Figure 8.11: Image of aerated geopolymer panels fire test setup

152

8.3 Results and Discussion

Average density of aerated geopolymer panels was 1246 kg/m3. Weight and

density for panels are summarised in Table 8.3

Table 8.3: Panels weight and density summary Panel Dimension Weight

(kg)

Volume

(m3)

Density

(kg/m3)

2.7 m (h) x 0.925 m (w) x

0.1 m (t) 315 0.25 1260

1246 1.35 m (h) x 0.925 m (w) x

0.1 m (t) 156 0.125 1248

1.075 m (h) x 0.925 m (w)

x 0.1 m (t) 123 0.1 1230

8.3.1 Room Temperature Test

8.3.1.1 Bearing Load Test

All three panels for bearing load test failed in the same mode of failure, i.e.,

failure of the edge. The failures were localised at the top and did not involve the steel

reinforcements or the rest of the panel. The failure is controlled by the shear and

tensile strengths of concrete. Photos of failure at the edge of the panels due to shear

and tensile strengths of concrete are shown in Figure 8.15. Average bearing capacity

for the panels is 244 kN. The results of the test are shown in Table 8.4. Load versus

displacement graphs for panels are shown in Figures 8.12 – 8.14.

Table 8.4: Results of Bearing Load Tests Panel Cast Date Test Date Max Load (kN)

1 21/3/2013 14/6/2013 296 244 2 23/4/2013 14/6/2013 190

3 22/4/2013 14/6/2013 247

153

Figure 8.12: Bearing load vs displacement graph for panel 1



0

50

100

150

200

250

300

350

0 1 2 3 4 5 6 7

Load

(kN

)

Displacement (mm)

Middle

Quarterfrom bottomQuarterfrom top

0

50

100

150

200

250

300

0 1 2 3 4 5 6 7

Load

(kN

)

Displacement (mm)

Middle

Quarter frombottom

Quarter fromtop

050

100150200250300350

-1 0 1 2 3 4 5 6 7

Load

(kN

)

Displacement (mm)

Middle

Quarter frombottomQuarter fromtop

154

Figure 8.15: Photos of panels’ shear failure after bearing load test

8.3.1.2 Axial Load Test

The Panel 1 failed by buckling of the wall as shown in the Figure 8.19. The

panels 2 and 3 had local failures similar to the bearing load tests failures except the

failures were localised at the bottom (Figure 8.20). The panel 1 had a small edge

failure around 420 kN, however, the failure was arrested from progressing and the final

failure occurred by buckling of the wall. This is likely to be due to some locally stronger

shear/tensile strength of the panel at the bottom. For the calculation of the variations,

the first minor failure observed at 420 kN will be taken. Average axial load capacity for

the panels is 408 kN. Figures 8.16 – 8.18 show load versus displacement graphs for

all panels. The results of the test are shown in Table 8.5.

155

Figure 8.16: Axial load vs displacement for panel 1


0

100

200

300

400

500

600

700

800

-1 0 1 2 3 4 5 6 7 8 9 10 11 12 13

Load

(kN

)

Displacement (mm)

Middle

Quarter frombottom

Quarter fromtop

Local failure at the bottom commenced at 420 kN but the failure did not progress.

0

100

200

300

400

500

600

700

800

-1 0 1 2 3 4 5 6 7

Load

(kN

)

Displacement (mm)

Middle

Quarter frombottom

Quarter fromtop

156


Table 8.5: Results of Axial Load Tests Panel Cast Date Test Date Max Load (kN)

1 17/4/2013 3/6/2013 420 (778) 408 2 16/4/2013 4/6/2013 363

3 18/4/2013 4/6/2013 442

Figure 8.19: Buckling failure of panel 1

0

100

200

300

400

500

600

700

800

-1 0 1 2 3 4 5 6 7

Load

(kN

)

Displacement (mm)

Middle

Quarter frombottom

Quarter fromtop

157

Figure 8.20: Localised failure at bottom of panels after axial load test

8.3.1.3 Corner Bearing Test

Corner bearing load test failure had local failures similar to the axial load tests

failures (Figure 8.21). The failure is controlled by the shear and tensile strengths of

concrete. Average corner bearing capacity of the panels was 8.7 MPa. Results of

corner bearing tests are presented in Table 8.6.

Table 8.6: Results of Corner Bearing Tests Panel (Test #) Cast Date Test Date Max Load

(kN) Max Bearing Stress

(MPa) 1(1) 19/3/2013 19/6/2013 170 9.4

8.7 1(2) 19/3/2013 19/6/2013 157 8.7 2(3) 10/4/2013 19/6/2013 150 8.3 2(3) 10/4/2013 19/6/2013 147 8.2

158

Figure 8.21: Photos of panels after corner bearing test

8.3.1.4 Flexural Load Test

The flexural load versus displacement for panels with Crust up and Crust down

are shown in Figures 8.22 and 8.23 respectively. The first cracks were determined

when the first load drop was observed. The strength of concrete in the extreme fibre

location was assumed to be the flexural strength. Most of the cracks occur at mid-span

and near mid-span as shown in Figure 8.24. Average flexural strength recorded for

aerated geopolymer concrete was 1.4 MPa. The flexural strength of the panels

presented in Table 8.7 was calculated using the load at first crack. There is no

significant difference between Crust down and Crust up. Therefore, the all six samples

are considered together.

159

Figure 8.22: Flexural load versus displacement for panels with Crust up

Figure 8.23: Flexural load versus displacement for panels with Crust down

Table 8.7: Results of Flexural Load Tests Panel Cast Date Test Date Crust Load at

1st crack (kN)

Max Load (kN)

Flexural strength of the concrete

(MPa) 1 6/3/2013 13/5/2013 Up 4.28 11.24 1.1

1.4

2 5/3/2013 13/5/2013 Up 4.95 8.93 1.3 3 4/3/2013 13/5/2013 Up 5.43 10.75 1.4 4 13/3/2013 14/5/2013 Down 6.79 7.91 1.8 5 12/3/2013 14/5/2013 Down 4.94 7.37 1.3 6 14/3/2013 14/5/2013 Down 5.46 8.75 1.4

0

2

4

6

8

10

12

0 20 40 60 80 100

Load

Displacement (mm)

Panel 1

Panel 2

Panel 3

0

2

4

6

8

10

12

0 20 40 60 80 100

Load

Displacement (mm)

Panel 4

Panel 5

Panel 6

160

The following calculation shows the method that was used to calculate the flexural

strength for Panel 1.

The Maximum Bending Moment, M = (P/2)*0.8 = (4.28/2)*0.8 = 1.712 kNm

The flexural strength, ft = M/Z

Where Z = bD2/6 = 925*1002/6 = 1541.67 x 103 mm4

ft = 1.712*106 /1541.67 x 103 = 1.1 MPa

The maximum load recorded is sensitive to the position and strength of the steel

reinforcement. It is less sensitive to the compressive strength of the concrete and

entirely unrelated to the flexural strength of concrete. The variability in maximum load

observed is likely to be due to the variability in the position of the steel reinforcement.

To demonstrate this, a sensitivity analysis is presented in Table 8.8 that shows the

calculated maximum load using AS3600 formulae for three positions of steel

reinforcements, namely, in the middle, 50 mm (where it is supposed to be), and 40 mm

and 60 mm (where a +/- 10 mm movement that can occur during construction). The

strength of steel is likely to be between 550 MPa and 600 MPa since the characteristic

strength of steel is 500 MPa. The compressive strength of concrete is likely to be

around 10 MPa, but both 20 MPa and 30 MPa results are also shown to indicate the

insensitivity of the maximum load to compressive strength of concrete. The steel

reinforcements mesh used is SL62 - 6 mm bars at 200 mm spacing.

Table 8.8: Calculated Maximum Loads for various d, f’c and fy fy = 600 MPa fy = 550 MPa

da (mm) f’ca (MPa) 40 50 60 40 50 60

10 7.28 9.38 11.48 6.76 8.68 10.61 20 7.84 9.94 12.04 7.23 9.15 11.08 30 8.03 10.13 12.23 7.39 9.31 11.24

aReinforcement depth, d (mm); Concrete strength, f’c (MPa)

161

Figure 8.24: Photos of panels’ cracks after flexural load test


8.3.2.1 Spalling

Figure 8.25 shows the aerated geopolymer panel condition after exposed to

hydrocarbon fire for 30 minutes. The test was shortened for 30 minutes due to findings

in Chapter 3, 4 and 7 that spalling occurs only within the first 30 minutes. Average of

4.7% of weight loss was recorded for the panels and moisture content of the panels

ranging between 3.8% - 4.1%. Table 8.9 presents the summary of moisture content

and panels spalling. No explosive spalling observed during the fire test period. Even

though there was no spalling observed, the surfaces that exposed to fire show that the

aerated geopolymer concrete softened when exposed to hydrocarbon fire. This is

because the geopolymer has a softening temperature around 680oC where it is shown

to flow (Pan et al., 2009). This softening is related to the glass transition behaviour of

geopolymer. The temperatures near the surface reached about 1000oC within the first 5

minutes of the fire. The softening of the geopolymer makes it more ductile and prevents

it from spalling. After the fire test, when the geopolymer cools down it regained the

stiffness and strength.

162

Figure 8.25: Aerated geopolymer panels after 30 minutes hydrocarbon fire exposure

Table 8.9: Aerated geopolymer panel spalling summary

No Weight (kg) Moisture

Content (%) Weight Loss (%) Before Test After Test

Panels

1 125 118 4.1 5.6

4.7 2 122 117 3.8 4.1 3 122 117 3.8 4.1 4 123 117 3.8 4.9

8.3.2.2 Temperature Results

Figure 8.26 shows the aerated geopolymer panel temperature for every 25 mm

depth of panel. From this graph, it is a clear indication that aerated geopolymer panel

has a high heat resistant due to low density and cellular structure of the panel.

Temperature increment at depth more at 25 mm from exposed surface shows high

increment after 15 minutes of fire test.

163

Figure 8.26: Temperature for aerated geopolymer panel

The temperature gradient across panel’s depth was high due to the low density, high

porosity and cellular structure of the concrete itself. Figure 8.27 shows the temperature

profile (and thermal gradient) of aerated geopolymer panel at first 10, 20 and 30

minutes of fire exposure. This shows that the aerated geopolymer concrete wall is a

very effective insulating wall without any spalling observed. The use of aeration instead

of aggregates avoids any thermal incompatibility between aggregates and paste

problem hence the risk of spalling is reduced.


0

200

400

600

800

1000

1200

1400

0 5 10 15 20 25 30

Tem

per

atu

re (o

C)

Time (min)

Temp. at 0 mm

Temp. at 25 mm

Temp. at 50 mm

Temp. at 75 mm

Temp. at 100 mm

0

200

400

600

800

1000

1200

0 25 50 75 100

Tem

pera

ture

(C)


10 minute

20 minute

30 minute

164

8.4 Conclusions

1 Under room temperature tests, aerated geopolymer concrete panel bearing

load capacity, axial load capacity, corner bearing strength and flexural strength

are 244 kN, 408 kN, 8.7 MPa and 1.47 MPa respectively. The average density

of the panel was 1246 kg/m3.

2 No spalling was observed during 30 minutes of hydrocarbon fire exposure.

Average weight loss recorded for the panel was 4.7%. The weight loss is

inclusive of the 3.8% - 4.1% moisture content loss. It proves that thermal

incompatibility between coarse aggregate and binder is the main factor of

spalling in concrete with coarse aggregates.

3 Aerated geopolymer concrete is very suitable to be used as wall panel and can

be classified as a fire resistant construction material due to the load capacity

under room temperature and performance when exposed to hydrocarbon fire

exposure.

165

References






Standardization.

Crozier, D. A. and Sanjayan, J. G. (2000). "Tests of load-bearing slender reinforced

concrete walls in fire." ACI Structural Journal 97(2): 243-251.


38(2): 103-116.

Kanema, M., Pliya, P., Noumowe, A. and Gallias, J. L. (2011). "Spalling, Thermal, and

Hydrous Behavior of Ordinary and High-Strength Concrete Subjected to

Elevated Temperature." Journal of Materials in Civil Engineering 23(7): 921-

930.



394-402.



40(2): 334-339.








Journal 80(2): 35.

166



Standard Australia, L. (2009). AS 3600-2009: Concrete Structures. Design Properties

of Materials. Sydney, Australia, SAI Global Limited. AS 3600-2009.

Wongkeo, W., Thongsanitgarn, P., Pimraksa, K. and Chaipanich, A. (2012).

"Compressive strength, flexural strength and thermal conductivity of autoclaved

concrete block made using bottom ash as cement replacement materials."

Materials & Design 35: 434-439.

167

CHAPTER 9

Summary, Conclusions and Recommendations

9.1 Summary

The performance of geopolymer concrete when exposed to hydrocarbon fire is

presented in this thesis. Full scale Portland cement high strength concrete wall panels

with the dimensions of 3.36 m (h) x 3.38 m (w) x 0.2 m (t) were fire tested under

hydrocarbon fire exposure for 2 hours. The results of these full scale tests are

presented in Chapter 3.

Further investigations of the specimen’s size, aggregate size and aggregate type effect

on the spalling of Portland cement high strength concrete were carried out. These

results are presented in Chapter 4. Variations of specimen’s sizes were tested by

testing large scale walls to small scale cylinders. For this work, 3.4 m square walls, 1 m

square walls, 400 mm square columns and 100 mm cylinders were tested. Aggregate

size effect was also tested by using 7 mm, 14 mm and 20 mm maximum aggregate

sizes. Two types of aggregates were tested, namely, basalt and granite types. A new

method of spalling parameter was introduced in this chapter namely spalling nominal

depth. Nominal depth is a parameter which represents the depth of spalled concrete

based on the amount of concrete spalling and the area of the specimen exposed to fire.

The severity of spalling for Portland cement high strength concrete was determined

and the next step of experimental work was to produce a concrete which have a good

fire resistance. Therefore, geopolymer concrete was selected due to its ceramic-like

168

properties. In Chapter 5, the fly ash which is the aluminosilicate source for geopolymer

concrete was determined based on workability judgement. Three types of fly ash from

three different sources were selected which are fly ash from Gladstone power station,

Tarong power station and Microash. Further investigations on fly ash physical and

chemical properties effect on fresh geopolymer workability were carried out.

Geopolymer is an aluminate and silicate reaction and not a hydration process like

Portland cement binder. Therefore, the strength of geopolymers in saturated and dry

conditions investigation is presented in Chapter 6. Several ratios of sodium silicate to

sodium hydroxide were tested in both saturated and dry condition. Two main factors

which are the leeching out of sodium silicate during saturation process and

geopolymer’s high strength were investigated.

In chapter 7, geopolymer high strength concrete was cast and exposed to hydrocarbon

fire. Geopolymer concrete mix was determined based on findings in Chapter 5 and 6.

The concrete specifications such as concrete dimensions, reinforcement details and

aggregate size were determined based on findings in Chapter 4. Geopolymer high

strength concrete was cast in three sizes which are 1.075 m (h) x 1.075 m (w) x 0.2 m

(t) panel, 150 mm (d) x 300 mm (h) cylinder and 100 mm (d) x 200 mm (h) cylinder.

The high strength geopolymer concrete wall panel casting procedures were also

presented in this chapter.

Further investigation on aerated geopolymer wall panels for fire application was carried

out (Chapter 8). Aerated geopolymer concrete can be widely used as wall panels due

to its lightweight. In this investigation, aerated geopolymer wall panels were exposed to

hydrocarbon fire.

9.2 Conclusions

Based on findings reported in this study, the following major conclusions are drawn:

1) Geopolymer concrete can be classified as a good fire resistance construction

materials based on spalling performance of both high strength and aerated

concrete when exposed to hydrocarbon fire. A maximum of 1% (excluding

water moisture loss) of spalling recorded for high strength geopolymer concrete

wall panel. No explosive spallings were observed for both high strength

geopolymer concrete and aerated geopolymer concrete.

2) Thermal diffusivity values of high strength geopolymer concrete ranged from

0.13 to 0.15 mm2s-1 when the temperature exceeds 120oC. This shows that the

insulating capacity of geopolymer concrete is larger than Portland cement high

169

strength concrete. Diffusivity values of Portland cement high strength for

temperature more than 120oC ranging from 0.5 to 0.7 mm2s-1. There were 6.7 –

38.5 MPa residual strength measured on high strength geopolymer concrete

after 2 hours of hydrocarbon fire exposure.

3) Portland cement high strength concrete spalled in an explosive manner when

exposed to hydrocarbon fire. Large panels spalled within the first 30 minutes of

fire exposure losing about 7% - 33% of the original weight. Thermal diffusivity

dropped sharply at temperatures which ranged from the temperatures of 110oC

and 155oC. This temperature can be described as phase changing steam

temperature. Maximum saturated steam pressure induced by the steam

corresponds to maximum saturated steam temperature of 155°C is 0.44 MPa.

This is significantly less that tensile strength of concrete which is estimated to

be 3 MPa. Hence, steam pressure itself is not a critical factor for concrete

spalling.

4) Specimen’s size and aggregate size do have effects on spalling of concrete.

The largest specimens with 3.36 m (h) x 3.38 m (w) x 0.2 m (t) dimension have

47.3 mm of average spalling nominal depth compared to 2.1 mm average

spalling nominal depth for 100 mm diameter cylinder. It shows that the lab scale

fire test is not representative of the severity of concrete spalling in actual fire

event. The aggregate size effect on spalling was prominent when the maximum

aggregate size increased from 7 mm to 20 mm. Specimens with 20 mm

maximum aggregate size spalling depth is less compared to specimens with 7

mm and 14 mm maximum aggregate size. In terms of aggregate type, there is

no clear trend observed on spalling for specimen with basalt and granite

aggregate.

5) Fly ash from Gladstone power station provides the best fresh geopolymer

workability compared to fly ash from Tarong power station and Microash.

Further investigations on physical properties of these fly ashes show that the

high porosity of Tarong fly ash due to particle agglomeration was the main

factor in low workability of the fresh geopolymer.

6) There is compressive strength reduction for geopolymer when exposed to

drying and saturation condition. Geopolymer specimens with 2.5 ratio of

Na2SiO3/NaOH and 0.4 alkaline solution to fly ash ratio provides the highest

170

average compressive strength of 92.8 MPa and the highest average strength

reduction of 17%. The strength reduction is caused by sodium silicate leeching

out from specimens when immersed in water. However, the strength reduction

of geopolymer is expected to be minimal when coarse aggregates and sand

were introduced to form geopolymer concrete.

9.3 Recommendations

There are several recommendations suggested for further research.

1) Future research needs to perform a hydrocarbon fire tests on a loaded

geopolymer high strength concrete. This will enhance the simulation of concrete

panel exposure to fire with the actual condition during a fire event.

2) Visualization of heat transfer by using heat camera during the fire test can

provides better knowledge on concrete thermal properties such as thermal

conductivity, thermal capacity and thermal diffusivity.

3) Workability improvement on high strength geopolymer concrete is an important

element in order to mass produce high strength geopolymer concrete. Better

workability of the concrete will enable geopolymer concrete to be mixed in

current concrete supplier facilities without altering any current Portland cement

mixing and delivering methods.

4) Utilization of fly ash other than the Gladstone fly ash such as Tarong fly ash,

brown coal ash and fly ash from power plants in Victoria can spread the

demand of high quality fly ash from Gladstone power plant. This is because the

demand from Gladstone fly ash expected to be very high in the near future.

Documents

Performance of geopolymer concrete in fire · Performance of Geopolymer Concrete in Fire . AHMAD ZURISMAN MOHD ALI . This thesis is presented as part of the requirement for the