Effect of Temperature-Time History on Concrete Strength in Mass ...docs.trb.org/prp/13-4969.pdf · 1 Effect of Temperature-Time History on Concrete Strength in Mass Concrete 2 Structures

Effect of Temperature-Time History on Concrete Strength in Mass Concrete 1

Structures 2

3

Tahsin Alper Yikici, M.Sc. 4

Research Assistant 5

Department of Civil and Environmental Engineering 6

College of Engineering and Mineral Resources 7

West Virginia University 8

PO Box 6103 9

Morgantown, WV 26506-6103 10

Ph: 304-293-4013 11

Fax: 304-293-7109 12

Email: [email protected] 13

14

Dr. Roger H.L. Chen (Corresponding Author) 15

Professor 16

Department of Civil and Environmental Engineering 17

College of Engineering and Mineral Resources 18

West Virginia University 19

PO Box 6103 20

Morgantown, WV 26506-6103 21

Ph: 304-293-9925 22

Fax: 304-293-7109 23

Email: [email protected] 24

25

26

Submission Date: November 15, 2012 27

Number of Words in Text: 3515 28

Number of Tables: 4 29

Number of Figures: 8 30

Total Equivalent Number of Words: 651531

TRB 2013 Annual Meeting Paper revised from original submittal.

Yikici and Chen

2

ABSTRACT

Concrete maturity method is a popular non-destructive testing method to estimate in-

place strength development of concrete structures. Many state highway agencies adopted

procedures for using maturity method to obtain better quality control while monitoring

in-place strength development in real time. In this study, maturity method was used to

estimate in-place strength of large concrete placements. Four 6-foot cube blocks were

constructed in four different locations and calibration curves for those concrete mixtures

were established using 6x12 inch cylinder specimens collected from the construction site.

Temperature sensors were embedded in specific locations throughout the depth of the

cubes, and the equivalent age of the in-place concrete was calculated. 4-inch diameter

core samples, with 6-foot in length, were taken from the cubes at four-day after

construction and the core strengths were compared with the predicted strengths using

maturity. In addition, activation energy values were determined in the laboratory and

used for equivalent age calculations as recommended in ASTM C 1074. According to

the test results, the concrete top surface strength prediction is always higher than the

actual core strength. For three cube constructions, core results from mid-section were

close to the predicted strengths and core results from the bottom section were higher than

the predicted values. Results show that in-place concrete strength is being influenced by

several factors other than temperature, including the location of the sample in the

structure, lack of compaction quality, higher air content and in-situ water-cement ratio,

so that establishing a reliable maturity and in-place strength relationship is rather difficult

within given circumstances. The results of this study provide useful information to

examine the accuracy of the maturity method used in the estimation of in-place concrete

strength in large structures.


Yikici and Chen

3

INTRODUCTION

The strength of properly batched, placed and vibrated concrete does not depend only on

the curing time, but also on the temperature-time history. This concept is known in the

concrete industry as maturity concept. According to the maturity concept, an empirical

relationship can be established between temperature-time history and strength

development of the concrete in order to predict strength of in-place concrete during the

curing period (1). ASTM C 1074 recommends maturity method as “a technique for

estimating concrete strength that is based on the assumption that samples of given

concrete mixture attains equal strength if they attain equal values of maturity index”(2).

The method assumes that the temperature-time history of concrete can be used to develop

a strength-maturity curve that is specific to each mix design. By preparing these

correlation curves, the strength development of in-place concrete can be estimated by

monitoring the concrete temperatures in real time. Consequently, this information can be

used to make decisions (e.g. time of formwork removal, or time of post-tensioning) that

save time and reduce the construction cost (1). Furthermore, monitoring concrete

temperatures, especially in mass concrete pours, can be used to prevent high internal

concrete temperatures and large temperature gradients that are specified by several state

agencies in order to reduce the possibility of thermal cracking.

Many state transportation agencies have already instituted procedures or are still

conducting research projects to implement the maturity method to predict in-place

concrete strength. According to the West Virginia Department of Highway (WVDOH)

survey results conducted in 2007, twenty-five out of thirty-six states used the maturity

concept mainly as a substitute for early cylinder compressive strength to allow formwork

to be removed or pavements to be opened to traffic (3). In 2008, Auburn University

employed maturity method on several precast, prestressed girders and a bridge deck.

They concluded that the method can be used accurately for estimating in-place concrete

strength up to an equivalent age of seven days (4). University of Washington researchers

reported in 2009 that the maturity method was used in three different Portland cement

concrete pavement (PCCP) projects in order to open traffic faster. Only one out of the

three trials was successfully conducted (3). Similarly, University of Maryland evaluated

maturity method for use in pavements and they concluded that the procedure is very

sensitive to the constituent materials and concrete mixtures. They recommended taking

extreme pre-cautions in order to obtain maximum accuracy when using maturity method

for field applications (5). Furthermore, the cross over effect (1) due to high temperature

curing has been shown to limit the applicability of maturity method in predicting the

behavior of concrete that has high early temperature, such as mass concrete construction.

One of the objectives of this study is to investigate the applicability of maturity

method to estimate the in-place concrete strength of large bridge sub-structure elements,

such as piers, pier footers, pier caps or abutments, using WVDOH approved Class B

concrete mixtures that are currently used in bridge projects. Class B concrete, as

described in WVDOH Standard Specifications, has minimum 3,000 psi (20 MPa) 28-day

design strength with optimum 4-inches slump and 7% target air. Class B concrete may

be designed using supplementary cementitious materials such as fly-ash, ground

granulated blast furnace slag (GGBFS) or micro-silica with 564 pound per cubic yard

(330 kg/m3) target cement content and 0.49 maximum water-cementitious ratio. In

addition, this study outlines the effect of the strength development from the temperature


Yikici and Chen

4

variations in concrete throughout the depth of 6-ft concrete cubes. This paper presents

test results from four different 6-ft cube constructions and the predicted in-place concrete

strength using a maturity function based on concrete equivalent age.

RESEARCH METHODOLOGY

Six-ft concrete cube blocks were constructed at different locations in West Virginia,

using Class B concrete from local ready-mix plants. Temperature sensors were

instrumented, fresh concrete properties were determined and 6x12 inch cylinders were

taken for the maturity test. Core samples were taken from the hardened concrete cube

block and the measured compressive strengths from the core samples were compared to

the predicted strengths from equivalent age calculations. Activation energy values for the

concrete mixtures were determined in the laboratory following ASTM C 1074 (2).

SIX-FT CUBE CONSTRUCTION

Six-ft concrete cubes were constructed at four different WVDOH districts (D1, D5, D6

and D9), located in Charleston (D1), Lewisburg (D9), Martinsburg (D5), and Wheeling

(D6), pouring approximately nine cubic-yards of concrete in each cube provided by local

ready-mix concrete plants. The concrete mix design for each casting is given in Table 1.

The cube blocks were constructed two feet in the ground on a two inch layer of #57

limestone. Each cube was instrumented with temperature loggers attached on a rebar

cage (Figure 1). A schematic of the sensor locations is given in Figure 2. Concrete was

poured directly from the mixer truck without pumping and then was subjected to

vibration in order to get sufficient compaction. Ordinary surface finish using wood-float

rubbing was applied on the top surface. The concrete surface was maintained completely

and continuously moist during the seven-day curing period. After the concrete placement

the top of the block was covered with white polyethylene sheeting. If necessary, concrete

blankets were used on top surface as well as around the formwork on the side surfaces.

One of the purposes of the six-foot cube constructions was to investigate strength

development of in-place concrete by monitoring the temperature distribution in concrete

and investigate the applicability and the limitations of the maturity concept for large

concrete placements throughout West Virginia.

(a) Instrumentation of the rebar cage

(b) Sensors attached to the rebar

FIGURE 1 Mounting of the temperature loggers.


Yikici and Chen

5

TABLE 1 One yd3 Theoretical Mix Design

Item

D1 D9 D5 D6

Class B

Fly-Ash Class B

Class B

GGBFS

Class B –

1 Extra

Cement (TYPE I/II), lbs 470 564 423 658

Fly-Ash (TYPE F). lbs 75 - - -

GGBFS (Grade 100), lbs - - 141 -

Water, lbs 245 262 276 260

Coarse Aggregate (#57), lbs 1775 1723 1815 1750

Fine Aggregate, lbs 1255 1299 1225 1111

Target Air Content, % 7.0 7.0 7.0 7.0

w/cm 0.45 0.45 0.49 0.40

NOTE: 1 lb = 0.454 kilograms (kg)

FIGURE 2 Schematic of the sensor locations.

6’-0”

6’-0”

SIDE VIEW

3

3’-0”

3’-0”

11

2’-0”

1’-0”

10 9

5

1 13

12

4

2

1’-6”

1’-6”

1’-0”

6

7

8

14


Yikici and Chen

6

Experimental Work

In order to establish the maturity-strength relationship of each mix, 6x12 inch cylinders

were cast during the cube constructions. Additionally, two 6x12 inch cylinders were

embedded with temperature loggers recording hourly temperature history. All cylinders

were placed inside insulated boxes to reduce the effect from the ambient conditions

overnight and then transported the next day to temperature controlled curing tanks at the

district material laboratory (Figure 3). Compressive strength of the concrete was

determined at 1, 3, 7, 14 and 28 days, testing at least two cylinders at each age.

In addition to that, at 4-days, 4-inch-diameter by 6-foot long core samples were

taken from the hardened concrete cube (Figure 5) and a total of six 4x8 inch cylinder

specimens were extracted from the core along the 6-ft length (Figure 6). The specimens

were prepared and tested immediately after coring to represent the in-place compressive

strength of the concrete cube at different depth. A schematic drawing that shows the cut

locations and specimen designations is presented in Figure 4.

(a) Concrete placement

(b) Maturity cylinders

(c) Cylinders on the field

FIGURE 3 D5-Cube construction and sampling.

FIGURE 4 Core specimen cut locations and designations.

8” 2” 4” 4” 8” 4” 8” 4” 8” 4” 8” 2” 8”

36” 36”

1C 2C 3C 4C 5C 6C 1R 2R 3R 4R 5R

1 foot Top of the core

(30.5 cm)


Yikici and Chen

7

(a) Coring machine

(b) 6-ft core

FIGURE 5 Six-ft cube coring.

FIGURE 6 Schematic of the coring locations.

X: Temperature sensors O: Coring positions

X X X X

O O

O

6’0”

X

2’0”

TOP VIEW

2’0”

1’0”

O

X

X

6’0”

28 days

56 days

4 days 28 days


Yikici and Chen

8

TEST RESULTS AND DISCUSSION

Determination of Activation Energy

The activation energy of concrete mixtures was determined experimentally following

ASTM C 1074-10 A1 procedure. It requires establishing the compressive strength versus

age relationship of 2-inch mortar cubes cured at three different temperatures (2). The

mortar was proportioned to have a fine-aggregate to cement- ratio equal to the coarse-

aggregate to cement ratio of the concrete mixture. Specimens were cured at three

different temperatures: high (104°F), low (50°F), and laboratory temperature (73°F).

Three cubes were tested at six different times in compression following the recommended

test schedule by Tank R. C. (6), based on equal temperature-time factors for different

curing temperatures.

Upon the completion of the compressive strength tests, hyperbolic equation

(1,2,7) was used to fit the set of data to determine the best fit regression parameters, such

as the limiting strength, Su, the rate constant of strength gain, k, and the dormant period

t0, for three different curing temperatures. After that, natural logarithm of the k-values

versus reciprocal curing temperature in Kelvin was plotted. From that, the negative slope

of the line is obtained which equals to the value of the activation energy divided by the

universal gas constant (R), also known as Q. This calculation is based on the Arrhenius

function that is being used to explain the temperature dependence of the rate constant, k

(1,7).

Apparent activation energy values for the Class B concrete containing fly ash and

GGBFS were determined according to ASTM C 1074 procedures. The calculation of this

value requires the determination of several parameters using the linear hyperbolic

equation:

)(1

)(

0

0

ttk

ttkSS u

where:

S= average strength of the cubes at age t

t= test age in hours

Su= limiting strength

t0= age when strength development assumed to begin

k=

rate constant.

It was found that the hyperbolic strength-age function can properly model the

strength development with the lowest goodness of fit R-square value 0.93 for each set of

experiment. The apparent activation energy values were calculated approximately 45,900

J/mol and 44,750 J/mol for Class B Fly-Ash and Class B GGBFS concrete mixtures,

respectively.


Yikici and Chen

9

Maturity Calculations

Equivalent age approach was used to establish maturity-strength relationship. The

actual age of the concrete was converted to its equivalent age at a specified temperature.

Equivalent age can be calculated according to the following “Arrhenius Equation”:

where:

te = equivalent age

Q = activation energy divided by the gas constant (R),

Ta = average temperature of the concrete during time interval,

Ts = specified (reference) temperature (typically 23°C)

Δt = the time interval

A calibration curve was prepared from strengths of the laboratory cured specimens using

the recorded temperature-time history of the cylinders. The calibration curve can be used

to estimate the in-place concrete strength if temperature history of the structure is known

(1,2,8). The calibration curve that represents the strength gain of the concrete was

modeled using the linear hyperbolic model suggested by ASTM C 1074-10 (2). The age

when strength development assumed to begin (t0) was set equal to the final setting time of

the concrete. Figure 8 shows the compressive strength versus equivalent age

relationships based on the cylinder compressive strength results. The equivalent age of

the D1 and D5 concrete were calculated using the measured activation energy values of

45,900 J/mol and 44,750 J/mol, respectively. The equivalent age of the D9 and D6

concrete were calculated using an assumed activation energy value of 41,800 J/mol based

on the model proposed by Han S.H. (9).

FIGURE 7 Concrete mix design calibration curve, strength vs equivalent age.


Yikici and Chen

10

Core Strength

The compressive strength results from the concrete cores obtained at 4-day from each 6-ft

cube are listed in Table 2. The test results show that there is a significant strength

difference along the depth, between the top (1C) position and the bottom (6C) position.

1C position appears to be the weakest and 5C and 6C positions are the strongest. This

appears to coincide with observation by other researchers that cores usually have lower

strength near the top surface and the strength increases with depth (10).

The core test results clearly indicate the variations from the conditions occurred

during concrete placement. During D9 cube construction concrete was delivered in two

separate trucks and the air content measured on the field was 7.8% and 9.5%,

respectively. The unexpected difference in air content may be the reason that shows a

large variation in strength between the cores 3C and 4C positions. During D6 cube

construction the slump of the fresh concrete was only 1.75 inches and vibration was very

difficult. Hence, honeycombing was observed at the mid-height section from the

concrete surfaces. The effect of the segregation and honeycombing on the core strengths

was detected between core samples 3C and 4C. Furthermore, there is a possibility of

strength reduction due to drilling operations. The coefficient of variation of strength

estimation using 4-in diameter cores was presented 4 to 5.5%. (11) However, it is really

difficult to separate out the errors due to on-site quality control issues, such as concrete

placement, compaction, air-content, actual water to cement ratio, etc. (10).

TABLE 2 Concrete Compressive Strength from the 6-ft Core at 4-day

1C 2C 3C 4C 5C 6C

Depth from the surface, inch 2”-10" 14"-22" 26"-34" 38"-46" 50"-58" 62"-70"

D1 Cube 3,160 4,670 4,830 4,690 4,850 4,930

D5 Cube 3,880 4,790 4,790 4,870 4,790 5,300

D9 Cube 2,420 2,660 * 3,620 3,670 4,010

D6 Cube 4,460 5,710 4,100 3,310 5,250 5,250

Note: 1 inch = 2.54 cm, * the core was broken at 3C position

In-Place Concrete Strength Prediction

Even though the maturity method is more reliable in predicting the relative strength than

the absolute strength (1), in this study, it was assumed that the 28-day (equivalent-age)

cylinder strength and the in-place concrete strength are same, hence, the developed

maturity-strength relationship is used to predict the concrete strength in the cube. In

order to estimate the in-place concrete strength, temperature sensors were installed at

critical locations in the 6-ft cubes. The locations were selected to be representative of the

temperatures at the locations of coring due to symmetry. In-place concrete strengths

were estimated using strength-equivalent age calibration curves (Figure 7) for each

concrete mixture. The equivalent ages of the concrete at three locations were calculated

based on the temperature-time history of three specific locations in the 6-ft cubes,

corresponding to sensor #6 (top section), #7 (mid-section) and #8 (bottom section) and


Yikici and Chen

11

given in Table 3. These sensors #6, #7 and #8 were located at 2”, 36” and 70” from the

cube top surface, respectively. Figure 8 shows a typical temperature-time history of

those three locations up to 14 days (from the D5 Cube).

In addition, the predicted strengths are compared to the 4-day core strength

results; “1C” representing the top position, “3C” and “4C” representing the center

position, and 6C representing the bottom position of the cubes. The results show that the

top surface predicted strength (#6) is always higher than the actual core strength at all

four cubes. For D1, D5 and D9 cubes, concrete strength at the mid-section (#7) were

close to the predicted strength, however the core strength results are higher than the

predicted values at the bottom section (#8). Essentially due to on-site quality issues

mentioned earlier, it is noticed that in D6 case the core strengths are lower than the

predicted strengths at each position.

TABLE 3 In-place Concrete Strength Prediction Compared with the Core

Strength Results

D1 D5 D9 D6

Sen

sor

Equ

ival

ent

age,

day

s

Pre

dic

ted

Str

eng

th, p

si

Co

re

Str

eng

th, p

si

Equ

ival

ent

age,

day

s

Pre

dic

ted

Str

eng

th,

psi

Co

re

Str

eng

th, p

si

Equ

ival

ent

age,

day

s

Pre

dic

ted

Str

eng

th,

psi

Co

re

Str

eng

th, p

si

Equ

ival

ent

age,

day

s

Pre

dic

ted

Str

eng

th,

psi

Co

re

Str

eng

th, p

si

#6 14.1 3,920 3,160 11.7 4,310 3,880 17.3 3,370 2,420 20.3 6,330 4,460

#7 21.2 4,060 4,760 18.2 4,970 4,830 31.1 3,500 3,620 24.5 6,400 3,710

#8 13.9 3,920 4,930 10.5 4,130 5,300 19.0 3,400 4,010 10.0 5,970 5,250

Note: 1 psi = 6.89 kPa

FIGURE 8 Measured concrete temperature-time history from D5 Cube.


Yikici and Chen

12

In addition to the 4-day core, there are 28-day and 56-day cores (Figure 6)

extracted from the cubes, and the compressive test results from these core specimens are

shown in Table 4. As expected, the long-term strength of the concrete from these core

specimens can not be predicted by the linear hyperbolic strength-maturity model (1,4).

Modification of the maturity method is needed for the prediction of long-term concrete

strength development, especially considering high-early temperature effects, such as

those seen in mass concrete.

TABLE 4 Compressive Strength from the 6-ft Core at 28-day and 56 day

1C 2C 3C 4C 5C 6C

Depth from the surface, inches 2"-10" 14"-22" 26"-34" 38"-46" 50"-58" 62"-70"

D1 CUBE

28 Days (center) 4,750 5,640 5,600 4,950 6,460 6,540

29 Days (corner) 4,370 5,600 5,640 5,490 6,070 5,900

56 Days 4,690 6,130 5,920 5,820 6,370 6,410

D5 CUBE

28 Days (center) 4,460 6,080 5,820 5,570 5,630 6,960

28 Days (corner) 4,510 4,800 5,150 6,040 5,700 6,590

76 Days* 4,180 5,750 5,580 5,310 6,090 7,430

D9 CUBE

28 Days (center) 2,960 2,670 2,520 3,710 3,630 4,120

28 Days (corner) 3,150 2,630 2,510 3,790 3,740 4,210

56 Days 3,350 2,730 2,640 4,000 3,840 4,330

D6 CUBE

28 Days (center) 6,010 6,440 5,150 6,490 6,210 6,230

28 Days (corner) 5,730 6,160 5,450 5,980 6,090 6,400

56 Days 5,390 6,530 6,160 6,590 6,630 6,440

Note: 1 inch = 2.54 cm

* Coring was delayed due to drilling equipment malfunction

SUMMARY AND CONCLUSIONS

Many states, including West Virginia are interested in using ASTM C 1074 Maturity

Method for the benefits of increasing quality control, accelerating construction time, or

reducing number and cost of sampling and testing standard cylinders. On the other hand,

accuracy, effectiveness, and reliability of the test method to estimate in-place concrete

strength have been a concern, especially for the concrete under high temperature

differential curing such as in the case of mass concrete. The purpose of this study is to

investigate the applicability of the maturity method on large concrete pours using regular

Class B Concrete in West Virginia. Four different concrete mix designs were

investigated in four 6-ft cube constructions in different districts. Maturity-strength

calibration curves for these mixes were established and concrete temperatures inside the

cubes were monitored in order to calculate equivalent concrete age using the measured

activation energy values. The in-place concrete strength was determined by testing core

samples extracted from cubes and results were compared with the predicted values.

Based on the test results the following conclusions can be made:


Yikici and Chen

13

1. Compressive strength-age development of concrete can be represented by testing

the corresponding mortar mixture following ASTM C 1074-A1. The hyperbolic

strength-age relationship can be used to model strength development at different

temperatures. Activation energy values for concrete mixtures including

supplementary cementitious materials were determined testing mortar cubes

prepared in the laboratory.

2. Test results show that the in-place concrete core strengths of the 6-ft cube close to

the concrete top surface were overly estimated using ASTM C1074 maturity

method. Effect of variable temperature curing in large structures cannot be

accurately predicted using the current maturity calculation with linear hyperbolic

equation. Further study is needed to modify the maturity calculation for its

application in mass concrete with high early-age temperature.

3. The error in estimating in-place concrete strength using equivalent age method is

unpredictable partly because the concrete in-place strength is highly dependent on

the quality control on-site. The variables include in-situ water-cementitious ratio,

air content, vibration/consolidation, and finishing.

ACKNOWLEDGEMENT

The authors acknowledge the support provided by the FHWA and West Virginia

Transportation Division of Highways for the project RP#257-Pre-liminary Analysis of

Use of Mass Concrete in West Virginia. Special thanks are extended to our project

monitors Michael A. Mance, Donald Williams and Ryan Arnold of WVDOH. The

assistance received from the Materials Control and Soils Testing Division (WVDOH

MC&ST), and WVDOH District 1, District 5, District 9 and District 6 Bridge and

Materials divisions are especially acknowledged.

REFERENCES

1. Carino N. J. “The Maturity Method,” Chapter 5 in Handbook on Nondestructive

Testing of Concrete, 2nd

Edition, Malhotra V. M. and Carino N. J., Eds., CRC

Press, Boca Raton, Fl, 2004.

2. ASTM C 1074-10. Standard Practice for Estimating Concrete Strength by the

Maturity Method. ASTM Standards, ASTM International, West Conshohocken,

PA.

3. Muench S., Pierce L.M., Kinne C., Uhlmeyer J.S. and Anderson K. W. Use of

Maturity Method In Accelerated PCCP Construction. WSDOT Research Report,

WA-RD 698.1. Washington State Department of Transportation, 2009.

4. Wade, S.A., Barnes R.W., Schindler, A.K. and Nixon J.M. Evaluation of the

Maturity Method to Estimate Concrete Strength in Field Applications. ALDOT

Research Report, Highway Research Center and Department of Civil Engineering

at Auburn University, 2008.

5. Hosten M. A., Johnson R. Implementation of the Concrete Maturity Meter for

Maryland. State Highway Administration Research Report, Report No. MD-11-

SP708B4K. Morgan State University, 2011.

6. Tank, R. C. The Rate Constant Model for Strength Development of Concrete.

Ph.D. Dissertation submitted at Polytechnic University of New York, June 1988,

209 pp.


Yikici and Chen

14

7. Brooks A. G., Schindler A. K. and Barnes R. W. Maturity Method Evaluated for

Various Cementitious Materials. Journal of Materials in Civil Engineering.

ASCE, December 2007, pp. 1017-1025

8. Poole, T.S. and Harrington P.J. An Evaluation of the Maturity Method (ASTM C

1074) for Use in Mass Concrete. Technical Report SL-96-16. U.S Army Corps of

Engineers, Vicksburg, MS, 1996.

9. Han S.H., Kim J.K., Park Y.D. Prediction of Compressive Strength of Fly Ash

Concrete by New Apparent Activation Energy Function. Cement and Concrete

Research, Vol. 33, 2003, pp. 965–971.

10. Neville, A. M. Properties of Concrete. Pearson Education Ltd, England, 2002.

11. Bartlett F. M., Precision of in-place concrete strengths predicted using core

strength correction factors obtained by weighted regression analysis, Structural

Safety, Vol. 19, Issue 4, 1997, pp. 397-410.


Documents

Effect of Temperature-Time History on Concrete Strength in Mass ...docs.trb.org/prp/13-4969.pdf · 1 Effect of Temperature-Time History on Concrete Strength in Mass Concrete 2 Structures