Determination of Temperature Rise and Temperature Differentials of CEMII/B-V Cement for 20MPa Mass Concrete using Adiabatic Temperature Rise Data

Experimental test was carried out to determine the temperature rise characteristics of Portland-Fly-Ash Cement (CEM II/B-V, 42.5N) of Blaine fineness 418.6m2/kg and 444.6m2/kg respectively for 20MPa mass concrete under adiabatic condition. The estimation on adiabatic temperature rise by way of CIRIA C660 method (Construction Industry Research & Information Information) was adopted to verify and validate the hot-box test results by simulating the heat generation curve of the concrete under semi-adiabatic condition. Test result found that Portland fly-ash cement has exhibited decrease in the peak value of temperature rise and maximum temperature rise rate. The result showed that the temperature development and distribution profile, which is directly contributed from the heat of hydration of cement with time, is affected by the insulation, initial placing temperature, geometry and size of concrete mass. The mock up data showing the measured temperature differential is significantly lower than the technical specifications 20°C temperature differential requirement and the 27.7°C limiting temperature differential for granite aggregate concrete as stipulated in BS8110-2: 1985. The concrete strength test result revealed that the 28 days cubes compressive strength was above the stipulated 20MPa characteristic strength at 90 days. The test demonstrated that with proper concrete mix design, the use of Portland flyash cement, combination of chilled water and flake ice, and good insulation is effective in reducing peak temperature rise, temperature differential, and lower adiabatic temperature rise for mass concrete pours. As far as the determined adiabatic temperature rise result was concern, the established result could be inferred for in-situ thermal properties of 20MPa mass concrete application, as the result could be repeatable on account of similar type of constituent materials and concrete mix design adopted for permanent works at project site.


INTRODUCTION
In large concrete elements, the interior is considered to be hydrating in an essentially adiabatic process. Because of the low thermal conductivity of concrete, the heat of hydration from its interior is prevented from being released into the environment, thus a negligible amount of heat is lost, compare to exterior of concrete element. The resulting large temperature differential between the interior and exterior of a concrete element may lead to thermal cracking. The term adiabatic refers to a process occurring without the gain or loss of heat. Adiabatic temperature measurement is used to simulate the condition in the interior of a large concrete element so that the maximum heat development potential in any concrete mix can be measured (Lee MH, Khil BS & Yun HD 2014). This data may then be used as input parameter for subsequent thermal modelling in the pursuit to determine the risk level and thermal cracking potential. Owing to data of this kind is not available locally; therefore adiabatic temperature rise test was carried out to determine the peak concrete core temperature, temperature differential and temperature gradient for this 20MPa mass concrete. This was determined through trial hot-box tests at prescribed initial placing temperature and anticipated concrete section thickness, rather than merely follow project contract specifications requirements (Technical Specifications 2000;2009, 2010& 2013, as stipulated below: (1) The fresh concrete placing temperature shall not exceeding 20°C; (2) Maximum core-surface temperature differential shall not exceeding 20°C To avoid surface cracking caused by heat generated in the concrete, European Standard ENV 206: 1992 suggests that the limit on the temperature differential between core and surface is 20°C and the average maximum temperature limit of concrete at core is 60°C. Based on UK experience (BS 8110-2: 1985), by limiting temperature differential to 20°C in gravel aggregate concrete, cracking can be avoided. This represents an equivalent restraint factor R of 0.36. However, MS523-2: 2011, Clause A9.3 has specified peak placing temperature of 30°C for mass concrete structure construction. As for the case of HTP project, the proposed quarried rocks for mass concrete production was granite aggregate whereby the limiting temperature differential adopted should be capped at 27.7°C instead of 20°C; BS 8110-2: 1985 (pp. 32), Table 3.2-Estimated limiting temperature changes to avoid cracking has specified that limiting temperature differential when R = 0.36 is 27.7°C when granite aggregate concrete was adopted for mass concrete production and construction. However, BS 8110-2: 1985 does not specify the fresh concrete placing temperature and peak temperature requirements in tropical environment.
Concrete temperature rise test under adiabatic condition is normally an approach in laboratory to evaluate the exothermic characteristics of the hydration of the concrete. In this case, 20MPa plain concrete trial hot-box with an array of thermocouples and strain gauges are used to assess the characteristics of the temperature rise, maximum temperature difference and temperature distribution profile in mass concrete were conducted with due consideration given on the selection of concrete block size and geometry, insulation, thermal couples and strain gauges locations, and the duration of temperature monitoring. Due to the reduced rate of hydration, the use of CEMII/B-V cement to replace OPC (CEM 1) is one of the recommended practices to considerably reduce the temperature rise in mass concrete. Since hydration occurs at the surface of cement particles, therefore finely ground cement will have a higher rate of hydration. As finely ground cement has a higher specific surface area, which means there is more area in contact with water for more complete hydration. The finer particles will also be more fully hydrated then coarser particles. However, the total heat of hydration at very late ages is not significantly affected.
This research is intended to overcome the above-mentioned 20MPa mass concrete specification requirements for the construction of Tembat Hydropower dam by adopting more prudent temperature control requirements through subsequent thermal modelling rather than arbitrary determined basing on designer's temperate country experience and practices, which is overly conservative, costly method, and non-sustainable due to carbon emission by running huge capacity of chillers and flake-ice plant to bring down the temperature of water for making concrete and aggregates substantially.
The experimental program was carried out in accordance with the procedure as shown in below

EXPERIMENTAL SET-UP
For Stage 1 testing, two (2) hot boxes were prepared as part of the trials. The first hot box (HB1) was prepared with full insulation (100mm thick MRB-32 Insulation Board) with block size of dimension 1.5m x 1.5m x 1.5m and two sets of thermocouples (thermocouples wire type K) installed, one set (3 thermocouple each) at the center and the other set (3 thermocouple each) at the corner of the hot-box and two strain gauges installed at the centre of the block. A second hot block (HB2) of dimension of 1.3m x 1.3m x 1.3m was prepared with two sets of thermocouples installed, one set (3 thermocouple each) at the center and the other set (3 thermocouple each) at the corner of the block. In this case, the base and sides were insulated but the top surface of the box was exposed to the environment. Based on the completed trials conducted on 11/2/2011 and 12/2/2011, an identical concrete mix design was selected for the hot box casting as given in Table 5 (see Page 3 of 11). For stage 2 testing, one hot box was prepared at Hulu Terengganu project site with similar block size and configuration with stage 1 HB, comply with full insulation (100mm thick MRB-32 insulation Board) and two sets each of thermocouples and strain gauges installed using identical concrete mix design as indicated in Table  3.3, whereby the trial was conducted on 21/6/2012.

Test Specimen
For the evaluation of the temperature rise in mass concrete trial block, a 1.5m x 1.5m x 1.5m concrete block was cast. Blended cement (CEM II/ B-V) containing 25% fly ash was used for the Grade 20 mass concrete hot block trial. The cement content and water/cement ratio was 200 kg/m3 and 0.70 (Stage 1 & 2), respectively. The schematic drawing of the block is described in Figures 1-3, with the locations of the thermocouples, 100mm insulation and timber formwork indicated. 100mm thick MRB-32 insulation board was adopted as the insulation to the concrete mass around the six sides. The top of the concrete was only covered with the polyfoam. Tc is the core temperature at center of the block while Ts and Tn represent the temperature at surface and corner location as indicated in the schematic drawing. The peak core temperatures obtained from stage 1-HB1, stage 1-HB2 and stage 2-HB1 were also used to determine the repeatability of the experiments. Compilation of testing results and data analysis. 8 Temperature analysis and interpretation of results. 9 Conclusions/Recommendations Temperature probing, slump measurements and concrete sampling using cube and cylinders as per testing plan carried out by established and accredidated testing laboratory.

Procedure
Casting of Hot-Block using raw materials established from step 1 above

Casting of Hot-Box
Hot-box is casted with 20Mpa temperature control concrete (average 25°C) with MSA 75mm whereby compaction was done by using 50mm vibrator pokers.

Data Acquisition
Campbell scientific CR10X data-logger was program automatically to record concrete temperature (°C) and VWSG's strain (με) data at every hourly interval simultaneously for a period of 7 days and the data are downloaded to laptop. The maximum temperature is defines as the readings when no further appreciably rise in temperature over a period of 3 -4 successive hours).

CONCRETE MIX DESIGN AND STRENGTH ANALYSIS
The physical properties and chemical compositions as well as the fineness of the Portland-fly-ash cement (PFAC) used in the study are shown in Tables 2-4 respectively. Table 5 presents the concrete mix design for grade 20 mass concrete applications for all three concrete hot-box tests. The measured initial placing temperature of the mixes was 23 o C to 25 o C respectively.      Based on the strength results shown in above Table 5, all concrete cubes tested at 7 days had exceeded the contract documents stipulated 90 days characteristic strength of 20MPa confirming that the required design strength of 20MPa is achievable within 28 day. Therefore, further optimization on content of cementitious material should be considered for lower heat of hydration. Cube compressive strength (see Table 6 In addition, internal cores were taken to study in-situ strength as shown in Figure 4. As expected with concrete containing flyash, the heat cycling had a beneficial effect on strength and the core strength is about 6.5 per cent higher comparing to the cube strength (see Table 7) measured at 56 days. The compressive strength test results confirm that the concrete is well above the specified requirements in terms of strength. It was also observed that the differential for Stage-1 & Stage-2 strength tests at effective age of 7 days for the compressive strength, tensile strength and elastic modulus are 1.82 N/mm 2 , 0.08 N/mm 2 , and 1.08 N/mm 2 respectively.  The constants and coefficients of heat generation for above item a), b), c), d), e) and f) are indicated in below Table 8. Table 8.

Constant & coefficient of heat generation of CEM II/B-V by method of CIRIA C660 (Bamforth 2007)
The modelled heat output (kj/kg) in the semi-adiabatic test and a Q ad exponential curve measured at Q 41 is 294.79 and 296.77 (See Figure 8b for exponential curve, Q ad ) respectively, which showing reasonably good correlation/ consistency has been obtained between prediction of temperature rise and the measured semi-adiabatic temperature rise derived from hot-box test, as graphically represented in below Figure 8a and 8b respectively.

Estimated Adiabatic Temperature Rise, Qad
The thermal model as provided in this paper uses the semi-adiabatic temperature rise data from published sources based on an extensive study carried out in University of Dundee, United Kingdom as the input. The hot-box tested as per the present trials was semi-adiabatic in nature as they are not perfectly insulated. This involves the establishment of different values of heat loss coefficient and adding incremental heat losses to the semi-adiabatic curve. A comparison between the estimated adiabatic curves is shown in below Figure 8a & 8b. The estimated maximum adiabatic temperature rise (MATR) measured at 72 hours is 23.72°C using data by method of CIRIA C660 could be used as input parameter for subsequent thermal modelling/ simulation i.e. heat of hydration analysis, from an initial placing temperature of 25°C. The aforesaid MATR could be further validated by using another empirical adiabatic hydration model ( Table 9. (Eq.2) Q ∞ = Ultimate adiabatic temperature rise, γ = Constant on rate of temperature rise, t = Age, in days Q(t) = Adiabatic temperature rise at an age of t, in °C The constant and coefficient of adiabatic heat generation using above Equation 2, is demonstrated in below Table 9.

Graphical Plots for Adiabatic Temperature Rise Monitoring Results (Adiabatic Temperature Rise Test Results 2011 & 2012)
From the adiabatic temperature rise monitoring test results (see Figure 9, 10 and 11), it could infer from VWSG's thermistor that the peak core temperature recorded for both phases of the trials i.e. Stage 1-HB1 @ point A2 (48.30°C) and Stage 2-HB1 (49.53°C) only showing minor differential of 1.53°C, whereby the established average peak core temperature is 48.915°C. This is well within the stipulated peak core temperature 50°C requirement. The respective maximum temperature of T c , T n and its corresponding time of occurrence for Stage 1-HB1, Stage 1-HB2 and Stage 2-HB1 are summarised in below Table10. T c = Temperature at core, at Point A2 (center core) T s =Temperature at surface, at Point A1 (near bottom surface) T cn =Temperature at center, at Point B2 (at center) T n = Temperature at corner, at Point B1 (near top corner) It has been demonstrated during mock-up test (Test Report for Mass Concrete, 2013); the core reaches its maximum temperature in about 72 hours after concrete placement.

Effects of Size of Concrete Block and Boundary Condition
The maximum temperature at center heavily depends on the insulation and the size of the hot-block.
The results indicate that the center of a 1.5m x 1.5m x 1.5m mass concrete hot-box with proper insulation could produce the maximum core temperature T c very close to that under full adiabatic condition. This implies that the maximum temperature at the center of a concrete block is little affected by the size of the concrete if the dimension is big enough.
The adiabatic temperature rise test simulation by method of CIRIA C660 (using identical mix design and input parameter as indicated in below Table 11) showing that hot-box size of 3m x 3m x 3m and greater having proper 50mm thick polyfoam insulation will yielded Tc at full adiabatic temperature. Furthermore, for a concrete hot-box of dimension 4m x 4m x 4m, the temperature recorded has exhibited full-adiabatic behaviour even without any insulation as the block size is big enough. However, this is non-practical for any laboratory mock-up test to be considered in such huge block dimension unless it is cast in-situ for thick raft foundation.  Figure 12. Temperature rise curve and differential in 1.5m x 1.5m x 1.5m hot box for Stage 1-HB1 Figure 12 presents the maximum temperature difference of (T c -T n ) and (T c -T s ) in a 1.5m x 1.5 x 1.5m concrete hot-box. The results reveal that the effective insulation using polyfoam of low conductance material plays a dominant role in controlling the maximum temperature difference across the block, even though the core temperature varies marginally. differentials (T c -T n ) and (T c -T s ) are affected to some degree between 1.79°C to 4.53°C. Whereby the recorded maximum temperature differentials between (T c -T n ) and (T c -T s ) are 4.66°C and 4.53°C respectively. As for the case of 1.3m x 1.3m x 1.3m hot box (Stage 1-HB2), the direct exposure of concrete surface to air is measured whereby the recorded maximum temperature differential between (T c -T n ) and (T c -T s ) is 9.22°C and 8.88°C respectively, which was about twofold than the fully insulated hot block as in the case of 1.5m x 1.5m x 1.5m hot box. Therefore, direct exposure of concrete surface to air or the use of steel formwork would expect even higher temperature differentials for (T c -T n ) and (T c -T s ).

Temperature Rise and Temperature Distribution for Stage 1 & Stage 2 Hot-Box Tests
The two hot-box trials have exhibited slightly different rates of temperature rise at some period during the experiments, whereby their peak core temperature rise values recorded by thermistors (strain gauges) only differed by around 1.225°C as shown in below Table 13. Likewise the difference in peak core temperature recorded by thermocouples is 1.25°C as shown in Table 14.
The differences in aforesaid peak core temperatures could be attributed to lower initial concrete placing temperature for Stage 1: HB1 hot box (average 23°C) as compare to Stage 2: HB1 hot box (average 25°C). This slight difference in temperature rise rate was also attributed to the rate of heat gain during phase 2 (Stage 2-HB1) is faster, resulting from different fineness and age of cement used during phase-1 and phase-2 test resulting in different rate of hydration. The various temperature rise curves, temperature differentials, and temperature distribution profiles are shown in Figures 12-16. Due to the greater heat loss rate at corner and side of the block, the temperatures at corner and side reach the peak before the maximum core temperature due to heat exchange to the environment. The center temperature begins to decrease when the heat generated in block is not enough to compensate the heat loss to the environment (as for the case of adiabatic condition). After the decrease of the core temperature, the temperature difference (T c -T n ) and (T c -T s ) gradually climb to the peak. This is normally regarded as the critical period for the surface 14

1234567890
International Conference on Materials Technology and Energy IOP Publishing IOP Conf. Series: Materials Science and Engineering 217 (2017) 012008 doi:10.1088/1757-899X/217/1/012008 concrete. Situations would become even worse if the formwork is removed during this stage. Also, the poor insulation could significantly reduce the time to reach the maximum temperature differences of (T c -T n ) and (T c -T s ). This infers that poorly insulated mass concrete could develop thermal cracks shortly after the placement.
The temperature at center is little affected by the environment variation due to the good insulation and big enough size of the hot box i.e. 1.5m x 1.5m x 1.5m, while the temperature difference (T c -T n ) and (T c -T s ) is affected in some degree. Essentially, during the course of insulation selection, it is necessary to consider the environment heat loss variation to minimize the effect. It is thus inferred that temperature gradient of actual concrete structures could be improved by additional insulation at edges and corners. The temperature gradient along the regions close to the side is much steeper than that in the central region. This suggests that the locations of the thermocouples in the concrete mock-up block should be carefully defined so that the comparable results can be obtained. Figures 12-14 showing how the temperature gradients formed after the concrete were cast into hot box. Temperature differential formed from the very beginning of the temperature rise in the concrete mass, whereby the temperature is also affected by the ambient temperature fluctuation, which is formed by the temperature change from daytime to night-time. The peak value of (T c -T n ) and (T c -T s ) occurred after T c .
This implies that the maximum temperature difference was encountered in the cooling down period of the concrete mass. The temperature differences of between centre core versus surface (T c -T s ) and between core versus corner (T c -T n ) were also demonstrated in below Figure 15 and Figure 16, which showing how the temperature at the corner of the block had the lowest value while the temperature at center gave the highest. (T c -T n ) is always greater than (T c -T s ). This may be explained as that the heat loss in the corner of the block is multi-dimensional. The peak value of (T c -T n ) and (T c -T s ) occurred after T c .   Figure 15. Comparison of temperature differential Figure 16. Comparison of temperature between Tc-Ts for Stage 1-HB1, Stage 1-HB2, differential between Tc-Tn for Stage 1-HB1 and Stage 2-HB1 and Stage 2-HB1

Temperature differential
The maximum differential temperature (Temperature recorded at point A1 minus temperature recorded at point A3) for stage 1-HB2 at the open top surface was found to be ~ 10.13°C, measured on 21st March 2011 at 09:00 am, about 72 hrs after the concrete placement. The top thermocouple was located 100mm below the surface. However an adjustment is required to determine the true differential between the centre and the surface. This adjustment is made by assuming the temperature profile follows a parabolic curve to the cooling surface as shown in Figure 17. With a peak temperature of 44.43°C as recorded, the difference between the surface and the point of location of the thermocouple at 100mm depth is about 2.3°C. This must therefore be added to the measured differential. Hence the true differential is 10.13 + 2.3 = 12.43°C ≈ 12.5°C. This is significantly less than the Contract Documents specified 20°C and the proposed differential of 27.7°C for granite aggregate concrete as in accordance with BS8110-2: 1985, Table 3.2-Estimated limiting temperature changes to avoid cracking. Figure 17. Adjustment for temperature differential at top surface of 1.3m x 1.3m x 1.3m hot-box, Stage 1-HB2