High Temperature Performance of Self-Compacting Concrete Containing Boron Active Belite Cement

In the present study, CEM I 42.5 R and CEM II B-M (P-LL) 32.5 R and BABC type of cements were used in accordance with TS EN 197/1 [37] standard. Physical, mechanical and chemical properties of three types of cement are given in Table 1. The main physical phase of cement paste, the calcium–silicate–hydrate (C–S–H)₂ phase, is a random complex composite at the nanometer scale. Ettringite is another phase that can appear in some quantity in some cement pastes, but probably only plays a minor elastic role [38, 39]. Belite cement contains the belite phase – 2CaO·SiO₂(C₂S) in the amount of 40 to 50%. Its main disadvantage is the low early strength due to the poor hydraulic activity of the belite. The hydration rate of the belite in Portland cement after 28 days may be even four times lower compared to the alite phase. The belite-phase consists of 5 polymorhps namely, α, α^ıH, α^ıL, β and γ. The stable-phase is γ–2CaO·SiO₂(γ-C₂S) at the temperature below 650°C. γ-phase is transformed into β-phase at the temperature of 675°C. However, β-phase is stable between the temperature range of 675°C and 1420°C. β-phase is changed into α-phase in the temperature range above 1420°C. β-phase is stable between the temperature range of 1410°C and 2130°C. The ability to react with water of belite phases is different from each other. The ability to react with water of β-phase is the best, which the γ-phase usually does not hydrate. β-C₂S phase or α-C₂S phase can be transformed into the γ-C₂S phase while the fired belite clinker is cooled. Therefore, the stabilization of those two phases is a very important issue to prevent their transformation into γ-C₂S phase [40, 41].

Polycarboxylic ether polymer based Draco-Levelcon FX10 hyperplasticizer was used in concrete mixtures at the maximum value of 2% at the end of preliminary tests to ensure the workability of all concrete samples. This product is particularly suitable for the production of high–performance concrete mixtures with high-water reduction capacity and excellent fluidity required for consistency-retaining.

The fine aggregate was 0–4 mm and the coarse aggregate was 4–16 mm and they conformed to the according to TS 706 EN 12620 [42] and TS 802 [43] standards. Maximum grain diameter of the aggregate used in the experiments is 16 mm. Aggregates were classified according to 0–4 and 4–16 for fine and coarse aggregates. The granulometry curve was prepared using coarse and fine aggregate in the volume rates of 38% and 62% in concrete mixtures, respectively. Since the properties of the concrete mix are significantly affected by aggregate granulometry, a single granulometry curve was used in the mixtures. Aggregate granulometry was used as a granulometry adjusted according to TS 802 standard [43]. The granulometry curve used in the concrete mixtures used in the experiments is presented in Figure 1.

Figure 1.

Water/binder ratio (w/b) was adopted to be 0.36 for all concrete mixes. The silica fume (SF) ratio used in the mixtures was taken by replacing with cement in the rate of 7.5%. Hyper plasticizer was used up to 2% of the cement weight in order to ensure sufficient fluidity in the mixtures.

Concrete mixes were prepared for three cement types in convenience with TS 802 [43] standards. The dosage of binder in the test mixtures was fixed, and the air volume was preferred as 1%. In addition, the total binder rate used was 500 kg and 7.5% SF were replaced with cement. Mixing ratios are given in Table 2. A laboratory type concrete mixer was used while preparing the mixtures. The mixing process in the concrete mixer was completed in 5 minutes according to ASTM C 192 [44].

No rodding or tamping was performed when placing concrete into the molds. The samples stored under laboratory conditions for 24 hours and after removing the casts were transferred in a cure pool saturated with lime and set at a temperature of 23 1°C. The samples were taken out of the cure pool one day before the tests and stored under laboratory conditions again for one day, and then the experiments were completed. For high temperature tests, the samples were taken out of water after 28 days of curing time and unit weight, UPV, splitting tensile strength and compressive strength tests were conducted after 3 hours of exposure to 100, 200, 400, 600 and 750°C. Three samples were produced from each sample in the experimental program, and the average of the results measured from the three samples was used in the relevant tables and graphics.

The flow diameter test was conducted in order to measure the self-flowing ability of concrete. In this method, the concrete is permitted to flow freely across a flow table by letting it fall through a cone with its own weight without any obstacles. In order to determine the fluidity of SCC and its ability to pass through reinforcements, the blocking risk of concrete was evaluated with the L-box test. V-funnel test was carried out in order to measure the viscosity and transition ability of SCC. With this experiment, the passing ability of the concrete was examined by providing flow with its own weight from a narrow section, as the greatest aggregate diameter below 20 mm.

The 100x200 mm cylindrical specimens were used conveniently with TS EN 12390-3 [45]. All samples were tested under a constant loading rate. The load on the samples increased consistently with a uniform rate of approximately 3.16 kg/cm²/s and the compressive strength related values were obtained from the digital display of the press when the sample was fractured. All results are the average of three samples for each group. While the ultrasonic waves were sent from one surface to the other surface of the concrete sample, they encountered gaps on their paths. Since these waves cannot cross the gap as fast as a solid section, they will go through the gap. Multiple repetitions of this action will increase the distance the ultrasonic waves must go between predetermined points. The decrease in the speed of ultrasound indicates strength of the material, also. The speedy ultrasonic wave means a full section without voids, which generally decreases the strength for conventional concretes. Direct measurement method was used to determine ultrasonic velocities.

Ultrasonic velocity measurements were conducted using 100x200 mm cylindrical specimens according to ASTM C597-09 [46].

Three cylindrical specimens with 100 x 200 mm dimensions were used in convenience with TS EN 12390-5 [47] in order to determine the splitting tensile strength, also.

High-strength SCC is of low fire resistance due to its relatively compact structure. Under high temperature, the free water in the body of the concrete sample and the chemically bonded water in the cement paste evaporates, but due to the dense structure of high-performance concretes, evaporation becomes relatively hard. As a result of this phenomenon, spills appear on the concrete surface, and the reinforcement steel cannot be protected from high temperatures. For high temperature tests, the relevant samples are taken out of the curing tank at the 27th day and at the 28^th day they are transferred to the furnace to exposure to the at 100, 200, 400, 600 and 750°C for 3 hours, then cooled to ambient temperature and subjected to unit weight, UPV, splitting tensile strength and compressive strength tests.

According to the concrete mixture types, flow diameter, flow time to 50-cm diameter, L-box, V-funnel flow time, and J-ring tests were performed immediately after the concrete mixing. Results are presented in Table 3. The effect of different cement types on workability was determined due to the use of different cements and hyper plasticizer in a fixed ratio in all mixtures. Although, the cement ratio was fixed, the variation of cement type influenced the mentioned flow diameters, as expected. Figure 2 shows the flow of the BABC mixes.

Figure 2.

It was found that SCC represents suitable viscosity properties within its general criteria. The low viscosity causes the concrete to undergo segregation on moving with its own weight, while the high viscosity makes it difficult for the concrete to move through the narrowest reinforcements. It was seen when the test results were evaluated that the highest viscosity value was obtained in the CEM II mixture and the lowest viscosity value was in the BABC mixture. Replacing SF with cement as a mineral additive also increased viscosity. This can be attributed to the lower specific gravity of SF, compared to cement. The material with lower specific gravity needs more water to provide the similar strength values.

The J-ring value in the BABC cement mixture was measured to be 10 mm, while 20 mm and 40 mm in CEM I and CEM II cement mixtures. It was stated that for the 300 – mm diameter J-ring proposed in EFNARC [48], J-ring values greater than 10 mm indicate the tendency to block in the mixture. Withal, there is no blocking tendency for the BABC mixes and there may be a blocking tendency for the others.

In this section hardened properties of the concrete samples, such as compressive strength, splitting tensile strength, and UPV were evaluated. The results of the mechanical properties of SCC produced in three different groups are given at Table 4. Lower w/b ratios increased high compressive strength values due to reasons like minimizing the pores by adding mineral additive, SF, within the concrete mixes. The observed 28^th days compressive strength values are obtained from greatest to the least is as BABC, CEM I and CEM II included samples, respectively. The compressive strength of the samples produced with BABC is 4.75% and 51.7% greater than the ones with CEM I and CEM II type cement, respectively.

The observed tensile splitting strengths for all three groups are presented at Table 4. The splitting tensile strength of the samples were found to be in the range of 5.13–3.40 MPa. Although the w/b ratio was the constant for all compared groups of samples. As a result, the SF addition improved the splitting tensile strength for all samples, as expected like the compressive strength values. The observed minimum tensile strength was for the samples including CEM II type cement, while the samples with BABC present the maximum. Withal, the samples with BABC presented the highest UPV result. The UPV results of the BABC samples are 4.81% and 9.56% greater than the ones with CEM I and CEM II cement, respectively. As shown in Table 4, the unit weights of the samples are very close to each other. The SF addition as a replacement of the cement, resulted in a decrease in unit weights. The relatively lower specific gravity of the SF comparing to the cement’s resulted in this behavior, as expected. Withal, the lowest and the greatest unit weights are observed for the samples manufactured with CEM II and BABC, respectively.

In this section, compressive strength, tensile splitting strength, UPV, and weight loss result from the SCC samples subjected to six different temperatures that were evaluated. In order to determine the effectiveness of BABC, CEM I and CEM II in concrete samples were subjected to 100°C, 200°C, 300°C, 400°C, 600°C and 750°C, at the age of 28^th days, in accordance with the relevant standards. The results are presented at Table 4. In addition, all samples exposed to high temperature are compared with the control samples at ambient conditions, 20°C.

Aydın et all. [21] reported that although the 3-day and 7-day compressive strengths of concrete with BABC were lower than the samples produced with Portland Cement, the 28-day and 90-day related ones with BABC were greater than the samples produced by the others. Figure 3 shows the compressive strength results of the samples exposed to high temperatures. As the temperature increased, the compressive strength values decreased, for all the samples, as expected. When the SCC samples produced with BABC cement were exposed to 100, 200, 300, 400, 600 and 750°C, the compressive strength values decreased by 3%, 22%, 23%, 28%, 55% and 82% compared to the control samples. However, the compressive strength values of the SCC samples with CEM I were decreased about 11%, 13%, 26%, 29%, 63% and 87%. Similarly, for the SCC samples with CEM II, the decrements were about 10%, 16%, 24%, 31%, 59% and 88%, respectively. Above-mentioned results indicate that, the BABC samples have less strength loss than the other ones. Among all the samples, the best performance was observed for the SCC mixture produced with BABC, and the strength losses of the concretes produced with the CEM I and CEM II were similar. Davraz et all. [49] stated that the compressive strength of concrete with BABC decreased in the temperature range of 500°C. Hence, it suddenly decreased at the 750°C. Liu and Zheng [50] stated that although the early strength of concrete (3-day and 7-day) with BABC was low, it increased compressive strength at 28 days.

Figure 3.

All SCC samples lost their strength to a great extent, especially when exposed to a temperature of 750°C. When compared to conventional concrete samples with the same w/b ratio and strength, the damage caused by high temperature in SCC is greater. This situation may be attributed to the destruction caused by the increased vapor pressure due to the high internal pressure under the effect of high temperature as a result of the compact microstructure of the SCC samples. The decrease in compressive strength occurring at temperatures exceeding 200°C can be attributed to the loss of water in the alumina and iron oxide components within the cement, causing a decrease in the strength values, especially for the compressive ones. In addition, the volatility of bound water which does not evaporate under normal conditions at temperatures exceeding 300°C, increases the degree of damage and decreases the sample’s compressive strength. As it is known, Ca(OH)₂ is one of the components found in cement paste in concrete. The conversion of Ca(OH)₂ to quick lime (CaO) occurs above 400°C. The transformation of lime in concrete into quicklime means a shrinkage of approximately 33%. Parasitic stresses that occur in concrete, which shrinks and expands in a short time, cause the damage to grow. At temperatures exceeding 400°C, it is seen that the C–S–H begin to be destroyed, the strength of the concrete decreases rapidly, and around 900°C the C–S–H structure is completely dissolved. Aygörmez et all. [51] showed that the conversion from Ca(OH)₂ to CO begins around 400°C and the C–S–H structures is quickly destroyed as it approaches 600°C.

Tang et all. [52] reported the tensile strength of concrete decreases after being subjected to the high temperature above 400°C and these characteristics are closely related to the expansion of mica in iron tailings and the dehydration and decomposition of cementing material hydration products, such as ettringite and C-S-H phase at high temperature. However, Cao et all. [53] and Abid et all. [54] stated that the dehydration of calcium hydroxide and damage of C-S-H bond at 400°C to 600°C lead to the decrease of strength and decomposition of calcium carbonate and C-S-H structure destroy the strength of cementitious composites completely.

The result that fly ash used as a substitute with cement in SCC production performs better at high temperatures compared to other mineral additives can be attributed to the fact that the tobermolite gel formation as a product of fly ash and lime at high temperatures and pressure is two or three times stronger and durable than C-S-H gel. SCC, which is among high-performance concretes, has a tighter micro structure than conventional ones. And therefore, the concrete cover discharges may occur as a result of the damage caused by the increased vapor pressure due to the high internal pressures under the effect of high temperature. It was observed that these cover discharges are relatively less for the BABC concrete samples.

The low tensile strength of concrete is the one of the most important problem for building structures. However, the researchers and civil engineers work to eliminate this disadvantage. The use of the BABC in the concrete mixture decreased the splitting tensile strength of conventional concrete. Figure 4 represents the tensile splitting strength results of SCC samples exposed to high temperatures. It is observed that as the temperature exposure of SCC specimen’s increases, splitting tensile strength decreases. When specimens with BABC cement are exposed to temperatures at 100°C, 200°C, 300°C, 400°C, 600°C and 750°C, the splitting tensile strength values decreased in the rates of 1%, 20%, 22%, 27%, 54% and 81% respectively. While this decrement is about 8%, 9%, 23%, 27%, 63% and 86%, for the CEM I including samples, respectively, it is understood that BABC samples have less strength loss than the others. Among all the samples, the SCC samples with BABC gave the best results. The strength losses of the concretes produced with CEM I, and CEM II were similar. This also indicates the parallel strength loss of the compressive and splitting tensile strength, as expected.

Figure 4.

The UPV test was employed due to assessments of the permeability and microstructure of samples for inspected cement types, CEM I, CEM II, and BABC. As known, the variations in velocity are related to physical changes in the microstructure of the samples. Figure 5 shows the UPV results of the SCC samples exposed to high temperatures. It was observed that UPV values decreased as the temperature increased. These reductions were found to be 1%, 8%, 12%, 31%, %59 and 63% respectively, for the samples exposed to 100, 200, 300, 400, 600 and 750°C compared to control samples. Withal, the control samples were 0%, 8%, 15%, 30%, 60% and 64%, respectively, compared to CEM I cement samples, and finally, the control samples were 0%, 12%, 20%, 26%, %62 and %69 respectively, compared to CEM II cement samples. As seen in Figure 5, UPV values decreased seriously at temperatures above 400°C. The wave velocity passing through the concrete is also affected by the cracks in the concrete at high temperatures. Another reason for the decrease in UPV with the increase of temperature is a measure of the progress of the cracks within the composite/material. Aygörmez et all. [51] and Öz et all. [55] showed that increasing the temperature have an important effect on the UPV values of concrete which affects unfavorably. As the water in the concrete evaporates with the increasing temperature, the volume of the pores in the microstructure increase. The voids that leads to lower UPV values are formed instead of evaporating water.

Figure 5.

The unit weight results of the SCC samples exposed to high temperatures are presented in Figure 6. It was observed that the unit weight values decreased as the temperature in the mixtures increased. In BABC cement samples exposed to high temperatures of 100, 200, 300, 400, 600 and 750°C, these reduction amounts were 0%, 1%, 3%, 5%, 6% and 8%, respectively, for CEM I cement samples 0%, 2%, 3%, 5%, 6% and 11% and for CEM II cement samples 0%, 2%, 3%, 4%, 8% and 9%, respectively. With the dehydration reaction, moisture in the matrix moves to the surface of the sample, causing internal damage to the microstructure and, consequently, weight loss. It was observed that a serious weight loss occurred in all samples as the temperature increased.

Figure 6.

The main reason for weight loss at the high temperature is the piece spalling the surface of concrete. However, the greatest weight loss at 600°C is in CEM II type cement. After being exposed to a temperature of 600°C, the weight loss of CEM I type cement is more than CEM II type cement. The best and the least weight loss performance in samples exposed to high temperature was obtained with BABC samples.

The weight loss of samples exposed to high temperature is close to each other up to 400°C. Nevertheless, being exposed to a temperature of 400°C, the difference between weight loss increases. Since the void ratio and impermeability of the samples produced with BAB cement is less, it has the compact structure. Therefore, the samples with BAB cement exhibited better weight loss performance than samples produced with conventional cements used. Aygörmez et all. [51] showed that the effect of high temperature on the moisture in the matrix decreased the weight loss of concrete. The biggest weight loss occurred between 400°C and 600°C. The most important point of this mechanism is due to the evaporation of free water and condensed hydroxyl groups [19, 51, 56].