This study conducted ACI mix design using a 0.5 water-cement ratio and a slump of 16 cm and created specimens with 0%, 1%, 3%, and 5% weight of cement replaced by nano-clay (numbered as W5N0, W5N1, W5N3, and W5N5, respectively). The mix proportions are shown in Table 1. Materials used included Type I Portland cement; coarse aggregate with maximum particle diameter of 3/4", 2.54 specific gravity with saturated surface dry, a 1.25% water absorption capacity, and 1475.4 kg/m³ dry-rodded unit weight; fine aggregate with a fineness modulus of 2.75, 2.53 specific gravity with saturated surface dry, and a 2.15% water absorption capacity. The nano-clay used in the tests was light yellow in color with purity greater than 95% and primarily consisted of SiO₂, Al₂O₃, Fe₂O₃, MgO, Na₂O, and TiO₂, with an average particle diameter smaller than 10 µm, pH value between 8 - 9, Portland suction amount of 142.5 mmol/100g, cation exchange capacity equal to 98, and moisture content less than 3.5%.

There are three test items:

1) Compressive strength of concrete at different curing ages

A ψ10 × 20 cm cylinder specimen is made according to the mix proportion, stripped on the following day and cured in saturated limewater. The compressive strength on Days 7, 14, 28, 49, 56 and 90 is measured respectively.

2) Compressive strength of concrete in hot environment

A ψ10 × 20 cm cylinder specimen is made according to mix proportion, stripped on the following day and cured in saturated limewater for 28 days for simulation test for the effect of different high temperatures on the compressive strength. The specimen is taken out of the water and put in 105˚C oven for 3 days, and then put in high temperature furnace, the specimen is exposed to 25˚C, 300˚C, 440˚C, 500˚C, 580˚C, 800˚C and 1000˚C respectively. It is indicated in literature that the compressive strength of concrete decreases mainly in one-hour delay under the effect of high temperature [5]. Therefore, there is a delay of one hour after the temperature reaches the preset temperature, the compression test is implemented when it is cooled to room temperature. All data are mean values of test results in triplicate.

The cement paste is extracted from the damaged specimen (including a small amount of fine aggregate in small particle size which cannot be removed) to test the ignition loss. The ignition loss test procedure is: 1) take 10 g sample, weighed W₂₅; 2) put in 105˚C oven for 24 hours, weighed W₁₀₅ after cooling; 3) put in full

electronic high temperature furnace, heated to 440˚C and delayed 8 hours, weighed W₄₄₀ after cooling; 4) heated in full electronic high temperature furnace to 580˚C and delayed 8 hours, weighed W₅₈₀ after cooling; 5) heated in full electronic high temperature furnace to 1007˚C and delayed 8 hours, weighed W₁₀₀₇ after cooling; 6) calculate ignition loss (IL), expressed as Equation (1).

3) Thermal conductivity of concrete in hot environment

A 15 × 15 × 24 cm square angle column specimen is made according to mix proportion, stripped on the following day and cured in saturated limewater for 28 days. This test only considers the longitudinal transfer of heat in concrete. Two 15 × 15 cm surfaces are heating surface and radiating surface, wrapped in refractory wool with good heat insulation to avoid the heat energy in the concrete dissipating from sides. The temperature is increased to 800˚C at heating rate of 5˚C/min. The temperature data acquisition unit records the temperature change per second. The heat transfer coefficient is calculated by Equation (2).

where k = heat transfer coefficient (W/m-˚K); Q = heat-flow rate (W); A = cross-section area normal to direction of heat flow (m²); dT/dx = temperature gradient of direction of heat flow (˚K/m).

The test results are shown in Table 2 and Figure 1. For normal concrete (W5N0), the concrete strength increases with the curing age, the compressive strength on Day 28 is 33.7 MPa, the strength on Days 7 and 14 is 73.6%

and 83.4% of the strength on Day 28 respectively. The compressive strength on Days 56 and 90 is increased by 9.2 and 12.5% respectively, meaning the strength grows rapidly during the first 28 days of curing, the strength develops slowly after 28 days. The concrete with a small replacement of nano-clay for cement (W5N1, W5N3 and W5N5) has a strength development trend similar to normal concrete, meaning a small amount of nano-clay has no significant impact on the time varying development trend of concrete strength.

Take the curing age of 28 days as an example, 0.1%, 0.3% and 0.5% nano-clay replace cement by weight, the strength is reduced by 11.0% and increased by 5.9% and 3.3% respectively. The strength of specimens at other ages has similar trends, meaning a proper amount of nano-clay replacing cement by 0.3% - 0.5% can increase the concrete strength, and the 0.3% replacement increases the concrete strength significantly, the 0.1% nano- clay replacement for cement reduces the concrete strength.

The test results are shown in Table 3 and Figure 2. For normal concrete (W5N0), there is a delay of one hour at 25˚C and 300˚C, the strength is increased by 13.4% and 9.5% respectively compared with the strength on curing Day 28. This result matches the result of Ref. [6] that the residual strength of concrete is not reduced apparently from normal temperature to 300˚C. However, when one hour delayed in 440˚C, 500˚C, 580˚C, 800˚C and 1000˚C environment, the strength is reduced by 14.2%, 26.7%, 46.9%, 76.3% and 93.8% respectively compared with the strength on curing Day 28. This result matches the result of Ref. [6] that 400˚C - 600˚C is the main strength reduction interval, the residual strength of concrete is mostly lower than 30% of the original strength when the temperature is higher than 700˚C.

Figure 2 shows that the effect of hot environment on the strength trend of nano-clay concrete is similar to that on normal concrete. In different hot environments, the 0.1% nano-clay replacement for cement can reduce the concrete strength, the 0.3% - 0.5% replacement for cement can increase the concrete strength, and the 0.3% replacement increases the concrete strength significantly.

Table 4 shows that the heat transfer coefficient of concrete decreases as the ambient temperature rises, matching the findings of Ref. [7]. Figure 3 shows that the maximum decreasing amplitude occurs at 200˚C - 300˚C, and then the decreasing amplitude decreases as the temperature increases. When the temperature exceeds 100˚C, the pore water in the concrete evaporates gradually, the heat transfer is slowed down. When the temperature is higher than 300˚C, the thermal strain difference between the aggregate and hardened cement paste results in microcracks, increasing the heat transfer paths, and the C-S-H colloid is decomposed gradually, the thermal conduction capability is reduced as a part of aggregate is damaged, so that the heat transfer coefficient decreases faster. When the temperature is higher than 600˚C, the heat is transferred between pores mainly by radiation [8]. The 0.1% nano-clay replacement for cement can reduce the heat transfer coefficient of concrete, and the 0.3% and 0.5% nano-clay replacement can increase the heat transfer coefficient of concrete.

Figure 4 shows the relationship between heat transfer coefficient and compressive strength. It is observed that the heat transfer coefficient decreases with the compressive strength. The decrease of the compressive strength in Figure 4 is resulted from high temperature. Therefore, hot environment reduces the compressive strength and heat transfer coefficient of concrete. According to Table 2 and Table 3, based on the compressive strength of concrete on the curing Day 28, when the ambient temperature is 800˚C, the compressive strength of W5N0, W5N1, W5N3 and W5N5 concrete is reduced by 76.1%, 80.1%, 75.4% and 75.4% respectively. According to Table 4, based on the heat transfer coefficient of concrete at ambient temperature of 200˚C, when the ambient temperature is 800˚C, the heat transfer coefficient of W5N0, W5N1, W5N3 and W5N5 concrete is reduced by 64.4%, 63.6%, 61.8% and 62.3% respectively.

The test results are shown in Table 5 and Figure 5. For normal concrete (W5N0), when the temperature is 300˚C to 580˚C, the mortar ignition loss decreasing amplitude is increased markedly. Because when the temperature is 300˚C - 500˚C, most of bonding water of C-S-H colloid has been lost, and Ca(OH)₂ begins to be decomposed at 500˚C - 580˚C [4]. The 0.1% - 0.5% nano-clay replacement for cement has no significant impact on the paste ignition loss trend at high temperature.

Figure 6 shows that the compressive strength increases with ignition loss. Because the ignition loss of paste reflects the degree of cement hydration, and the hot environment results in chemical reaction of cement hydration product. When the temperature is 200˚C, the bonding water begins to deviate from C-S-H colloid. The paste has lost most water when the temperature is 500˚C, and Ca(OH)₂ begins to be decomposed when the temperature is 500˚C - 580˚C, so that the ignition loss of paste is reduced, namely, the degree of cement hydration is reduced, the higher the ignition loss is, the higher is the degree of cement hydration.

When the water-cement ratio of paste is 0.5, and the nano-clay replacement for cement is 0% - 0.5%, the ignition loss is approximately in exponential relationship to the compressive strength when the ignition loss is lower than 6%, y = 1.267e^0.443IL, the correlation factor (R²) is 0.969; when the ignition loss is higher than 6%, the ignition loss is approximately in logarithmic relationship to the compressive strength, y = 22.832ln(IL) - 22.003, the correlation factor is 0.920. Therefore, the compressive strength of concrete can be calculated by ignition loss, or the ambient temperature of concrete can be calculated by ignition loss (Figure 5).