Experimental Study on Mechanical Properties and Compressive Constitutive Model of Recycled Concrete under Sulfate Attack Considering the Effects of Multiple Factors

Abstract

To investigate the mechanical properties and a compressive constitutive model of recycled concrete under sulfate attack considering the effects of multiple factors, two waste concrete strengths (i.e., C30 and C40), four replacement ratios of recycled coarse aggregates (i.e., 0, 30%, 50% and 100%), and two water–cement ratios (i.e., 0.50 and 0.60) were considered in this study, and a total of 32 recycled concrete specimens were designed and tested. The results indicated that the failure processes and patterns of recycled concrete were not significantly influenced by the replacement ratio of recycled coarse aggregates, the waste concrete strength, the water–cement ratio, or sulfate attack. The higher the replacement ratio of recycled coarse aggregates and the water–cement ratio and the lower the waste concrete strength, the more obvious the reduction in cubic compressive strength, with a maximum reduction of 38.48%. A prediction model for the cubic compressive strength of recycled concrete under sulfate attack was proposed. The higher the replacement ratio of recycled coarse aggregates and the water–cement ratio and the lower the waste concrete strength, the more significant the reduction in axial compressive strength, with a maximum reduction of 37.82%. A prediction model for the axial compressive strength of recycled concrete under sulfate attack was established. A compressive constitutive model of recycled concrete under sulfate attack considering the effects of the replacement ratio of recycled coarse aggregates, the waste concrete strength, and the water–cement ratio was established. The pore structure of recycled concrete was significantly destroyed by the expansion stress generated by Na₂SO₄ crystals: a large number of Na₂SO₄ crystals were attached to the surface of concrete matrix, and the concrete matrix became loose. The research results can provide a theoretical basis and data support for engineering applications of recycled concrete.

1. Introduction

In recent years, the process of urbanization has accelerated worldwide, and the world construction industry has entered a stage of rapid development. However, a lot of construction waste is generated during this process. Due to the problems of recycling technology and cost, serious resource, environmental, and economic problems are brought by the low recycling rate of construction waste [1,2]. The majority of construction waste is waste concrete, so how to realize the reuse of waste concrete has become a problem that countries all over the world compete to study and solve [3,4,5,6,7,8,9]. The emergence of recycled concrete technology reduces the emission of construction waste and the dependence on natural resources, and promotes the sustainable use of resources [10,11]. Therefore, recycled concrete technology plays a positive role in promoting the development of the construction industry [12,13,14,15,16], and great development prospects and application scales are characterized in the application of recycled concrete in building materials and structures.

Mechanical properties and durability are directly related to the whole life cycle of recycled concrete materials and structures. Although recycled concrete has great potential in the field of sustainable construction, its performance still lags behind that of ordinary concrete. The damage to recycled aggregates is caused by the mechanical crushing process of waste concrete to some extent, and the old mortar is attached to the surface of the recycled aggregates. Therefore, compared with natural aggregates, recycled aggregates are characterized by defects of large porosity, high water absorption, and low strength [17,18,19,20]. The pore structure and interface transition zone of recycled concrete are badly affected by these defects, so the mechanical properties and durability of recycled concrete are lower than those of ordinary concrete under the same conditions. Therefore, the feasibility of recycled concrete in engineering applications is restricted [21,22]. The strength and stress–strain curves of recycled concrete are the main parameters for evaluating the bearing capacity and deformation characteristics of recycled concrete components and structures, which directly affect the accuracy of the calculation results [23]. Nearly three-quarters of the Earth’s surface is covered by oceans, and long coastlines and large areas of saline land are located in many countries. China has a vast sea area with tens of thousands of kilometers of coastline, and a large number of salt lakes and saline–alkali land are distributed in the coastline, which contain relatively high concentrations of sulfate ions [24]. The mechanism of sulfate attack is complex and very destructive to buildings [25]. Marine and lake environments are affected by dry–wet cycles, and water movement is very complex, which accelerates the corrosion and destruction of buildings [26]. According to statistics, at the end of the last century, many reinforced concrete buildings in developed countries were forced to retire early due to severe sulfate attack [27]. Therefore, the application of recycled concrete in these areas with significant sulfate attack will face great risks. If recycled concrete is widely used, its mechanical properties under sulfate attack must be investigated and a corresponding constitutive model should be established.

The mechanical properties of recycled concrete under sulfate attack have been studied by some researchers. Wang et al. [28] studied the microstructure of the interface transition zone of recycled concrete after sulfate attack, and the results indicated that the higher the strength of recycled concrete, the higher the micro-strength and the smaller the width of the interface transition zone. Thus, the resistance to sulfate attack of the interface transition zone can be improved by increasing the strength of recycled concrete. Qiao et al. [29] studied the effects of the replacement ratios of recycled fine aggregates and recycled coarse aggregates on the resistance to sulfate attack of recycled concrete. Jia et al. [30] carried out sulfate dry–wet cycle tests on recycled concrete and ordinary concrete where the apparent appearance, quality, and strength of concrete were tested, and the results showed that the resistance to sulfate attack of recycled concrete decreased with the increase in the replacement ratio of recycled coarse aggregates. Du et al. [31] prepared recycled concrete with different replacement ratios of recycled coarse aggregates (0%, 30%, and 50%) and recycled fine aggregates (0% and 15%), and the resistance to sulfate attack of recycled concrete was evaluated by measuring the change in macroscopic mechanical indicators after sulfate dry–wet cyclic tests. Tian [32] prepared recycled concrete with 30% recycled coarse aggregate content and water–cement ratios of 0.35, 0.45, and 0.55, and the effect of water–cement ratio on the resistance to sulfate attack of the recycled concrete was studied through macroscopic and microscopic tests. Fu et al. [33] designed four groups of concrete specimens with different replacement ratios of recycled coarse aggregates, and sulfate attack tests were conducted. By measuring mass, relative dynamic modulus, and compressive strength, the effect of replacement ratio of recycled coarse aggregates on the resistance to sulfate attack of recycled concrete was studied. Yan [34] conducted experimental research on the resistance to sulfate attack of recycled concrete, and the results showed that reducing the water–cement ratio and adding fly ash not only improved the pore structures of recycled concrete but also enhanced its sulfate resistance. Liu et al. [35] studied the cubic compressive strength, the splitting tensile strength, and the mass loss rate of recycled concrete under sulfate attack, considering the factors of replacement ratio of recycled coarse aggregates and water–cement ratio, and a prediction model for the service life of recycled concrete under the dry–wet cycles of sulfate was established. It is clear that most of the current research about the mechanical properties of recycled concrete under sulfate attack is based on experimental results considering the influence factors of replacement ratio of recycled aggregates and water–cement ratio. It is well known that many factors, such as replacement ratio of recycled aggregates, waste concrete strength, and water–cement ratio, simultaneously affect the resistance to sulfate attack of recycled concrete, and a constitutive model of recycled concrete under sulfate attack considering the effects of multiple factors has yet to be proposed. Therefore, the mechanical properties and a constitutive model of recycled concrete under sulfate attack should be investigated by considering the replacement ratio of recycled coarse aggregates, the waste concrete strength, the water–cement ratio, and so on, which can provide strong support for the design, theoretical calculation, and finite element analysis of recycled concrete components and structures.

In order to study the mechanical properties and compressive constitutive model of recycled concrete under sulfate attack, two waste concrete strengths, four replacement ratios of recycled coarse aggregates, and two water–cement ratios were considered, and a total of 32 recycled concrete specimens were designed and tested. Firstly, the failure patterns of recycled concrete under sulfate attack during the failure process of compression were analyzed. Secondly, the effects of the replacement ratio of recycled coarse aggregates, the waste concrete strength, and the water–cement ratio on the cubic compressive strength of recycled concrete under sulfate attack were studied through tests, and prediction models of the cubic compressive strength and the corrosion resistance coefficient of recycled concrete considering the effects of multiple factors were proposed by fitting the experimental data. Thirdly, the effects of the replacement ratio of recycled coarse aggregates, the waste concrete strength, and the water–cement ratio on the axial compressive strength of recycled concrete under sulfate attack were investigated through tests, and a prediction model of the axial compressive strength of recycled concrete considering the effects of multiple factors was established by fitting the experimental data. Fourthly, based on the experimental data of compressive stress–strain curves and the constitutive relation of ordinary concrete, a compressive constitutive model of recycled concrete under sulfate attack was established considering the effects of the replacement ratio of recycled coarse aggregates, the waste concrete strength, and the water–cement ratio. Finally, scanning electron microscopy (SEM) was used to observe the microscopic morphology of recycled concrete under sulfate attack, and the degradation mechanism of mechanical properties of recycled concrete under sulfate attack was investigated.

2. Experimental Program

2.1. Raw Materials

The cement used in the test was ordinary PO 42.5R Portland cement produced by Huarun Cement Manufactory Co., Ltd., in Dongguan of China, and the chemical compositions of cement that were provided by the cement manufacturing plant are listed in

Table 1. Chemical compositions of cement (%).

SiO2	Al2O3	Fe2O3	CaO	MgO	SO3	Loss
22.25	4.98	3.47	64.84	0.81	0.81	2.84

. Natural river sand was used as fine aggregate with a fineness modulus of 2.8, apparent density of 2640 kg/m³, and mud content of 1.1%. Continuously graded artificial crushed stone with a particle size of 5–20 mm was adopted as natural coarse aggregate, and the physical properties are listed in

Table 2. Physical properties of natural coarse aggregates.

Apparent Density (kg/m3)	Mud Block Content (%)	Mud Powder Content (%)	Needle and Flake Particle Content (%)	Crushing Index (%)
2650	0.2	0.6	7	14

. The mixing water was tap water.

2.2. Preparation of Recycled Coarse Aggregate

The strengths of waste concrete were C30 and C40, which were taken from an abandoned concrete bridge. According to the China Standard GB/T 25177 [36], after manual crushing, cleaning, and grading of waste concrete, recycled coarse aggregates with a particle size of 5–20 mm were obtained, as shown in Figure 1. Compared with the natural coarse aggregates, the surface of the recycled coarse aggregates was rougher, more angular, and attached old mortar. Compared with the recycled coarse aggregates from the waste concrete with a strength of C40, the recycled coarse aggregates from the waste concrete with a strength of C30 had more corners and were rougher. The distribution curves of pore diameters of recycled coarse aggregates were measured by a high-performance automatic mercury injection instrument, as illustrated in Figure 2. The indexes of physical properties of recycled coarse aggregates from waste concrete with strengths of C30 and C40 are given in

Table 3. Physical properties of recycled coarse aggregates.

Recycled Coarse Aggregate	Apparent Density (kg/m3)	Crushing Index (%)	Water Absorption (%)	Porosity (%)
C30	2495	14	6.4	13.8
C40	2467	10	4.5	9.9

.

2.3. Preparation of Specimens

In this test, 32 recycled concrete specimens under different working conditions were designed. Among them, the cubic specimens with sizes of 100 mm × 100 mm × 100 mm were used to test the cubic compressive strength, and prismatic specimens with sizes of 100 mm × 100 mm × 300 mm were adopted to test the axial compressive strength and the compressive stress–strain curves. The specimens are detailed in

Table 4. Design of recycled concrete specimens.

Specimen Number	Waste Concrete Strength	Water–Cement Ratio	Replacement Ratio of Recycled Coarse Aggregates (%)	Corrosive Environment
NAC–N0.50	—	0.50	0	None
NAC–S0.50				Sulfate solution
NAC–N0.60		0.60	0	None
NAC–S0.60				Sulfate solution
RAC30–R30–N0.50	C30	0.50	30	None
RAC30–R30–S0.50				Sulfate solution
RAC30–R50–N0.50			50	None
RAC30–R50–W0.50				Tap water
RAC30–R50–S0.50				Sulfate solution
RAC30–R100–N0.50			100	None
RAC30–R100–S0.50				Sulfate solution
RAC30–R30–N0.60		0.60	30	None
RAC30–R30–S0.60				Sulfate solution
RAC30–R50–N0.60			50	None
RAC30–R50–W0.60				Tap water
RAC30–R50–S0.60				Sulfate solution
RAC30–R100–N0.60			100	None
RAC30–R100–S0.60				Sulfate solution
RAC40–R30–N0.50	C40	0.50	30	None
RAC40–R30–S0.50				Sulfate solution
RAC40–R50–N0.50			50	None
RAC40–R50–W0.50				Tap water
RAC40–R50–S0.50				Sulfate solution
RAC40–R100–N0.50			100	None
RAC40–R100–S0.50				Sulfate solution
RAC40–R30–N0.60		0.60	30	None
RAC40–R30–S0.60				Sulfate solution
RAC40–R50–N0.60			50	None
RAC40–R50–W0.60				Tap water
RAC40–R50–S0.60				Sulfate solution
RAC40–R100–N0.60			100	None
RAC40–R100–S0.60				Sulfate solution

, where “NAC” indicates ordinary concrete specimens, “RAC30” means the recycled concrete specimens made by recycled coarse aggregates from waste concrete with a strength of C30, “R30” is the recycled concrete specimens with a replacement ratio of recycled coarse aggregates of 30%, “0.50” indicates the recycled concrete specimens with a water–cement ratio of 0.50, and “N”, “S” and “W” are the non-corrosive environment, sulfate attack, and dry–wet cycles of water, respectively. For example, RAC40–R100–S0.60 refers to the recycled concrete specimen under sulfate attack, with recycled coarse aggregates from waste concrete with a strength of C40, with a replacement ratio of recycled coarse aggregates of 100%, and with a water–cement ratio of 0.60. The mix proportions of recycled concrete are listed in

Table 5. Mixing proportions of recycled concrete.

Water–Cement Ratio	Cement (kg/m3)	Water (kg/m3)	Sand (kg/m3)	Coarse Aggregates (kg/m3)
0.50	390	195	635.25	1179.75
0.60	325	195	658	1222

. After 24 h of pouring and molding, the specimens were demolded and placed in the laboratory for standard curing until 28 days.

2.4. Sulfate Attack Test

All specimens were taken out after curing for 28 days. Then, the specimens were soaked in Na₂SO₄ solution with a mass fraction of 5%. Dry–wet cyclic tests were simulated by soaking and natural air-drying the specimens, as shown in Figure 3. The dry–wet cycle lasted for a total of 14 days, with Na₂SO₄ solution soaking for 7 days, followed by natural air-drying for 7 days. The test process was planned to take 12 cycles, that is, the duration of the test was 168 days. The mechanical properties of recycled concrete may be affected by water and ${\mathrm{SO}}_4^{2-}$ in the solution [37,38]. In order to clarify whether the mechanical properties of recycled concrete were mainly affected by water or sulfate ions, the specimens under dry–wet cycles of water were compared in tests, with tap water-soaking for 7 days followed by natural air-drying for 7 days, and this process was repeated for a total of 12 cycles.

2.5. Compressive Strength, Compressive Stress–Strain Curve, and Scanning Electron Microscopy (SEM) Tests

A pressure testing machine with an extensometer installed was adopted in the test. According to the China Standard GB/T 50081 [39], the cubic compressive strength tests were carried out on cubic specimens, the tests of axial compressive strength and stress–strain curves were carried out on prismatic specimens, and the axial loading and vertical deformation during the tests were collected automatically by the computer. After the strength tests were completed, the samples for SEM tests were taken from the specimens, and the sizes of the samples were all within 5 mm. The selected samples were glued to the sample table with conductive adhesive, gold-plated, and then the samples were viewed by SEM. The samples were viewed after the instrument had been vacuumed to an appropriate degree, and the images of the samples were scanned and saved.

3. Results and Discussion

3.1. Analysis of Failure Patterns

Failure patterns of recycled concrete are shown in Figure 4, Figure 5 and Figure 6, indicated that no obvious differences were found between the failure process and patterns of recycled concrete specimens under various working conditions. At the initial stage of loading, no obvious cracks appear in the specimens. However, with the increase in load, the specimens begin to crack, finally with a “bang” sound, and the specimens are suddenly broken. The fracture zones are formed at roughly diagonal positions.

The failure patterns of prismatic specimens under sulfate attack are similar to those of the prismatic specimens under a non-corrosive environment, both of which form crack zones roughly along the diagonal position. However, compared with the prismatic specimens under a non-corrosive environment, the prismatic specimens under sulfate attack are accompanied by more small cracks under uniaxial compression, and the propagation of cracks is also more obvious. The reason for this phenomenon may be that sulfate reacts with hydration products in recycled concrete (such as calcium hydroxide, calcium aluminate, etc.) to produce expansive products such as gypsum and ettringite, and the volume of these products is larger than the hydration products, resulting in the increased internal stress. As a result, the growth and quantity of pores and micro-cracks are accelerated and increased, respectively, leading to the destruction of the internal structure of recycled concrete [31]. In addition, the dry–wet cycles may lead to the repeated entry and exit of moisture inside the recycled concrete, which will cause the stress changes within the recycled concrete and accelerate the speed of cracking [40]. During the wetting stage, water infiltration brings in more sulfate, increasing the likelihood of chemical reactions. During the drying stage, the evaporation of water may lead to the deposition of substances and an increase in internal pore pressure, further promoting the development of cracks. Therefore, more small cracks are developed in the prismatic specimens under sulfate attack during the uniaxial compression failure. In the axial compression tests, prismatic specimens experienced the most significant shear stress along diagonal directions. When the applied load exceeds the ultimate load, internal micro-cracks propagate along the diagonals until they extend through the entire specimen. Although sulfate salt wet–dry cycles alter the material properties, they do not significantly affect the overall failure mode. Hence, the failure patterns of prismatic specimens under sulfate attack are similar to those of the prismatic specimens under a non-corrosive environment, both of which form crack zones roughly along the diagonals.

The damage sounds of prismatic specimens under the non-corrosive environment and sulfate attack are compared, and the results indicate that the sounds of the prismatic specimens under sulfate attack are slight. The reason for this phenomenon may be that the gypsum and calcium sulfoaluminate hydrate and other volumetric expansion products are developed inside the recycled concrete, resulting in the formation and expansion of micro-cracks in recycled concrete, and although the overall strength may decrease, the presence of micro-cracks can create a “buffering” effect under compression, making the material exhibit more “ductility”. As a result, these micro-cracks reduce the possibility of energy being released in the form of sound waves.

For the prismatic specimens under dry–wet cycles of water, their failure patterns are similar to those under the non-corrosive environment and sulfate attack. Compared with the prismatic specimens under the non-corrosive environment, the prismatic specimens under dry–wet cycles of water develop more small cracks when they are damaged under uniaxial compression, but the propagation of cracks is obviously inferior to that of the specimens under sulfate attack, and the damage sound characteristics are similar between the specimens under the non-corrosive environment and sulfate attack. The reason for this phenomenon may be that the wet–dry cycles of water primarily affect concrete through physical mechanisms. Repeated wetting and drying alter the internal porosity of the concrete, but since there is no chemical erosion involved, only physical volume changes occur, leading to relatively limited crack propagation. Although wet–dry cycles do induce some micro-cracks, they produce fewer than sulfate wet–dry cycles, resulting in a more limited “buffering” effect of these micro-cracks under compression. As a result, the prismatic specimens under dry–wet cycles of water have more small cracks when they are damaged under uniaxial compression, and the damage sound characteristics are similar between the specimens under the non-corrosive environment and sulfate attack.

3.2. Analysis of Cubic Compressive Strengths

The relationships between the cubic compressive strength of recycled concrete and the replacement ratio of recycled coarse aggregates are shown in Figure 7, where “C30” represents the specimens with waste concrete strength of C30, “0.50” is the specimens with a water–cement ratio of 0.50, and “N” and “S” are the specimens under the non-corrosive environment and sulfate attack, respectively. For example, C40–S0.60 refers to the specimens with a water–cement ratio of 0.60 and waste concrete strength of C40. Compared with the specimens under the non-corrosive environment, the cubic compressive strengths of the specimens under sulfate attack are lower. The cubic compressive strengths of ordinary concrete with water–cement ratios of 0.50 and 0.60 are reduced by 0.81% and 22.13%, respectively. The cubic compressive strengths of recycled concrete with a water–cement ratio of 0.50 and waste concrete strength of C30 are reduced by 3.56%, 10.00%, and 16.17% when the replacement ratios of recycled coarse aggregates are 30%, 50%, and 100%, respectively. The cubic compressive strengths of recycled concrete with water–cement ratio of 0.60 and waste concrete strength of C30 are reduced by 25.82%, 32.44%, and 38.48% when the replacement ratios of recycled coarse aggregates are 30%, 50%, and 100%, respectively. The cubic compressive strengths of recycled concrete with water–cement ratio of 0.50 and waste concrete strength of C40 are reduced by 1.94%, 9.62%, and 15.46% when the replacement ratios of recycled coarse aggregates are 30%, 50%, and 100%, respectively. The cubic compressive strengths of recycled concrete with water–cement ratio of 0.60 and waste concrete strength of C40 are reduced by 23.31%, 31.77% and 33.59% when the replacement ratios of recycled coarse aggregates are 30%, 50%, and 100%, respectively. Apparently, the effects of replacement ratio of recycled coarse aggregates and water–cement ratio on the cubic compressive strength of recycled concrete are significant. The higher the replacement ratio of recycled coarse aggregates and water–cement ratio, the more obvious the strength reduction, which is consistent with the results of Xie [41]. While the waste concrete strength decreases, the decline in cubic compressive strength of recycled concrete also increases. However, the effect of waste concrete strength is obviously not as significant as the replacement ratio of recycled coarse aggregates and water–cement ratio. The sulfate reacts with calcium hydroxide and tricalcium aluminate in recycled concrete to produce expansive products such as gypsum and ettringite, and the internal volume expansion of recycled concrete is caused by the formation of these products. As a result, micro-cracks develop. During the dry–wet cycles, the pore structure of recycled concrete will be repeatedly expanded and contracted due to the evaporation and absorption of water. Especially in the drying stage, the sulfate solution is crystallized in the pores of the recycled concrete, and the internal pressure is generated with the increase in crystalline volume, further exacerbating the damage of recycled concrete. In addition, the micro-cracking or damage is contained by recycled coarse aggregates, and their surfaces are also attached to the old mortar. Under sulfate attack, the erosion process is also accelerated by the damage and old mortar. The higher the replacement ratio of recycled coarse aggregates, the more micro-cracks and damage areas develop in recycled concrete, and as a result, the compressive strength is significantly reduced. The higher the water–cement ratio, the higher the porosity and the lower the overall compactness of recycled concrete, and more channels are provided for the sulfate attack, which significantly reduces the compressive strength. The lower the waste concrete strength, the higher the water absorption and the more micro-cracks corresponding to the production of recycled aggregates, and more pores in the concrete are introduced by these defects, which is beneficial for the sulfate to permeate, resulting in a reduction in compressive strength. However, the waste concrete strength used in this experiment is not much different, and hence the reduction in compressive strength caused by waste concrete strength is not as significant as the replacement ratio of recycled coarse aggregates and water–cement ratio.

The comparisons of the cubic compressive strengths of recycled concrete under the non-corrosive environment and dry–wet cycles of water are depicted in Figure 8. Compared with the recycled concrete under the non-corrosive environment, the cubic compressive strength of the recycled concrete under dry–wet cycles of water with a water–cement ratio of 0.50, waste concrete strength of C30, and recycled coarse aggregate replacement ratio of 50% increases by 9.88%. The cubic compressive strength of the recycled concrete under dry–wet cycles of water with a water–cement ratio of 0.60, waste concrete strength of C30, and recycled coarse aggregate replacement ratio of 50% increases by 11.45%. Also, the cubic compressive strength of recycled concrete under dry–wet cycles of water with a water–cement ratio of 0.50, waste concrete strength of C40, and recycled coarse aggregate replacement ratio of 50% increases by 20.96%. The cubic compressive strength of recycled concrete under dry–wet cycles of water with a water–cement ratio of 0.60, waste concrete strength of C40, and recycled coarse aggregate replacement ratio of 50% increases by 10.29%. The reason is that the wet–dry cycles of water primarily affect concrete through physical mechanisms. Repeated wetting and drying alters the internal porosity of the concrete, but since there is no chemical erosion involved, only physical volume changes occur, leading to relatively limited crack propagation. As a result, the effect of dry–wet cycles of water on the deterioration of recycled concrete is much less than that of sulfate attack. Under the effect of water, micro-cracks in recycled concrete may be partially healed by the rehydration of cement, which reduces the spread of cracks, and the bonding force between mortar and aggregates can also be enhanced. Therefore, the compressive strength of recycled concrete is improved. Under the dry–wet cycle period of 168 days, the deterioration effect of dry–wet cycles of water on recycled concrete is not as great as the rehydration effect. The decrease in cubic compressive strength of recycled concrete due to the deterioration in dry–wet cycles of water cannot match the increase due to rehydration. Therefore, the cubic compressive strength of recycled concrete increases under dry–wet cycles of water. This further eliminates the interference of dry–wet cycles of water with the analysis of the results of this experiment.

In this study, corrosion resistance coefficient represents the ratio of the cubic compressive strength of recycled concrete under sulfate attack to the strength under the non-corrosive environment. The corrosion resistance coefficient is adopted as an evaluation index to analyze the resistance of recycled concrete to sulfate attack under different working conditions. The relationships between the corrosion resistance coefficient and the replacement ratio of recycled coarse aggregates are shown in Figure 9, where “C30” represents the specimens with a waste concrete strength of C30, and “0.50” is the specimens with a water–cement ratio of 0.50. For example, C30–0.50 refers to specimens with a water–cement ratio of 0.50 and a waste concrete strength of C30.

As illustrated in Figure 9, when the water–cement ratio is 0.50, compared with ordinary concrete, the corrosion resistance coefficients of recycled concrete with waste concrete strength of C30 are reduced by 2.77%, 9.26%, and 15.48% when the replacement ratios of recycled coarse aggregates are 30%, 50%, and 100%, respectively. When the water–cement ratio is 0.60, the corrosion resistance coefficients of recycled concrete with waste concrete strength of C30 are reduced by 4.74%, 13.24%, and 21.00% when the replacement ratios of recycled coarse aggregates are 30%, 50%, and 100%, respectively. When the water–cement ratio is 0.50, the corrosion resistance coefficients of recycled concrete with waste concrete strength of C40 are reduced by 1.14%, 8.88%, and 14.77%, respectively, as the replacement ratios of recycled coarse aggregates are 30%, 50%, and 100%. When the water–cement ratio is 0.60, the corrosion resistance coefficients of recycled concrete with waste concrete strength of C40 are reduced by 1.52%, 12.38%, and 14.71%, respectively, as the replacement ratios of recycled coarse aggregates are 30%, 50%, and 100%. The corrosion resistance coefficient of ordinary concrete is higher than that of recycled concrete with the same water–cement ratio. The corrosion resistance coefficient of recycled concrete decreases linearly with the increase in the replacement ratio of recycled coarse aggregates. The higher the replacement ratio of recycled coarse aggregates, the more obvious the decrease in the corrosion resistance coefficient. With the increase in water–cement ratio, the corrosion resistance coefficient of recycled concrete decreases significantly, that is, the resistance to sulfate attack of recycled concrete can be significantly improved by reducing the water–cement ratio. With the increase in waste concrete strength, the corrosion resistance coefficient of recycled concrete increases, that is, the resistance to sulfate attack of recycled concrete can be improved by the increase in waste concrete strength. The reason is that the interface transition zone exists in recycled concrete is generally not as tight or uniform as that of ordinary concrete. Such loose interface transition zones can easily become infiltration channels of sulfate attack. In addition, recycled coarse aggregates are characterized by higher water absorption and porosity. When the replacement ratio of recycled coarse aggregates increases, the overall porosity of recycled concrete also increases, the possibility of sulfate solution leaching into the concrete is promoted, and hence the resistance to sulfate attack is reduced. In addition, the higher water–cement ratio can increase the porosity of recycled concrete and reduce its compactness significantly. Therefore, the resistance to sulfate attack is reduced. Recycled coarse aggregates made from waste concrete with higher strength are often of better quality, and as a result, with the increase in strength of waste concrete, the resistance to sulfate attack of recycled concrete becomes stronger.

Considering the effects of the replacement ratio of recycled coarse aggregates, the strength of waste concrete, and the water–cement ratio, a prediction model for the corrosion resistance coefficient of recycled concrete is proposed using Python 3.8 software to conduct multiple polynomial regression analysis on the experimental data:

$$K=0.0027C-2.1427W+0.2645r^2-0.5099r+2.0678\left(R^2=0.98,MSE=0.0001,MAPE=1.16\%\right)$$

(1)

where K represents the corrosion resistance coefficient of recycled concrete, r is the replacement ratio of recycled coarse aggregates, C is the waste concrete strength, and W is the water–cement ratio.

The comparisons between the experimental values and predicted values of K are depicted in Figure 10. It can be seen that the relative errors of the experimental values and predicted values of K are all below 3%, and the lowest is only 0.18%. That is to say, the accuracy of the prediction model of K mentioned in Equation (1) is high enough.

Considering the effects of the replacement ratio of recycled coarse aggregates, the strength of waste concrete, and the water–cement ratio, as well as the corrosion resistance coefficient under sulfate attack over 168 days, a prediction model for the cubic compressive strength of recycled concrete is proposed using Python 3.8 software to conduct multiple polynomial regression analysis on the experimental data:

$$f_{\mathrm{cu}}=\left(0.7195K^2+1.9478K+0.4471\right)\left(\begin{array}{l}0.1488C-23.4767W-32.3261r^3\\ +60.8858r^2-36.4891r+28.7039\end{array}\right)\left(R^2=0.99,MSE=0.79,MAPE=2.16\%\right)$$

(2)

The comparisons between the experimental values and predicted values of f_cu are shown in Figure 11. The relative errors of the experimental values and predicted values of K are all below 6%, and the lowest is only 0.32%. This shows that the accuracy of the prediction model of f_cu proposed in Equation (2) is high.

3.3. Analysis of Axial Compressive Strength

The relationships between the axial compressive strength of recycled concrete and the replacement ratio of recycled coarse aggregates under different working conditions are shown in Figure 12. Compared with recycled concrete under the non-corrosive environment, the axial compressive strengths of ordinary concrete with water–cement ratios of 0.50 and 0.60 under sulfate attack are reduced by 1.15% and 21.74%, respectively. The axial compressive strengths of recycled concrete with water–cement ratio of 0.50 and waste concrete strength of C30 decrease by 3.78%, 10.35%, and 16.80% when the replacement ratios of recycled coarse aggregates are 30%, 50%, and 100%, respectively. Also, the axial compressive strengths of recycled concrete with water–cement ratio of 0.60 and waste concrete strength of C30 decrease by 25.91%, 33.17%, and 37.82% when the replacement ratios of recycled coarse aggregates are 30%, 50%, and 100%, respectively. The axial compressive strengths of recycled concrete with water–cement ratio of 0.50 and waste concrete strength of C40 decrease by 2.74%, 5.65%, and 14.79% when the replacement ratios of recycled coarse aggregates are 30%, 50%, and 100%, respectively. The axial compressive strengths of recycled concrete with water–cement ratio of 0.60 and waste concrete strength of C40 decrease by 24.58%, 25.55%, and 28.91% when the replacement ratios of recycled coarse aggregates are 30%, 50%, and 100%, respectively. The axial compressive strength of the recycled concrete is significantly influenced by the replacement ratio of recycled coarse aggregates and water–cement ratio, and the waste concrete strength is less significant than the two factors. The higher the replacement ratio of recycled coarse aggregates and water–cement ratio and the lower the waste concrete strength, the more obvious the reduction in strength. The reason is the same as that mentioned in Section 3.2.

The comparison of the axial compressive strengths of recycled concrete under the non-corrosive environment and dry–wet cycles of water is shown in Figure 13. Compared with recycled concrete under the non-corrosive environment, the axial compressive strength of recycled concrete under dry–wet cycles of water increases. Compared with recycled concrete under the non-corrosive environment, the axial compressive strength of recycled concrete with water–cement ratio of 0.50, waste concrete strength of C30, and recycled coarse aggregate replacement ratio of 50% under dry–wet cycles of water increases by 0.11%. The axial compressive strength of recycled concrete with water–cement ratio of 0.60, waste concrete strength of C30, and recycled coarse aggregate replacement ratio of 50% under dry–wet cycles of water increases by 0.22%. The axial compressive strength of recycled concrete with water–cement ratio of 0.50, waste concrete strength of C40, and recycled coarse aggregate replacement ratio of 50% under dry–wet cycles of water increases by 17.05%. The axial compressive strength of recycled concrete with water–cement ratio of 0.60, waste concrete strength of C40, and recycled coarse aggregate replacement ratio of 50% under dry–wet cycles of water increases by 0.71%. The reason is the same as that mentioned in Section 3.2.

Considering the effects of the replacement ratio of recycled coarse aggregates, the strength of waste concrete, and the water–cement ratio, as well as the corrosion resistance coefficient under sulfate attack over 168 days, a prediction model for the axial compressive strength of recycled concrete is proposed using Python 3.8 software to conduct multiple polynomial regression analysis on the experimental data:

$$f_{\mathrm{c}}=\left(3.0467K^2-4.5888K+3.8012\right)\left(\begin{array}{l}0.2266C-49.6568W-49.0073r^3\\ +90.7507r^2-54.0487r+43.541\end{array}\right)\left(R^2=0.97,MSE=1.47,MAPE=4.03\%\right)$$

(3)

The comparison between the experimental values and predicted values of f_c is shown in Figure 14. The relative errors of the experimental values and predicted values of K are all below 9%, and the lowest is only 0.52%. It can be seen that the accuracy of the prediction model of f_c proposed in Equation (3) is relatively high.

3.4. Compressive Stress–Strain Curve and Compressive Constitutive Model

In order to facilitate the comparisons and analyses of the compressive stress–strain curves of recycled concrete with different waste concrete strengths, different water–cement ratios, and different replacement ratios of recycled coarse aggregates, results are compared as groups with the same replacement ratio of recycled coarse aggregates, as shown in Figure 15. Compared with recycled concrete under the non-corrosive environment, the peak stress of recycled concrete under sulfate attack decreases, which indicates that the bearing capacity of recycled concrete is significantly reduced after sulfate attack. This is consistent with the variation in the peak stress of ordinary concrete under sulfate attack [42]. The reason is that sulfate reacts with calcium hydroxide and calcium silicate hydrate in recycled concrete to produce expansion products such as ettringite and gypsum [43]. Volume expansion inside the recycled concrete is caused by these products, which leads to the formation and expansion of micro-cracks [31,40,43]. In the process of dry–wet cycles, the sulfate crystals dissolve and deposit repeatedly in the pores of recycled concrete, resulting in crystallization pressure, which further intensifies the expansion of micro-cracks. The internal pores of recycled concrete further increase, the internal structure of recycled concrete is damaged, the compressive strength is significantly reduced, and as a result, the peak stress decreases.

The comparison of the compressive stress–strain curves of recycled concrete under the non-corrosive environment and dry–wet cycles of water is shown in Figure 16. Compared with recycled concrete under the non-corrosive environment, the peak stress of recycled concrete under dry–wet cycles of water increases, which shows that the bearing capacity of recycled concrete increases after dry–wet cycles of water.

In this study, combined with the results measured by tests, the obtained compressive stress–strain data are normalized and analyzed, and the two-stage full-curve equation proposed by Guo [23] is selected for fitting using Origin 2018 software, as written in Equation (4). The fitting parameters a and b are tabulated in

Table 6. Values of the fitting parameters.

Specimen Number	a	b
NAC–S0.50	1.5391	3.1324
NAC–S0.60	2.1523	1.7400
RAC30–R30–S0.50	1.7950	2.3455
RAC30–R50–S0.50	2.2512	1.9512
RAC30–R100–S0.50	3.1005	2.1476
RAC30–R30–S0.60	2.3873	1.4549
RAC30–R50–S0.60	3.0773	1.4337
RAC30–R100–S0.60	4.9639	1.2775
RAC40–R30–S0.50	1.7581	2.3700
RAC40–R50–S0.50	1.9416	2.7759
RAC40–R100–S0.50	2.4141	2.8049
RAC40–R30–S0.60	2.3580	1.5698
RAC40–R50–S0.60	2.4668	1.5476
RAC40–R100–S0.60	3.4459	2.1859

, and the corresponding coefficients of determination (R²) are all above 0.99:

$$\left\{\begin{array}{l}y=ax+(3-2a)x^2+(a-2)x^3\hspace{1em}\hspace{1em}\hspace{1em}x\le 1\\ y={\displaystyle \frac{x}{b{(x-1)}^2+x}}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{0.17em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}x\ge 1\end{array}\right.$$

(4)

where a and b are the undetermined parameters, x = ε/ε_p, y = σ/f_c, ε and ε_p are the strain and peak strain, and σ and f_c are the stress and peak stress.

Python 3.8 software was adopted to conduct multiple polynomial regression analysis on the values of a and b in Table 6, and the calculation formulas for a and b considering the replacement ratio of recycled coarse aggregates, the waste concrete strength, the water–cement ratio, and the corrosion resistance coefficient of recycled concrete are obtained:

$$a=\left(\begin{array}{l}11.3283K^4-36.7753K^3+44.538K^2\\ -23.837K+4.7758\end{array}\right)\left(\begin{array}{l}-1.5876C+450.121W+134.4869r^3\\ -331.83r^2+324.3459r-169.79\end{array}\right)$$

(5a)

$$b=\left(\begin{array}{l}-14.8989K^4+46.6304K^3-54.0987K^2\\ +27.6128K-5.2317\end{array}\right)\left(\begin{array}{l}2.8433C+1143.8364W-869.323r^3\\ +1803.5183r^2-1010.2508r-381.4235\end{array}\right)$$

(5b)

The values of coefficient of determination (R²) of Equation (5a,b) obtained by fitting are 0.99 and 0.96, respectively. The values of coefficient of mean square error (MSE) of Equation (5a,b) obtained by fitting are 0.01 and 0.01, respectively. The values of coefficient of mean absolute percentage error (MAPE) of Equation (5a,b) obtained by fitting are 3.66 and 4.89, respectively. The compressive constitutive model of recycled concrete under sulfate attack over 168 days considering the replacement ratio of recycled coarse aggregates, the waste concrete strength, and the water–cement ratio is shown in Equations (4) and (5a,b).

The comparisons between the experimental curves and predicted curves are shown in Figure 17. It can be seen that the experimental curves are in good agreement with the predicted curves. Therefore, the constitutive model introduced in this study can better reflect the mechanical properties of recycled concrete under sulfate attack.

3.5. Analysis of Microscopic Morphology

The microscopic morphologies of recycled concrete are shown in Figure 18, Figure 19 and Figure 20. The figures depict that the compactness of recycled concrete under the non-corrosive environment is greater, and a large amount of flocculated calcium silicate hydrate (C-S-H) can be observed to fill its internal pores. As for the recycled concrete under dry–wet cycles of water, more hydration products such as C-S-H are created and the internal pores are filled more densely, so the mechanical properties are improved, which validates the explanation of rehydration of recycled concrete in Section 3.2. Therefore, under dry–wet cycles of water, the mechanical properties of recycled concrete are improved. On the contrary, as for the recycled concrete under sulfate attack, it is not difficult to see that the pore structure is significantly destroyed by the expansion stress generated by Na₂SO₄ crystals on the pore walls, and thus the formation of micro-cracks in the recycled concrete is accelerated. This is consistent with the research results of Du et al. [31]. Moreover, a large number of Na₂SO₄ crystals are attached to the surface of concrete matrix and the concrete matrix becomes loose, indicating that damage has occurred inside the recycled concrete, and the bonding ability between the recycled aggregates and the mortar decreases. Therefore, the mechanical properties of recycled concrete under sulfate attack are reduced.

Figure 18, Figure 19 and Figure 20 also show that micro-cracks of varying lengths are distributed in disorderly fashion in recycled concrete, which provides more channels for sulfate attack, and this phenomenon can better explain the reason that the resistance to sulfate attack of recycled concrete is lower than that of the ordinary concrete. The higher the replacement ratio of recycled coarse aggregates and the lower the waste concrete strength, the more micro-cracks in the recycled concrete under sulfate attack and the more incompact the concrete matrix. This reason can well explain the phenomenon that the higher the replacement ratio of recycled coarse aggregates and the lower the waste concrete strength, the higher the decrease in the cubic and axial compressive strengths of recycled concrete under sulfate attack, that is, the easier for its mechanical properties to deteriorate.

4. Conclusions

In this study, the mechanical properties and compressive constitutive model of recycled concrete under sulfate attack considering the effects of multiple factors have been analyzed, and the main conclusions are as follows.

(1)

The higher the replacement ratio of recycled coarse aggregates and water–cement ratio and the lower the waste concrete strength, the more obvious the reductions in the cubic compressive strength, with a maximum reduction of 38.48%. Compared with the recycled concrete under the non-corrosive environment, the cubic compressive strength of recycled concrete under dry–wet cycles of water with water–cement ratio of 0.50, waste concrete strength of C40, and recycled coarse aggregate replacement ratio of 50% increases by 20.96%.

(2)

The corrosion resistance coefficient of recycled concrete decreases linearly with the increase in the replacement ratio of recycled coarse aggregates. With the increase in water–cement ratio and decrease in waste concrete strength, the corrosion resistance coefficient decreases, with a maximum decrease of 21.00%. Considering the effects of the replacement ratio of recycled coarse aggregates, waste concrete strength, and water–cement ratio, a prediction model for the corrosion resistance coefficient and a prediction model for the cubic compressive strength of recycled concrete under sulfate attack are proposed.

(3)

The higher the replacement ratio of recycled coarse aggregates and water–cement ratio and the lower waste concrete strength, the more significant the reductions in the axial compressive strength, with a maximum reduction of 37.82%. Compared with recycled concrete under the non-corrosive environment, the axial compressive strength of recycled concrete with water–cement ratio of 0.50, waste concrete strength of C40, and recycled coarse aggregate replacement ratio of 50% under dry–wet cycles of water increases by 17.05%. A prediction model for the axial compressive strength and a compressive constitutive model of recycled concrete under sulfate attack considering the effects of the replacement ratio of recycled coarse aggregates, waste concrete strength, and water–cement ratio are established.

(4)

The pore structure of recycled concrete is significantly destroyed by the expansion stress generated by Na₂SO₄ crystals on the pore walls, the concrete matrix becomes loose, and the bonding ability between the recycled aggregates and the mortar decreases.

(5)

In this study, the exploration of the mechanical properties of recycled concrete under sulfate attack is conducted on a material level, and the research results can provide theoretical basis, data support, and design methods for durability in engineering applications using recycled concrete under sulfate attack. Further research can be conducted on the mechanical properties of recycled concrete components or structures under sulfate attack, taking into account the effects of more factors.