Research Article  Open Access
Influence of Water Pressure on the Mechanical Properties of Concrete after FreezeThaw Attack under Dynamic Triaxial Compression State
Abstract
This study aims at determining the effect of water pressure on the mechanical properties of concrete subjected to freezethaw (FT) attack under the dynamic triaxial compression state. Two specimens were used: (1) a 100 mm × 100 mm × 400 mm prism for testing the loss of mass and relative dynamic modulus of elasticity (RDME) after FT cycles and (2) cylinders with a diameter of 100 mm and a height of 200 mm for testing the dynamic mechanical properties of concrete. Strain rates ranged from 10^{−5}·s^{−1} to 10^{−3}·s^{−1}, and FT cycles ranged from 0 to 100. Three levels of water pressure (0, 5, and 10 MPa) were applied to concrete. Results showed that as the number of FT cycles increased, the mass loss rate of the concrete specimen initially decreased and then increased, but the RDME decreased. Under 5 MPa of water pressure and at the same strain rate, the ultimate compressive strength decreased, whereas the peak strain increased with the increase in the number of FT cycles. This result is contrary to the variation law of ultimate compressive strength and peak strain with the increase in strain rate under the same number of FT times. With the increase in FT cycles or water pressure, the strain sensitivity of the dynamic increase factor of ultimate compressive strength and peak strain decreased, respectively. After 100 FT cycles, the dynamic compressive strength under all water pressure levels tended to increase as the strain rate increased, whereas the peak strain decreased gradually.
1. Introduction
In the vast cold regions of the world, hydraulic concrete structures are subjected to freezethaw (FT) cycles due to day and night replacement and seasonal changes. The performance of concrete after FT deteriorates to varying degrees. Furthermore, these structures bear the impact of dynamic loads, such as vehicles, waves, and earthquakes, which seriously affect the longterm use and safe operation of concrete structures. The mechanical behaviour of concrete under dynamic load is different from that under static load [1, 2].
Numerous studies have been conducted to investigate the effect of strain rate on concrete characteristics, such as compressive strength, peak strain, elastic modulus, and flexural and tensile strength. Wang et al. [3] reported the effect of water pressure on the mechanical properties of concrete and established failure modes, which were affected by strain rate and water content. Yan and Lin [4] evaluated concrete dynamic strength and deformation properties under initial static stress and found that total deformation of concrete was greatly influenced by initial static loading history. Xiao et al. [5] reported that, when the strain rate reaches approximately 10^{−1}·s^{−1}, the compressive strength increased linearly as the strain rate increased. Furthermore, the strength increased more strongly at high strain rates [6]. Ahmad et al. [7] concluded that, compared with elastic modulus, compressive strength is more sensitive to strain rate. This view was also supported by [8]. Scholars have found that peak strain increases with strain rate [9–12]. Chen et al. [5, 13, 14] studied the tensile strength of cementbased materials and compared the strain rate sensitivity of concrete under tension and pressure. They found that dynamic tensile strength increased as strain rate increased, while the dynamic increase factor was less influenced by the watercement ratio.
Aside from focusing on conventional concrete, substantial research has also been conducted on the dynamic mechanical properties of other types of concrete, such as recycled aggregate concrete (RAC), steel fibrereinforced concrete (SFRC), and ceramsite concrete. Several studies emphasised the mechanical behaviour of RAC under the effect of strain rate. Researchers [15–19] found that elastic modulus and peak stress increase as strain rate increases. Moreover, the strain rate sensitivity of elastic modulus and peak stress of RAC are more remarkable than those of RAC with carbonated RAC due to the Stefan effect [19]. Deng et al. [20] reported that the compressive strength of RAC under triaxial loading is much higher than that under uniaxial loading. The stress ratio and replacement rate of RCA affect strength and strain greatly. Wang and Xiao [10] investigated the influence of RCA replacement ratio and strain rate on the mechanical performance of confined recycled aggregate concrete (CRAC) and proposed a constitutive model for CRAC, which may be used for dynamic analysis of RAC. Some researchers conducted experiments to evaluate the mechanical behaviours of SFRC at a high strain rate. They concluded that dynamic compressive strength, critical strain, and energy absorption increased as strain rate and steel fibre content increased. Furthermore, the growth of strength and strain is less significant than that of energy absorption because increases in strength and strain are ultimately reflected in increases in energy absorption [21–24]. Chen et al. [25] tested the dynamic performance of ceramsite concrete with different numbers of FT cycles and ceramsite volume fraction. They suggested that the increase in FT cycles or the volume fraction of ceramsite is harmful to the compressive strength and damage degree of ceramsite concrete.
Several studies on the dynamic uniaxial compressive performance of concrete after FT cycles can be found in the literature. Wang et al. [26] studied the coupled influence of strain rate and FT on the uniaxial dynamic compressive performance of fully graded concrete. Furthermore, they proposed a failure mechanism of fully graded concrete, considering the influence of FT cycle number and strain rate. Yi et al. [27] conducted a uniaxial compressive experiment with different FT cycles and loading rates to verify the rationality of the model of porous asphalt mixtures. Zhou et al. [28] tested five groups of sandstone dynamic mechanical behaviours under the effect of FT cycles. They suggested that the void ratio rises and the structure of pores changes considerably as the number of FT cycles increase. Dynamic peak stress and elastic modulus are reduced, but peak strain increases with the process of FT cycles. Lu et al. [29] conducted a uniaxial compressive experiment on concrete specimens subjected to the combined effects of FT cycles and fatigue loading. They found that concrete dynamic compressive strength increases with the increase in strain rate, and the strain rate effect of concrete becomes apparent after FT cycles. Li et al. [30] conducted a test on the dynamic mechanical properties of sandstone samples exposed to FT cycling and reported that, as FT cycles increased, peak strain and apparent damage increase, but peak stress is reduced.
Studies on the dynamic performance of concrete under multiaxial compression have also been conducted. Under multiaxial stress loading, compressive strength in dynamic state is higher than that in static state [31–33]. Wang et al. [34] evaluated the coupled effects of water content and loading rate on the properties of concrete. They reported that a high water content of concrete corresponds to high sensitivity; this finding is consistent with that of [3]. However, few studies have been conducted on the dynamic performance change of concrete under triaxial compression stress after FT cycles, hence proving the importance of this research.
In this study, the dynamic mechanical properties of concrete subjected to FT cycles under triaxial compressive loading were investigated. The strain rate, number of FT cycles, and water pressure level ranged from 10^{−5}·s^{−1} to 10^{−3}·s^{−1}, 0 to 100, and 0 to 10 MPa, respectively. The microstructure was measured via scanning electron microscopy (SEM). The effects of strain rate, number of FT cycles, and water pressure on ultimate compressive strength, peak strain, and dynamic increase factor were analysed systematically.
2. Materials and Methods
2.1. Materials and Mix Proportions
In this investigation, Chinese standard [35] P.O. 42.5R ordinary Portland cement supplied by the Shanxi Qinling cement plant in China was used. The chemical composition of the cement used for the test is listed in Table 1. Weihe River sand was used as fine aggregate, with a fineness modulus of 2.35. The coarse aggregate of natural river pebble with diameters of 5–20 and 20–30 mm was used, constituting half of the total coarse aggregate. Tap water, airentraining agent, and polycarboxylic superplasticiser were used. The mix proportions of concrete are shown in Table 2.


2.2. Casting and Curing of Specimens
Concrete specimens were cast in steel moulds with dimensions of 100 mm × 100 mm × 400 mm and Φ100 mm × 200 mm. The specimens were removed from the moulds after 24 h of casting and cured for 28 days under standard conditions (relative humidity was higher than 95%, and temperature was 20 ± 2°C). Prism specimens with dimensions of 100 mm × 100 mm × 400 mm were used to measure the loss of mass and relative dynamic modulus of elasticity (RDME) after FT cycles. Cylinder specimens with dimensions of Φ100 mm × 200 mm were used to assess the triaxial mechanical properties of concrete after FT cycles at different strain rates. Three specimens were used in each test, and the average value was used for the test results.
2.3. Apparatus and Testing Methods
In accordance with the Chinese standard GB/T 500822009 [36] (similar to ASTM C666/C666M03), a rapid FT cycle test was conducted by using an FT apparatus. During the test, each FT cycle was completed within 4 h, and the thawing time was more than one quarter of the freezing and thawing time. The minimum and maximum temperatures of the specimen center were controlled within (–18 ± 2) °C and (5 ± 2)°C, respectively. The cooling time from 3°C to −16°C was more than onehalf of the freezing time, and the heating time from –16°C to 3°C was more than onehalf of the thawing time. FT cycles were repeated 100 times, and the specimens were evaluated every 25 cycles for the calculation of RDME.
The testing instrument used in this test was a servohydraulic, static, and dynamic triaxial testing system in the hydraulic laboratory of Xi’an University of Technology, as shown in Figure 1. The axial load capacity is 2000 kN for concrete materials, and its accuracy is 0.1%. The target values of water pressure and loading rate were set before loading. Then, the electromagnetic directional valve was opened to connect the pressurised water cylinder and the confining pressure channel. Afterwards, the pressure servo oil source was opened; the servo valve controlled the cylinder plunger forward, and the water pressure increased to the target value, which was detected by the pressure sensor. Once the deviation occurred, the servo valve controlled the pressure drive cylinder piston to swing back and forth to maintain the set value of the water pressure. The axial strain rate in the test was set at 1 × 10^{−5}, 1 × 10^{−4}, 2 × 10^{−4}, 5 × 10^{−4}, and 1 × 10^{−3}·s^{−1}.
(a)
(b)
3. Results and Discussion
3.1. Surface Scaling
The surface scaling of the concrete specimens after 0, 25, 50, 75, and 100 FT cycles is shown in Figure 2. Unlike the samples that did not undergo an FT cycle, the local surface of the samples became rough and the cement mortar on the epidermis spalled off after 50 FT cycles. After 75 and 100 FT cycles, the rough surface and flaking area of the specimen were enlarged, and microholes appeared.
3.2. Mass Loss Rate
Figure 3 illustrates the change in mass loss of the concrete after 100 FT cycles. The mass loss rate of the concrete specimens generally showed a trend of initial decline followed by a gradual increase. The mass loss rates at 0, 25, 50, 75, and 100 FT cycles were 0, −0.05%, 0.08%, 0.10% and 0.24%, respectively. After 0–25 FT cycles, the loss rate of concrete mass showed a downward trend due to the microcracks inside the specimen after the FT cycle. The concrete specimen absorbed water from the surrounding environment, and water turned into ice due to freezing. Pores expanded and resulted in additional pores because of frost heaving. Therefore, water was immersed in the pores, and the internal water content continuously increased. Simultaneously, the flaking of the mortar on the surface worsened with the increase in FT cycles, resulting in the reduced quality of the specimen. Therefore, during the early stage of FT cycles, the mortar peeled off less, and the increase in mass caused by water absorption dominated, thereby indicating that the overall quality of the specimen increased. After 25–50 FT cycles, the spalling of mortar aggravated due to intensified FT degradation. Given that the mass loss caused by the apparent spalling of cement mortar of the concrete specimen was greater than the water absorption of the internal crack, the mass loss rate of the specimen shows an upward trend. Under the effect of the FT cycle, the water flowing into the concrete pores turned into ice, causing the volume of concrete to expand and the width of microcracks to increase. Moreover, with the formation of some new pores, the amount of water absorbed by the cracks in the specimen was equal to the amount dropped by mortar, and the mass loss rate after 50–75 FT cycles was relatively flat. However, after 75–100 FT cycles, the drop of particles on the surface of the specimen played a major role, and the mass loss rate continued to rise.
3.3. RDME
As shown in Figure 4, the RDME of concrete specimens tends to decrease with the increase in the number of FT cycles. The value of initial dynamic elastic modulus was 42.07 GPa. The RDME of concrete specimens at 0, 25, 50, 75, and 100 FT cycles was 100%, 99.46%, 98.67%, 97.69%, and 95.20%, respectively. The loss rate of RDME increased with the increase in FT cycles due to the formation of frost heaving pressure caused by FT cycles. Furthermore, the microstructure of concrete was affected, the degree of damage aggravated, the cracks multiplied, and density was reduced. During this process, water accelerated into the interior of the void, which aggravated the damage of specimens under the action of FT cycles, resulting in increased damage of the internal structure of the concrete.
3.4. Dynamic Mechanical Properties
3.4.1. Ultimate Compressive Strength
Figure 5 shows the relationship between the dynamic ultimate compressive strength and the number of FT cycles under a constant water pressure of 5 MPa. Test results revealed that the ultimate compressive strength of concrete decreased with the increase in the number of FT cycles at the same loading rate, and different trends occurred at various stages of the FT cycles. The gradient of the ultimate compressive strength of concrete at 0–50 FT cycles is smaller than that at 50–100 FT cycles. This result may have occurred because during the early FT cycles, the fine particles inside the concrete were squeezed and compactness increased, thereby making it difficult for water to enter the pores. This condition limited the expansion of the pores. These results agreed with those suggested by Li et al. [30] that, during the FT cycles, mineral particles are compacted and the cohesive forces between them are gradually strengthened, which can availably offset part of the frost heave force, thus limiting pore expansion. Therefore, the internal damage of the concrete was weakened, and the influence of the FT cycle on ultimate compressive strength was reduced. At 50–100 FT cycles, the pore structure inside the concrete changed greatly as the degree of FT deterioration increased. The width of some small pores increased and evolved into microcracks, and the pores that connected with each other caused the rapid decrease in the ultimate compressive strength of the concrete under dynamic loading.
The relationship between ultimate compressive strength of concrete and strain rate at a water pressure of 5 MPa is shown in Figure 6. Concrete is a material with high strain rate sensitivity. With the same number of FT cycles, the ultimate compressive strength of concrete increased with the increase in strain rate. This view is also supported by [26, 29].
Griffith theory and the principle of subcritical crackle expansion can be applied to explain these trends. On the one hand, according to Griffith theory, when the flaw size of brittle material is larger than the critical size, fracture will occur. Subcritical cracks have enough time to expand at lower strain rates, and then failure occurs under a lower load. When the strain rate is higher, the subcritical crackles have less time for extension. At this time, the material structure can withstand a relatively larger load before the occurrence of failure. Furthermore, on the basis of knowledge of fracture mechanics [37–39], a microcrack process area is believed to be present before flaw expands. The speed of crackle growth also determines the size of the process area, as proposed by [40]. Therefore, at a high strain rate, the propagation velocity of flaw becomes faster, and the process area becomes larger. Thus, crack expansion requires more energy, while concrete has insufficient time to store energy at a higher strain rate. The extra energy can be released by only increasing the stress, which finally manifests as an increase in strength.
On the other hand, free water viscosity affects concrete. Kaplan et al. [41] believed that, when a specimen is subjected to compression stress, the pores and channels perpendicular to the loading direction in the concrete tend to close, allowing free water to flow inside the specimen. The presence of viscosity and liquidphase pressure gradients causes pore water pressure to increase at a high loading rate, which delays the occurrence of excessive cracks in the solid phase and eventually increases compressive strength. Through experimental observations and theoretical assumptions, Rossi et al. [8, 42–44] indicated that the mechanical properties of concrete under high strain rate could be explained by the Stefan effect and material cracking process interaction. Moreover, for the conetype concrete failure mode, lateral deformation was restricted to viscous stress. Thus, the damage of specimens decreased. The effect of damping caused by the viscous stress could also slow down the expansion of microcracks for the concrete failure mode of slant shear. Finally, the strength of concrete increased [3].
Viscous fluid exists between two flat parallel plates with a distance of h. When two parallel plates are separated or closed at a certain speed, their motion will be affected by the viscous resistance generated by the viscous fluid. This phenomenon is called the Stefan effect. Moreover, a great velocity corresponds to a great reaction force, as shown in Figure 7. Pores and cracks of varying sizes in concrete contain free water. At this point, cement matrix and water can be regarded as a microdisc system, and pore water will generate viscous resistance under dynamic loading, as shown in Figure 8. This condition is an important factor that affects the strain rate sensitivity of concrete [8]. From this perspective, more free water will immerse into the pores after the FT cycles. As the strain rate increases, the number of viscously resistant microelements increases, and the influence of free water on viscous resistance becomes greater. Therefore, the damage to the concrete is lessened and concrete exhibits a higher ultimate compressive strength value, which agrees with Li et al. [19], who reported that the strength increase in RAC samples is greater than that of CRAC due to higher viscous resistance.
3.4.2. Peak Strain
Similarly, the relationship between the peak strain and the number of FT cycles under a water pressure of 5 MPa is shown in Figure 9. At the same strain rate, the peak strain of concrete tended to increase gradually as the number of FT cycles increased, because during the FT cycles, the water in the pores would freeze and swell. Under the repeated action of the frost heaving force, the pores gradually increased, the concrete volume increased, and the compactness gradually decreased. However, in the compression test, the concrete first underwent the compaction process after the FT cycles. Therefore, the peak strain of the concrete, as well as its ductility, increased.
Figure 10 illustrates the relationship between peak strain and strain rate under a water pressure of 5 MPa. With the same number of FT cycles, the peak strain of the concrete tended to decrease with the increase in strain rate. According to the theory of minimum energy consumption, the destruction of a system is always operated with the minimum energy consumption per unit time. A previous research showed that cracks mostly expand from interfacial zones between aggregates and mortars, and few cracks directly passing through weak aggregates due to the strengths of a mortar matrix and interfacial transition zone are lower than the strength of a natural coarse aggregate [16, 19]. Fracture pattern depends on aggregate shape and other factors, such as aggregate properties and matrixaggregate interface. Therefore, at a relatively low strain rate, the crack had enough time to select a fracture path, that is, the crack started at the weak surface between the aggregate and the mortar and then expanded and passed through the mortar along the weak surface. At a relatively high strain rate, the crack became more dispersed. The crack had insufficient time to select the weakest failure interface but expanded along the path with the fastest energy release immediately. It would break directly through the aggregate, resulting in the failure path becoming shorter, thereby decreasing the deformation of the concrete.
3.4.3. Dynamic Increase Factor (DIF)
The DIF of ultimate compressive strength or peak strain is defined as the ratio of ultimate compressive strength or peak strain at each strain rate to ultimate compressive strength or peak strain at a strain rate of 1 × 10^{−5}·s^{−1}.
The relationship between DIF of ultimate compressive strength and strain rate under different numbers of FT cycles is shown in Figure 11. As the strain rate increased, the DIF of the ultimate compressive strength of each number of FT cycle increased. However, the increase rate varied under different FT cycles. The relationship between strength increment and strain rate was established by the following fitting test data:where is the ultimate compressive strength in dynamic loading, is the ultimate compressive strength in quasistatic loading, is the current strain rate ( = 1 × 10^{−4}, 2 × 10^{−4}, 5 × 10^{−4}, and 1 × 10^{−3}·s^{−1} in this study), is the quasistatic strain rate ( = 1 × 10^{−5}·s^{−1}), and is a parameter.
As the number of FT cycles increased, the slope of the fitting curve gradually decreased; that is, the increase in the number of FT cycles reduced the rate sensitivity of the ultimate compressive strength of the concrete, because as the number of FT cycles increased, the capillary microcracks inside the concrete gradually developed, and the amount of water absorbed from the surrounding environment increased. On the one hand, the presence of water in concrete, with the increase in strain rate, enhanced water viscous action and increased the sensitivity of concrete accordingly. On the other hand, under the repeated action of frost heaving force, the internal pores of concrete further increased, and the wedging effect of water enhanced. Thus, the slope of the fitting curve grew slowly. At the same time, the crack expanded further due to concrete damage during the test loading stage, and the water in the concrete spread around the crack under the action of the pressing force to form a water wedge effect. However, the water content of concrete at 0 FT cycle was lower than those at 50 and 100 FT cycles. The wedge effect can be ignored. Thus, the slope of 0 FT cycle was higher than those of 50 and 100 cycles. Concrete rate sensitivity decreased as the number of FT cycles increased.
The relationship between the DIF of peak strain and strain rate under different numbers of FT cycles is shown in Figure 12. With the increase in strain rate, the DIF of peak strain decreased in each cycle. However, the magnitude of reduction in the fitted curves was not the same. To analyse the relationship between peak strain increment and strain rate, the relationship between peak strain and strain rate was established by fitting test data, as shown below:where is the peak strain in dynamic loading, is the peak strain in quasistatic loading, is the current strain rate ( = 1 × 10^{−4}, 2 × 10^{−4}, 5 × 10^{−4}, and 1 × 10^{−3}·s^{−1} in this study), is the quasistatic strain rate ( = 1 × 10^{−5}·s^{−1}), and is a parameter.
Figure 12 shows that, under the same number of FT cycles, the peak strain growth rates were all less than 1 as the strain rate increased, that is, the peak strain gradually decreased. Moreover, the variation curve of specimens with more FT cycles was relatively gentle, which means that the increase in FT cycles weakens the sensitivity of the peak strain rate of concrete. The reason is that, with the increase in the strain rate, the internal microcrack of concrete cannot be fully compressed due to the short time of load action, and the crack will expand in a manner that consumes less energy, that is, directly passing through the aggregate, to cause damage. These results are consistent with those of Wang et al. [45, 46], who demonstrated that cracks propagate by cutting through aggregates and aggregates are fractured into many small pieces with the increase of strain rate. More cracks passed through the aggregate rather than bypass the expansion and compression closure of the aggregate. Eventually, the peak strain decreased gradually with a growth rate of less than 1. However, many intricate and different sizes of internal cracks exist in concrete with more severe FT deterioration. The crack that originally broke through the aggregate changed the damage mode and increased the length of the damage path. Thus, the decrease in peak strain was small.
3.5. Influence of Water Pressure on Dynamic Mechanical Properties
Each test was repeated three times, and the average ultimate strength of concrete subjected to 100 FT cycles under triaxial compression at different strain rates and standard deviation δ are shown in Table 3. The concrete’s ultimate strength under uniaxial compressive loading was lower than that under triaxial compressive loading due to the constraint effect of triaxial compression. This condition can explain why the further expansion of the crack was limited by the water pressure. This finding is close to the results of [34, 47–49], which also stated that the grading of aggregates may be one of the factors that influence the value of ultimate strength [47]. The strength of concrete decreased as water pressure increased at a strain rate of 10^{−5}·s^{−1} [3], which is contrary to the findings of this study. This phenomenon may be primarily due to the different concrete conditions before the triaxial test. In this study, specimens were cured in water, whereas in [3], concretes were placed in a natural environment for 1200 days. Researchers [50, 51] noted that water will flow to the crackle tip and destroy the capillary structure of the dry concrete matrix under the action of water pressure. However, only a small amount of water will immerse into the watercured concrete because of waterfilled pores, thereby resulting in slower crack propagation and little or even negligible damage to the matrix.

Figure 13 shows the relationship between the compressive strengths of concrete and water pressure, where is the water pressure and is the ultimate compressive strength in quasistatic loading. The dynamic strength of concrete tended to increase with the increase in strain rate. When the constant lateral pressure was in the lower range (0–5 MPa), the dynamic ultimate compressive strength of concrete had a remarkable tendency to increase with the increase in strain rate. However, when the water pressure was higher (5–10 MPa), the increase in dynamic ultimate compressive strength of concrete became muted with the increase in strain rate. In addition, some researchers [33] believe that when the confining pressure exceeds a certain value and this value is the uniaxial compressive strength of concrete, the compressive strength of concrete no longer considers the effect of strain rate.
Figure 14 shows the effect of water pressure on the relationship between DIF and strain rate. The DIF of dynamic compressive strength and peak strain trend is shown at different strain rates. Figure 14(a) illustrates the relationship between the DIF of dynamic compressive strength of concrete and the strain rate under different water pressures. At the same strain rate, the slope of the fitted line decreased gradually with the increase in water pressure after 100 FT cycles. This result indicates that the presence of water pressure reduced the sensitivity of concrete strength. This finding is consistent with the view that the effect of strain rate decreases as water pressure increases [2, 33], but it diverges from the data by [3, 34]. Concrete in these papers [2, 3, 33, 34] was not exposed to FT cycles. Moreover, under 5 and 10 MPa of water pressure, the increment speed was nearly the same, because the water content of concrete under 5 MPa is equal to that under 10 MPa. As previous research [3] indicated, water is important in the improvement of concrete strength under the same strength of concrete skeleton at the same strain rate. Figure 14(b) also reveals that the peak strain of specimens increased with the increase in water pressure. This result can be due to the fact that, when water pressure was applied, the concrete specimens were restrained by lateral pressure, and the appearance and expansion of internal cracks were delayed. The compressive strength and deformation capacity of concrete improved, and the deformation property changed from brittleness to ductility. Consequently, the DIF of strength and peak strain of concrete under conventional triaxial stress state was strongly dependent on the value of water pressure.
(a)
(b)
3.6. Microstructure Characterisation
The damage of concrete exposed to the combined effect of FT cycles and strain rate under triaxial compression is a complex process that comprises physical and chemical aspects. Evaluation of the microstructure of damaged concrete is a common method to reveal the cause of failure. In this study, SEM was used to understand the coupled effect of FT cycles and strain rate on the concrete dynamic compressive strength at a water pressure of 5 MPa; the micrographs of concrete are shown in Figure 15. Figures 15(a)–15(c) demonstrate that at the strain rate of 1 × 10^{−4}·s^{−1}, a relatively tight and dense structure and only one microcrack inside the concrete specimens exist without the effect of FT cycles, which can be explained by shrinkage deformation of cement paste during hydration reaction [52, 53]. However, different degrees of FT damage occurred after being subjected to 50 and 100 FT cycles. The length, width, and number of microcracks increased. Microcracks connected to each other can be easily seen in Figure 15(c), which is good evidence that at the same strain rate, and the value of ultimate compressive strength decreased with the increase in FT cycles.
(a)
(b)
(c)
(d)
A comparison of Figures 15(c) and 15(d) reveals that, under the same number of FT cycles (100), the width of microcracks with a strain rate of 1 × 10^{−3}·s^{−1} was smaller than that with a strain rate of 1 × 10^{−4}·s^{−1}, because the speed of microcrack propagation increased with the increase in strain rate, and microcracks did not have enough time to expand fully. This view is consistent with the results indicating that the ultimate compressive strength at the strain rate of 1 × 10^{−3}·s^{−1} is higher than at 1 × 10^{−4}·s^{−1} after 100 FT cycles under a water pressure of 5 MPa.
4. Conclusions
In this study, dynamic compressive experiments for concrete subjected to different numbers of FT cycles (0, 50, and 100) and under various water pressure levels (0, 5, and 10 MPa) were conducted at different strain rates (1 × 10^{−5}, 1 × 10^{−4}, 2 × 10^{−4}, 5 × 10^{−4}, and 1 × 10^{−3}·s^{−1}). The following conclusions can be drawn from the test results:(1)As the number of FT cycles increases, the mass of concrete specimen initially increases and then decreases, whereas the relative dynamic elastic modulus of concrete tends to decrease.(2)With the increment of FT cycles at the same strain rate and a water pressure of 5 MPa, the dynamic ultimate compressive strength of concrete decreases, but the peak strain of concrete tends to increase. The decrease in compressive strength is due to the dramatic change in pore structure under the action of FT cycles. However, the corresponding ductility increases as the porosity of concrete increases, thereby improving peak strain.(3)With the same number of FT cycles under a water pressure of 5 MPa, the dynamic ultimate compressive strength of concrete increases with the increase in strain rate, whereas the peak strain of the concrete tends to decrease.(4)The strain rate sensitivity of the DIF of ultimate compressive strength and peak strain decreases with the increase in FT cycles. The same results apply to the influence of water pressure on the strain rate sensitivity of the DIF of ultimate compressive strength and peak strain. A formula that describes the relationship between the DIF of compressive strength and peak strain and strain rate under different numbers of FT cycles is proposed based on the experimental results. The calculated results are in good agreement with the experimental data using this equation.(5)After 100 FT cycles, with the increase in strain rate, the dynamic compressive strength under every water pressure level tends to increase, and its value under triaxial compressive loading is higher than that under uniaxial compressive loading. However, peak strain decreases gently, especially with a smaller variation under 10 MPa water pressure.(6)The SEM images of concrete at the same strain rate after FT action reveals more microcracks and less compacted microstructures, which coincide with the results of ultimate compressive strength. Moreover, cracks in concrete structures are relatively narrower under a high strain rate.
For future work, more FT cycles and a higher strain rate range may be tested to understand the precise research of strength and deformation under dynamic loading. Moreover, other elements, such as mineral admixture, watercement ratio, and specimen size, affect the dynamic mechanical properties of concrete under triaxial compression. Therefore, additional experiments are needed.
Data Availability
The data used to support the findings of this study are available upon request from the corresponding author.
Conflicts of Interest
The authors declare no conflicts of interest regarding the publication of this paper.
Acknowledgments
This study was financially supported by the National Natural Science Foundation of China (51679197).
References
 L. Bjerkeli, J. J. Jensen, and R. Lenschow, “Strain development and static compressive strength of concrete exposed to water pressure loading,” American Concrete Institute Structural Journal, vol. 90, no. 3, pp. 310–315, 1993. View at: Publisher Site  Google Scholar
 Z. Chen, Y. Hu, Q. Li, M. Sun, P. Lu, and T. Liu, “Behavior of concrete in water subjected to dynamic triaxial compression,” Journal of Engineering Mechanics, vol. 136, no. 3, pp. 379–389, 2010. View at: Publisher Site  Google Scholar
 Q. Wang, Y. Liu, and G. Peng, “Effect of water pressure on mechanical behavior of concrete under dynamic compression state,” Construction and Building Materials, vol. 125, pp. 501–509, 2016. View at: Publisher Site  Google Scholar
 D. Yan and G. Lin, “Influence of initial static stress on the dynamic properties of concrete,” Cement and Concrete Composites, vol. 30, no. 4, pp. 327–333, 2008. View at: Publisher Site  Google Scholar
 S. Xiao, H. Li, and P. J. M. Monteiro, “Influence of strain rates and loading histories on the compressive damage behaviour of concrete,” Magazine of Concrete Research, vol. 63, no. 12, pp. 915–926, 2011. View at: Publisher Site  Google Scholar
 L. E. Malvern, Dynamic Response of Concrete and Concrete Structures, University of Florida, Gainesville, FL USA, 1986.
 S. H. Ahmad and S. P. Shah, “Behavior of hoop confined concrete under high strain rates,” Journal of the American Concrete Institute, vol. 82, no. 5, pp. 634–647, 1985. View at: Publisher Site  Google Scholar
 P. Rossi and F. Toutlemonde, “Effect of loading rate on the tensile behaviour of concrete: description of the physical mechanisms,” Materials and Structures, vol. 29, no. 2, pp. 116–118, 1996. View at: Publisher Site  Google Scholar
 X. Chen, S. Wu, and J. Zhou, “Experimental and modeling study of dynamic mechanical properties of cement paste, mortar and concrete,” Construction and Building Materials, vol. 47, pp. 419–430, 2013. View at: Publisher Site  Google Scholar
 C. Wang and J. Xiao, “Evaluation of the stressstrain behavior of confined recycled aggregate concrete under monotonic dynamic loadings,” Cement and Concrete Composites, vol. 87, pp. 149–163, 2018. View at: Publisher Site  Google Scholar
 L. Jin, W. Yu, X. Du, S. Zhang, and D. Li, “Mesoscale modelling of the size effect on dynamic compressive failure of concrete under different strain rates,” International Journal of Impact Engineering, vol. 125, pp. 1–12, 2019. View at: Publisher Site  Google Scholar
 F.Q. Gong, X.F. Si, X.B. Li, and S.Y. Wang, “Dynamic triaxial compression tests on sandstone at high strain rates and low confining pressures with split Hopkinson pressure bar,” International Journal of Rock Mechanics and Mining Sciences, vol. 113, pp. 211–219, 2019. View at: Publisher Site  Google Scholar
 X. D. Chen, S. X. Wu, and J. K. Zhou, “Experimental study on dynamic tensile strength of cement mortar using split hopkinson pressure bar technique,” Journal of Materials in Civil Engineering, vol. 26, no. 6, Article ID 04014005, 2014. View at: Publisher Site  Google Scholar
 X. Chen, J. Bu, and L. Xu, “Effect of strain rate on postpeak cyclic behavior of concrete in direct tension,” Construction and Building Materials, vol. 124, pp. 746–754, 2016. View at: Publisher Site  Google Scholar
 J. Xiao, L. Li, L. Shen, and J. Yuan, “Effects of strain rate on mechanical behavior of modeled recycled aggregate concrete under uniaxial compression,” Construction and Building Materials, vol. 93, pp. 214–222, 2015. View at: Publisher Site  Google Scholar
 J. Xiao, L. Li, L. Shen, and C. S. Poon, “Compressive behaviour of recycled aggregate concrete under impact loading,” Cement and Concrete Research, vol. 71, pp. 46–55, 2015. View at: Publisher Site  Google Scholar
 L. Li, J. Xiao, and C. S. Poon, “Dynamic compressive behavior of recycled aggregate concrete,” Materials and Structures, vol. 49, pp. 4451–4462, 2016. View at: Publisher Site  Google Scholar
 Y. Lu, X. Chen, X. Teng, and S. Zhang, “Dynamic compressive behavior of recycled aggregate concrete based on split Hopkinson pressure bar tests,” Latin American Journal of Solids and Structures, vol. 11, pp. 131–141, 2014. View at: Publisher Site  Google Scholar
 L. Li, C. S. Poon, J. Xiao, and D. Xuan, “Effect of carbonated recycled coarse aggregate on the dynamic compressive behavior of recycled aggregate concrete,” Construction and Building Materials, vol. 151, pp. 52–62, 2017. View at: Publisher Site  Google Scholar
 Z. Deng, Y. Wang, J. Sheng, and X. Hu, “Strength and deformation of recycled aggregate concrete under triaxial compression,” Construction and Building Materials, vol. 156, pp. 1043–1052, 2017. View at: Publisher Site  Google Scholar
 B. W. Xu and H. S. Shi, “Correlations among mechanical properties of steel fiber reinforced concrete,” Construction and Building Materials, vol. 23, pp. 3468–3474, 2009. View at: Publisher Site  Google Scholar
 X. Sun, K. Zhao, Y. Li et al., “A study of strainrate effect and fiber reinforcement effect on dynamic behavior of steel fiberreinforced concrete,” Construction and Building Materials, vol. 158, pp. 657–669, 2018. View at: Publisher Site  Google Scholar
 L. Jin, R. Zhang, Y. Tian, G. Dou, and X. Du, “Experimental investigation on static and dynamic mechanical properties of steel fiber reinforced ultrahighstrength concretes,” Construction and Building Materials, vol. 178, pp. 102–111, 2018. View at: Publisher Site  Google Scholar
 X. Hou, S. Cao, W. Zheng, Q. Rong, and G. Li, “Experimental study on dynamic compressive properties of fiberreinforced reactive powder concrete at high strain rates,” Engineering Structures, vol. 169, pp. 119–130, 2018. View at: Publisher Site  Google Scholar
 J. Y. Chen, X. Qiu, G. Y. Shi, B. F. Chen, and X. S. Huang, “Effect of freezethaw cycles on mechanical behaviors of ceramsite concrete under impact loading,” Applied Mechanics and Materials, vol. 784, pp. 508–513, 2015. View at: Publisher Site  Google Scholar
 H. T. Wang, W. X. Fan, T. Y. Liu, and Z. Q. An, “Experimental study on properties of fullygraded concrete under uniaxial dynamic compressive stress state after freezing thawing cycles,” Journal of Hydraulic Engineering, vol. 46, no. 6, pp. 732–738, 2015. View at: Google Scholar
 J. Yi, S. Shen, B. Muhunthan, and D. Feng, “Viscoelasticplastic damage model for porous asphalt mixtures: application to uniaxial compression and freezethaw damage,” Mechanics of Materials, vol. 70, no. 1, pp. 67–75, 2014. View at: Publisher Site  Google Scholar
 K.P. Zhou, B. Li, J.L. Li, H.W. Deng, and F. Bin, “Microscopic damage and dynamic mechanical properties of rock under freezethaw environment,” Transactions of Nonferrous Metals Society of China, vol. 25, pp. 1254–1261, 2015. View at: Publisher Site  Google Scholar
 J. Lu, K. Zhu, L. Tian, and L. Guo, “Dynamic compressive strength of concrete damaged by fatigue loading and freezethaw cycling,” Construction and Building Materials, vol. 152, pp. 847–855, 2017. View at: Publisher Site  Google Scholar
 J. Li, R. B. Kaunda, and K. P. Zhou, “Experimental investigations on the effects of ambient freezethaw cycling on dynamic properties and rock pore structure deterioration of sandstone,” Cold Regions Science and Technology, vol. 154, pp. 133–141, 2018. View at: Publisher Site  Google Scholar
 J. K. Gran, A. L. Florence, and J. D. Colton, “Dynamic triaxial tests of highstrength concrete,” Journal of Engineering Mechanics, vol. 115, no. 5, pp. 891–904, 1989. View at: Publisher Site  Google Scholar
 K. Fujikake, K. Mori, K. Uebayashi, T. Ohno, and J. Mizuno, “Dynamic properties of concrete materials with high rates of triaxial compressive loads,” Structures under Shock and Impact, pp. 511–522, 2000. View at: Google Scholar
 D. M. Yan, “Experimental and theoretical study on the dynamic properties of concrete,” Ph.D. thesis, Dalian University of Technology, Dalian, China, 2006. View at: Google Scholar
 H. Wang, L. Wang, Y. Song, and J. Wang, “Influence of free water on dynamic behavior of dam concrete under biaxial compression,” Construction and Building Materials, vol. 112, pp. 222–231, 2016. View at: Publisher Site  Google Scholar
 GB 175–2007, Common Portland Cements, General Administration of Quality Supervision, Inspection and Quarantine of the People’s Republic of China and Standardization Administration of the People’s Republic of China, Beijing, China, 2008.
 National Standard of the People’s Republic of China, GB/T500822009 Standard for Test Methods of LongTerm Performance and Durability of Ordinary Concrete, China Architecture and Building Press, Beijing, China, 2009, in Chinese.
 X. X. Zhang, R. C. Yu, G. Ruiz, M. Tarifa, and M. A. Camara, “Effect of loading rate on crack velocities in HSC,” International Journal of Impact Engineering, vol. 37, no. 4, pp. 359–370, 2010. View at: Publisher Site  Google Scholar
 F. H. Wittmann, P. E. Roelfstra, H. Mihashi, Y. Y. Huang, X. H. Zhang, and N. Nomura, “Influence of age of loading, watercement ratio and rate of loading on fracture energy of concrete,” Materials and Structures, vol. 20, no. 2, pp. 103–110, 1987. View at: Google Scholar
 R. John, S. P. Shah, and Y.S. Jeng, “A fracture mechanics model to predict the rate sensitivity of mode I fracture of concrete,” Cement and Concrete Research, vol. 17, no. 2, pp. 249–262, 1987. View at: Publisher Site  Google Scholar
 H. W. Reinhardt, H. A. W. Cornelissen, and D. A. Hordijk, “Tensile tests and failure analysis of concrete,” Journal of Structural Engineering, vol. 112, no. 11, pp. 2462–2477, 1986. View at: Publisher Site  Google Scholar
 S. A. Kaplan, “Factors affecting the relationship between rate of loading and measured compressive strength of concrete,” Magazine of Concrete Research, vol. 32, no. 23, pp. 79–88, 1980. View at: Publisher Site  Google Scholar
 P. Rossi, J. G. M. Van Mier, F. Toutlemonde, F. Le Maou, and C. Boulay, “Effect of loading rate on the strength of concrete subjected to uniaxial tension,” Materials and Structures, vol. 27, no. 169, pp. 260–264, 1994. View at: Publisher Site  Google Scholar
 P. Rossi, “Influence of cracking in the presence of free water on the mechanical behaviour of concrete,” Magazine of Concrete Research, vol. 43, no. 154, pp. 53–57, 1991. View at: Publisher Site  Google Scholar
 P. Rossi and C. Boulay, “Influence of freewater in concrete on the cracking process,” Magazine of Concrete Research, vol. 42, no. 152, pp. 142–146, 1990. View at: Publisher Site  Google Scholar
 C. Wang, W. Chen, H. Hao, S. Zhang, R. Song, and X. Wang, “Experimental investigations of dynamic compressive properties of roller compacted concrete (RCC),” Construction and Building Materials, vol. 168, pp. 671–682, 2018. View at: Publisher Site  Google Scholar
 J. Cui, H. Hao, and Y. Shi, “Numerical study of the influences of pressure confinement on highspeed impact tests of dynamic material properties of concrete,” Construction and Building Materials, vol. 171, pp. 839–849, 2018. View at: Publisher Site  Google Scholar
 H.L. Wang and Y.P. Song, “Biaxial compression behaviour of different aggregate graded concrete,” Magazine of Concrete Research, vol. 61, no. 6, pp. 457–463, 2009. View at: Publisher Site  Google Scholar
 D. Yan, S. Xu, G. Chen, and H. Li, “Biaxial behaviour of plain concrete subjected to dynamic compression with constant lateral stress,” Structural Concrete, vol. 15, no. 2, pp. 202–209, 2014. View at: Publisher Site  Google Scholar
 D. M. Yan and G. Lin, “Dynamic behavior of concrete in biaxial compression,” Magazine of Concrete Research, vol. 59, no. 1, pp. 45–52, 2007. View at: Google Scholar
 I. Yurtdas, N. Burlion, and F. Skoczylas, “Triaxial mechanical behaviour of mortar: effects of drying,” Cement and Concrete Research, vol. 34, no. 7, pp. 1131–1143, 2004. View at: Publisher Site  Google Scholar
 L. Chen, G. Duveau, and J. F. Shao, “Modelling of plastic deformation and damage in cementbased material subjected to desiccation,” International Journal for Numerical and Analytical Methods in Geomechanics, vol. 35, no. 17, pp. 1877–1898, 2010. View at: Publisher Site  Google Scholar
 M. Sahmaran, M. Lachemi, K. M. A. Hossain, and V. C. Li, “Internal curing of engineered cementitious composites for prevention of early age autogenous shrinkage cracking,” Cement and Concrete Resarch, vol. 39, no. 10, pp. 893–901, 2009. View at: Google Scholar
 W. Zhang, S. Chen, N. Zhang, and Y. Zhou, “Lowvelocity flexural impact response of steel fiber reinforced concrete subjected to freezethaw cycles in NaCl solution,” Construction and Building Materials, vol. 101, no. 10, pp. 522–526, 2015. View at: Publisher Site  Google Scholar
Copyright
Copyright © 2019 Ruijun Wang et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.