Study on Pore Structure and Mechanical Property of Expansive Soil under Different Dehydration Conditions

Abstract

To study the influence of dehydration conditions on the pore structure and the mechanical property of expansive soil, two experimental conditions with high-temperature drying (temperature 50 °C-humidity 10%) and high-temperature humid (temperature 50 °C-humidity 70%) were carried out. Taking the remolded expansive soil in the province of Anhui in China as the research object, this paper used mercury intrusion porosimetry (MIP) and scanning electron microscopy (SEM) to analyze the pore size distribution and pore structure of remolded expansive soil samples under different dehydration conditions and duration. After these tests, their mechanical properties were further obtained by uniaxial compression tests. The results showed that the distribution of pore structure of expansive soil was various under different dehydration conditions. Under high temperature and dry environment, the volume of large pores decreased first and then increased with the time of dehydration, while the medium pore volume decreased until the dehydration was stable. Under high temperature and humid environment, the volume of large pore and medium pore both showed a trend of decrease until the dehydration kept stable. The pore volume of expansive soil did not change significantly under the two dehydration conditions. The uniaxial compressive strength (UCS) of remolded expansive soil samples in high temperature and dry environment reached the highest on the 5th day of dehydration, and then the soil strength decreased slightly until it stabilized. The UCS of remolded expansive soil reached the highest on the 15th day of dehydration under high temperature and humidity environment, and the soil strength changed little after continual dehydration. These tests showed that the UCS of dehydrated expansive soil samples under the condition of high temperature and humidity is higher than that of dehydrated expansive soil samples under the condition of high temperature and dry environment.

1. Introduction

Expansive soil is a kind of special soil with strong water sensitivity and structural instability, so it has problems of swelling-shrinkage, fissure and over consolidation [1]. A lot of studies have been carried out on the macroscopic behavior of expansive soil such as swelling-shrinkage and its treatment [2,3,4]. Nevertheless, more studies on the microscopic structure of expansive soil under different dehydration environments are needed before the treatment. The Pishihang irrigation area in the Anhui province in China is located in the temperate monsoon region [115.4° E 32.7° N, 117.9° E 31.2° N], which spans a large expandable soil and has a distinct climate throughout the year. It is characterized by high temperature and humidity in summer, dry in autumn, and low temperature and little rain in spring and winter. The expansive soil located in the study area is in an environment of significant temperature and humidity change all year round, and the moisture dehydration is severe, which leads to the cracking and deformation of roadbed and slope in different degrees, and seriously restricts the local traffic and the operation of irrigation. Therefore, it is of great significance to carry out experimental research on the microscopic pore structure and mechanical properties of expansive soil under different dehydration conditions.

Many studies showed that the occurrence of soil fissures was closely related to the rate of dehydration, and the secondary fissures generated by the dehydration became a new water transport channel [5,6,7]. In the process of dehydration, the water content of soil decreased, and the internal pore structure of soil was constantly adjusted, resulting in the change of mechanical properties of soil [8,9,10]. According to the experimental studies of mercury intrusion porosimetry (MIP) and scanning electron microscopy (SEM), Zhao et al., found that temperature and humidity had significant effects on the pore size distribution and micro-structure of the large pore and medium pore of the expansive clay, but little effect on the small pore [11]. Tang et al., adopted MIP, SEM, computed tomography (CT) and some other methods to study the evolution of soil micro-structure under wetting-drying cycles. It was found that the decrease of soil volume was mainly caused by the shrinkage of pores between soil aggregates in the drying process. The number and size of pores inside and between soil aggregates increased gradually during wetting, and the changes of soil structure were not completely reversible during wetting-drying cycles [12].

The change of small pore structure led to macro cracks in soil, which affected the bearing capacity of soil [13,14]. Zhu et al., measured the characteristics of mechanical strength attenuation of expansive soil under freezing-thawing cycles through triaxial test [15]. Pei et al., proposed the safety factor of expansive soil slope with different fracture forms by analyzing the hydraulic mechanical properties of expansive soil [16]. Xu et al., studied the development process of dry-shrinkage cracks in soil during dehydration, and found that water evaporation is the premise of soil cracking, and the dry-shrinkage cracks were the result of internal stress of soil [17]. For expansive soil, the fracture morphology and pore structure caused by shrinkage were different under different dehydration, which results in different mechanical properties of soil, and the dehydration rate was controlled by the dehydration environment and the mineral composition of soil itself. Most of the existing studies considered the evolution of macroscopic expansion and contraction deformation and microstructure of soil under single environment when soil is hygroscopic or dehydrated, but the research on the evolution process and mechanical properties of microscopic pore structure of expansive soil under different dehydration environments is relatively rare.

The expansive soil samples used in this study were remolded from the monsoon region. The testing flow started with the basic property tests, including liquid and plastic limited water content, free expansion rate, shrinkage factor, specific gravity, particle distribution curve, mineral composition, compaction curve, etc. The mineral types and contents of soil samples were obtained by XRD experiments. Then, the effects of high temperature-drying and high temperature-humidity on pore volume and pore size distribution of these samples were evaluated based on MIP test. According to the SEM test, the changes of expansive soil microstructure under different dehydration conditions were analyzed. Moreover, the mechanical properties of expansive soil under different dehydration conditions and duration were obtained through uniaxial compression test.

2. Materials

The expansive soil used in this paper comes from Pishihang, Anhui province, which is located in the sub-humid monsoon region. Its temperature and rainfall are obvious seasonal, resulting in the seasonal variations of cracking and failure of expansive soil foundation and slope, with frequent destruction in summer and autumn, and fewer disasters in spring and winter.

According to the Technical Code for Building in Expansive Soil Region (GB50112-2013), the expansion grade of our samples was defined as weak expansive soil. The basis property of expansive soil samples such as free swelling ratio, liquid and plastic limit and shrinkage factor are tested and shown in

Table 1. Basic physical properties of expansive soil samples.

Natural Moisture Content (%)	Natural Density (g/cm3)	Free Swelling Ratio (%)	Liquid Limit (%)	Plastic Limit (%)	Plastic Index	Shrinkage Factor	Specific Gravity Gs
26.78	1.98	55	46.4	18.8	27.6	0.56	2.75

. More properties of samples, including mineral composition, particle size distribution and compaction curve are also tested and shown in

Table 2. Mineral composition of expansive soil.

Mineral Content (%)				Clay Mineral Relative Content (%)
Quartz	Orthoclase	Plagioclase	Clay Mineral	Illite-Smectite Formation	Illite	Kaolinite
51.4	2.6	10.7	35.3	80	16	4

, Figure 1 and Figure 2, respectively. The soil samples were prepared into cylindrical shape with 90% compacted degree, 100 mm height and 50 mm diameter. Before the experiment, the prepared cylindrical expansive soil samples were dealt with vacuum extraction and saturation in a saturator.

3. Experimental Methods and Tests

3.1. Dehydration Test

The dehydration test for the saturated cylindrical expansive soil was conducted under two environmental conditions with temperature-humidity of 50 °C-10% and 50 °C-70%, respectively. The samples were weighed and recorded every day until the mass was stable (changed less than 0.2 g for two consecutive days), and then calculated the variation of moisture content of the soil samples with the time length of dehydration. In the process of dehydration, the soil samples were wrapped with latex film, and exposed the upper surface to air, which simulated the unidirectional dehydration process of soil.

3.2. MIP Test

In this study, the MIP test was carried out with a high-performance automatic mercury injection apparatus of Michael AutoPore V 9600. Place the above saturated soil samples in the temperature-humidity of 50 °C-10% and 50 °C-70% environments, respectively, to carry out the dehydration test. Then take out a small piece of expansive soil sample for MIP test to measure the distribution characteristics of pore structure in different stages of the experiment. The MIP tests were performed at ambient pressures ranging from 0.5 to 33,000 psi, pore sizes ranging from 350 μm to 5 nm, and temperatures ranging from 20 °C to 22 °C. To keep the pore structure unchanged, the soil samples were frozen and air-dried with liquid nitrogen [18], which can sublimate the frozen water in the sample. Moreover, the pre-dried samples were vacuumized to remove water vapor and gas from the sample before injecting mercury into the sample dilatometer.

The principle of MIP method to measure the microscopic pore structure of expansive soil is based on Washburn equation [19], and the relationship between mercury pressure P and capillary radius R is obtained as follows:

$$P=-\frac{2\sigma cos\phi }{R}$$

(1)

where σ is the surface tension of mercury (take σ = 0.485 N·m⁻¹); φ is the wetting angle between the measured porous material and mercury (take φ = 130° for the test).

The soil sample placed in the 50 °C-10% environment was dehydrated for 15 days to reach equilibrium. Then take out small pieces of expansive soil samples under different dehydration duration including initial state, 1 day, 3 days, 5 days, 7 days, 10 days and 15 days, respectively, and conduct the MIP tests after freezing and air drying. The soil samples placed in 50 °C-70% environment was dehydrated for 26 days to reach equilibrium. Then take out small expansive soil samples under different dehydration duration including initial state, 1 day, 3 days, 5 days, 7 days, 10 days, 15 days, 20 days and 26 days, respectively, and conduct MIP tests after freezing and air drying. On the other hand, the MIP method can also measure the porosity of soil samples based on the relationship:

Porosity = Mercury Injection Volume/Sample volume

(2)

3.3. SEM Test

Czech TESCAN SEM was used to scan and photograph the soil samples under different dehydration conditions and states to observe the morphological changes of expansive soil. The saturated soil samples above were placed in an environment of 50 °C-10% humidity and 50 °C-70% humidity for dehydration. After stabilization, small pieces of expansive soil samples were taken out for freeze-drying treated by liquid nitrogen. SEM tests with 1 kx magnification were carried out to observe the microstructure distribution of large pores and small pores.

3.4. Uniaxial Compression Test

Cylinder samples of vacuum-saturated expansive soil (height: 100 mm, diameter: 50 mm) were placed in environments of 50 °C-10% humidity and 50 °C-70% humidity for dehydration, respectively. Take out the soil samples with different dehydration duration and then carry out the uniaxial compression tests (Standard for Soil Test Method GB/T 50123-1999) to obtain the change rule of expansive soil mechanical strength under different dehydration environments. The loading rate of uniaxial compression is 2% strain per minute.

The soil sample placed in the 50 °C-10% environment was dehydrated for 15 days to reach equilibrium. Then take out small pieces of expansive soil samples under different dehydration duration including initial state, 1 day, 3 days, 5 days, 7 days, 10 days, and 15 days, respectively, to conduct the uniaxial compression test. The soil sample placed in 50 °C-70% environment was dehydrated for 26 days to reach equilibrium. Then take out small expansive soil samples under different dehydration duration including initial state, 1 day, 3 days, 5 days, 7 days, 10 days, 15 days, 20 days and 26 days, respectively, to conduct the uniaxial compression test.

4. Results and Discussion

4.1. Moisture Content Variation

The initial saturated expansive soil sample was dehydrated at 50 °C-10% humidity and 50 °C-70% humidity. After a period of time, the mass of the expansive soil samples reached stability (Figure 3). The results showed that the dehydration stability time and moisture content of expansive soil samples were different under different dehydration conditions. When the soil samples were placed in the environment of 50 °C-10%, the moisture content began to change gently on the 10th day of dehydration. After 15 days, the moisture content of soil samples reached a dynamic balance, and the value was 4.15% when dehydration was stable. While the moisture content of soil samples placed in the environment of 50 °C-70% tended to be flattening out on the 15th day of dehydration. When the soil mass reached a dynamic balance on the 26th day of dehydration, the moisture content of the soil is stable and 6.30%. Compared with the variation characteristics of moisture content under other conditions [17], the high temperature environment has a great influence on the dehydration rate.

4.2. Pore Size Distribution

Based on the MIP test, the curves of cumulative pore volume distribution and pore size distribution of expansive soil under different dehydration conditions and duration were obtained, as shown in Figure 4 and Figure 5.

Figure 4 shows that the pore size distribution range of expansive soil measured by MIP test was 5–10⁶ nm. Seen from Figure 5, the initial state of pore size distribution curve presented a three-peak distribution, and the corresponding distribution ranges of peak values were 5–100 nm, 500–4000 nm, and 4000–10,000 nm, respectively. The corresponding pore size range of peak values kept changing with the duration of dehydration. Referring to the classification method of pores by Zhang et al., and Ding et al. [20,21], and based on the tri-modal of pore distribution curve of expansive soil, the pores of expansive soil were divided into small pores (5 nm ≤ d < 300 nm), medium pores (300 nm ≤ d < 4000 nm) and large pores (d ≥ 4000 nm). The pore size of the samples used in this study are characterized by the medium pore range, which is different from the soils in western and southern China [20].

Figure 6 shows that under different dehydration environments, the pore volumes of large, medium and small pores varied with the duration of dehydration. In the dehydration environment of high temperature and dry (50 °C-10% humidity) environment, the large pore volume of expansive soil samples decreased first and then increased with the dehydration process. In the initial state, the large pore volume was 0.093 mL/g, and decreased to 0.021 mL/g on the 7th day of dehydration. When the soil mass reached dynamic balance after 15 days of dehydration, the porous volume of soil was 0.44 mL/g. In the initial state, the volume of medium pore in the sample was 0.107 mL/g. On the 5th day of dehydration, the volume of medium pore decreased to 0.042 mL/g. Thereafter, the change of the volume of medium pore in the soil sample fluctuated little. The volume of small pore changed little, whose initial pore volume was 0.066 mL/g and the final pore volume was 0.057 mL/g when dehydration reached a stable state.

Under high temperature and humidity (50 °C-70%) environment, the volume of large and medium pores of expansive soil samples both decreased with dehydration until stable, while the volume of small pores changed little. In the initial state, the volume of large pores was 0.093 mL/g, and it decreased to 0.025 mL/g on the 5th day of dehydration. Thereafter, the volume of large porous soil fluctuated little until the mass of soil sample reached dynamic balance on the 26th day of dehydration. The initial pore volume was 0.107 mL/g, and decreased to 0.039 mL/g on the 5th day of dehydration. Thereafter, the pore volume of soil samples fluctuated little. However, under the high temperature and humidity environment, the pore volume of soil sample changes a little, whose initial pore volume was 0.066 mL/g, and the final pore volume was 0.059 mL/g when dehydration reached a stable state.

According to the results of pore size distribution, we can find that the dehydration rate of expansive soil samples is fast in the initial stage due to the environmental influence of high temperature and dry. The soil shrinks because of the rapid water loss, resulting in the rapid shrinkage of the large and medium pores in the soil, and then the volume content of pores shows a decreasing trend. This result is supported by the study of influences of the drying rates on shrinkage [22]. As we all know, the occurrence of cracks in soil is usually controlled by mineral composition and structural composition. Then the distribution of water content in soil is always nonuniform in the process of rapid dehydration due to the difference of minerals and structure in soil samples, which leads to the emergence of tensile cracks. At the same time, dehydration leads to the gradual connection of some medium pores in the soil, which forms new large pores. Therefore, with the process of dehydration, the large pores in the soil samples show a trend of decreasing first and then increasing, while the medium pores show a trend of decreasing until dehydration is stable. On the other hand, in the evaporation environment, the small pore volume decreases slightly, and the overall change is not big. The main reason is that small pores are controlled by the degree of clay minerals consolidation. The particles of clay minerals are small, and the water molecules attached to the surface mostly exist in the state of bound water. In the dehydration process, the free water molecules in soil are the first to be dehydrated, and the energy required for dehydration of bound water is much higher than that of free water molecules, so it is more difficult to become dehydrated. Therefore, under the environment of 50 °C-10%, water molecules in small pores are relatively less affected by the environment, so the volume of small pores does not change much in the process of dehydration.

As a result of high moisture content under the environment of high temperature and humidity, the dehydration rate of expansive soil is relatively slow. This slow process makes the soil moisture content change relatively evenly in the soil, resulting in the relatively uniform shrinkage. In the process of dehydration, large and medium pores become small or even closure due to the dehydration action. Therefore, the volume of large and medium pores decreased until dehydration is stable. The water molecules inside the pores mostly exist in the state of bound water, and the dehydration environment of 50 °C-70% humidity has little influence on the bound water, so the fluctuation of the small pore volume is small.

Based on the MIP test, the variation trend of expansive soil porosity with the duration of dehydration in different environmental conditions was also obtained (Figure 7). According to Figure 7, under both high-temperature dry and high-temperature humidity dehydration conditions, the porosity of expansive soil samples decreased firstly, then increased and finally tended to be stable during the dehydration process. This process is opposite to the result of freeze-thaw cycle which kept the same dehydration condition [15]. The porosity of the expansive soil was 39.5% in the initial state, and decreased to 23.4% after dehydrating for a period of time under the environment of 50 °C-10% humidity. The porosity increased to 26.4% when the dehydrated mass of the sample reached dynamic balance. When the dehydration test was carried out at the condition of 50 °C-70% humidity, the porosity of the expansive soil sample decreased from 39.5% to 22.8%, and then increased to 28.6% when the dehydration of the soil sample reached dynamic equilibrium.

In this paper, we can find that the microscopic pore structure of expansive soil samples undergo a process of shrinkage and deformation due to the dehydration process under the conditions of high temperature drying and high temperature humidity, so the porosity decreases in the initial stage of dehydration. On the other hand, the uneven dehydration process is caused by the difference in mineral structure of the sample itself, which leads to the formation of tensile cracks in the soil. In this way, the newly formed tensile cracks become the channel of water migration, which further intensifies dehydration and leads to the increase of porosity.

4.3. Microstructure Feature

Based on SEM tests, the changes of the microstructure of expansive soil samples were obtained in two environmental conditions: high-temperature drying (50 °C-10%) and high-temperature humid (50 °C-70%), as shown in Figure 8.

As shown in Figure 8, the microstructures of expansive soil samples under different environmental conditions are different and they changed significantly before and after dehydration. Firstly, Figure 8a shows the SEM image of the expansive soil sample under initial saturation. At this time, the overall pore structure of the expansive soil is relatively loose, and there are many loose pores and fissures among the clay particles. The large pore size is more than 25 μm, and the small pore size is less than 4 μm. Moreover, there are no obvious connected cracks and clay particle aggregates in the expansive soil samples under initial saturation state. The size of microstructures of initial expansive soil is similar to previous study [23].

Figure 8b shows the SEM image of the microstructure of the expansive soil sample under the stable dehydration condition in the environment of 50 °C-10%. Seen from the figure that after dehydration in the high temperature drying environment, the microstructure of the expansive soil sample changes from the original loose state to the layered and tightly stacked state, and some clay mineral particle aggregates can also be recognized from the SEM image. The shrinkage cracks are distributed between the aggregates and the surrounding layered stacked particles, and some of the cracks are cross-connected, enlarging the size of cracks. Under the high temperature drying environment, the initial saturated expansive soil sample is dehydrated rapidly and unevenly due to the difference of mineral composition and internal structure of the sample. The uneven dehydration rate among soil particles results in uneven shrinkage and cracking in soil. Moreover, the originally loose clay mineral structure becomes dense due to water loss, and the SEM image presents the shape of stacking and dense agglomerate. With the dehydrating, the shrinkage cracks between clay minerals continue to expand and gradually form partial connected cracks.

Figure 8c shows that the microstructure of expansive soil samples becomes more compact after dehydration and stabilized under high temperature and humidity environment. Although there are some fractures with small size, there are no large granular aggregates and obvious connected cracks. Under the influence of high humidity, the initial saturated expansive soil dehydrated slowly, resulting in the slow shrinkage of the lattice structure of clay minerals. Then the internal stress of soil is adjusted to the new soil-water interaction state. In this way, the process of water dehydration is uniform, so the microstructure of soil is compact, and there are no large particle aggregates or obvious connected cracks. Compared with initial samples [23], the size of microstructures are smaller, which is close to that of treated samples [4].

4.4. UCS

The experimental study on the mechanical characteristics of expansive soil were carried out in high temperature-drying (50 °C-10%) and high temperature-humidity (50 °C-70%) conditions. Figure 9 shows the mechanical properties of expansive soil in different dehydration environments varying with dehydrated time. Figure 10 shows the variation of UCS of expensive soil with dehydration time.

Seen from Figure 9, both the environmental condition and dehydration duration have significant effects on mechanical properties of expansive soil. In the high temperature dehydration environment, the axial stress of its deformation and failure is very large, and the failure mode is no longer just creep mode [24]. The detailed changes are as follows.

According to Figure 9a, in the high temperature drying environment, the failure mode of the expansive soil sample gradually changed from creep failure to brittle failure with the process of dehydration. In the initial state, the expansive soil sample began to produce obvious swelling deformation when the axial strain was 6.9%, and the axial stress of the sample was 20.0 kpa, presenting typical creep failure characteristics. In this environment, a mechanical experiment was carried out on the soil sample after one day of dehydration. When the axial strain was 4.2%, the sample began to produce obvious bulging deformation, and the axial stress of the soil sample was 401.2 kpa. After 7 days of dehydration, the expansive soil sample failed at the axial strain of 2.1%, forming an obvious shear plane, and the axial stress of the sample was 3096.7 kpa. After 15 days of dehydration, shear failure occurred at the axial strain of 1.5%, and the axial stress of soil sample was 2886.6 kpa.

As shown in Figure 9b, the failure mode of expansive soil samples gradually changed from creep failure mode to brittle failure mode with the process of dehydration in high temperature and humidity environment. In the initial state, the swelling deformation of the expansive soil sample began to occur when the axial strain was 6.9%, and the axial stress of the sample was 20.0 kpa, showing typical creep failure characteristics. Mechanical experiments were carried out after one day dehydration in this environment. The samples began to produce obvious bulging deformation at the axial strain of 4.5%, and the axial stress of the soil sample was 150.6 kpa. After 7 days of dehydration, shear failure occurred when the axial strain was 3.8%, and the axial stress of the sample was 4417 kpa. After 26 days of dehydration, brittle shear failure occurred when the axial strain was 2.6%, and the axial stress was 5660.1 kpa.

Figure 10 shows that the UCS of expansive soil samples varies with dehydration time under different environmental conditions. In the high temperature dehydration environment, the UCS is much larger compared to that of samples under other conditions [2,3].

Under the high temperature and dry (50 °C-10%) environment, the UCS of the sample increased first, then decreased, and finally reached a stable trend with the process of dehydration. The maximum UCS of soil samples in the initial state was 27.8 kpa. With the dehydration process, the maximum UCS of soil sample was 3277 kpa after 5 days of dehydration. After 10 days of dehydration, the UCS of the soil sample decreased to 2546 kpa. The UCS was 2886.6 kpa until the mass of soil sample reached stability after 15 days of dehydration.

Under high temperature and humidity (50 °C-70%) environment, the UCS of the sample increased first and then reached the dynamic equilibrium with the dehydration process. In the initial state, the maximum UCS of the soil sample was 27.8 kpa. When the soil was dehydrated for 15 days, the maximum UCS of the sample reached 5921 kpa. Thereafter, the UCS of the sample changed little. After 26 days of dehydration, the UCS of the sample was 5658 kpa.

According to the results above, we found that the expansive soil samples underwent three stages of rapid dehydration, slow dehydration and stable dehydration when the dehydration tests were carried out under high temperature and drying environment. In the first stage, due to rapid water loss, the sample shrinks, resulting in the porosity decreasing and the soil compactness increasing. In the microstructure, the volume of large and medium pores decreased, while the strength increased in the mechanical properties. In the second stage, with further dehydration, some cracks caused by uneven shrinkage gradually expanded and even cut through, which was manifested by the increase of large pore volume and the formation of aggregate and connected cracks. In terms of mechanical properties, the strength decreased. The third stage is stable dehydration. In this stage, the dehydration rate decreased, and the soil gradually reached a dynamic dehydration equilibrium, hereafter the pore volume and mechanical strength of the soil tended to be stable.

In the dehydration test under high temperature and humidity environment, the expansive soil sample underwent two stages of slow dehydration and stable dehydration. In the first stage, affected by the high humidity, the moisture migration rate of soil sample was slow, the internal stress of soil had been adjusted to the new state of soil-water. Therefore, in the process of dehydration, the structure of expansive soil gradually became dense, which was manifested as the volume of large and medium pores gradually decreased, and the strength gradually increased in mechanical properties. In the second stage, due to the volume decrease of the large pore and medium pore in soil, the water migration channel was reduced, and the dehydration gradually reached a stable state. After that, the pore volume and mechanical strength of soil gradually tended to be stable.

5. Conclusions

This research adopts the weighing method, the MIP test, SEM test and unconfined compression experiment to study the pore structure and mechanical properties of expansive soil under high temperature-drying and high temperature-humidity dehydration conditions, and then obtain the change rule of water content, pore volume distribution, micro-pore structure and mechanical properties. The conclusions of these studies are as follows:

(1) The moisture contents of expansive soil samples after dehydration under two environmental conditions are different, and the time for dehydration to reach stability is different.

(2) The environmental conditions have a significant effect on the volume distribution of large and medium pores, but little effect on small pores. Under high temperature and drying conditions, the volume of large pores decreases first and then increases, while the volume of medium pores decreases first and then stabilizes. Under high temperature and humidity conditions, the volume of large pores and medium pores decreases first and then keeps stable.

(3) Under high temperature drying and high temperature humidity, the porosity of expansive soil samples decreases first and then fluctuates little.

(4) The dehydration environment has a significant effect on the microscopic pore structure of expansive soil. After dehydration and stabilization in the high temperature drying environment, the microstructure of expansive soil presents a dense stack and a dense aggregate, and there are connected cracks. However, after dehydration and stabilization in the high temperature and humidity environment, the microstructure of expansive soil is dense and no obvious connected cracks are observed.

(5) The environment and duration of dehydration have significant effects on the mechanical properties of expansive soil. With the process of dehydration, the mechanical failure characteristics of expansive soil samples change from creep deformation to brittle shear failure. The UCS of expansive soil under high temperature and drying environment increases first, then decreases and finally stabilizes, while the UCS of expansive soil under high temperature and humidity environment firstly increases and then stabilizes.