Failure Mechanisms and Protection Measures for Expansive Soil Slopes: A Review

Abstract

Due to the significant hydrophilicity and cracking properties of expansive soils, expansive soil slopes are prone to destabilization and landslides after rainfall, seriously threatening the safety of buildings, highways, and railroads. Substantial economic losses often accompany the occurrence of expansive soil slope disasters; thus, it is of great significance to understand the slope failure mechanisms experienced by expansive soil slopes and to prevent expansive soil slope disasters. In this paper, the current research status of the landslide failure mechanism of expansive soil slopes is systematically reviewed based on three research methods: field test, model test, and numerical simulation. The failure mechanisms of expansive soil slopes and the main influencing factors are summarized. Based on the failure mechanisms, three protection principles (waterproofing and water blocking, swelling–shrinkage deformation limitation, and crack inhibition and strength enhancement) that can be followed for disaster prevention of expansive soil slopes are proposed. The research status and advantages and disadvantages of these protection methods are reviewed, and future researchable directions of the stability of expansive soil slopes and slope protection methods are explored. Based on the previous work, a new flexible ecological slope protection system with a double waterproof layer is proposed for expansive soil slopes to realize ecological, efficient, and long-term protection. This paper thus aims to provide technical reference for the prevention and control of slope engineering disasters in expansive soil areas.

1. Introduction

Expansive soil is a potentially catastrophic soil that is widely distributed around the globe and has been found in more than 60 countries and regions across six continents [1]. Expansive soils have significant water sensitivity. They are susceptible to swelling and softening after water absorption, and shrinkage and cracking after water loss. This results in extremely poor engineering properties of expansive soils. In the United States, expansive soils cause $15 billion in economic damage annually, which is more than the damage caused by earthquakes, floods, hurricanes, and tornadoes combined [2,3,4]. China has one of the widest distributions of expansive soils (Figure 1), encompassing approximately 600,000 square kilometers [1,5]. In recent years, due to rapid economic development, a large number of roads, railroads, and other transportation facilities have been constructed globally. Many of these transportation facilities will likely face the challenge of construction in expansive soil areas. These constructions could result in many newly excavated expansive soil slopes, requiring adequate slope protection to ensure the quality of the projects. The failure of expansive soil slopes is potential, repetitive, and long-term, and the resulting losses and repair difficulties are more remarkable than is the case with general slopes. Accordingly, expansive soil is also known as the “cancer” of the engineering world. During the initial period of operation of the Nanning-Kunming Railway in China, the annual cost of repairing expansive soil embankments and managing slope failures amounted to more than 3.6 million US dollars [6]. In India, South Africa, and many other countries, there are also large quantities of expansive soils that cause severe serious damage to transport infrastructure [7]. Repeated failures of expansive soil slopes in states such as Mississippi and Texas in the United States have seriously affected highway operations [8]. These engineering accidents have forced many scholars to research expansive soil slopes [8,9,10]. From the perspective of engineering safety and economy, it is vital to analyze the failure mechanism of expansive soil slopes comprehensively and systematically to improve the protection methods of such slopes. This paper can provide a comprehensive and systematic technical reference and basis for the study of the failure mechanism of expansive soil slopes and the protection measures for those slopes.

The relatively easy destabilization of expansive soil slopes is closely related to the three properties of expansive soils: swelling–shrinkage, cracking, and over-consolidation. Swelling–shrinkage refers to the dramatic expansion and shrinkage in expansive soil volume with the absorption and evaporation of moisture [11]. Water films form around the mineral particles when expansive soils interact with water. Changes in the soil moisture can cause the water film to thicken or diminish. As a result, the distance between the particles will thus increase or decrease, manifesting macroscopically as swelling or shrinking of the soil [12,13]. Expansive soils are rich in hydrophilic minerals, such as montmorillonite. Montmorillonite has a significantly high specific surface area, which attracts a large amount of water [14,15]. This means expansive soils have a high free expansion index [14]. The swelling–shrinkage of expansive soils can severely alter the stresses and geometry of slopes, which can induce slope instability [16]. Some scholars found that the higher the montmorillonite content, the lower the shear strength of the soil [17]. Therefore, a high montmorillonite content seriously reduces the stability of expansive soil slopes. Cracking of expansive soil occurs when the soil shrinks or swells with water loss or absorption. Based on their connecting states, the cracks have been classified into interconnected and isolated categories [18]. The volume of isolated cracks remains almost unchanged, while the volume of interconnected cracks increases substantially when the water content of expansive soils decreases [18]. Rainwater can infiltrate expansive soil slopes through the cracks, softening and reducing the strength of the slope soils [19,20]. The higher the montmorillonite content in expansive soils, the more serious the crack development and the worse the slope stability [16]. Over-consolidation refers to the fact that expansive soil has been subjected to loads higher than the current stress level in the course of geological history and that the primary cause of their formation is the action of unloading from above [17,19]. Although there are few studies on the over-consolidation of expansive soils, it is reported that over-consolidation causes horizontal stresses to be greater than vertical stresses and that slope excavation in such soils has a greater unloading effect, which is detrimental to slope stability [19,21].

Swelling–shrinkage causes considerable changes in the volume and morphology of expansive soils and is the leading cause of cracking. Cracking will destroy the integrity of expansive soils, reduce soil strength, and provide an unobstructed channel for rainwater infiltration. Therefore, cracking is also one of the main factors contributing to the failure of expansive soil slopes. Over-consolidation can promote the development of cracks and exacerbate the softening nature of expansive soils. It can be considered that swelling–shrinkage is the internal factor affecting the mechanical properties of expansive soils and the stability of the slope, cracking is the key factor, and over-consolidation is the promoting factor [22]. Therefore, in practical engineering, the failure of expansive soil slopes is usually caused by a combination of factors.

In general slope engineering, slope stability can be improved by reducing the slope gradient. However, such measures are ineffective for expansive soil slopes, and they will encroach substantially on adjacent agricultural land and increase the construction project costs [23,24,25]. Additionally, if the treatment of newly excavated expansive soil slopes is delayed, erosion damage can form due to long-term rainfall, as shown in Figure 2a. The replacement method, where the expansive soil is replaced with cohesive soil, is the most direct measure to improve the stability of expansive soil slopes, but with significant costs [26]. To reduce the construction cost, Yang et al. [6,27,28] developed a treatment technique to construct embankments with weak expansive soil. Although this embankment type has shown good stability characteristics, such treatment requires the weak expansive soil to meet certain mechanical indexes. Slope support based on concrete materials has shown satisfactory results for general cohesive soil slopes but limited protection for expansive soil slopes (Figure 2b,c). From an ecological perspective, vegetation slope protection is gradually becoming a trend, yet landslides may still occur at the early stage of protection, as shown in Figure 2d. Consequently, the landslide failure mechanism of expansive soil slopes should be thoroughly understood so that targeted and suitable protective measures for expansive soil slopes can be investigated.

To date, the published literature on expansive soil has mainly focused on improving soil properties, and few articles have analyzed the failure mechanisms experienced by expansive soil slopes and slope protection measures. Expansive soil slope disasters seriously disrupt and affect engineering construction; therefore, it is of great significance for society and the economy to prevent and mitigate the occurrence of such disasters. In this paper, the landslide failure mechanism of expansive soil slopes is analyzed and reviewed according to three research methods, and the key factors causing the failure of expansive soil slopes are identified. Various protection principles for expansive soil slopes are summarized according to these influencing factors. Corresponding protection methods under different protection principles are proposed, and the advantages and disadvantages of each are compared and analyzed. Finally, future research directions of expansive soil slope failure mechanisms and protection methods are proposed.

2. Mechanical and Hydraulic Properties of Unsaturated Soils

Expansive soil is a kind of unsaturated soil with significant swelling and shrinkage. Thus, expansive soil has both the general properties of unsaturated soil and its own special properties. It is important to ascertain the soil–water characteristics of unsaturated expansive soils to understand the failure mechanism of expansive soil slopes.

2.1. Effective Stress Principle and Shear Strength Theory

In 1936, Terzaghi [29] proposed the effective stress principle based on saturated soils, revealing the relationship between the stress transmitted by soil particles and pore water pressure in saturated soils:

$${\sigma }^{\prime }=\sigma -u_w$$

(1)

where σ′ is the effective normal stress; σ is the total normal stress; and u_w is the pore water pressure.

However, unsaturated soils usually consist of a solid phase, a liquid phase, and a gas phase. The above effective stress equations for saturated soils ignore the effect of pore gases in the soil. For unsaturated soils, the effective stress equation proposed by Bishop (1959) [30] is widely used:

$${\sigma }^{\prime }=\left(\sigma -u_a\right)+\chi \left(u_a-u_w\right)$$

(2)

where u_a is the pore gas pressure; and χ is the effective stress parameter, which is related to the saturation degree of the soil. When the soil is completely saturated, χ = 1; when the soil is completely dry, χ = 0 [31]. Due to the simplicity of the parameters as well as the conceptual clarity of Bishop’s effective stress equation, it has become a representative of the effective stress equation for unsaturated soils.

The matrix suction (ψ) of unsaturated soils is defined as the difference between the pore air pressure and the pore water pressure in the soil (u_a − u_w). The suction of the shallow soil of the expansive soil slopes varies continuously with the wet–dry cycles. Changes in suction will affect the soil’s shear strength, which in turn affects the stability of the slope. Understanding the change in shear strength of unsaturated soils is essential for analyzing the stability of slopes. Bishop and Blight [32] proposed a shear strength equation for unsaturated soils based on the effective stress principle:

$${\tau }_f=c^{\prime }+\left[\left(\sigma -u_a\right)+\chi \left(u_a-u_w\right)\right]\tan {\phi }^{\prime }$$

(3)

where τ_f is the shear strength; c′ is the effective cohesion; φ′ is the effective internal friction angle; and (σ − u_a) is the net normal stress. With the development of the understanding of unsaturated soils, Fredlund et al. [33] proposed a shear strength formula based on the Mohr–Coulomb strength theory that takes into account the two stress variables (σ − u_a) and (u_a − u_w):

$${\tau }_f=c^{\prime }+\left(\sigma -u_a\right)\tan {\phi }^{\prime }+\left(u_a-u_w\right)\tan {\phi }^b$$

(4)

where φ^b is the internal friction angle associated with matrix suction. The value of φ^b can be determined from the slope of the shear strength curves corresponding to different matrix suctions (u_a − u_w), while tan(φ^b) can represent the rate of matrix suction contribution to the shear strength [34].

The shear strength theory mentioned above shows that the shear strength of unsaturated soils consists of three main components: the effective cohesive c′, the shear strength caused by the net normal stress (σ − u_a), and the shear strength caused by the matrix suction (u_a − u_w). The first two components can reflect the contribution of the soil skeleton to the shear strength.

The suction is closely related to the strength of unsaturated expansive soil. From the effective stress principle and shear strength theory, when rainwater infiltrates the expansive soil slope along the cracks, the pore water pressure of the expansive soil rises, causing a decrease in the effective stress, which leads to a decrease in the shear strength. Simultaneously, rainwater infiltration will also lead to the reduction or even loss of matrix suction, and the soil will swell and soften, which is extremely detrimental to the stability of expansive soil slopes.

2.2. Soil–Water Characteristic Curve and Permeability Function

The matrix suction (ψ) of unsaturated soils varies with the water content of the soil, and the soil–water characteristic curve (SWCC) can reflect the relationship between matrix suction and volumetric water content (θ_w) of unsaturated soils. The SWCC is an indispensable and essential parameter for analyzing the movement of water in soils. Currently, direct measurement of the SWCC faces the problems of being time-consuming and high-cost, and having a limited determination range. Its measurement accuracy is also disturbed due to the influence of test conditions and the operation process [35].

The SWCC is crucial for the numerical simulation calculation of expansive soil slopes and is an indispensable prerequisite. The van Genuchten model is one of the most commonly used SWCC empirical models for expansive soils [36]:

$${\theta }_w={\theta }_r+\frac{{\theta }_s-{\theta }_r}{{\left[1+{\left(\frac{\psi }{a}\right)}^n\right]}^m}$$

(5)

where θ_r is the residual volumetric water content; θ_s is the saturated volumetric water content; and a, n, and m are curve-fitting parameters. Parameter n is usually related to parameter m, n = 1/(1 − m). The corresponding permeability function is [36]

$$K_w=K_s\frac{{\left[1-\left(a{\psi }^{\left(n-1\right)}\right)\left(1+{\left(a{\psi }^n\right)}^{-m}\right)\right]}^2}{{\left[{\left(1+a\psi \right)}^n\right]}^{\frac{m}{2}}}$$

(6)

where K_w is the hydraulic conductivity; and K_s is the saturated hydraulic conductivity.

The SWCC and the permeability function show that the matrix suction is crucial for the stabilization of unsaturated expansive soil slopes, and the matrix suction decreases with the increase of volumetric water content. The decrease in matrix suction will lead to an increase in the hydraulic conductivity of the soil. Additionally, the permeability of expansive soils is generally low, highlighting the contribution of cracks to reducing soil suction and strength. It is difficult for rainwater to infiltrate into the soil in non-cracked areas, and there is a significant hysteresis in the reduction of suction and strength [37]. The soil in the crack areas is susceptible to rainwater infiltration, which will cause a sudden drop in the suction of the soil at the cracks. The decrease in suction makes the expansive soil more susceptible to expansion and deformation after wetting. In contrast, when the soil water content decreases due to evaporation, the soil suction gradually recovers, and the soil tends to shrink gradually. The increase and decrease of suction have an important contribution to the shrinkage and expansion of expansive soil. Within a certain suction range, a 1% change in the water content of expansive soils can cause swelling and shrinkage deformations [38].

In addition, under rainfall conditions, the lower permeability of the non-cracked expansive soils leads to a significant hysteresis in the change of parameters such as water content and pore pressure in the soil layer inside the slope; the deeper the soil layer, the higher the degree of hysteresis. The swelling of the soil also has a certain hysteresis, and the hysteresis time grows with the increase of depth [16].

Due to the special properties of expansive soils, the SWCC of expansive soils in the wetting and drying processes show apparent differences, and there is an obvious hydraulic hysteresis characteristic of unsaturated expansive soils in the original state in the engineering area [38]. Generally, the drying curve is located above the wetting curve, and the SWCC of the wetting process has more obvious hysteresis than that of the drying process [39]. In addition to hydraulic hysteresis, volume change also affects the hysteresis of expansive soils on the drying and wetting paths [40,41]. An increase in vertical stress attenuates the hysteresis phenomenon due to the inhibition of the volume change (shrink-swell) in expansive soils by the applied stress [40]. This laterally reflects that changes in pore volume in expansive soils can complicate the hysteretic behavior of soils [40,42]. Moreover, expansive soils containing reactive clay minerals exhibit higher hysteresis than other soils [43].

The increase in wet–dry cycles also affects the hysteresis of the SWCC of expansive soils [42,44]. The wet–dry cycles have a significant effect on the SWCC of the drying process, while there is no significant effect on the SWCC of the wetting process [44]; with the increase of the number of wet–dry cycles, their effect on the hysteresis of the SWCC gradually decreases, while the hysteresis loop also decreases [44,45,46]. After three wet–dry cycles, their effect on hysteresis gradually weakens [44].

The above analysis denotes the significance of the SWCC in the stability analysis of expansive soil slopes. This indicates that when conducting stability analysis of expansive soil slopes, different SWCCs should be used for expansive soil slopes in different states to ensure the accuracy of numerical simulation results. For example, the SWCC of the wetting process can be used in rainfall infiltration, and the SWCC of the drying process can be used for evaporation on sunny days. Simultaneously, it is also worthwhile to distinguish the SWCC of expansive soils under different numbers of wet–dry cycles in numerical simulation. In conclusion, the reasonable use of simulation parameters can accurately assess the stability of expansive soil slopes and provide data references for monitoring and early warning in slope engineering.

3. Failure Mechanisms of Expansive Soil Slopes

The primary forms of expansive soil slope damage include landslide, slump, and skidding slump, with landslide being the most common [3]. Landslides on expansive soil slopes have the characteristics of shallowness, traction, recurrence, group distribution, and rainy season frequency and are the most serious maladies experienced by highways in expansive soil areas [5]. The main research methods on the stability of expansive soil slopes presently include field tests, model tests (centrifugal and physical model tests), and numerical simulations; the advantages and disadvantages of these three methods are listed in

Table 1. Research methods regarding the failure mechanism of expansive soil slopes.

Research Methods	Advantages	Disadvantages
Field test	1. High accuracy and reliability of results. 2. In-situ (undisturbed) test soil. 3. Test environment close to the actual project site conditions.	1. Long cycle [47]; non-repeatable. 2. High cost [47]. 3. Difficulty in controlling the test conditions, with many constraints.
Centrifugal model test	1. The stress and strain of the model and the prototype are equal, with similar deformation [48,49]. 2. The use of an artificial gravity field can compensate for the loss of stress in the test and reproduce the characteristics of the prototype in a short time [50,51]. 3. The small size of the model saves labor and material resources [52]. 4. Test conditions can be controlled [50].	1. High equipment cost. 2. High requirements for material selection. 3. Particle size effect and boundary effect have an impact.
Physical model test	1. Economical and intuitive. 2. Relatively low equipment requirements. 3. Test conditions can be controlled; not subject to the limitations and influence of environmental conditions [53].	1. The stress level of the model is much lower than that of the prototype, making it difficult to reproduce the characteristics of the prototype. 2. The test soil is remodeled soil [2]. 3. Boundary effects have an impact.
Numerical simulation	1. Low cost and short time. 2. Easy to operate and repeatable. 3. Can be used in combination with field tests and model tests [53]. 4. It is easy to analyze the stability of expansive soil slopes under different operating conditions by changing the variables [8]. 5. Can be used with access to various parameters of the slope [54].	1. Need to make more simplifications and assumptions, resulting in a certain degree of error [55]. 2. Incorrect parameters, conditions, and assumptions can lead to wrong conclusions.

. In this paper, the landslide failure mechanisms of expansive soil slopes will be analyzed and discussed based on these research methods.

3.1. Field Test

Numerous scholars have conducted in-depth research on the failure mechanisms experienced by expansive soil slopes through field tests. Liu et al. [56] used field monitoring to establish that the statistical mean value of the vertical crack lengths found in expansive soil slopes was consistent with that of the thickness of the individual landslide body when the slope was destabilized. Gong et al. [57] reported that in an engineering investigation, it was found that expansive soils in certain areas often had cracked surfaces inside. Furthermore, the soil at the crack surfaces had a lower natural density, higher water content, and lower strength than the soil on either side, which were detrimental to slope stabilization. Additionally, research has shown that a large number of cracks develop at the site of instability and failure in expansive soil slopes [9,58,59]. The above studies indicate that shallow failure of expansive soil slopes can be closely linked to soil cracks.

The mechanism of cracking properties influencing the failure of expansive soil slopes is complex. Rainwater infiltrates the soil through the cracks during rainfall, and the soil expands and softens as it absorbs water. This leads to an increase in the pore ratio and pore water pressure, a decrease in matrix suction, and a decrease in the effective stress and shear strength of the soil, thus promoting shallow failure of such expansive soil slopes [23,60,61]. Based on a study of an expansive soil foundation pit, Li et al. [59] reported that the main factor leading to slope failure was the continuous infiltration of rainwater along the cracks. Ng et al. [62] monitored the variation of soil parameters at different depths (h) of expansive soil slopes through field tests. They found that rainfall led to increased water content and pore water pressure, decreased matrix suction, and increased expansive deformation of the soil (Figure 3), all of which are detrimental to slope stability. Rainwater infiltration has a greater impact on the stability of expansive soil slopes in the initial stages of rainfall. Later in the rainfall period or after the rainfall has ended, the softening of the soil has a greater effect on slope stability [63].

From the above research, it is clear that the unfavorable changes in expansive soil slopes are closely related to rainfall. Thus, rainfall is considered the most direct factor leading to slope deformation, while evaporation is a precondition for slope deformation [64]. Rainfall duration has a more significant effect on soil water content and properties than rainfall intensity [60]. This implies that different rainfall types have varying degrees of weakening on slope stability, and the total amount of rainfall in a rain event also affects slope stability. Zhan et al. [23] found that low-intensity and prolonged rainfall weakened expansive soil slope stability more than high-intensity and short-duration rainfall. Most experimental research has focused on the effects of rainfall infiltration on expansive soil slopes. However, before infiltration occurs, the rainfall erodes the surface soil on the slope; as the rainfall intensity increases, the severity of the erosion increases correspondingly [60]. The deformation of expansive soil slopes increases with the number of rainfall events [48]. However, higher soil densities show a more vital ability to resist deformation, and thus, expansive soil slopes experience higher stability [49]. Therefore, slope protection is necessary to minimize soil erosion and can be affected by appropriate slope compaction during construction. Moreover, some scholars have proposed that increasing the slope gradient can accelerate the runoff rate, thereby reducing the influence of rainfall infiltration and achieving the effect of maintaining water and slope stability [65,66]. Nevertheless, the stability of expansive soil slopes with steep gradients requires further verification.

The cracking of expansive soil and rainfall are internal and external factors of slope failure, respectively. Cracks can weaken the integrity of expansive soil while accelerating rainwater infiltration and soil erosion on the slope surface. Consequently, measures can be taken to inhibit rainwater infiltration and crack development during construction to improve slope stability effectively.

3.2. Model Test

Model tests include both centrifugal and physical model tests. The centrifugal model test can reduce the model’s size, saving a lot of labor, material, and financial resources compared to field tests. The effects of cracks and water infiltration on slope deformation and stability can be comprehensively considered compared with the physical model test [67]. Rao et al. [68] verified that the centrifugal model test could be applied to the stability study of expansive soil slopes with similar results to numerical simulation analysis. Physical model tests have the advantage of controllable test conditions compared to field monitoring and have a lower cost than the centrifugal model test. However, the soils used in model tests are remodeled soils, which cannot precisely simulate undisturbed (in situ) soil properties.

As noted, the swelling–shrinkage characteristics of expansive soil have a critical effect on slope stability. The volume shrinkage of expansive soil after water loss promotes the development of cracks, which in turn influences slope stability [20]. In contrast, expansive soil absorbs water and swells in volume, which generates expansion forces, thereby increasing soil stress [67]. Expansive soil is also prone to strength failure because of the considerable displacement constraints in the downslope direction [48,69]. Destabilizing failure surfaces are formed in expansive soil slopes at the wet–dry and saturated–unsaturated interfaces; stress concentration zones form in the soil at these interfaces [53,70]. Water absorption and swelling of the soil result in a tendency for the soil cracks to close. Rainwater infiltration depth is thus limited due to the reduced downward progression of the cracks and the lower permeability of the soils [71]. Wet–dry and saturated–unsaturated interfaces are typically found within 2–4 m of the slope surface [16,23,62,72,73], destabilizing expansive soil slopes at shallow depths. In addition to the expansive soil cracking conditions, the depth of the wet–dry and saturated–unsaturated interfaces is also related to factors such as region and vegetation on the slope surface.

Wet–dry cycles also have an important effect on slope stability. Wang et al. [67] used a centrifugal model test to investigate the reaction of slope soil under the action of wet–dry cycles. They found that with an increasing number of cycles, settlement and horizontal displacement of the slope occurred, as shown in Figure 4. Zhang et al. [74] used indoor model tests and found that the formation of cracks was closely related to the number of wet–dry cycles. Their results showed that surface expansive soil under wet–dry cycles produced a large number of residual cracks and cumulative damage, and continuous rainfall infiltration softened the surface soil and accelerated slope destabilization. Studies have shown that wet–dry cycles promote the development of crack depth and width, leading to significant changes in the water content of the shallow soil layers of the slope, which is unconducive to slope stability [37,75,76,77,78,79,80,81]. Furthermore, soil in cracked areas experienced a sudden drop in matrix suction and strength decay after water absorption, with a significant variation in strength between the soil in cracked and non-cracked areas. It has been found that the range of slope sliding is closely related to the development of cracks in the soil [37]. A few studies found a relationship between crack development and slope gradient [82]. For cracked expansive soil slopes, reinforcement at the mid and toe of slope areas can effectively increase the safety factor [83,84,85]. However, when reinforcing at the slope toe, care must be taken regarding the blockage of groundwater drainage [84,85]. Studies have also shown that wet–dry cycles are vital factors that contribute to cumulative deformation and soil loss from slopes [86,87,88].

3.3. Numerical Simulation

Although field tests align better with actual site conditions, they are costly and time-consuming. In contrast, the numerical simulation method is low-cost, easy to operate, and reproducible. Numerical simulation results can be combined with field and model test results for analysis [68]. Therefore, numerical simulation analysis has become the most commonly used method to analyze the stability of expansive soil slopes. Expanded analysis using numerical simulation methods is very convenient and can be used to analyze the stability of expansive soil slopes under different operating conditions by changing the variables [8]. Numerical simulation also provides rapid access to various slope parameters, such as pore water pressure, displacement, stress, and slope safety factors [54]. Therefore, numerical simulation can quickly determine the overall state of expansive soil slopes. Because of the significant advantages of the numerical simulation method, it also has a broader application prospect in practical engineering.

Numerical simulation can mimic the effect of rainfall on slopes with greater accuracy. Changes in the property of the soil in different parts of the expansive soil slope affect slope stability to different degrees. The variations in suction at the top of expansive soil slopes tend to be greater than at the toe of the slope [8]. However, failure at the toe of the slope is the critical factor in the failure of expansive soil slopes [89,90,91]. Scholars have used numerical simulation to show that continuous rainfall softens the soil at the toe of the slope, with the result that this area is most prone to plastic failure [54]. The failure surface typically begins at the toe of the slope and progressively develops and penetrates deeper into the soil, forming a progressive traction slide on expansive soil slopes [92]. Sliding surfaces of expansive soil slopes usually form in soil layers with sufficiently developed cracks [55]. In summary, strengthening waterproofing, water insulation, and drainage measures at the toe of slopes is a crucial method to enhance the stability of expansive soil slopes [24,47,93].

In addition to rainfall, evaporation also affects the stability of expansive soil slopes. Evaporative water loss from expansive soil can lead to the deepening and widening of existing cracks and the formation of new cracks. The degree of crack development increases with the number of evaporations (wet–dry cycles) [49]. Li et al. [72] found that the slope stability worsened due to the increased crack rate of expansive soil. Larger and deeper cracks reduced the slope integrity and stability, accelerated rainwater infiltration, and deteriorated soil properties [94,95,96,97]. Expansive soil slopes are relatively stable during evaporation and susceptible to destabilization during rainfall [98]. Construction of expansive soil slope works should be avoided during the rainy season.

The slope safety factor can be used to measure the ability of a slope to resist sliding and slumping. Numerical simulation studies have shown that wet–dry cycles significantly reduce the factor of safety of expansive soil slopes, contributing to shallow and progressive landslide damage [55,98,99,100]. Ikeagwuani et al. [101] found through numerical modeling that the higher the degree of over-consolidation, the greater the effect of dilatancy on the safety factor of expansive soil slopes. Cracks tend to form in expansive soils exposed to wet–dry cycles. Increasing the number of wet–dry cycles can lead to the expansion of a substantial number of secondary cracks and damage to the slope [102,103]. Rainwater infiltration similarly affects the slope safety factor and contributes to slope failure [10,104,105,106,107]. Landslide damage can also occur on low-gradient expansive soil slopes due to wet-dry cycles, indicating that the soil’s strength can dramatically affect slope stability [100,108]. Through SEEP/W and SLOPE/W, Boluk et al. [109] proved that soil cracking and strength reduction caused by wet–dry cycles are the key factors in the sudden reduction of the safety factor of expansive soil slopes. Additionally, the effects of human activities, such as top-of-slope loading, can reduce the slope factor of safety [103,110]. Slope protection should be carried out on newly excavated expansive soil slopes in a timely manner to avoid damage to the slopes caused by wet–dry cycles.

Zhao et al. [111] conducted a comparative analysis of the implications of including or disregarding the influence of moisture expansion on the maximum displacement and the safety factor of a slope; the results are shown in Figure 5. It was found that the number of wet–dry cycles increased the maximum displacement and decreased the safety factor, with more significant changes when moistening expansion was included in the analysis. Although numerical simulation is widely used in the stability analysis of expansive soil slopes, it requires precise parameter definitions and theoretical assumptions to prevent erroneous conclusions due to inaccurate data.

Based on research, the main factors that cause destabilization and failure of expansive soil slopes include external factors such as rainfall, wet–dry cycles, and human activities, as well as internal factors such as cracking, swelling–shrinkage, and soil properties. These factors affect the slope stability while interacting with each other (

Table 2. The effect of different factors on the stability of expansive soil slopes.

Factor Properties	Influencing Factors	Effects on Slopes
External factors	Rainfall	1. Soil swelling and softening [62,63]. 2. Pore water pressure increases and effective stress decreases [23,62]. 3. Soil shear strength decreases [23]. 4. Soil self-weight increases, sliding force increases. 5. Slope erosion damage [60].
	Wet–dry cycles	1. Increased crack development destroying the integrity of the soil [74]. 2. Significant changes in soil water content [80]. 3. Generate cumulative deformation [67]. 4. Reduction of the slope safety factor [10,54].
	Human activities	1. Excavation unloading [17,19]. 2. Loading on the top of the slope, accelerating instability [103,110].
Internal factors	Cracks	1. Provision of a direct pathway for rainwater infiltration [59,74]. 2. Reduction in soil strength [57,74]. 3. Tends to the formation of weak or sliding surfaces [56,57]. 4. Acceleration of soil erosion.
	Swelling–shrinkage	1. Production of uneven deformation and acceleration of crack development. 2. Causes uneven soil stresses inside the slope; production of stress concentration [53,70]; triggering factor of strength failure.
	Soil properties	1. Increased expansion rate increases the potential for destabilization. 2. Increased hydrophilic mineral content of the soil increases the potential for induced cracking [14]. 3. Decreased soil density and increased degree of weathering increase the likelihood of erosion damage. 4. Increased degree of over-consolidation decreases the safety factor of expansive soil slopes [101].

). Under repeated wet–dry cycles, slope cracks develop, and soil deformation damage accumulates due to swelling–shrinkage of the expansive soils. This destroys the integrity of the slope while providing a channel for rainwater infiltration. Under rainfall, rainwater infiltration causes the soil to expand and soften. As a result, there is an increase in pore water pressure and a decrease in matrix suction and soil strength, which leads to increased sliding forces. Localized failure can also occur when the local shear stress of the slope exceeds the shear strength of the soil. With the continuation of rainfall, the cracked area gradually forms a sliding surface that penetrates the lower soil layers. When the downward sliding force is greater than the sliding resistance force of the soil, a shallow progressive landslide will occur on the expansive soil slope. The failure mechanisms and the destabilization processes of an expansive soil slope are shown in Figure 6 and Figure 7, respectively. It can be concluded that these changes are closely related to the water sensitivity of expansive soil. Therefore, the moisture change of expansive soil slope is the most fundamental and critical factor that triggers slope instability. Accordingly, the protection of expansive soil slopes should be carried out against the water sensitivity of expansive soil to reduce the impact of the interaction between expansive soil and water, thereby effectively preventing disasters such as landslides in expansive soil slopes. In addition to the factors mentioned above, the slope shape, groundwater level, loading, and drainage system may all impact the stability of expansive soil slopes, indicating that the failure of expansive soil slopes is an extremely complex process. Accordingly, slope stability should be evaluated in construction projects in the context of the site conditions.

4. Protective Measures for Expansive Soil Slope Failures

The critical factors affecting the stability of expansive soil slopes are soil swelling–shrinkage deformation, crack development, and strength reduction caused by rainfall infiltration and wet–dry cycles. Therefore, protective measures for expansive soil slope failures can be addressed by the following principles:

(1) Waterproofing and water-blocking: By applying a protective layer on the slope surface and taking waterproofing and drainage measures, the water content of expansive soil slopes can be stabilized.

(2) Limitation of swelling-shrinkage deformation: Physical methods are used to limit the swelling–shrinkage deformation of the slope to avoid destabilizing damage caused by excessive deformation of the slope.

(3) Crack inhibition and strength enhancement: By improving the basic properties of expansive soils, the complex changes in expansive soils after water absorption or water loss are suppressed, thus inhibiting the development of cracks on the slope surface and increasing the strength of expansive soils.

4.1. Waterproofing and Water Blocking

The purpose of waterproofing and water blocking is to prevent contact between the slope surface and external water and to inhibit the evaporation of water from the slopes, thereby stabilizing the water content of the soil and improving the stability of the slopes.

The underwater portion of expansive soil channel slopes is often protected in this way. Liu et al. [86,112,113] used soilbags filled with expansive soil to strengthen expansive soil channel slopes. The change in the water content of the soilbags and the expansive soil below was effectively reduced due to the small permeability coefficient of the soilbag combination (Figure 8). Soilbags can provide drainage and seepage prevention, reduce evaporation, and stabilize the toe of the slope through their self-weight, effectively improving the safety factor of expansive soil slopes [114,115]. A schematic diagram of soilbag stabilization is shown in Figure 9 [116]. Geosynthetic clay liners (GCLs) and composite geomembranes can effectively prevent rainwater infiltration and improve slope stability; however, GCL protection is only suitable for expansive soil channel slopes, whereas composite geomembranes are suitable for other types of expansive soil slopes [117,118,119]. GCLs consist of two layers of non-woven geotextile fabrics with an intermediate layer of expansive soil powder, which expands dramatically when exposed to water, decreasing the GCL’s permeability and thus preventing the infiltration of rainwater. Hewitt and Daniel indicated that after freeze–thaw cycling, the GCL retained its protective properties for expansive soil slopes [120].

The research was conducted on a double-layer soil system composed of soils with different permeabilities; the results showed that its drainage efficiency was more than 90%, indicating that this system would effectively ensure the stability of expansive soil slopes [121]. Capillary barrier systems (CBS) follow a similar approach and are widely used in landfill and slope protection; therefore, this system can be considered suitable for application to expansive soil slopes [122,123,124,125]. Ma et al. [71,80] used polymer waterproof coating (PWC) to protect expansive soil slopes, effectively inhibiting rainwater infiltration and moisture evaporation and reducing erosion damage. Vegetation slope protection can significantly attenuate rainwater erosion, reduce runoff from the slope surface, and inhibit slope deformation. However, this may increase rainwater infiltration due to the vegetative root systems [60,126].

Waterproofing and water blocking reduce the impact of rainfall and water evaporation on expansive soil slopes. Except for vegetation protection, the application of these measures is immediately effective in improving the stability of expansive soil slopes and requires minimal long-term maintenance. Therefore, this approach could present a future developmental direction for expansive soil slope protection.

4.2. Limitation of Swelling–Shrinkage Deformation

The limitation of swelling–shrinkage deformation refers to the use of physical reinforcement or strengthening measures, such as retaining walls, anchors, anti-slip piles, and geosynthetics, to limit the swelling deformation of expansive soils and to reduce the adverse effects of swelling–shrinkage on slopes. These measures include both flexible and rigid limitation methods. Geogrid reinforcement is an example of a flexible limitation measure, which can inhibit the displacement generated by moisture absorption of expansive soil slopes and allow a moderate amount of deformation [127,128]. Small geogrid mesh sizes are recommended, providing better slope reinforcement [129]. Some scholars have used geo-cells flexible retaining walls to effectively limit swelling–shrinkage deformation of expansive soil slopes [130]. The previously described soilbag stabilization method prevents rainwater infiltration and limits the lateral deformation of expansive soil [115]. Xu et al. [131,132] proposed an anchor-reinforced vegetation system (ARVS) consisting of vegetation, anchors, and high-performance turf reinforcement mats (HPTRMs). The results indicated that this system effectively prevented large fluctuations in soil water content, restrained the swelling–shrinkage deformation of slopes (Figure 10), and provided superior slope protection [133]. The ARVS reinforcement system is shown in Figure 11 [132]. The replacement of expansive soils with non-expansive soils can effectively inhibit expansion and improve slope stability. However, this method has disadvantages, such as high costs and potential environmental damage [134,135,136].

Various rigid limitation methods have been investigated. Hou et al. [104] proposed a reinforcement system consisting of anchor cables, anchor rods, and soil nails to improve the stability of expansive soil slopes. Cheng et al. [137] stated that pre-stressed anchor cables were an effective measure to improve slope stability. Although steel spiral anchors have a simple structure, they are prone to corrosion. In contrast, glass fiber-reinforced plastic anchors possess better corrosion resistance and can rotate into the soil, thus effectively stabilizing expansive soil slopes [138]. Anti-slip piles and retaining walls rely on their inherent stiffness to resist soil stresses, limit the deformation of expansive soils, and achieve certain slope stabilization effects [139,140,141].

Rigid limitation measures are stricter in controlling the displacement of expansive soil slopes than flexible measures. However, these elements are prone to structural failure in practical engineering due to excessive internal forces. AbdelSalam et al. [142] reported that using geofoam inclusions in combination with retaining walls can effectively reduce the possibility of structural failure. In contrast, flexible limitation measures have shown excellent reliability for long-term stability and have become the current development trend. Combining rigid and flexible limitation measures could present a promising development direction [143].

4.3. Crack Inhibition and Strength Enhancement

The principle of crack inhibition and strength enhancement refers to mixing a certain material with expansive soil to modify the adverse engineering properties of expansive soil and reduce the probability of slope failure. Research on such protective measures has progressed from the modification of a single material to mixed modification of multiple materials, and from modification under simple conditions to modification under complex conditions such as wet–dry and freeze–thaw cycles. Current research focuses on material modification, such as using industrial and agricultural waste products, and biological modification to improve economy and environmental protection. According to this principle, modified expansive soil should be reused in engineering.

4.3.1. Material Modification Method

Industrial waste materials such as lime, cement, fly ash, marble dust, and dolomite can increase the strength of expansive soil and reduce the expansion potential but require long-term maintenance to achieve the desired modification effect [50,144,145,146,147,148,149,150,151]. Mahedi et al. [152] compared the modification effect of three materials on expansive soil after 90 days of maintenance; the test results are shown in Figure 12. Lime provided the most remarkable strength enhancement, followed by cement and fly ash, yet the lime-modified soil had the highest expansion rate. Moreover, prolonged immersion of lime-modified expansive soils significantly decreased shear strength [153]. As sulfates can weaken the modification effect by chemically reacting with the cement or lime, neither cement nor lime are suitable for modifying high sulfate expansive soils [154]. Low-CaO (calcium oxide) materials such as fly ash, slag, and metakaolin can reduce the involvement of sulfate in the reaction and are suitable for modifying high-sulfate expansive soils [152,155]. Therefore, appropriate modifying materials should be selected according to the use environment to avoid causing adverse reactions.

Agricultural waste materials have also been used for expansive soil modification, with the advantage of increased environmental protection value. Bagasse fiber, bagasse ash, and rice husk can improve mechanical properties such as shear strength and compressive strength of expansive soils and reduce expansion potential and crack development [156,157,158]. In addition, biochar can inhibit the expansive deformation of such soils [159]. Numerous studies have shown that a combination of modified materials is more effective than a single material in improving the properties of expansive soil. Such combinations include wheat straw and silica fume, rice husk ash and phosphogypsum (a by-product of fertilizer), lime and perlite, and bagasse fiber and lime [160,161,162,163,164]. In addition to strengthening the soil and reducing expansion deformation, mixing calcium lignosulfonate and lignin fibers can also effectively improve the seepage resistance of expansive soils [165,166].

The effectiveness of expansive soil modification materials under adverse environmental influences, such as wet–dry cycles and freeze–thaw cycles, should also be considered. Mixtures of octadecyl trimethyl ammonium chloride and Potassium chloride, iron tailing sand and calcium carbide slag, polypropylene fiber and silica fume, and nano-silica and electric arc furnace slag still exhibit good modification effects of expansive soil after wet–dry or freeze–thaw cycles [167,168,169,170].

The material modification method reduces the influence of cracks, swelling–shrinkage, and other internal causes of slope instability of expansive soil, and effectively improves the engineering properties of expansive soil. This method has multiple optional materials that can modify expansive soil by selecting the most suitable material combinations according to local conditions. However, this process requires difficult mixing, with the risk of changing the soil pH value, polluting the soil environment, and unfavorably affecting vegetation growth.

4.3.2. Biological Modification Method

Research has been done on environmentally friendly biomodification methods to treat expansive soils, with a number of specific measures for biomodification of expansive soils emerging in recent years. Microbial modification of expansive soils has long been theoretically feasible [171]. Actinomycetes extracted from soil can proliferate and grow in large numbers in expansive soils, increasing the internal friction angle and decreasing the expansion rate and force (Figure 13) [172,173,174]. The decrease in the cohesion of expansive soils under wet–dry cycles has the greatest effect on soil strength [54,175]. However, actinomycetes have not played a role in improving the cohesion of expansive soils [174]. Microbially induced calcite precipitation (MICP) has attracted more attention as a new soil modification method, as the swelling–shrinkage, compressibility, hydrophilicity, and soil strength of expansive soils can be improved using this technology [176,177,178,179]. Nevertheless, research is still needed regarding the required nutrition of strains in engineering applications and the long-term effects of soil modification.

Bio-enzymes can increase the compressive and shear strength of expansive soils, reduce expansiveness, and stabilize the properties of expansive soils for engineering purposes [180,181,182]. Although the biological modification method of expansive soil is environmentally friendly, there are challenges to guaranteeing the survival of microbes and increasing the long-term effectiveness. At present, the biological modification method is less frequently applied in practical engineering and requires further research. Combining the material and biological modification methods could play a complementary role and provide an improved modification effect [183].

Using these three principles to prevent expansive soil slope disasters can achieve certain improvements. These improvements can reduce the influence of internal and external factors such as wet–dry cycles, rainfall, cracking, and swelling–shrinkage on slope stability.

Table 3. Preventive methods againts the failure of expansive soil slopes.

Protection Principles	Methods and Materials	Effects Analysis	Disadvantages
Waterproofing and water blocking	1. Geosynthetics reinforcement (e.g., Soilbags [86,112,113], GCL [117,120], geotextiles, geomembranes [119]). 2. Double-layer soil structure protection [121]. 3. Polymer waterproof coating protection [71,80]. 4. Vegetation protection [126].	1. Effectively prevent rainwater infiltration and maintain soil moisture stability [71,80,117,120]. 2. Reducing water evaporation, crack development, and soil erosion on slopes [71,80]. 3. Most of the methods can obtain ideal slope stabilization effects with short-time maintenance. 4. Some methods apply to the reinforcement of the underwater portion of expansive soil channel slopes [118].	1. Complicated construction process and high cost. 2. Difficult to guarantee the weather resistance and durability of geosynthetics. 3. Poor initial stress resistance of vegetation.
Limitation of swelling–shrinkage deformation	1. Flexible limitation method (e.g., Geogrids [128], geo-cells [130], soilbags [115], ARVS [131,132]). 2. Rigid limitation method (e.g., anchors [137,138], soil nails [104], retaining walls [140,141], anti-slide piles).	1. Effectively reinforcing the soil of the slope and suppressing swelling–shrinkage deformation [130]. 2. Flexible materials allow a certain deformation of the slope [115,131,132]. 3. The combination of flexible and rigid limitation methods can better improve slope stability [143].	1. Rigid structures are prone to structural damage. 2. Difficulty in ensuring the weather resistance and durability of geosynthetics. 3. Difficult to inhibit rainwater infiltration and soil erosion effectively.
Crack inhibition and strength enhancement	1. Industrial waste material modification (e.g., lime [93,144,154], cement [148,152], fly ash [149,152,155], marble dust [150], dolomite). 2. Agricultural waste material modification (e.g., bagasse fiber [163], bagasse ash [157], rice husk [161], biochar [159]). 3. Biological modification (e.g., microbial [176,179], bio-enzymes [180,181,182]). 4. Mixed modification of various materials [160,161,162,163,164].	1. Various types of modified materials are available and waste materials can be effectively utilized [149,152,155]. 2. It can improve the strength and seepage resistance, reduce the expansion potential of expansive soil, and improve the adverse properties of expansive soil [50,144,145,146,147,148,149,150,151,159]. 3. Specific materials can be mixed to treat expansive soil with better modification effects [152]. 4. Biological modification is environmentally friendly [176,179,180,181,182].	1. The mixing method is troublesome and pollutes the soil environment. 2. Most of the modifications require long-term maintenance [144,154]. 3. It is necessary to add expansive soil in a reasonable amount to achieve the best balance between improvement effect and economy. 4. It is difficult to guarantee the long-term effectiveness of the biomodification method.

summarizes the specific methods corresponding to the various principles and their characteristics.

5. Discussion and Prospects of Expansive Soil Slope Research

5.1. Study of Slope Failure Mechanism

Numerous scholars have concluded that the stability of expansive soil slopes is affected by external and internal factors, among which rainfall, wet–dry cycles, and cracking have the most significant influence. The adverse changes that occur on expansive soil slopes are closely related to changes in the water content of the soil. The variation in water content causes cracking and strength reduction of the soil, which reduces the stability of the slope. Therefore, slope protection should be undertaken timeously on newly excavated expansive soil slopes, and construction should be avoided in the rainy season. Studies on the effect of water content change on crack development are mostly limited to the surface of expansive soils. However, some scholars have successfully investigated crack development within the soil using X-rays [18,184].

Research has shown that corresponding improvements in strength measurement of expansive soils and numerical simulation methods can lead to more accurate research results. For example, the triaxial test can reflect the effect of cracking on soil strength with better accuracy than the straight shear test [185]. Compared to conducting wet–dry cycles on soil samples, taking samples of expansive soil slopes after wet–dry cycles and measuring the soil strength can eliminate the disturbance and size effects, and provide a closer representation of the in situ conditions [186,187,188].

A prerequisite for ensuring accurate numerical simulation results is reasonable and correct computational assumptions. Considering multiple factors such as hydro-thermal, hydro-mechanical, and saturation-expansion coupling can provide a more reasonable, realistic, and accurate analysis of the stability of expansive soil slopes than considering only a single factor [72,89,189,190]. The slope model with differentiated material settings represents site conditions more accurately than the model with a single material setting. More cracks in the shallow layers of expansive soil slopes result in different soil strengths in the shallow and deep layers. Therefore, using the strength reduction method in numerical simulation will improve the accuracy of the results [55,90,191]. One such approach was proposed by Liu et al. [55], where the slope was divided into a layer with fully developed cracks, a transition layer, and a layer without cracks (Figure 14), with different soil parameters set for each layer. Additionally, consideration of matrix suction, expansion, soil softening, and creep will significantly impact the study results [10,52,92,106,108,189,192,193]. Accordingly, the influence of multiple factors in the field should be considered as far as possible in the numerical modeling process to obtain a more realistic failure mechanism for expansive soil slopes.

5.2. Research Direction for Slope Failure Mechanism

At present, some factors affecting the stability of expansive soil slopes still lack detailed studies. Dynamic loads, such as vehicle loads, can cause plastic deformation of expansive soil roadbed slopes, but the related slope failure mechanism has rarely been studied. Some scholars have investigated the change rule of soil temperature of slope but have not reported the effect of soil temperature on slope failure [50,87,126]. Freeze–thaw cycles will cause the skeleton of expansive soil to be damaged and cracks to be further developed, which will affect the stability of slopes. Nevertheless, there are fewer related experiments, and further research is needed to provide technical references for the construction of expansive soil slopes in alpine regions [51,82,194]. Moreover, the unloading effect and expansive soil over-consolidation will affect the stability of expansive soil slopes, but its action mechanism on slope failure is unclear and requires further study [21]. Additionally, expansive soil channel slopes are affected by the cyclic rising and falling of river water in addition to rainfall; it is of great significance to conduct an in-depth study on the failure mechanism under the dual action of rainfall and river water.

5.3. Protection Measures for Expansive Soil Slopes

At present, the best and most effective way to treat expansive soil slopes in engineering is to completely excavate them or replace them with non-expansive soils, thus solving the problem at the root. However, these methods are costly and time-consuming. According to the failure mechanism of expansive soil slope, the protective measures against expansive soil slope failure can follow three principles: waterproofing and water blocking, limitation of swelling–shrinkage deformation, and crack inhibition and strength enhancement. Modifying and reusing expansive soil can effectively increase soil strength and reduce cracking. However, the addition of modified materials should be appropriate. The large-scale utilization of modified expansive soil in a project will increase the construction period and workload. Furthermore, the effect of such modifications on the environment requires further study. Using soilbags, geosynthetics, and vegetation for slope protection can better intercept rainwater and reduce scouring. It is currently considered effective and has been applied in large quantities to protect expansive soil slopes. Nevertheless, geogrids have a diversion effect on rainwater. Rainwater can infiltrate along the geogrid into the interior of expansive soil slopes, which is not conducive to the water stabilization of slopes. The process of geogrid reinforcement of slopes is complicated, and the lap joint of the geogrid is relatively weak and low-strength. Geomembrane and geocell have weak bonds with expansive soil layers and are prone to relative slip, and geocell is unsuitable for the protection of steep slopes. In addition, geomembrane is prone to perforation and joint damage, which cannot prevent seepage after damage. Soilbags have poor weather resistance and are prone to aging and breakage [80]. The durability of geosynthetics is unsupported by adequate field data, and the construction process is cumbersome. Vegetation protection is less stable in the initial stage and requires long-term maintenance and subsequent management. Vegetation roots increase soil permeability, and dead vegetation’s decaying root system may affect slope stability [66]. The conventional method of rigidly limiting displacement suffers from the disadvantages of a poor ability to coordinate deformation as well as an inability to withstand inhomogeneous deformation. Therefore, conventional rigid displacement-limiting structures are prone to stress concentrations and damage. It is recommended that a combination of rigid and flexible limitation methods be used to ensure structural safety and optimal slope stability [195].

With the current expansive soil slope protection methods, it is difficult to simultaneously achieve the desired effective anti-infiltration, anti-evaporation, and anti-scouring results. An analysis of the failure mechanism of expansive soil slopes indicates that the change in slope moisture content is the most critical factor causing the instability. Therefore, model tests were conducted on expansive soil slopes protected with PWC. Five wet-dry cycles were carried out, and the results showed that PWC could effectively maintain the stability of slope moisture and reduce deformation and erosion (Figure 15) with remarkable effects [80,196,197]. Although protection of expansive soil slopes based on all three principles can achieve some effectiveness, cross and combined use should be studied in a focused manner.

5.4. Development Trends of Protection Methods for Expansive Soil Slopes

The research solutions obtained from laboratory studies should be extended and validated in field conditions, and only then can they be reliably used in the practical engineering of expansive soil slopes [198]. In complex scenarios, such as long-term wet–dry and freeze–thaw cycles, the durability of protective structures or materials requires further exploration. Based on the foundation of the previous work and the three protection approaches discussed proposed in this paper, a new flexible ecological slope protection system with a double waterproof layer is proposed for expansive soil slopes (Figure 16) to provide ecological, efficient, and long-term protection. The root system of the vegetation layer can grow into the backfill soil layer, forming a reinforcing effect with the protective layer of a three-dimensional vegetation network–polymer waterproof coating (TDVN–PWC). A small amount of rainwater will infiltrate the backfill soil layer and be absorbed and utilized by the vegetation root system. The second PWC protective layer will isolate the rainwater infiltration completely and inhibit evaporation from the slope. This will maintain the slope water stability and inhibit swelling–shrinkage deformation and crack development. Depending on site conditions and requirements, the PWC layer can be replaced with a concrete lining or a concrete lining-PWC layer. Furthermore, this novel protection system can be used in combination with geosynthetics and anchors to provide the highest level of safety for a project. If the long-term stability of expansive soil slopes can be guaranteed, increasing the initial installation cost during the construction of the protective layer will be worthwhile. To this end, anti-seepage and moisturizing are expected to be key research directions for the future development of expansive soil slope protection.

Field monitoring and early warning systems are also crucial for managing expansive soil slopes. Slope monitoring and early warning systems now incorporate a wide range of new technologies. Traditional methods, such as rainfall thresholds, cannot accurately predict landslides due to a lack of information on the necessary conditions. Yang et al. [35] proposed Bayesian back analysis, which has the potential to enhance landslide warning tools to predict future responses to rainfall. Fang et al. [199] successfully predicted the time of slope failure via the inverse velocity method. Huang et al. [200] pointed out that the GNSS monitoring technique with multi-source parameters avoids the unreliability of relying on a single monitoring information to assess the failure of expansive soil slopes. It is difficult to accurately obtain the development of fissures by traditional technical means. Ye et al. [201] proposed an early warning technique for landslides in expansive soils using a 3D laser scanner based on fissure properties. With the development of science and technology, intelligence gradually penetrates all fields, geotechnical engineering being no exception. For expansive soil slope disasters, it is possible to transition from traditional monitoring methods to novel intelligent monitoring methods [202,203,204]. Puppala et al. [198] successfully incorporate visualization tools and unmanned aerial vehicle platforms for studying and monitoring the health of civil infrastructure built on expansive soils. By combining the protective structure of an expansive soil slope with an intelligent monitoring system, it is possible to understand the dynamic information of the slope in time and provide early warning of potential slope disasters [26]. Intelligent monitoring can reduce the workload of on-site monitoring and maximize the life safety of on-site personnel, which is a very meaningful research direction.

6. Conclusions

A comprehensive review of the work related to the failure mechanism and protection methods of expansive soil slopes is presented in this paper, current research is discussed and concluded, and some directions for future research are proposed. The main conclusions and recommendations are as follows:

(1)

The failure mechanism of expansive soil slopes is complex. Repeated wet–dry cycles cause expansion and contraction deformation of the expansive soil, leading to cracking and strength reduction of the soil. Then, the rainwater infiltrates along the cracks under rainfall conditions, resulting in expansion and softening of the soil, increase of pore water pressure, and further reduction of soil strength. Simultaneously, rainwater infiltration generates seepage force, decreasing the slope safety coefficient. The expansive soil slope will experience a step-by-step retrogressive tractive shallow landslide when the downward sliding force exceeds the slip resistance force. The failure of expansive soil slopes is closely related to changes in soil moisture.

(2)

Obtaining the real strength of expansive soil is crucial to ensure the accuracy of slope stability studies. Meanwhile, numerical simulation should consider the influence of multiple factors and their coupling effects and analyze them in combination with model and field tests to obtain a more realistic slope failure mechanism. Notably, factors such as cracks, expansion forces, matrix suction, and soil softening should not be ignored in numerical simulations.

(3)

Three slope protection principles are proposed, including waterproofing and water blocking, swelling–shrinkage deformation limitation, and crack inhibition and strength enhancement. Nevertheless, there are still some shortcomings in slope protection based on these three principles, such as the impact on the ecological environment, the construction’s complexity, and the material’s durability and resistance. Additionally, the existing protection methods cannot simultaneously achieve effective anti-infiltration, anti-evaporation, and anti-scouring effects.

(4)

At present, few studies are related to protecting expansive soil slopes under complex scenarios such as long-term wet–dry and freeze–thaw cycles. The effectiveness of protective structures or materials in complex environments deserves further study. The failure mechanism of expansive soil slopes under complex environments will also be the focus of future research.

(5)

A novel flexible ecological protection system based on polymer waterproof coating with a double waterproofing layer can be considered to protect expansive soil slopes, which can reduce infiltration and evaporation and inhibit the cracking of slopes while being environmentally friendly. Low-carbon environmental protection, being environmentally friendly, and anti-seepage and moisturizing are vital points to be considered in future slope protection work and development trends. Finally, the combination of intelligent monitoring, early warning systems, and expansion soil slope protection is a critical research direction.