Healing of Undisturbed Slide Zone Soil: Experimental Study on the Huaipa Landslide in Sanmenxia City

Abstract

The occurrence and development of a landslide is a gradual process of destruction, causing huge losses to people’s lives and property. The shear strength of the shear zone gradually decreases to the residual state during the sliding process but it can recover to a certain extent during a relatively stable period. In a landslide, more than one slide usually occurs; however, after the first slide stops, it can enter a dormant period. The sliding surface can then experience a self-healing strength recovery phenomenon; this self-healing phenomenon has a significant impact on the reactivation of the landslide. Relevant studies have shown that the strength of the sliding surface is slightly greater than the residual strength when a landslide is reactivated; however, the explanations provided by these studies have not been sufficiently systematic. In this study, focusing on the undisturbed slide zone soil of the Huaipa landslide in Sanmenxia City, the “shear–pause–shear–pause–shear” test scheme is adopted. The soil is subjected to 3 h, 6 h, 12 h, 24 h, and 72 h healing shear tests, and combined with the SEM microstructure characteristics of the shear surface, to explore the internal mechanism of self-healing. The results show that the landslide soil exhibits strong self-healing strength recovery characteristics; however, these strength recovery characteristics decrease rapidly after experiencing a very small displacement. The strength recovery was strongly correlated with the vertical stress and healing time. With increasing vertical stress, the strength recovery value of the soil increases. Under low pressure, the strength recovery is small, and under high pressure, the strength recovery is obvious. With increasing healing time, the strength recovery increases; however, the increase in the amplitude diminishes and ultimately approaches zero with increasing healing time. A “healing phenomenon” occurs in the shear surface of slide zone soil after a short period of time. A shear strength value greater than the residual strength can be used to check the landslide design, which can effectively reduce costs.

1. Introduction

Currently, the most accepted method for obtaining the residual strength are reversal direct shear and ring shear apparatuses. Reversal direct shear apparatus is easy to operate, has a low cost, and is favorable for undisturbed sample testing. This process has become the most commonly used method of residual strength testing; however, in the shear process, the sample shear surface area continues to change, resulting in uneven stress on the sample, and the sample is limited by the shear displacement. The incomplete reorientation of the particles leads to a deviation in the test results; this is the biggest disadvantage of this type of instrument [1]. Consequently, in rigorous academic research, it is not recommended to use reversal direct shear apparatus.

Compared with reversal direct shear apparatuses, ring shear apparatus has been developed with different properties and characteristics by scholars worldwide since the 1920s and has been successfully applied to geotechnical research. A comparison of the characteristics of reversal direct shear and ring shear instruments is provided in

Table 1. Comparison of characteristics of reversal direct shear and ring shear instrument.

Characteristic	Reversal Direct Shear	Ring Shear
Shear method
Mass of sample	Less	More
Sample state	Undisturbed sample	Remolded sample
Shear displacement	Intermittent	Progression
Sliding surface	Continue to change	Remain constant
Soil extrusion during shearing	Less	More
Orientation of particles at the sliding surface	Straight	Rotary

Note: The instrument applicable sample state refers to the situation defined in favor of sample testing. Both reversal direct shear and ring shear instruments can be applied to undisturbed and remolded samples.

.

The most classical ring shear apparatuses are the Bishop ring shear apparatus and the Bromhead ring shear apparatus. The former is divided into upper and lower shear boxes, and the shear surface is located at the contact between these boxes. During the shear process, the upper shear box does not move while the lower shear box rotates. Meanwhile, the shear boxes of the latter apparatus are not divided. During the shear process, the torque is applied to the load plate at the top of the sample via air pressure or voltage, such that the sample produces torsional shear, and the shear surface is located at the upper part of the sample. Different ring shear apparatuses have their own advantages and disadvantages; however, it is indisputable that ring shear apparatuses have greater advantages with respect to simulating the shear characteristics of landslides under large-displacement conditions. The particles at the shear surface are parallel to the shear direction to the greatest extent after shearing; during the shearing process, the shear surface area remains constant, such that the sample is subjected to a uniform force, and the residual strength is more accurate [2]. In this study, the German WILLE ARS-3 ring shear apparatus is used to simulate the residual strength characteristics of the slide zone soil under large displacement shear conditions.

When a landslide occurs, subsequent revival movement will shear along the existing sliding surface which is controlled by the residual strength. In the design of slope preventive measures, the residual strength is generally considered. If the landslide is revived and recovers to a state of residual strength within a short time, this strength could be used as a remedy for landslide design and remediation. The recovered strength is greater than the residual shear strength, which increases the resisting strength. As a result, an increased safety factor reduces the cost of remedial measures. It is extremely important to study the recovery strength under the state of residual shearing. Therefore, it is necessary to study the mechanism of strength recovery and its influencing factors before using it in landslide design and remediation. Several studies have been conducted on the recovery characteristics of the shear strength.

D’Appolonia [3] showed that the shear surface of colluvial clay soil experiences “healing”, resulting in the shear strength of the pre-existing failure surface becoming greater than the residual value. Chandler [4] found that, before a landslide reactivates, the shear strength of the slide zone recovers during the stable period. Angeli et al. [5,6] studied the strength recovery mechanism of different normally consolidated clays and found that the recovery strength increased with time whether using the direct shear test or the ring shear test. Stark and Hussain [7] conducted laboratory ring shear tests of recovery strength on four natural soils, which demonstrated significant strength recovery at low effective vertical stress of 100 kPa or less. In highly plastic soil, with effective vertical stress of 100 kPa or less and liquid limits between 80% and 112%, the recovered strength is closer to the fully softened shear strengths after a long recovery period. This is due to the small difference between the fully softened shear strength and the residual strength of low-plasticity soil. Eigenbord [8] studied the self-healing in fine-grained soil during infiltration. Miao et al. [9,10] performed experimental analyses of the strength recovery of slide zone soil of landslide along a reservoir bank and showed that there was a significant strength recovery of the peak strength after a short resting time; however, with changing shear displacement, attenuation occurred at a short displacement. It was thought that this might be the result of the soil particle interactions having an influence on the strength recovery of the slide zone soil. Yan et al. [11,12] studied the self-healing phenomenon of a red-bed landslide. The relationship between the recovery strength of the shear surface and the healing time was analyzed by a reversal shear test, and the law of the recovery strength under different vertical stress was then analyzed. It was found that the healing time has a significant effect on the recovery strength compared with the vertical stress. The influence of healing time on recovery strength is more significant than that of vertical stress, and the recovery strength increases with the increase in time. Jiang et al. [13] studied the self-healing of a landslide using laboratory ring shear tests with different shear rates and grain sizes. They found that the peak strength of the landslide that had undergone accelerated creep damage was greater than the previous peak strength. Wang et al. [14] investigated the strength healing law with different vertical stresses and recovery times via direct shear tests on loess with specialized properties. Zheng et al. [15] believed that, after a short rest of one day, the reactivated residual strength increased with the increase in OCR under a given normal effective stress, but it was lost after a small shear displacement. The overconsolidation effect on reactivated residual strength at a lower stress was more prominent than that at a higher stress. Tan et al. [16] explored the strength recovery mechanism of remolded lime soil and found that the lime improved the recovery strength, and the structure of the shear surface was analyzed via microscopy. Miao and Wang [17] believed that, after long period of reconsolidation, the shear zone volume changes as a result of the influence of secondary compression, resulting in significant strength recovery.

Studies have also shown that plasticity has an important effect on strength healing. Gibo et al. [18] concluded that samples composed of silt have characteristics of strength recovery. Chen et al. [19] also found that the strength recovery of low-plasticity slide zone soil dominated by silt and sand is more obvious under lower effective vertical stress. Ramiah et al. [20] investigated the strength recovery characteristics of remolding normal consolidated kaolinite and bentonite with a healing time of 4 days and showed that the strength recovery of high-plasticity clays (bentonite) is more pronounced even with shorter healing time. Stark et al. [21] performed a 230-day strength recovery ring shear test on two soils with different plasticity. The results showed that the strength recovery increases with increasing plasticity and that this strength recovery characteristic decreases rapidly after experiencing a very small displacement. Carrubba and Del Fabbro [22] used a healing time of 30 days and found that Montona flysch had greater recovery strength than Rosazzo flysch because of the strong plasticity of Montona flysch, which is similar to the findings of Stark et al. [21]. Bhat et al. [23,24] conducted strength recovery tests for three soils with different plasticity for healing times of 1, 3, 7, 15, and 30 days. The results showed that the high-plasticity soil had higher recovery strength than the low-plasticity soil and that the friction angle of the soil did not increase within a healing time of 3 days, while the friction angle of the soil only increased by 1° after a healing time of 30 days.

The formation of the Huaipa landslide was accompanied by soil creep characteristics [25]. The creep characteristics include the destruction and recovery between particles. These two effects determine that the long-term strength of soil should be considered in the actual engineering design, and the long-term strength reflects the self-healing phenomenon of the shear surface, which is reflected in the value of the increase in the shear strength relative to the residual strength. In this study, undisturbed slide zone soil of the Huaipa landslide in Sanmenxai City is examined, and the ring shear apparatus is used to control the healing time, vertical pressure, and other factors. The generating factors of the healing effect of the landslide sliding surface are then analyzed, and the internal mechanism of the strength recovery is further explored using scanning electron microscopy (SEM) characteristics.

2. Project Profile

The Huaipa landslide is located on the mountain to the southeast of the Huaipa Yellow River water-raising project in Chencun Township, Mianchi County, Sanmenxia City, Henan Province, west of Xipo gully, south of the Coal Kiln gully fault and the steep wall of the mountain front to the south, and north of the loess terraces on the banks of the Yellow River. The Huaipa landslide has a length of 550 m from north to south and a width of 450 m from east to west, with a total volume of approximately 400 × 10⁴ m³. The elevation of the trailing edge is 437 m, and the lowest elevation of the leading edge is 295 m (Figure 1). The landslide includes seven small coal and bauxite kilns and shafts, the lower part of which is mostly excavated by mining and is relatively gentle. The large pits formed by mining have resulted in a gentle topography at the leading edge and a slightly steeper topography in the center and rear, with a natural slope angle of approximately 30°. The upper colluvium deposit strata consist of haphazard gravelly soil strata with uneven gravel content, locally high gravel contents, and incorporated rolled stones. The main slip direction of the landslide is approximately 324°, and the gully (Xipo gully) on the west side is “V”-shaped in cross-section, with deeper cuts, and the south side of the Coal Kiln gully fault on the upstream side of the gully has been completely blocked by mine slag. On the east side, a shallow gully has developed, with a bottom elevation of 301–350 m.

Cracks of the Huaipa landslide were mostly 18–106°, with width ranging from 0.5 m to 5.0 m. Most were transverse tensile cracks perpendicular to the main sliding direction.

At the leading edge of the Huaipa landslide, because of the antisliding effect of the rock strata, the shearing outlet is exposed (Figure 2a), the ground is bulging (Figure 2b), and the top strata of rock and soil is obviously disintegrated, with densely distributed cracks appearing in clusters. In the central part of the Huaipa landslide, there are mainly shearing cracks and first-shear-then-tension cracks in the northeast area, with many soil strips and soil wedges, and the relative displacement on both sides of the cracks is obvious (Figure 2c,d). In the eastern area, there are many large longitudinal cracks, which are first-shear-then-tension and parallel to the direction of sliding, with many soil strips, which are wider at the top and narrower at the bottom (Figure 2e). The trailing edge of the landslide contains mainly tensile cracks, and drunk men woods and bent trees are distributed in this area (Figure 2f,g). The surface cracks are crisscrossed, with tensile cracking forming a soil lattice and soil wedge; the cracks are wide at the top and narrow at the bottom (Figure 2h) and are distributed in a branching shape.

The slide zone soil of the Huaipa landslide is fully weathered mudstone and sandstone strata located in the soil–rock contact zone, which is 0.2–0.5 m thick. The Huaipa landslide slide surface is shown in Figure 3. There are many cypress trees growing on the slope, and the stratigraphic lithology is mostly residual slope sediments and fully weathered bedrock, with the inclination angle of the slide zone (bed) ranging from 17.8° to 52°, mostly around 35°; the thickness of the slide zone is from 13 m to 29 m. The sliding mass is a Quaternary slope colluvial and alluvial gravel soil with bedrock of interbedded anticlinal sandstone and mudstone. A profile of the Huaipa landslide is shown in Figure 4.

3. Basic Physical Properties of Slide Zone Soil

The sample was selected from undisturbed slide zone soil of the Huaipa landslide, and physical properties were tested, with a liquid limit of 34.4%, a plasticity limit of 19.8%, and a plasticity index of 14.6. The physical indices of sample are shown in

Table 2. Basic physical property indices of sample.

Natural Water Content (%)	Natural Density (kg/m3)	Liquid Limit (%)	Plastic Limit (%)	Plasticity Index	Saturated Water Content (%)
18.20	2.12	34.40	19.80	14.60	24.10

. The particle size distribution curve of the undisturbed sample was obtained via a laser particle sizer, as shown in Figure 5. The slide zone soil sample was found to be low liquid limit silty clay.

The rock composition of sample are shown in

Table 3. Rock composition of sample.

Type and Content of Minerals/%					Clay Mineral/%
Quartz	Potassium Feldspar	Plagioclase	Calcite	Dolomite
57.3	1.6	6.5	10.1	0.9	23.6

. The mineralogical composition of sample are shown in

Table 4. Mineralogical composition of sample.

Relative Clay Mineral Contents/%				Interstratified Ratio/%
I/S	Illite	Kaolinite	Chlorite	SI/S	SC/S
28	43	11	18	50	/

. The rock composition of the Huaipa sample is dominated by quartz, with a content of approximately 57.3%, And clay minerals, with a content of approximately 23.6%, are dominated by illite, with a content of approximately 43%. Clay minerals have strong water absorbing and swelling characteristics as well as water softening characteristics and are susceptible to weathering and disintegration, which affects the engineering characteristics of the rock–soil mass. The Huaipa landslide is affected by heavy rainfall, and clay minerals are easily gathered near the slide zone, which affects the engineering mechanical characteristics of the soil and is extremely detrimental to the stability of the landslide.

4. Test Scheme

To analyze and discuss the recovery strength, the ring shear tests of the slide zone soil under the initial state and after a healing time were compared. The test was performed in accordance with a “shear–pause–shear–pause–shear” cycle. First, the slide zone soil was consolidated and sheared to the residual strength state after consolidation was completed. Then, under the condition in which the vertical stress and the shear stress remained unchanged, a pause state was maintained. After reaching the specified healing time, the sample was again sheared to the residual strength. Then, the process was repeated. The shear strength and residual strength of the first shearing and the shearing following different healing times were recorded during the test.

Using an undisturbed sample, a ring shear test was conducted to simulate the possible strength recovery of the Huaipa landslide under large-displacement shear conditions. Because of the long duration of the test, effective measures were taken to prevent water evaporation. The specific test plan was as follows.

(1)

After the consolidation of the slide zone soil was stable, shearing was applied at a rate of 0.2 mm/min. When the sample reached the residual strength for the first time, the vertical stress and shear stress were maintained, such that the undisturbed stress conditions during the reactivation of the Huaipa landslide could be reproduced. The sample was paused for 3 h and then was sheared to a new stable residual strength state. Then, the sample was paused for 6 h and sheared to the next residual state. Subsequently, it was paused for 12 h, sheared to the next residual state, paused for 24 h, sheared to the next residual state, paused for 72 h, and sheared to the final residual state.

(2)

The above shear cycle state was completed at three different vertical stresses: 50 kPa, 100 kPa, and 200 kPa.

The primary shearing is performed to obtain the residual shear state of the sample. The residual state is confirmed when the shear value is minimum and the data observed by the computer are basically stable. Then, the strength recovery test is carried out. When the shearing reaches the residual state, the shearing stopped, the sample is kept static in the ring shear instrument, and the vertical stress is constantly applied during the process. Because the sliding mass of the field is still affected by the shear stress after movement, the shear strength remains constant throughout the pausing time during the strength recovery test to simulate the field conditions. During the period, the change in shear strength is observed by the computer to prevent any reduction of shear strength. Whether it is consolidation process, shear process, or pausing process, data changes can be observed through the computer so as to visually observe the strength recovery.

The strength recovery value and recovery rate are defined in this study. The strength recovery value is the difference between the peak shear strength obtained by reshearing and the residual strength of the previous shearing. The recovery rate is the ratio of the strength recovery value to the healing time.

5. Characteristics of the Slide Zone

After the “shear–pause–shear–pause–shear” ring shear cycle test was conducted, the shear box was extracted from the ring shear instrument as a whole, the screw was removed, the shear box was opened, and the upper and lower shear rings containing the sample were extracted. After repeated shearing, the upper and lower parts of the sample were completely separated, and the characteristics of the shear zone were clear and complete, as shown in Figure 6. The shearing gap divided the sample into two, as shown in Figure 6a; the shear zone is obvious, and the cracks of the upper shear box sample, which develop gradually from the shear surface to the upper surface, are fine and dense. Figure 6b shows that the lower sample has an overall smooth shear surface, and the scratches after large-displacement shear can be clearly seen; the shear direction is obvious, and there is local unevenness.

6. Influencing Factors of the Strength Recovery

6.1. Relationship between the Vertical Stress and the Strength Recovery

The undisturbed soil samples were sheared after the completion of consolidation at 50 kPa, 100 kPa, and 200 kPa under natural water content, and then, after sequential pauses of 3 h, 6 h, 12 h, 24 h, and 72 h, they were sheared again. The stress–displacement curves were then obtained with different vertical stress, as shown in Figure 7, Figure 8 and Figure 9. The slide zone soil exhibits obvious signs of strength recovery; there is a clear peak at very small displacement when the sample starts shearing again, and when the displacement increases further, the strength decreases rapidly to the residual strength. The initial peak shear strength and residual strength under the three pressures are 65.20 kPa and 60.60 kPa, 74.13 kPa and 68.59 kPa, and 125.74 kPa and 113.00 kPa, respectively. The results indicate that the shear strength increases with increasing vertical stress.

Under the same vertical stress, as the healing time increases, the shear strength becomes higher than the residual strength of the previous shear, indicating that the slide zone soil exhibits a strength recovery phenomenon. The recovered strength is brittle and disappears quickly after experiencing a very small displacement. The shear curves of the samples after a 3 h healing do not have consistent regularity. This may be because, after the initial shear is completed, a healing time of 3 h is relatively short and the pores of the soil particles and the cohesion at the shear zone are not fully recovered, such that the particles are still in a relatively disordered state, making the strength recovery not obvious. Ikari et al. [26] showed that there was no obvious regularity between the recovery strength value and the healing time when the healing time was short. Therefore, the results of the test with a 3 h healing were not analyzed and discussed, whereas the data for healing of more than 3 h were analyzed and are discussed in detail.

The peak strengths under vertical stress of 50 kPa, 100 kPa, and 200 kPa at 6 h, 12 h, 24 h, and 72 h are listed in

Table 5. Peak strength (kPa).

Vertical Stress (kPa)	50	100	200
Healing Time (h)
6	61.80	75.62	143.45
12	66.78	84.34	151.82
24	72.95	89.70	162.13
72	93.10	113.72	180.23

. The peak strength consistently increased compared with the previous shearing, and the change in the vertical stress had an effect on the recovery of the shear strength. However, the recovery value of the strength did not depend entirely on the increase or decrease in vertical stress. If the vertical stress increases, the contact area between the soil particles increases. Simultaneously, there is a continuous reduction in the porosity as a result of the continual squeezing and filling of pores by fine soil particles. In addition, particle rearrangement occurs, optimizing the arrangement mode at the sliding surface, leading to a reduction in the porosity and an increase in the compactness. The interaction force between the particles continuously increases, resulting in a recovery of the shear strength upon reshearing. However, the compactness of the soil is limited and the number of pores and pore interconnectors first increases, then decreases, and finally stabilizes [27,28]. Once the contact area and porosity are saturated, it is highly likely that the peak strength will increase with increasing vertical stress and then stop increasing and stabilize. Therefore, the effect of the vertical stress on the recovery of the soil strength is limited.

Table 6. Recovery strength value and recovery rate indices.

Vertical Stress (kPa)	Indices	Healing Time (h)
		3	6	12	24	72
50	Strength recovery value	3.55	1.20	6.18 0.52	12.35 0.51	32.60 0.45
	Recovery rate	1.15	0.85	0.52	0.51	0.45
100	Strength recovery value	0	7.03	15.75	21.11	45.13
	Recovery rate	0	1.17	1.31	0.88	0.63
200	Strength recovery value	18.20	30.45	38.42	49.13	67.23
	Recovery rate	6.20	5.07	3.26	2.05	0.93

summarizes the strength recovery value and recovery rate corresponding to different healing times for different vertical stresses.

Figure 10 shows the relationship between the vertical stress and strength recovery value given different healing times. It can be surmised that, under vertical pressure of 50 kPa, 100 kPa, and 200 kPa, the strength recovery values were 1.2 kPa, 7.03 kPa, and 30.45 kPa given 6 h of healing time; 6.18 kPa, 15.75 kPa, and 38.82 kPa given 12 h of healing time; and 12.35 kPa, 21.11 kPa, and 49.13 kPa given 24 h of healing time, respectively. The strength recovery values were 32.6 kPa, 45.13 kPa, and 67.23 kPa given 72 h of healing time. Given the same healing time, the strength recovery value increased with increasing vertical stress. At low pressure, the strength recovery value increased less, while the strength recovery value increased significantly at high pressure.

With increasing vertical stress, the curves of the strength recovery values for each healing time show an increasing trend, which indicates that the vertical stress has an effect on the strength recovery. Meanwhile, the relationship between the healing time and the strength recovery value given the same vertical stress indicates that, for increasing healing time, the strength recovery value increases; the strength recovery value reached the highest value for a healing time of 72 h.

The relationship between the vertical stress and the strength recovery rate given different healing times is shown in Figure 11, which shows that the strength recovery rate increases with increasing vertical stress given the same healing time and exhibits a positive correlation. This is mainly due to the increase in the vertical stress, which increases the density of the soil particles on the shear surface and the occlusion force between the particles. With increasing vertical stress and healing time, the slope gradually becomes smaller. When the vertical stress increases from 50 kPa to 200 kPa, the recovery rate increases by 4.22 kPa/h, 2.72 kPa/h, 1.53 kPa/h, and 0.48 kPa/h given healing time of 6 h, 12 h, 24 h, and 72 h, respectively. This indicates that the increment value of recovery rate gradually decreases with increasing vertical stress and healing time. The increase in the vertical stress and the healing time inevitably leads to an increase in the contact area and the contact time between the soil particles, an enhancement of the interactions between surfaces, an increase in the friction between particles, and an increase in the shear strength. On the other hand, the pores between the soil particles are determined by the minerals and the property of the soil itself. So, the soil porosity ratio and the degree of compactness of the contact surface are limited. Even given a higher vertical pressure and longer healing time, the strength recovery value is limited under a gradual change in compactness. When the soil particles at the shear surface reach absolute compactness, the contact area of the particles reaches saturation. Even if the healing time is longer, the strength recovery value is saturated and the strength recovery rate will be reduced or even stop. In summary, with the increase in vertical stress, the strength recovery rate shows an increasing trend, but the increasing amplitude tends to be stable with the increase in healing time.

Figure 12 shows the SEM (50×) characteristics of the shear surface after completing the “shear–pause–shear–pause–shear” cycle. Given the same healing time, the shear surface of the sample under low pressure is relatively rough and obviously undulating, with many distributed cracks and pores, and the soil particles are misaligned with each other, resulting in relatively inconspicuous and misaligned sliding traces. To some extent, the porosity near the shear surface is increased because of the volume expansion that occurs between the particles under low pressure and the compactness is lower than that of the high-pressure sample. The coarse particles are rolled and pushed, resulting in the roughness of the shear surface.

With increasing pressure, the coarser particles are disintegrated to form finer particles. The higher pressure causes the clay minerals to complete their directional arrangement faster, resulting in a smoother, tighter shear surface and more obvious directional slip marks.

6.2. Relationship between the Healing Time and the Strength Recovery

Ikari et al. [26] showed that the healing time has a significant effect on the recovery of landslides. After a certain healing time, there is an obvious healing phenomenon. The frictional healing and holding time has a log–linear relationship; with longer healing time, the rate of healing tends to stabilize and saturate. That is, the healing phenomenon is not significant at a short healing time, which gives a plausible explanation for the strength recovery phenomenon of the 3 h healing in this paper.

A regression analysis was performed on the strength recovery value obtained from shearing for healing time of 6 h, 12 h, 24 h, and 72 h, as shown in Figure 13. With increasing healing time, the strength recovery value at the shear surface of the sample continues to increase. The increase in healing time means that the consolidation and hardening time of the sample increase, showing increases in strength recovery value, indicating the dependence of the strength recovery on healing time. The increase in shear strength given different healing times indicates that the slide zone soil has a certain strength healing characteristic, which is manifested in the significant increase in the strength recovery value with increasing healing time. In addition, the strength recovery value under 200 kPa is significantly higher than those under 50 kPa and 100 kPa, and the strength recovery values under 50 kPa and 100 kPa are relatively close to each other, such that the strength recovery under high vertical stress is better than that under low vertical stress.

For low-plasticity soil, the characteristic of a gradual increase in shear strength with healing time is not infinite. The shear strength increases up to a certain extent, until the healing time exceeds 100 days, and then gradually converges to the fully softened strength value [21]. The magnitude of the shear strength is determined by the structure of the soil itself. Both the peak strength and the residual strength show a certain degree of recovery, reflecting the structural recovery of the soil during the healing time. After the sample is sheared to the residual strength, the particles produce cementation under a certain vertical pressure and healing time. In particular, the soil particles at the shear surface will interact with each other, and the shear strength will recover and increase under the action of continuous stress. The amount of shear strength recovery during the healing time is related to the mineral composition and content of the soil itself, and the magnitude of recovery rate varies given the same stress and healing time.

A nonlinear fitted regression analysis of the healing time and the recovery rate was performed, as shown in Figure 14. It can be seen that, with the increasing of the healing time, the recovery rate gradually slowed and eventually stabilized. The 6 h healing time had the highest recovery rate. As the healing time increased, the contact area of the soil particles increased, the pores between the mineral particles filled with fine soil particles, the sample porosity decreased, the occlusion force between the particles increased, the degree of cementation of the samples increased, and the soil gradually densified. However, the reduction of pore ultimately reaches a saturated state. Therefore, in the process of the healing time increasing, the soil particles of the shear surface also reach saturation and the recovery rate at the shear surface gradually decreases. Therefore, the strength recovery rate gradually decreases with increasing healing time and the strength recovery value saturates, eventually stabilizing and stopping recovery.

From the above analysis, it can be seen that when the soil reaches the residual strength state, the particles have formed a directional arrangement. However, in fact, when the shearing is generated again, the strength produces a certain degree of recovery, which is directly related to the face-to-face arrangement of the minerals. The strength recovery mechanism involves the van der Waals force and thixotropy. Smooth shear surfaces exhibit more van der Waals attraction than rough shear surfaces, and smooth plate-like clay particles with directional arrangements have a greater van der Waals attraction than randomly arranged clay particles. Mitchell and Soga [29] believed that clay particles absorb cations under certain environmental conditions (e.g., pressure, temperature, and water content). The net negative charge on the surface of the clay particles is neutralized by cations of water [30]. The exchange reaction generally depends on the valence state of the cations and the relative concentration of cations of water, whereas the strength recovery of the shear surface is likely to cause cation exchange. Thixotropy refers to the isothermal, reversible, and time-varying material hardening process that occurs in soil particles with constant composition and volume. That is, although the soil shear reaches a residual strength, in fact, the soil particles have not internally reached a state of complete equilibrium. The unbalanced state is primarily caused by the high repulsive force formed by the close face-to-face contact of the clay minerals. Under this unbalanced state, the soil particles will readjust to achieve a new strength value.

Figure 15 shows the SEM (20,000×) characteristics of the shear surface at 100 kPa given healing time of 3 h, 6 h, 12 h, 24 h, and 72 h. Given a 3 h healing time, the soil particle unit is primarily in the form of face-to-face contact, with larger pores between particles, less compact contacts, and poorer uniformity; the soil particles are in a loose state. Given a 6 h healing time, the soil particle unit is still dominated by the face-to-face contact form but the macropores are slightly reduced and the connections between the particles begin to increase. Given a 12 h healing time, the macropores again decrease, the connections between the soil particles increase, and the soil becomes denser. Given a 24 h healing time, there are almost no macropores and the soil is denser. It is shown again that the strength recovery is directly related to the structure of shear surface. Given a 72 h healing time, there are almost no large or small pores and the soil is denser.

7. Conclusions

This study examined undisturbed soil from the Huaipa landslide in Sanmenxia City and adopted the “shear–pause–shear–pause–shear” cyclic method to perform ring shear tests to study the strength recovery characteristics. The following preliminary understanding was obtained.

• When the sample reaches the residual strength after the first shearing and is then sheared again with the shear stress and vertical stress kept constant, the shear strength exhibits an increase; however, this strength recovery feature decreases rapidly after experiencing a very small displacement.
• Given the same healing time, the strength recovery value increases with increasing vertical stress. Under low pressure, the strength recovery value is small, and under high pressure, the strength recovery value is obvious.
• Given the same vertical stress, the recovery strength increases with increasing healing time. As the healing time increases, the rate of recovery, although consistently increasing, increases in progressively smaller magnitudes and ultimately approaches zero.