Introduction

Neogene soft rock is widely distributed in the northeastern margin of the Tibetan Plateau in China and in the Duero Basin in Spain (Caterina et al. 2010, 2013; Yenes et al. 2009, 2015). Several landslides have occurred in this type of rock groups. A typical shear zone in the Neogene soft rock includes two groups of horizontal boundary faults and several internal secondary shear planes (Hart 2000; James 2001). The clay mineral content in the Neogene soft-rock shear zones is greater than 30% (e.g., Xin et al. 2018; Yenes et al. 2009, 2015). Picarelli et al. (2006) observed specimens exhibiting small shear fissures, which can be observed as groups of shear structures exhibiting similar orientations in hard-soil shear zones. The structural development characteristics of the shear zones would change gradually during large shear displacement. Therefore, a detailed study of shear zone development is considered to be crucial for understanding the development of landslides in the case of Neogene soft rocks.

The deformation behavior of the shear zone during large shear displacement is an important factor associated with mass movements involving residual and steady-state strength. Skempton (1985) observed the development of a low-angle failure surface in pure montmorillonite and illite, with the internal friction angles reaching 5° and 10°, respectively. Lupini et al. (1981) observed that the occurrence of horizontal faults in London clay can be determined based on its clay content under drained shear tests with a displacement of lower than 100 cm (Fig. 1a–c). When the clay content was greater than 40%, a 1.5-mm-thick shear zone was parallel to the shear direction (Fig. 1b). However, two sets of shear planes can be clearly observed when the content of clay was greater than 48%. The shear zone is approximately 2.5-mm thick, and the angle between the clay mineral orientation and shear direction is <45° (Fig. 1c). Yenes et al. (2009) considered that the shear zone, including the boundary fault planes, central shear planes, and a Riedel plane with high calcium carbonate content, exhibited brittle failure in accordance with the Riedel shear criterion (Fig. 1d). The boundary faults and cleavage planes could be observed in a ductile shear zone comprising an abundant amount of clay (Figs. 1e–f). The particles in the shear zone will be arranged in layers when the shear displacement in the shear zone increases from 20 cm to 10 m (Muhammad et al. 2004; Wang et al. 2000). For example, when silica sand was continually sheared, the grain size decreased and the thickness of the core zone reduced from 4 to 2.2 cm; fine particles were observed to be concentrated near the slip surface (Fig. 1i) (Muhammad et al. 2004). Picarelli et al. (2006) believed that the platy clay minerals in the core zone would be arranged along a plane parallel to the shear plane and that the content of these minerals would continually increase under continuous shearing. However, the relation between the mineral content and water in any region of the Neogene soft-rock shear zone, which comprises clay particles, has not been addressed. Quantitative analysis of a geological section would be beneficial to reveal the development processes in the slide mass.

Fig. 1
figure 1

Mechanical modeling of the shear zone and rearrangement of the internal clay minerals (Lupini et al. (1981), Yenes et al. (2009), and Muhammad et al. (2004)): a cataclastic failure with a clay fraction of 34%; b cataclastic failure with a clay fraction of 40%; c cleavage failure with a clay fraction of 48%; d Riedel shear fracture with high calcium carbonate content; e ductile shear fracture with an abundant amount of clay; f ductile–brittle fracture; g initial structure of a sandy shear zone; h progressive rupture of a sandy shear zone; and i layered deposition of particles

The distribution characteristics of clay minerals are crucial to understand the development of the Neogene soft-rock shear zones. Previously, the ring shear tests were conducted to investigate the evolution of the shear zone structure (e.g., Lupini et al. 1981; Wang et al. 2000). However, the confining pressure and shear displacement in experiment cannot reach the values obtained in situ in case of landslides (Carey and Petley 2014; Skempton 1970; Xin et al. 2016, 2018). To address these discrepancies, the distribution characteristics of clay minerals and the exchange process between clay minerals and pore water in case of actual landslides were investigated.

Geological profiles and testing methods

Seven landslides among the 110 landslides on the northeastern margin of the Tibetan Plateau in China were selected to analyze the distribution of clay minerals and their interaction with water (e.g., Xin et al. 2016, 2018). These landslides can be divided into two groups, i.e., short-distance landslides with a movement distance of less than 100 m (such as Caijiapo landslide and Songjiayao landslide) and long-distance landslides with a movement distance of more than 100 m (such as Bojishan landslide, Gaojiaya landslide, Caizigou landslide, Xinguancun landslide, and Liujiazhuang landslide).

Geological profile and structure

The shear zones are characterized by layers exhibiting cataclasis, cleavage, and argillation in the core and are described below.

Cataclastic layer

The cataclastic layers in shear zones are in the initial stage of failure, and their structure is consistent with the Riedel fracturing. The sliding zone of the Caijiapo landslide, which occurred in 1932, is approximately 0.8-m thick and 320-m long. The shear zones of the Caijiapo landslide (Fig. 2(a) and (b)) and the Songjiayao landslide (Fig. 2(c) and (d)) comprise two sets of boundary faults and central shear planes, above which a cataclastic zone can be observed.

Fig. 2
figure 2

Photos and geological profiles of the shear zone in landslides considered in this study. a–b cataclastic structures in shear zone of Caijiapo landslide; c–d photos of the shear zone in Songjiayao landslide; e–h cleavage structures in shear zone of Liujiazhuang landslide; i–j plastic structures in shear zone of Caizigou landslide

Cleavage layer

In the cleavage layer, clay minerals are arranged in a network comprising individual or stacked plates and are characterized with bright and reflective layers that may exhibit scratches in the core. The slip zone is approximately 0.6-m thick in the Liujiazhuang landslide. Plastic and cleavage zones developed near the upper shear plane, with a spacing of less than 0.2 cm between the cleavage planes (Fig. 2(e) and 2(f)). Similar features can also be observed in the Gaojiaya landslide (Fig. 2(g) and 2(h)).

Argillaceous layer

In the argillaceous layer, the soft rock has undergone weathering, and the cementation structure is completely deteriorated. The observed structure comprises powdered particles that have undergone plastic deformation in fairly thin layers. A film of water has also been observed on some of the shear planes. For example, in the Caizigou landslide, the thicknesses of the mud layer, cleavage zone, and cataclasis slip zone are approximately 2 cm, 0.3 m, and 0.75 m, respectively.

Testing methods and procedures

The cumulative displacements of the core and boundaries of the shear zones are different and exhibit different characteristics (Muhammad et al. 2004; Wang et al. 2000). The vertical profiles in Fig. 3 were quantitatively studied from top to bottom in terms of the material composition, pore structure, and water properties of each sample location to understand the changes in the structure of the shear zone. Testing proceeded via three steps after the onsite borehole drilling. The first step involved the preparation of thin sections from samples along the vertical profile of core. Subsequently, the mineral content and the arrangement of minerals at each position were determined using a polarizing microscope. Step two includes testing the quantitative characteristics of the clay minerals, pore structure, and water content along the identified shear zone. Finally, the sliding process of the unsaturated samples under undrained conditions and long shear displacement were analyzed.

Fig. 3
figure 3

Sampling in landslides from the northeastern margin of the Tibetan Plateau: a borehole ZK1 in the Caizigou landslide; b borehole ZK3 in the Xinguancun landslide; c borehole ZK2 in the Bojishan landslide; d sampling of the Caijiapo landslide; e sampling of the Gaojiaya landslide; and f sampling of the Liujiazhuang landslide

Fifty samples are selected for testing from among six geological profiles with different forming times and stress environments. The sample core ZK1 from the Caizigou landslide, cores ZK3 and ZK4 from the Xinguancun landslide, and core ZK2 from the Bojishan landslide were obtained from the center of their respective landslide masses. The sampling interval in ZK1 (Caizigou landslide) was in the range of 1–6 m, as shown in Fig. 3a. For ZK3 and ZK4 (Xinguancun landslide), samples were obtained at intervals of 0.5–2 m, as shown in Fig. 3b. Figure 3c shows the sampling scheme at an interval of 0.1–0.2 m for ZK2 (Bojishan landslide). The horizontal shear zones are prominently exposed in the contraction zones of the front edge of a translational sliding body. Because of the good accessibility of such zones in the Liujiazhuang, Gaojiaya, and Caijiapo landslides, they were selected for the case study in which the sampling interval was set as 0.1 m, with the sampling positions shown in Fig. 3d–f.

Thin-section analysis for identifying the mineral and estimating the content in samples from shear zones with different structures was performed in accordance with the national standard: Identification Manual of Transparent Mineral Thin Section and Classification and Nomenclature Schemes of the Rocks—Classification and Nomenclature Schemes of the Sedimentary Rock (GB/T 17412.2 1998).

The quantitative analysis of material composition and pore structure included tests for determining the clay and calcium carbonate content, natural water content, pore ratio, plastic limit, and specific surface area (Table 1). The tests of water content, fine-grained soil particle size, plastic limit, and calcium carbonate content comply with SL-237-002, SL-237-005, and SL-237 in the Standard for Soil Test Method (GB/T 50123 1999). The specific surface area tests conform to the methods outlined in the Determination of the Specific Surface Area of Solids by Gas Adsorption Using the BET Method (GB/T 19587 2004).

Table 1 Sample composition and clay mineral data from the shear zones at the foot of landslides

The clay mineral content, such as montmorillonite and illite, can be obtained via X-ray diffraction. The hydrologic properties of montmorillonite and those of a mixed-mineral layer of illite and smectite are considerably different, and the hydrologic environment affects the distribution of the aforementioned clay minerals. Three preparation methods were used to analyze 19 triplicated samples. These samples are obtained from the cataclasis, cleavage, and argillation shear zone in landslides, including the Caijiapo, Gaojiaya, and Liujiazhuang landslides (Fig. 3d–f). The first method was used to extract the sample suspension (less than 2 μm) and prepare directional tablets. The second method treated the directional tablets via ethylene glycol saturation. In the third method, the directional tablets were heated for 2 h at 550°C before testing. The X-ray diffraction patterns were determined using the computer information processing software to analyze the quantitative relation between the area of diffraction peak and the clay mineral content, the mineral peak type of a mixed layer, and the mineral ratio of mixed layer and obtain the composition, relative content, and ratio of clay minerals. A Rigaku Rotary Anode X-ray diffractometer (D-Max 2400 Series) running at 12 kW with a tube voltage of 35 Kv, current of 20 mA, stability within +0.03%, focusing area of 0.5 mm (+10 mm), and a goniometer diameter of 185 mm was used to analyze the clay samples. During the experiment, a scanning mode of 2θ/θ linkage was adopted using a scanning range of 3°–35° and a scanning speed of 2°/min. The testing equipment was a Bruker AXS D8-Focus X-ray diffractometer (the Netherlands). The test procedure complied with the Oil and Gas Industry Standards of the People’s Republic of China (SY/T 5163 2010).

The pore structure and hydraulic properties of the shear zone changed gradually with the development of fractures and shearing planes. The tests were conducted on samples from the undisturbed zone and core zone of the shear zone in the Bojishan landslide to quantitatively measure its pore. First, all the samples would be dried. Then, they would come into contact with water. Next, the amount of water absorbed by the samples will be estimated. Finally, the curves of water absorption in samples were plotted based on the tests. The curves included the water absorption capacity (Q) and water absorption time (t).

The ring shear tests were conducted to simulate the sliding process of the unsaturated samples under undrained conditions and long shear displacement. The average annual rainfall is <700 mm in the northeastern margin of the Tibetan Plateau in China, and the groundwater level changes by 0.2–2.5 m around the shearing zone (e.g., Xin et al. 2016). The landslides are triggered by the rising groundwater level and water pressure (e.g., Xin et al. 2016). However, there is insufficient water for the soil to attain saturated liquefaction at the aforementioned low water pressure. According to the above conditions, two steps were devised in the test for two samples obtained from the borehole ZK2 in Bojishan landslide. First, the natural samples were filled and compressed in the ring shear box for approximately 1 h under 180 kPa and 160 kPa. CO2 gas was supplied for approximately 1 h to expel ambient air. The initial water pressure was maintained at approximately 10 kPa. Second, the samples were sheared at speeds of 0.02 and 0.05 mm/min. The equipment was a DTA-138 ring shear apparatus (Seiken Inc., Japan).

The residual strength of the samples from the Bojishan landslide could be estimated via repeated shear tests. The test procedure complied with the Standard for Soil Test Method of the People’s Republic of China (GB/T 50123 1999).

Results

Reorganized structure of the shear zone

Neogene soft rocks are clastic rocks in which clay minerals, quartz, and mica constitute approximately 80% of the mineral assemblage. In this study, the observed clay content in soft rocks ranged from 20 to 64%. The interaction between clay minerals and water alters the structure of the shear zone. In the fragmentation zone, the soft rock has a blocky structure and is filled with autogenous clasts along the fracture (Fig. 4(a)). In contrast, in the core zone, clay minerals and calcite are aligned in the same direction and have a parallel cleavage domain (Fig. 4(b) and (c)). Clay minerals, such as illite, are arranged in a continuous and directional manner. In the shear zone, there are obvious dispersed water films on the clay minerals, which can be attributed to the interaction of water and rock (Fig. 4(b) and (c)). Under intense deformation episodes, the platy clay minerals gradually changed from flocculent to scaly and were superimposed.

Fig. 4
figure 4

The structure of shear zone in the Liujiazhuang landslide: (a) Photos of the shear zone; (b–c) the microstructure of the cataclastic layer in upper zone; (d–e) photos of microstructure of the cleavage layer in the core zone; (f–g) photos of microstructure of the argillaceous layer in below zone

Distribution characteristics of clay minerals in different structures

The content of clay minerals in the shear zone of short-distance landslides is less than that in case of long-distance landslides. For example, the borehole ZK1 of the Caizigou landslide indicates that the slip zone is cataclastic, and a strong variation zone can be observed at 50.3–51.4 m (Fig. 5a). Two samples from this zone yield total clay contents of 39.64% and 38.42%, with montmorillonite contents of 26.26% and 25.34%, respectively. The measured contents of calcium carbonate are 4.23% and 5.11%, and the specific surface areas relevant to the pore size of rock are 232.74 and 205.66 m2/g (Table 1). The lowest clay mineral content could be observed within the fine particulate matter aggregates on the upper boundary surface of the shear zone (Table 1). The thickness of the sliding zone is approximately 2.1 m in borehole ZK3 of the Xinguancun landslide. The clay mineral contents toward the top, center, and bottom of the sliding zone are 73.39%, 49.85%, and 63.84%, respectively (Fig. 5b). The calcium carbonate contents and montmorillonite concentrations for these sampling positions are 7.86%, 8.83%, and 9.43% and 11.80%, 19.60%, and 12.90%, respectively. In contrast, the montmorillonite content in the core from borehole ZK4, which is within the shear zone of the Xinguancun landslide, is 20.3%.

Fig. 5
figure 5

Characteristics of clay particles in boreholes from geologically young landslides from the northeastern margin of the Tibetan Plateau: a borehole ZK1 in Caizigou landslide; b borehole ZK3 in Xinguancun landslide; and c borehole ZK2 in Bojishan landslide

In case of long-distance landslides, the content of fine particulate matter, such as clay, is slightly greater than that in the vicinity of the core zone, whereas the clay content is the highest in the shear zone and the lowest in the fragmentation zone. In the Bojishan landslide, the depth of the 2.2-m slip zone is 64.4–66.6 m. The upper boundary at 64.4 m is 0.2–0.3-m thick and experiences the greatest deformation. A middle shear plane is defined between 65.2 and 65.6 m with a thickness of 0.4–0.6 m, and a 0.49-m-thick lower boundary can be observed between 65.51 and 66.6 m. The dynamic failure of soft rock affects the clay mineral content. The Bojishan landslide (Fig. 5c) has a clear mechanical structure in which the soft rock in the upper boundary is muddy, the soft rock in the central region contains a shear plane, and the lower boundary displays cleavage. The contents of clay in these three regions are 62.60%, 62.46%, and 58.04%, respectively, corresponding to calcium carbonate contents of 1.36%, 1.33%, and 1.18%, respectively. The contents of water in the above three regions are 28.82%, 25.34%, and 25.52%, respectively.

The contents of total clay minerals and montmorillonite are high near the shear surface with cleavage and mud structures, although the calcium carbonate content is visibly low and montmorillonite is concentrated in the upper part of the slip zone. The fragmentation slip zone of the Caijiapo landslide exhibits a thickness of 30 cm and includes a strong fragmentation zone and a disturbance zone. The clay, montmorillonite, and calcium carbonate contents are 50.36%, 25.29%, and 1%, respectively. However, the clay content is 34.08%–39.80%, the montmorillonite content in the mixed-mineral layer is 23.23%–22.78%, and the calcium carbonate content is 2.21%–4.77% within the slide disturbance zone.

The contents of clay minerals, such as montmorillonite and illite, gradually increase because of the hydration reactions (Shuzui 2001; Wen and Aydin 2003). When dynamic fragmentation is strong, the numerous structural planes in the sliding zone promote strong soil–water interaction. The results of a chemical constituent test indicate that the HCO3− content is 358.15–373.04 mg/l, SO42− content is 13.91–27.65 mg/l, and Ca2+ content is 75.73–77.82 mg/l in the Bojishan and Gaojiaya landslides (Table 2). The local groundwater contains high levels of HCO3− and Ca2+, providing an environment for the hydration of clay minerals and the conversion reaction of montmorillonite and illite. Shuzui (2001) observed that the stability of kaolinite improved when the concentration of HCO3− was as high as 30 mg/L. When the concentration of HCO3− exceeded 40 mg/L, the groundwater near the slip surface was in equilibrium with montmorillonite, providing a favorable environment for the dissolution of kaolinite and formation of montmorillonite. With the increasing HCO3− concentration, the active ion exchange of calcium ions was separated, promoting the formation of montmorillonite.

Table 2 Chemical constituents and groundwater characteristics of the Gaojiaya and Bojishan landslides

As noted above, the clay minerals in the slip zone primarily contain illite and montmorillonite. The clay mineral content can be identified via X-ray diffraction analysis. Thus, it was determined that the ratio of montmorillonite in mixed-mineral layers near the shear plane is higher than the values obtained for the remaining zones (Fig. 6, Table 1). The clay mineral content and the proportion of montmorillonite in the mixed-mineral layer are 62.60% and 80%, respectively, in the core of the shear zone in the Bojishan landslide, whereas they were 58.04% and 65%, respectively, in other parts of the shear zone. As the shear displacement in the slip zone increases, the proportion of montmorillonite in the mixed-mineral layer gradually increases. The proportion of montmorillonite in the mixed layer of illite and montmorillonite in the Gaojiaya landslide becomes 67%, whereas the absolute content of illite and montmorillonite is 25.5%. In contrast, the montmorillonite in the relatively young Liujiazhuang landslide has a maximum value of 55% in the mixed-mineral layer. The amount of montmorillonite in the mixed-mineral layers in the shear zone of these two landslides is lower than that in the shear zone of the Bojishan landslide, indicating that the content of montmorillonite in a landslide shear zone is affected by its dynamic environment.

Fig. 6
figure 6

X-ray diffractograms from oriented slides of <2-μm fractions from ab the shear zone of the Caijiapo landslide and cd the shear zone of the Bojishan landslide. ① Normal sample; ② glycolated sample; ③ sample heated to 550°C. I, illite; I/S, illite and montmorillonite mixed-layer; K, kaolinite; C, chlorite; Q, quartz; and Ca, calcium carbonate

Change of shear strength

Any change in the mineral composition of the slip zone would affect the permeability of the water within this zone. Water will accumulate or flow in the cracks, pores, and shearing planes under hydrostatic pressure. Pore water penetrates the fractures when the groundwater pressure increases. The results indicate that absorption rates are 0.24–0.3 ml/h in the shear zones with cleavage and shear planes and 0.13–0.14 ml/h in the fractured zones. Although the soft rock is generally impervious to water, free water will rapidly infiltrate along the shear zones (Fig. 7). The fracture and pore in shear zone would be full of water under high pressure. For instance, the average of pore water pressure in the Bojishan landslide is 92 kPa. The natural water content of the upper boundary in the strong deformation fault is 24.7%–26.8%, that near the middle slide zone is 18.7%–22.9%, and that near the lower surface is only 18.7%–25.1% (Table 3). The dilatancy and grain crushing in shear zone are the main reasons for the increase of cracks and pores (Muhammad et al. 2004).

Fig. 7
figure 7

Water absorption–time curve of the slip zone and the surrounding rock for samples from the Bojishan landslide

Table 3 Relation between clay content and strength of the shear zone in Bojishan landslide

The seepage of water and its interaction with minerals can be attributed to the development of the shearing surfaces. Usually, the undisturbed soft rock is not conducive to the flow of water. The change in volume of soft rock occurs in three different stages when it is compressed and sheared. In the first stage, the samples are compressed; this is accompanied by an increase in pore water pressure. In the second stage, the pore water pressure gradually decreases to negative values accompanied by a gradual decrease in shear stress. In the third stage, the pore water pressure progressively increases and subsequently experiences another reduction (Fig. 8).

Fig. 8
figure 8

Ring shear tests for samples 7910 and 7911 from the borehole ZK2 in the Bojishan landslide

The accumulation of shear deformation and water could change the strength of the soft rock. The core of the shear zone has a high natural water content. The internal friction angle at the upper boundary of the fault, that of the main slip zone in the central zone, and that of the lower boundary fault are estimated to be 9.2°–12.6°, 15.8°–17.4°, and 15.8°–16.1°, respectively, through repeated tests of the undisturbed samples from the shearing zones in the Bojishan landslide (Table 3).

Discussion

The distribution of montmorillonite in shear zones is not uniform, and clay particles can be observed as aggregates along the fracture surface. The cataclasis of Neogene soft rock results in a gradual transformation of the materials in the shear zone through fragmentation, formation of cleavage, plasticization, argillation, and the accumulation of clay particles (Wu et al. 2020). The argillaceous layer has the lowest residual strength because it contains the most number of cracks and exhibits the lowest water content. However, it remains unclear which processes cause the accumulation of mixed-mineral layers and which processes are responsible for the notable increase in montmorillonite content near the shear surface.

Relation between the structure and clay minerals

The coarse and fine particles within the shear zone segregated during motions, and a parallel orientation structure developed during the steady-state deformation (Muhammad et al. 2004). According to the data from the rock samples obtained from the Longzhong Basin (Table 4), water content near the shear surface varies from 10.12 to 19.75%, and the average content of montmorillonite is 13.26%, with a maximum of 23.74%. The average ratio of illite–montmorillonite mixed-mineral layer in these mudstones is 38%, with the highest clay and internal water content of 65.23% and 14.53%, respectively (Table 4).

Table 4 Clay mineral content in the shear zones in northwestern China

Based on statistical analysis, higher clay contents tend to correspond to higher mixing layer ratios in the sliding zone. The illite/montmorillonite mixed-mineral layer that accumulates near the shear surface exhibits strong hydrophilicity (Fig. 9). The clay contents in the argillation layer reach 62%, and the water content exceeds the plastic limit at 28.82%. The mixed-mineral layer ratio ranges from 80 to 100%; however, the mixed-mineral layer ratio in the unaffected original soft rock is only 50%. Thus, as noted above, soft rock is originally impervious to water; however, it can be easily infiltrated in the post-slide state.

Fig. 9
figure 9

Relation between the mixed-mineral layer ratio and clay particle content (a) and water (b)

The arrangement of minerals and particles in the shear zone are controlled by physical friction. In the cleavage zone, fine and delicate clayey particles are arranged in an orderly network. Den Brok (1998) concluded that compression through pressure dissolution is an important factor associated with cleavage formation. Pressure dissolution can be observed at the boundary of particles in the direction of maximum vertical compression. The dissolved substances migrate and accumulate from the area with the higher chemical potential energy to the low-stress region. Because quartz and feldspar dissolve in the direction of vertical compression, their particles become lenticular or form long strips. Pressure dissolutions continuously progress on the boundary of the particles or the interface of layers in the direction of vertical compression. Insoluble residues, such as clay or mica, become enriched, and the sheet minerals, such as mica, progressively rotate and align in the direction of maximum vertical compression, forming cleavage domains.

The deformation of shear zone

The penetration of water would induce soil–water interaction along shear plane. The deformation of the shear plane primarily includes creep and rapid slip stages. The creep stage is dominated by viscous deformation, as indicated by the continuous creep slip of the Hongluo Village landslide in Tianshui from 2016 to 2017 (Fig. 10a). The upper boundary fault plane of the horizontal slip zone varies most intensely, and its cumulative displacement is the largest. In the Bojishan landslide from 2012 to 2016, the cumulative displacements of the #2 and #3 main sliding zones were 150.73 and 152.05 mm, whereas those of the #1 lower fracture surface and #4 upper fracture surface were 0 and 99.75 mm, respectively (Fig. 10b). The rapid slip of the shear zone extends along the upper shear plane. In addition, the clay mineral content tends to be the largest in zones with the maximum displacement (Massey et al. 2013, 2016; Petley et al. 2002; Palladino and Peck 1972). The fracture in the clay-rich soft rock allowed the shear plane to become a pathway for free water; thus, there is a relation between physical mechanisms and the rock–water chemical reactions of clay that should be addressed in future studies.

Fig. 10
figure 10

Variations in displacement landslides: a displacement of the Hongluocun landslide from 2016 to 2017 and b displacement of the shear zone in the Bojishan landslide from 2012 to 2015

Conclusions

In this study, seven landslides were selected to analyze the structural development process of shear zone. Several vertical profiles were quantitatively studied from top to bottom in terms of the material composition, pore structure, and water content to understand the changes in the cementation and frictional properties of the shear zone. Even though no sliding surface liquefaction occurred, the slipping of clayey particles could be observed.

The shear zones are characterized with layers exhibiting cataclasis, cleavage, and argillation in the core. With the increasing displacement of the shear zone, the clay particles accumulate in the core of the shear zone, surrounded by the remaining mineral particles. Montmorillonite aggregates can be observed along the shear boundaries of the shear surface. The montmorillonite content is the highest near the core zone. The differences in montmorillonite content in the three fracture surfaces in the shear zone can be attributed to their varying cumulative displacements.