Keywords

1 Introduction

Understanding the shear strength behavior of landslide soil is important to study the rapid and long-traveling landslides in Sri Lanka. Peak and steady-state shear strength parameters of the landslide soil are usually required for stability analysis, flow path prediction, and mitigation designs. To obtain the shear strength behavior of soil numerous laboratory and field experiments have been used such as direct shear test, triaxial test, ring shear test etc.

During the past decade, limited studies have been done to observe the peak and ultimate relations between the shear stress, s′ and the effective mean normal stress t′ of landslide soil in Sri Lanka using triaxial tests (Dias et al. 2014). Some experimental studies have been conducted using ring shear apparatus for the soil samples from the Aranayake landslide area in Sri Lanka to obtain peak and residual strength parameters required for LS-RAPID to simulate the initiation and downward movement of landslide mass (K. Konagai et al. 2022).

Ring shear test is a very common method used to determine the residual shear strength of landslide soil materials. The significant advantage of the ring shear apparatus is that it can develop large shear deformations in the specimen, and there is no change in the shear plane area during shearing. Therefore, a ring shear apparatus is a major tool used in the analysis of the residual shear strength of the soil.

As stated by D. H. Loi et al. (2022), ring shear test was introduced by Bishop et al. (1971), and it was improved by Bromhead (1979), Savage and Sayed (1984), Sassa (1984), Hunger and Morgenstern (1984), Tika (1989), and Garga and Infante Sedano (2002). Sassa et al. (1997) developed and improved a series of ring shear apparatus that can used to study landslide dynamics by simulating the entire process of the landslide.

In this study, we assess the shear behaviors of landslide soil material by performing undrained ring shear tests under different effective normal stress conditions. Landslide soil samples with different fine contents have been used to study the effect of fine content on the shearing behavior using the undrained ring shear apparatus, ICL-2.

2 Testing Program

2.1 Materials

The test series consists of several samples collected from the Athwelthota landslide area. This landslide occurred in the Athwelthota area in Baduraliya DS division, Kalutara District, on May 26, 2017, at around 0500 h due to heavy precipitation. The field investigations indicate that the landslide was initiated at two locations on the mountain crest of Paru Pana Mukulana Kanda. For this study, we collected undisturbed box samples from landslide crest and middle areas as shown in Fig. 2.

Sieve and hydrometer analysis, Atterberg limits and specific gravity tests were conducted on these soil samples to determine the index soil properties. Figure 1 shows the grain size distribution curves. The soil samples’ fine content was 26, 42, 47, and 55 percent by weight (termed F26, F42, F47, and F55, respectively) (Fig. 2). According to the BS soil classification, the soil can be categorized as Silty Sand or Sandy SILT. Table 1 summarizes the liquid limit, plastic limit, plasticity indices, and particle densities of the soil samples.

Fig. 1
A multiline graph plots percentage passing versus particle size in millimeters. The plotlines are for F 26, F 42, F 47, and F 55. The plotlines decline from approximately 10 millimeters to 0.001 millimeters. The plotline for F 26 is the lowest whereas F 55 is mostly high with F 42 higher at the end.

Grain size distribution of soil samples

Full size image
Fig. 2
A google map of Athwelthota landslide area with sampling locations. Sample F 55 is near S 1, F 42 near S 3, F 26 near S 4, and F 47 near S 2.

Athwelthota landslide area with sampling locations

Full size image
Table 1 Physical properties of the soil
Full size table

2.2 Ring Shear Apparatus

This paper’s ring shear test results are obtained using ICL-2 undrained ring shear apparatus, donated to the National Building Research Organization, Sri Lanka, under the Science and Technology Research Partnership for Sustainable Development (SATREPS) project in 2022.

According to D. H. Loi et al. (2022), Sassa and his colleagues in the Disaster Prevention Research Institute (DPRI), Kyoto University, and International Consortium on Landslide (ICL) have developed nine designs dynamic loading ring shear apparatus since 1984 (DPRI-1, DPRI-2, DPRI-3, DPRI-4, DPRI-5, DPRI-6, DPRI-7, ICL-1, and ICL-2). The details of the ICL-2 ring shear apparatus and basic ring shear experiments have been discussed comprehensively in his study. ICL-2 undrained ring shear apparatus is having a maximum loading capacity of 1000 kPa.

In this study the residual shearing behavior of landslide soil was examined by performing a series of undrained ring shear tests. For the soil sample with a fine content of 26% (F26), monotonic shear-stress control, undrained ring shear tests were conducted in saturated condition, with shear stress increment rate of 0.25 kPa/sec under effective normal stress of 100 kPa, 200 kPa, 300 kPa, 400 kPa and 500 kPa.

For the soil samples with different fine contents (F26, F42, F47, and F55) undrained ring shear tests been conducted to study the influence of fine particles on the shear behavior. Table 2 summarizes the test conditions used for each test.

Table 2 Ring shear test conditions
Full size table

Due to the size of the shear box, dry soil passing a 2 mm sieve is used, as shown in Fig. 3. About 0.8 kg of dry soil is placed in a plastic container, and de-aired water is poured into it up to 3 mm above the soil surface. It is placed in the vacuum tank to remove entrapped air in the soil to saturate the soil sample until the air bubbles cease escaping, as shown in Fig. 4.

Fig. 3
A photograph of soil sample in 2 trays and a sieve. There is granulated soil in 1 tray whereas fine soil in the other. The sieve has large chunks of soil.

Sample preparation

Full size image
Fig. 4
A photograph of a vacuum tank with a liquified soil sample in a cylindrical container at the center.

De-airing the soil sample

Full size image

The shear box is set up in the ring shear apparatus, and the first empty shear box is filled with CO2 and then de-aired water to remove entrapped air in the shear box. Then the de-aired soil sample is slowly built in the shear box and the sample surface is leveled, as shown in Fig. 5.

Fig. 5
A photograph of an apparatus for the leveling of the sample surface. The apparatus is circular with the sample inside. A person operates the procedure.

Leveling the sample surface

Full size image

After setting up the sample in the ring shear apparatus, de-aired water circulates through the shear box to remove trapped air. Bd valve is checked to measure the degree of saturation; if it is greater than 0.95, the sample is in the saturated state.

After the saturation, soil specimen is consolidated to simulate the initial stress conditions of the sample and the shearing is carried out under the undrained condition through a servo-controlled motor at a 0.25 kPa/sec stress-controlled condition. Pore-water pressure, shear displacement, shear stress, normal stress, and vertical displacement of the specimen are monitored during the ring shear test to study the soil’s shear behavior.

3 Test Results

For the soil sample with 26% fines, undrained ring shear test results under normal stresses of 100 kPa, 200 kPa, 300 kPa, 400 kPa and 500 kPa are shown in Fig. 6. For different normal stress conditions, the effective stress path was plotted and the residual failure line was obtained.

Fig. 6
5 line graphs of results of stress path from undrained ring shear test. They plot shear stress in kilo pascal versus effective normal stress. The plotline inclines and then declines forming concave down declining curves. The theta m values are as follows respectively from a to d. 48, 44, 44, 41.

Results of stress path from undrained ring shear test under different normal stresses on silty sand with 26% fines, RFL-residual failure line

Full size image

Undrained ring shear tests results for the soil samples with different fine contents F26, F42, F47, and F55 are presented in Figs. 7, 8, 9, and 10.

Fig. 7
3 line graphs. a, shear stress versus effective normal stress. The plotline inclines and then declines forming a concave down declining curve with theta m at 44. b, stress-displacement versus time. Pore water pressure and shear displacement are inclining. c, shear stress versus shear displacement. The plotline inclines and drops.

Results of undrained ring shear test. (a) Stress path, (b) Time series of data, (c) Shear stress and shear displacement relationships. Test conditions: sample: silty sand with 26% fine content, BD = 0.94, normal stress = 500 kPa, shear stress increment rate: 0.25 kPa/s

Full size image
Fig. 8
3 line graphs. a, shear stress versus effective normal stress. The plotline inclines and then declines forming a concave down declining curve with theta m at 41. b, stress-displacement versus time. Pore water pressure and shear displacement are inclining. c, shear stress versus shear displacement. The plotline inclines and drops.

Results of undrained ring shear test. (a) Stress path, (b) Time series of data, (c) shear stress and shear displacement relationships. Test conditions: sample: sandy silt with 42% fine content, BD = 0.93, normal stress = 500 kPa, shear stress increment rate: 0.25 kPa/s

Full size image
Fig. 9
3 line graphs. a, shear stress versus effective normal stress. The plotline inclines and then declines forming a concave down declining curve with theta m at 41. b, stress-displacement versus time. Pore water pressure and shear displacement are inclining. c, shear stress versus shear displacement. The plotline inclines and drops.

Results of undrained ring shear test. (a) Stress path, (b)Time series of data, (c) Shear stress and shear displacement relationships. Test conditions: sample: sandy silt with 47% fine content, BD = 0.95, normal stress = 500 kPa, shear stress increment rate: 0.25 kPa/s

Full size image
Fig. 10
3 line graphs. a, shear stress versus effective normal stress. The plotline inclines and then declines forming a concave down declining curve with theta m at 40. b, stress-displacement versus time. Pore water pressure and shear displacement are inclining. c, shear stress versus shear displacement. The plotline inclines and drops.

Results of undrained ring shear test. (a) Stress path, (b) Time series of data, (c) shear stress and shear displacement relationships. Test conditions: sample: sandy silt with 55% fine content, BD = 0.96, normal stress = 500 kPa, shear stress increment rate: 0.25 kPa/s

Full size image

From Fig. 7(a), we obtained mobilized friction angle at failure (Øm =440), steady state shear resistance (τss =118 kPa) for the sample with 26% fines. The shear stress and shear displacement relationship, Fig. 7(c) shows shear displacement at the start of the strength reduction (DL=6 mm) and shear displacement at the start of steady state (DU=80 mm).

From Fig. 8(a), we obtained mobilized friction angle at failure (Øm =410), steady state shear resistance (τss=61 kPa) for the sample with 42% fines. The shear stress and shear displacement relationship, Fig. 8(c) shows shear displacement at the start of the strength reduction (DL=8 mm) and shear displacement at the start of steady state (DU=200 mm).

From Fig. 9(a), we obtained mobilized friction angle at failure (Øm =410), steady state shear resistance (τss =52 kPa) for the sample with 47% fines. The shear stress and shear displacement relationship, Fig. 9(c) shows shear displacement at the start of the strength reduction (DL = 9 mm) and shear displacement at the start of steady state (DU =300 mm).

From Fig. 10(a), we obtained mobilized friction angle at failure (Øm =400), steady state shear resistance (τss=73 kPa) for the sample with 55% fines. The shear stress and shear displacement relationship, Fig. 10(c) shows shear displacement at the start of the strength reduction (DL = 4 mm) and shear displacement at the start of steady state (DU=250 mm).

Table 3 contains the corresponding residual shear strength parameters for all test specimens in tabular form.

Table 3 Summery of undrained ring shear test results
Full size table

4 Discussion

Many factors influence the residual shear behavior of the soil. In this study, the effect of normal stress was observed by conducting several sets of undrained ring shear tests. Most residual failure envelopes were reasonably linear for the range of normal stresses used in this study. Based on the test results, the residual failure envelopes often show cohesion values equal to zero or small values of cohesion as presented in Fig. 11. In the lower normal stress conditions, residual strength envelopes give small values of cohesion, with a maximum value of c equal to 4 kPa. A constant value of residual friction angle with zero cohesion was achieved at high normal stresses.

Fig. 11
A multiline graph of shear stress versus effective stress in kilopascals. The plotlines incline and then decline. The maximum effective stress is 500 kilopascals with shear stress at 0. The lowest effective normal stress is at 100 with shear stress at 0. The highest shear stress is 220. Approximated values.

Results of stress path from undrained ring shear test under different normal stresses on silty sand with 26% fines

Full size image

The effect of fine content on the residual shear behavior was also studied and when comparing the stress path of the four different soil specimens, it shows that fine content have some effect on the shear behavior as shown in Fig. 12(a). The effective stress paths obtained from the undrained ring shear tests shows the mobilized friction angle at failure decreases with the increase of the fine content. Figure 12(b) shows the shear strength tends to decreases as the fine content increases. The soil specimen with less fines, tend to show greater peak strengths than the specimens with more fines. When subjected to large shear displacements the excess pore water pressure developed within the shear zone increases with the increase of fine content as shown in Fig. 12(c).

Fig. 12
3 multiline graphs. a, shear stress versus effective normal stress with F42 the highest at (290, 220). It declines to (510, 0). b, shear stress versus shear displacement has F 26 with the highest shear stress of 125 at shear displacement 10000 millimeters. c, pore water pressure versus shear displacement with inclining trend.

Results of undrained ring shear tests on soil samples of different fine contents (26%, 42%, 47% and 55%) (a) Effective stress paths; (b) Shear resistance versus shear displacement; and (c) pore-water pressure versus shear displacement

Full size image

As shown in Fig. 12(a), each sample in this undrained ring shear test series undergoes some amount of reduction in the shear strength after the failure, and this reduction differs for different samples. Therefore, the brittleness index (Bishop 1967) is used to analyze the consequences of shear failure.

$$ IB=\frac{\tau p-\boldsymbol{\uptau} \mathbf{s}}{\tau s} $$
(1)

Where τp is peak shear strength (kPa) and τs is shear strength at steady state (kPa)

The brittleness index is commonly used to evaluate the strain-softening behavior of soil. Brittle material often observed the highest strength loss in first time failure. The test results showed that the brittleness index increases with the increase of fine content and decreases with the increase of effective normal stress (Fig. 13).

Fig. 13
2 scatterplots. a plots brittleness index versus fine content percentage. The trend is inclining with the lowest data point at (26, 0.45) and highest at (47, 0.77). b plots brittleness index versus effective normal stress. The trend is declining with (200, 0.48) the lowest and (400, 0.65) the highest. Approximated values.

Brittleness index relationships (a) Brittleness index versus fine content; (b) Brittleness index versus effective normal stress

Full size image

5 Conclusion

A Series of ring shear tests were conducted for the soil samples collected from the Athwelthota landslide area to study the shear behavior of landslide soil. The effects of the effective normal stress and fine content on the shear behavior in undrained monotonic loading conditions were examined based on the test results. The following conclusions were obtained.

  1. 1.

    Many of the landslide soil exhibited a small cohesion at low values of effective normal stress, but the mobilized friction angle at failure achieved a constant value at higher normal stresses.

  2. 2.

    With the increase of the fine content, the mobilized friction angle at failure decreases.

  3. 3.

    When the fine content increased high excess pore water pressures were developed within the shear zone when subjected to large shear displacement.

  4. 4.

    With the increase of fine content, the brittleness index is increased and it decreases with the increase of effective normal stress.