Landslide Stability Analysis Utilizing Shear Strength of Slip Surface Soil: Asato and Tyunjun Landslides, Okinawa, Japan

Size: px

Start display at page:

Download "Landslide Stability Analysis Utilizing Shear Strength of Slip Surface Soil: Asato and Tyunjun Landslides, Okinawa, Japan"

Ethelbert Mason
5 years ago
Views:

1 1246 Landslide Stability nalysis Utilizing Shear Strength of Slip Surface Soil: sato and Tyunjun Landslides, Okinawa, Japan Sho Kimura 1, SeiichiGibo 2, Shinya Nakamura 3, Shriwantha Buddhi Vithana 1 1 Graduate Student, The United Graduate School of gric. Sci., Kagoshima University 2 Professor, Faculty of griculture, University of the Ryukyus 3 ssociate Professor, Faculty of griculture, University of the Ryukyus, Senbaru 1, Nishihara, Okinawa, 30213, Japan; ; s-naka@agr.u-ryukyu.ac.jp BSTRCT In the stability analysis of sato and Tyunjun landslides, the average shear strength acting along the slip surface has been calculated utilizing results of laboratory shear tests of the slip surface soils. For the sato landslide, which is a large-movement along the whole slip surface, it is considered that most of the slip surface zone has been lying at residual state and the remaining parts at fully softened state. verage shear strength parameter, (c, ), acting along the slip surface has been determined as c = 0 kn/m 2 and = Thus, the average shear strength was close to the residual strength, which is reflected in the well-defined slickenside on the slip surface. For the Tyunjun landslide, the slip surface consists of a fractured-mudstone zone in the toe part and slickenside in the upper/middle part. In the stability analysis, it is considered that in the upper/middle part of the slip surface residual strength has been mobilized, while in the toe part peak strength of the fractured-mudstone has been utilized. Thus, the calculated average cohesion for entire slip surface is greater than zero. INTRODUCTION Landslides have several contributory causes, and among them, nature of soil is important for a landslide and the shear strength of the soil at the slip surface is directly related to sliding. For effective control works for landslides, what is important in the safety evaluation of landslides is to understand shear strengths of the soil and rock along the slip surface and decide an appropriate average shear strength parameter acting along the slip surface (average cohesion c, average angle of shearing resistance ). However, in stability analyses with the

2 1247 application of peak, fully softened or residual strengths obtained from laboratory shear tests, the calculated factor of safety tends to be overestimated or underestimated (Skempton, 1964; Skempton, 1985; Mizuno, 19). This paper discusses the shear strengths along the slip surfaces of sato and Tyunjun landslides. In the stability analysis utilizing shear strength, the average shear strength acting along the slip surface is calculated through laboratory shear tests of the soil samples of the two landslides, which are located in the area of Shimajiri-mudstone in Okinawa, Japan. CONSPECTUS OF LNDSLIDES Shimajiri mudstone The Neogene Shimajiri-mudstone is widely distributed in the central and southern areas of the Okinawa Island and along the marine terrace in the North Eastern part of Miyako Island (Kizaki and Takayasu, 1976) and Shimajiri-mudstone formation consists of Tomigusuku, Yonabaru and Shinzato geological strata (Sunagawa and Uehara, 1983). The Tomigusuku stratum is the alternation of strata with a majority of sand rock and mudstone. Yonabaru stratum constitutes mostly of Simajiri-mudstone, which is exposed in the ground and Shinzato stratum is similar to Yonabaru stratum and slightly distinguishable by the tuff and sand rock stratum of it. The geological structure of Shimajiri-mudstone is controlled by the uplift, where the strike is in a general NE-SW direction and dips in a NE direction with a slight flexure. The fault system of Shimajiri-mudstone, which was uplifted by Nakagusuku dome are in the directions of NW-SE, NE-SW and E-W. Landslides abruptly occur in those areas controlled by the landform and geological structure of the mudstones (Gibo et al, 1986; Sasaki et al., 19; Nakamura et al., 2004). sato landslide The landslide that occurred on June 10, 2006 was of fairly large proportions and the trigger for the slide was due to the rise of soil pore water pressures, which was directly related to the antecedent rainfall of 509 mm (May 12 - June 10) and the 86 mm per day (June 10). t the time of the first-time activation, the slope inclination of the upper part from the main scarp of old landslides is about 23 degrees and in the lower part with the old landslides is approximately 12 degrees. The type of movement of the landslide varied with time (Gibo et al., 2006). The crown cracks of the landslide head were first noted in the morning of June 10 and the prefectural road 35 on the slope breast was uplifted during the same day afternoon. t

3 :00 on June 10, the slope began to slide, while the prefectural road was totally disrupted by the sliding. The width of the main slide body that occurred on June 10 was about 120 m, the length being approximately 2 m and the main scarp height being about 30 m (Figure 1). The amount of movement of head length, slope breast and toe were 68 m, 77 m and 77 m, respectively, which indicates that the movement was relatively large. This large movement was partly related to the old landslides that existed in the lower part of the slide mass. In addition, two secondary landslides in the right wing of head occurred on the nights of June 12 and June 13 and the main body, which had been dormant, was reactivated by these landslides. The final dimensions of the landslide were approximately 260 m and about 500 m in width and length, respectively. Bv-4 Bv-3 Bv Bv-5 Bv Bv-2 Bv-7 Bv Bv Bv Bv Bv Bv-15 Bv Bv N 30 Main scarp m Main slide body Moving body Figure 1. Plan of the sato landslide

4 1249 Tyunjun landslide Tyunjun landslide is located in the Kitanakagusuku-village in the central area of Okinawa Island. It is determined that the landslide was triggered by a rapid increase in pore water pressure from the antecedent rainfall of 586 mm (September 6 October 5) and the 160 mm per day (October 5). Inclinations of slope at the time of first-time activation were 30 degrees in the upper part and 15 degrees in the lower part, and decreased to 12 degrees after sliding. It is clear that an old landslide also existed in the slope breast of this landslide. Since the slip surface of the slope breast, which is 20.8 m, was deep, the slip surface of the previous slide did not transform into being a part of the slip surface of the succeeding slide. Sliding increased on October 5 and reached a total cumulative distance of 10 m in the head, 15 m in the breast and 3 m in the toe. The damages caused by the movement included the following; total destruction of one house, partial destruction of two houses and m of roadway and near complete destruction of a park facility. The toe of the slide reached the community center and partially damaged the building. Fortunately, no one was injured (Nakamura et al., 2004). The width of the slide block that occurred on October 5 was about 120 m, the length being approximately 155 m and the main scarp height being about 15 m (Figure 2). 4 Boreholes 1 19 Village road Village road Community center N Main scarp Social center 0 20m Moving body Figure 2. Plan of the Tyunjun landslide ddition to Nakamura et al., 2004

5 1250 STBILITY NLYSIS UTILIZING SHER STRENGTHS Experimental procedures Soil core samples were taken using a rotary core and were subjected to a triaxial test in an undisturbed condition. The disturbed soils were passed through a 425 µm sieve and subjected to physical and mineralogical analyses. The reconstituted samples of the < 425 µm soil fraction were subjected to a ring shear test in the ring-shear testing apparatus designed by one of the authors (Gibo, 1994). The shear strengths of undisturbed soil were measured by a saturated-consolidated-undrained compression test with pore water pressure measurements in a triaxial apparatus (CU) (The Japanese geotechnical society, 2000). In the determination of residual strength, the slip surface soil was placed in a reconstituted condition as demonstrated by Bishop et al. (1971) following the concept that the residual strength was unaffected by the initial structure of the soil. The < 425 µm soil samples were packed into a shear box with and 60 mm outer and inner diameters, respectively, then consolidated at different normal stresses. To achieve the full dissipation of excess pore water pressures, the shear rate of 0.01 mm/min was chosen for displacement before the fully softened state was attained and also in the residual stage of shear (Bishop et al., 1971; Gibo, 1979). The fully softened strength was reached at an early stage of shearing, and large displacements were required for the strength to drop to the residual value. sato Landslide For the sato sample, the peak strength parameters of cohesion intercept (c f ) and shear angle ( f ) obtained in the CU test were kn/m 2 and , respectively, where the difference of c f is larger than the difference of f. The fully softened and residual stages were estimated by Coulomb s law. Cohesion, c, was determined by the method described by Skempton (1964, 1985). The parameters of fully softened strength ( sf ) and residual strength ( r ) were estimated to be 27.8 and 10.8, respectively. Figure 3 shows the cross section and the result of stability analysis at the time of recession of the landslide. The groundwater level was measured as the maximum water level and factor of safety was assumed to be The c tan relationship established by the back calculation method is illustrated in Figure 3. It is reasonable to apply different strength parameters in the stability analysis, based on the conditions of the slip surface, which were investigated through bore samples. In this landslide, which was a large movement along the entire slip surface, slickensided and fully softened clay were found in the middle/upper and

6 1251 toe parts of the slip surface zone, respectively. The residual strength r = 10.8 was mobilized in the middle/upper area and the fully softened strength sf = 27.8 was mobilized in the toe area. The stability analysis was performed using the method in which the residual factor is incorporated (Gibo et al., 1986; Gibo, 1996). The average shear strength parameter (c, ) acting along the slip surface and the residual factor R were determined as 0 kn/m 2, 12.3 and 0.92, respectively. The 0.92 suggests that most of the slip surface zone is lying at the residual state and the remaining area at the fully softened state. The 12.3 is close to the residual strength of 10.8, and it is reflected by the well-defined slickenside on the slip surface. s a result, both the residual and fully softened strengths mobilized in the slide could be expressed clearly for the whole slip surface zone, which is represented as line a-a in the cross section shown in Figure 3. ccording to the division of the slip surface zone into different shear strengths mobilized along the slip surface, the following results were expected; there is a possibility of effective piling, and, the evaluation of effective of subsurface drainage works become more exact. Bv-1 Bv-11 Bv-10 Bv-9 Bv-13 Bv-12 a = Ground water 0 50(m) a Slip surface c (kn/m 2 ) c = tan (Fs = 1.00) Residual strength (tan10.8 0) B Fully softened strength (tan27.8 0) IP verage shear strength (tan12.3 0) IP B tan Figure 3. The cross section and c - tan for stability analysis at the time of recession of sato landslide

7 1252 Tyunjun Landslide For the Tyunjun sample, the peak strength parameters were c f = kn/m 2 and f = The fully softened and residual strength parameters were sf = 29.5 and r = 9.3, respectively. The cross section and the result of stability analysis at the time of recession of the landslide are shown in Figure 4. The middle/upper parts of the slip surface zone were a large movement and was found to be a well-defined slickenside. In the toe area, the movement was small and the ground surface was uplifted and the slip surface was in fractured condition. n estimated residual strength of r = 9.3 was mobilized in the middle/upper area and the peak strength of the fractured-mudstone c f = 50.0 kn/m 2, f = 35.5 were mobilized in the toe area. The stability analysis was performed by using the method in which the residual factor is incorporated. The c, and R were determined as 6.7 kn/m 2, 13.3 and 0.87, respectively. The 0.87 suggests that most of the slip surface zone is lying at the residual state and the Ground water Bv-4 Bv-1 a Bv-7 Bv-2 = = Bv-3 a Slip surface 0 50(m) B Residual strength (tan9.3 0) B Fractured peak strength (tan ) IP verage shear strength (tan ) c (kn/m 2 ) c = tan Fs = IP tan Figure 4. The cross section and c - tan for stability analysis at the time of recession of Tyunjun landslide

8 1253 remaining area at the fractured-mudstone strength. ccordingly, both the residual and peak strengths mobilized at the slide could be expressed clearly for the whole slip surface area, which was represented as line a-a in the cross section shown in Figure 4. CONCLUSION In the stability analysis of sato and Tyunjun landslide, the average shear strength parameter acting along the slip surface has been calculated using the results of laboratory shear tests of the slip surface soils. For the sato landslide, which is a large-movement along the whole slip surface, it is considered that the most of the slip surface area has been at the residual state and the remaining area at the fully softened state. verage shear strength acting along the slip surface has been determined as c = 0 kn/m 2, = 12.3 and as being close to the residual strength, which is reflected in the well-defined slickenside on the slip surface. For the Tyunjun landslide, in the stability analysis, it is considered that in the upper/middle part of the slip surface residual strength has been mobilized, while in the toe part peak strength of the fractured-mudstone has been utilized. Thus, the calculated average cohesion for entire slip surface is greater than zero. REFERENCES Bishop,.W., Green, G.E., Garga, V.K., nderson,., Brown, J.D. (1971). " new ring shear apparatus and its application to the measurement of residual." Géotechnique, 21(4): Gibo, S. (1979). "Sudies on determination of residual strength in clay." PhD thesis, Kyushu University (in Japanese with English abstract). Gibo, S. (1994). "Ring shear apparatus for measuring residual strength and its measurement accuracy." the Japan. Landslide Soc. J. 31(3): (in Japanese with English abstract). Gibo, S. (1996). "Stability analysis method in which the residual factor is incorporated pplication to the landslide in the Shimajiri mudstone of Okinawa-." the Japan. Landslide Soc. J. 31(3): (in Japanese with English abstract). Gibo, S., Sasaki, K., Yoshizawa, M., Ida, S. (1986). "Control of geological structure and parameters of average shear strength along the slip surface in a landslide in the

9 1254 neogene mudstone in Okinawa." the Japan Landslide Soc. J. 23(3): (in Japanese with English abstract). Gibo, S., Sasaki, K., Zhou, Y., Nakamura, S. (2006). "Kitauebaru landslide caoused by continual rainfall in Nakagusukusu Village, Okinawa Prefecrure, on June 10, 2006." the Japan. Landslide Soc. J. 43(2): (in Japanese). The Japanese Geotechnical Society (2000). "Explanation and method of the soil test." The Japanese Geotechnical Society: (in Japanese). Kizaki, K., Takayasu, K. (1976). "Outline of geohistory of the Ryukyu rc." Marine Science (in Japanese with English abstract). Mizuno, K. (19). "The rapid-type landslide in mudstone of the Teradomari and Shiiya formations in Niigata prefecture, Japan consideration on shear strength of landslide mass-." Japanese Geomorphological Union. J. 11(1): 29- (in Japanese with English abstract). Nakamura, S., Gibo, S., Hayashi, Y. (2004). "Tree-dimensional stability analysis." Trans. of JSIDRE, 229: (in Japanese with English abstract). Sasaki, K., Yoshizawa, M., Gibo, S., Egashira, K. (19). "Landslides in neogene Shimajiri mudstone in Okinawa -Geological background-." the Japan. Landslide Soc. J. 27(2): (in Japanese with English abstract). Skempton,. W. (1964). "Long-term stability of clay slopes." Géotechnique, 14(2): Skempton,. W. (1985). "Residual strength of clays in landslides, folded strata and the laboratory." Géotechnique, 35(1): Sunagawa, T., Uehara, H. (1983). "Engineering properties of the Shimajiri group Yonabaru formation." The Japanese Geotechnical Society, 31(4): (in Japanese).

FINITE ELEMNT ANALYSIS FOR EVALUATION OF SLOPE STABILITY INDUCED BY CUTTING

FINITE ELEMNT ANALYSIS FOR EVALUATION OF SLOPE STABILITY INDUCED BY CUTTING Toshinori SAKAI Department of Environmental Science and Technology, Mie University, Tsu, Japan Tadatsugu TANAKA Graduate School