| dc.description.abstract | The Hungtsaiping (HTP) landslide is a gigantic, deep-seated landslide that was triggered by 1999 Chi-Chi earthquake. In previous studies, researchers used several methods to understand the sliding mechanism and to evaluate the stability of this area. However, the most important factors of the soil slope stability analysis are strength parameters, environment conditions, and the selection of the profile to be analyzed. The objective of this study is to understand the velocity-dependent frictional characteristics through the low-to-medium shear rate (2.1×10-7 – 1.3×10-2 m/s) rotary shear tests and the kinematical characteristics during Chi-Chi earthquake by Newmark displacement analysis for circular sliding surface. Discussing the reason on the movement of 20-30 m landslide at HTP landslide area during Chi-Chi earthquake.
The rotary-shear experiment results can be divided into two part: (1) The single constant velocity rotary-shear experiments under normally-consolidated, immersed in water conditions show that frictional characteristics are velocity-neutral (friction coefficient not varied with shear velocity) at low slip rate (μ=0.25-0.27) when slip rate smaller than 2.1×10-5 m/s. When slip rate in the range 4.0×10-5 - 2.1×10-4 m/s, μ decrease rapidly (μ=0.11-0.13). When slip rate reach 1.2×10-3 m/s, μ even drop to very low value (μ=0.02-0.03). The velocity-stepping experiments indicate that when slip rate changed, the friction coefficients are velocity-independent, which with an average value of 0.20, which is comparable to the value of single constant shear velocity experiments under low slip rate. (2) The single constant velocity rotary-shear experiments under over-consolidated, natural water content conditions show that frictional coefficients are 0.27, 0.05 and 0.08 at slip rates are 2.1×10-5, 1.2×10-3 and 1.3×10-2 m/s, respectively. The friction coefficients of velocity-stepping experiment are largest in all of the shear experiments at velocity of 2.1×10-7 to 2.1×10-4 m/s (μ=0.36-0.41).
This study use STABL 5M to calculate the critical acceleration under different friction coefficient. The relation between critical acceleration and friction coefficient can express as Ac=0.63μ-0.23, which is used in Newmark displacement analysis. The analysis results have divided into two part: (1) If friction coefficient is 0.366 (the average μ of over-consolidated, natural water content velocity-stepping experiment), that slightly higher than the friction coefficient, 0.36 (the μ when F.S equal to 1.0 without earthquake occurrence), the Newmark displacement analysis yields a result that closest to the reality measured average horizontal displacement, 24.7 m. However, the maximum slip velocity calculated by Newmark displacement analysis during earthquake almost reach 1.0 m/s. At this velocity level, steady-state friction coefficient will weaken to about 0.1 in previous studies and this study. Once the friction coefficient weakens to 0.1, the HTP landslide won’t stop. (2) This study incorporated velocity-dependent friction law of Vajont landslide gouges (mineral composition similar with HTP landslide gouges) under room-humidity condition with Newmark displacement analysis. Three of four strong motion stations’ analysis results show that HTP area also have rapid, long distance landslide occurred. According to the aforementioned, we speculate that the resistance force contributed by the artificial structures at toe of HTP landslide area is the important reason to stop the rapid movement of HTP landslide. | en_US |