1. MODELING OF SEISMIC SLOPE BEHAVIOR WITH SHAKING TABLE TEST Meei-Ling Lin and Kuo-Lung Wang Department of Civil Engineering, National Taiwan University The Next Generation of Research on Earthquake-induced Landslides An International Conference in Commemoration of 10th Anniversary of the Chi-Chi Earthquake, 2009

2. Outline Shaking table test Specimen preparation and law of similarity Test result Particle Image Velocimetry (PIV) analysis Displacement behavior Summary

3. Objectives The initiation of landslide and the development of slip surface for landslides induced by earthquakes Run-out distance and slope recession caused by landslide Identification of affected area of potentially unstable slope

4. 4 Model Slope Shaking Table Test System calibration m: 39838 kg k: 551.9kN/m c: 174.3 kN.sec/m Insignificant amplification were found from accelerometer and LVDT measurements

5. Law of Similitude Assumptions (Iai, 1989) –soil skeleton is regarded as continuous medium –deformation is assumed to be small so that the equilibrium equation remains the same before and after the deformation –the strain of the soil skeleton is small 8.9 Hz of loading frequency was applied for a scale factor of 20 based on 1-g, equivalent density, and strain conditions Meymand(1998)

6. Model Granular Slope Shaking Table Test Specimen Preparation Sand –Uniform, medium, poorly graded sand (SP, Unified Soil Classification System) Preparation method –Compaction –Dry pluviation Pluviation equipment attached on the model box Sieve Sampling container

7. 7 Study on boundary effects 1:1:1 2:1:2 3:1:3 30 cm base thickness is chosen after analysis

8. 8 Specimen preparation - compaction slope modeling tool

9. 9 Instrumentation and loading sequence First loading sequence

10. 10 input amplitude from 0.4g to 0.5g input amplitude from 0.5g to 0.6g

11. 11 input acceleration amplitude of 0.6g

12. 12 linear non linear

13. Specimen Preparation - Pluviation 30cm 50cm 30° 100 cm 253.4 cm SpecimenUnit weight (kN/m 3 ) Preparation Method Boundary Treatment A15.3PulviationGrease B15.8PulviationGrease w/ plastic sheet C15.8PulviationGrease w/ plastic sheet D15.5Excavation after Pulviation Grease w/ plastic sheet E15.5Excavation after Pulviation Grease w/ plastic sheet

14. Instrumentation and Measurement Reference frame Shaking table Accelerometers Acceleration by accelerometers Image video recordings, particle displacement, particle velocity Mapping of slip and deposit surface, mapping of run-out and recessional line Video camera (CCD)

15. Loading Sequences Specimen Maximum Loading (H, V) in gLoading History A1 st - (0.35, 0.00) 2 nd - (0.58, 0.00) B1 st - (0.28, 0.00) 2 nd - (0.43, 0.00) C1 st - (0.34, 0.08) 2 nd - (0.41, 0.20) D1 st - (0.26, 0.10) 2 nd - (0.32, 0.19) E1 st - (0.38, 0.12) 2 nd - (0.38, 0.12)

16. 16 Specimen B

17. 17 Instrumentation and loading sequence First loading sequence

18. 18 possible sliding

19. 19 Specimen C

20. Instrumentation and loading sequence First loading sequence

21. 0.3g

22. 22 Vertical acceleration response

23. 23 slipping with separation

24. Test Result – the Initiation of Slip Surface change detection via particle image velocimetry (PIV) –Particle moving direction and magnitude –The initiation of slope surface slip Identification of slip initiation from acceleration history –The initiation of subsurface slope slip Specimen C crest

25. Test Result – Video Recording and PIV Analysis Processed with PIVview2C Note: play video files Crest

26. Test Result – the Initiation of Slip SpecimenLoading amplitude (g) from PIV Loading amplitude (g)/Time (sec) from acceleration history A0.110.35g, 11sec BN/A0.28g, 4.04sec C0.090.32g, 22.04sec D0.150.24g, 25.2sec EN/A0.24g, 9.85sec Initiation of surface slip Initiation of subsurface slip

27. Runt-out and Recessional Distances Specimen A Specimen B Specimen C CrestToe unit: cm

28. Run-out and Recessional Distances Specimen D Specimen E CrestToe unit: cm

29. Relationship Between Distances and Loadings SpecimenUnit weight (kN/m 3 ) Maximum Loading Amplitude (H, V) in g Loading Period (sec) Crest Displacement Max/Min (cm) Toe Displacement Max/Min (cm) A15.31 st - (0.35, 0.00) 2 nd - (0.58, 0.00) 1 st - 21 sec 2 nd - 11 sec 43.5/22.128.6/19.1 B15.81 st - (0.28, 0.00) 2 nd - (0.43, 0.00) 1 st - 14 sec 2 nd - 14 sec 12.7/8.618.5/10.3 C15.81 st - (0.34, 0.08) 2 nd - (0.41, 0.20) 1 st - 32 sec 2 nd - 32 sec 26.7/21.845.6/22.6 D15.51 st - (0.26, 0.10) 2 nd - (0.32, 0.19) 1 st - 32 sec 2 nd - 32 sec 12.6/8.017.2/6.3 E15.51 st - (0.38, 0.12) 2 nd - (0.38, 0.12) 1 st - 16 sec 2 nd - 16 sec 11.4/7.715.8/5.7 Values after prolong loading sequence

30. Comparing specimens (A, B) and (C, D) Recessional and run-out distance increased with increasing maximum loading amplitude The set of data with larger displacement were subjected to higher vertical loading coupled with higher horizontal loading Comparing specimens (B, C) Recessional and run-out distance increased with additional vertical loading amplitude Comparing specimens (D, E) and (C, E) Higher vertical loading resulted in higher recessional and run-out distances Maximum distances Minimum distances

31. The Relationship Between Crest Recession and Toe Run-out versus Loading Amplitude Maximum distances Minimum distances

32. Potentially Affected Zone Restricted exploitation area Sand, gravel  The 0.58g loading amplitude of specimen A resulted in highest recessional distance  The vertical acceleration for specimen C could result in larger run-out and recessional distances at the crest ToeCrest Slope height, H=50 cm 1/2H H H A B C D E

33. Newmark’s analysis – specimen A K hy =0.29 Jibson et al. (2000)

34. Input acceleration history – specimen 2

35. Exceeding acceleration and integrated velocity

36. Integrated displacement 0.532 cm 10.64 cm for prototype 10.64 cm for prototype

37. 37 Parameters and boundary conditions

38. Numerical Analysis of Slope Responses Hardin & Drnevich (1972) Assimaki et al. (2000) Shear wave velocity

39. Amplification factor from base to the crest Measured: 1.1

40. Modulus adjustment – amplification factor at 0.4g Variations of amplification factor with degradation of shear modulus under amplitude of 0.4g

41. Amplification factors

42. Summary The initiations of slope slip took place from the surface of slope, and then the subsurface slip initiated with increasing loading amplitude. Larger loading amplitude resulted in larger recessional and run-out distances. Additional vertical loading amplitude resulted in larger recessional and run-out distances than without vertical loading condition. The maximum recessional and run-out distances could reach as far as the height of slope. The initiation of slips measured from PIV analysis are smaller than those measured from accelerometers buried in the specimens. The potentially affected zone caused by earthquake can be estimated using loading amplitude versus normalized slope height regression equation.

43. Thank you for your attention