NUMERICAL STUDY OF SATURATED SANDY SOIL CYCLIC BEHAVIOR AS A RESULT OF CYCLIC SIMPLE SHEAR TEST

• Main Author: Rafidah, Ismi
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/49749
• Institution: Institut Teknologi Bandung

The behavior of sandy soil in receiving stress is based on its initial condition, namely relative density and initial stress condition. A saturated sandy soil that is subjected to cyclic loading is potential to fail, the mechanism called liquefaction. In this study, loose, medium, and dense sand will be studied through numerical studies using laboratory test result of cyclic simple shear (CSS) which are modelled through 2 constitutive models: fully-coupled (UBCSAND) and loosely-coupled (Finn-Byrne) effective stress in FLAC-2D (v.7). The input soil parameters are calibrated, so that the modelling results are in accordance with the laboratory test. Modelling also varied in the amount of frequency (0.25 Hz and 1.0 Hz) to see the effect on soil behaviour. Each model will show cyclic load in time domain and number of cycles, ru, ?????????,????, and modulus reduction. The results of numerical modeling are evaluated against results of the laboratory test to analyse the characteristics of the model. Two different models showed sufficient soil behaviour according to the laboratory test results with the use of proper calibration. The use of the stiffness factor, hfac1 in UBCSAND and hysteresis damping in the Finn-Byrne model has an important role in producing representative cyclic behaviour. Keyword: Liquefaction, cyclic behaviour, numerical analysis, cyclic simple shear, UBCSAND, Finn-ByrneThe behavior of sandy soil in receiving stress is based on its initial condition, namely relative density and initial stress condition. A saturated sandy soil that is subjected to cyclic loading is potential to fail, the mechanism called liquefaction. In this study, loose, medium, and dense sand will be studied through numerical studies using laboratory test result of cyclic simple shear (CSS) which are modelled through 2 constitutive models: fully-coupled (UBCSAND) and loosely-coupled (Finn-Byrne) effective stress in FLAC-2D (v.7). The input soil parameters are calibrated, so that the modelling results are in accordance with the laboratory test. Modelling also varied in the amount of frequency (0.25 Hz and 1.0 Hz) to see the effect on soil behaviour. Each model will show cyclic load in time domain and number of cycles, ru, ?????????,????, and modulus reduction. The results of numerical modeling are evaluated against results of the laboratory test to analyse the characteristics of the model. Two different models showed sufficient soil behaviour according to the laboratory test results with the use of proper calibration. The use of the stiffness factor, hfac1 in UBCSAND and hysteresis damping in the Finn-Byrne model has an important role in producing representative cyclic behaviour.