Evaluation of synthetic and recorded ground motions for soil liquefaction analyses

Abstract

In performance-based earthquake engineering, simulated earthquake ground motions are often generated for time history analyses of structures and geotechnical systems. Liquefaction refers to dramatic loss of stiffness and strength of soils during earthquake shaking. As a complicated nonlinear process associated with accumulation of excessive pore pressure and decrease of effective stress due to soil-fluid interaction, whether synthetic motions can be appropriately used in liquefaction assessment still remains a grand challenge. In this study, various synthetic ground motions are generated using different simulation techniques, whose performances are compared with real recorded motions through nonlinear time history analyses of liquefiable grounds with different soil densities and ground inclinations. One-to-one comparison of results in this study clearly demonstrates the importance of preserving accumulative intensity measures of synthetic motions in liquefaction assessment. Strong linear relationships are also found in a log-log space between the maximum ground displacement and evolutionary Arias intensity during post-liquefaction stage for different liquefiable grounds, and the predictive relationships seem not to be affected by the types of synthetic ground motions.