Multidimensional constitutive and numerical modeling in geotechnical seismic analysis and design

点击翻译按钮获取中文摘要

The future of modeling in geotechnical earthquake engineering will be based almost exclusively on Performance-Based Seismic Design philosophy. PBSD requires a paradigm shift toward a displacement or deformation-based design as opposed to the traditional force-based design. It is more rational than the conventional analysis and design methods, and leads to more cost effective construction or safer design. PBSD demands an adequate measure of reliability of computed deformations. A realistic simulation of deformation of geo-structures subjected to irregular earthquake loading is very complex. The simulation tools should handle 3D dynamic analysis based on effective stress formulation, fully coupled solid-pore fluid interaction, and parallel/cloud computing features for handling large problems. The most important issue is proper constitutive modeling of the soil matrix. Current soil models used in engineering practice do not include many of the very important factors that affect the stress-strain relationship. It is also extremely challenging to address multidirectional cyclic loading of earthquakes. From discussion made with major practitioners, such as BC Hydro who deal with earth dams, it becomes clear that their main concern is with the different responses given by the different models that are available; in some cases these differences are significant enough to cause great concern. ***The proposed research will address several challenging aspects of constitutive and numerical modeling of sand, silt, and clay under complex multidirectional loading. A series of related laboratory element test data will be collected and studied in detail. The information will be used in development of several novel constitutive modeling features for multidirectional cyclic loading of sands, silts and clay. The highlight would be on development of the SANISAND-Z model with zero elastic range, refined inherent and evolving fabric for multidirectional loading, and provisions for accumulation of large post-liquefaction shear deformation. Also the important feature of rate-dependency will be added to the recently developed SANICLAY-B model. The developed models will be robustly implemented in FLAC3D, OPENSEES and/or DIANA. The models and developed tools will be extensively verified and validated using element test and centrifuge test results. On the application side among many relevant problems focus will be given to the very challenging and important problems of (i) multidimensional seismic site response analysis, and (ii) seismic deformation analysis of earth dams, which gives the loss of freeboard in extreme events, and also study of the efficiency of remedial measures for the dam stability. The developed modeling tools and simulations lead to revisiting conventional seismic analysis and design with PBSD, which is of significant relevance to the reduction of seismic risk in Canada.**