NEESR-SG: Experimental and Micromechanical Computational Study of Pile Foundations Subjected to Liquefaction-Induced Lateral Spreading

点击翻译按钮获取中文摘要

Abstract: Critical research will be conducted on the problem of pile foundations subjected to liquefaction-induced lateral spreading in sloping ground and near waterfronts. This is a major cause of earthquake damage with costs in the billions of dollars. A large knowledge gap has been identified in current engineering methods to evaluate bending response of pile foundations to the lateral loading applied by the liquefied sand (limit equilibrium (LE) and p-y methods). One source of uncertainty is the area over which the lateral pressure must be applied for pile groups, the sum of the areas of the piles alone, or also soil between the piles. The difference between these two assumptions can produce pile-bending moments varying by a factor of three. Another source of uncertainty is the importance of soil permeability, not included in current LE methods, but which full scale and centrifuge shaking tests indicate can produce a variation of bending moments by a factor of seven. Centrifuge and full-scale shaking table tests, as well as insitu testing and preliminary analyses, suggest a complicated response of the liquefied sand in the neighborhood of individual piles and pile groups. Included is stiffening of the liquefied soil due to excess negative pore water pressures caused by soil-pile interaction; water flowing from the free field with eventual expansion; and, softening of the liquefied soil near the foundation, in a complicated time-dependent manner. A fundamental and comprehensive approach, made possible by NEES, will be used to tackle the problem using a range of available experimental and analytical tools. The objectives of the research are to understand the phenomenon of sand liquefaction during lateral spreading near pile foundations; to solve the engineering questions of how to design pile foundations against lateral spreading both in simplified terms as well as in terms of providing a basic understanding for appropriate analytical platforms; and, to clarify the correct way to use centrifuge modeling in future research and engineering applications. This is done through an experimental research plan using the University at Buffalo 1g 2D shaking, 6m-tall laminar box container NEES facility and the RPI NEES centrifuge facility, including use of novel advanced sensors to measure soil accelerations, deformations and pressures during shaking; and micromechanical Discrete Element (DEM) numerical experiments and finite element (FE) analyses at RPI, UC San Diego and Tulane University; all integrated by an appropriate identification and analysis framework including system identification and visualization capabilities. The DEM numerical experiments, validated by the 1g and centrifuge physical tests, aim at providing for the first time detailed information at the level of individual soil grains and groups of grains, on the response of the liquefied soil close to pile foundations including strain localization, shear banding and flow around the piles, calculated from basic micromechanical principles. The DEM insights will be used to refine important unsolved aspects of the constitutive relations used in the FE analyses. A Year 3 Workshop with the participation of a broad range of engineers and researchers from the US, Japan (including Japanese collaborators) and other countries, will evaluate the results of the research. The project will exercise a new set of research tools and ways to integrate them. The Interactive Web Site of the project will be periodically updated with results and interpretations. Graduate and undergraduate involvement in research, with emphasis on the recruitment of under-represented students, term projects in capstone and similar courses, preparation of 1-2 Educational Modules at the end of the project, and specific efforts toward K-12 students and their teachers, are aimed at intensifying the impact of the project.