Development of a Soil-Specific Liquefaction Assessment Framework: Role of Soil Gradation

点击翻译按钮获取中文摘要

Assessment of liquefaction susceptibility and its consequences is a major part of geotechnical earthquake engineering. The practice relies on a case-history-based relation between in-situ penetration tests and cyclic response of soils. Apart from a poorly understood fines content correction, soils with different compositions and gradations and widely varied mechanical properties are essentially viewed as uniformly graded quartz sand. The existing methods would not have predicted liquefaction in almost any of the sites where liquefaction of widely graded gravelly soils resulted in widespread damage during the 2008 Wenchuan, China earthquake. Soil gradation has a first order effect on dilatancy and compressibility, as well as particle breakage; factors that control the penetration resistance and cyclic response of soils in different ways. Previous research suggests that wider gradation has a marginal effect on cyclic resistance, but imposes a more pronounced increase on penetration resistance. Consequently, widely graded soils with similar penetration resistance to uniform sands can have a significantly weaker cyclic response. The short-term goal of the research program is to develop a mechanistic framework that explains the general effects of gradation on the cyclic response of cohesionless soils and the penetration resistance used for its in-situ evaluation. The work will provide a scientific explanation of the existing fines content correction, and expand the concept to incorporate all widely graded soils including those with significant gravel content. The long-term goal is developing practical, well-validated, soil-specific liquefaction criteria that take into account independently measurable soil properties. Four specific components aim at (i) quantifying the influence of gradation on dilatancy, compressibility, particle breakage, and cyclic response, (ii) quantifying the influence of field vs. laboratory soil fabric on behaviour, (iii) modelling the Cone Penetration Test and investigating the influence of gradation on its measurements and (iv) developing general gradation corrections for the case-history-based method and validation. By integrating the influence of soil properties into the current liquefaction assessment practice, the research will reduce the significant uncertainty in dealing with soils that deviate from the idealised uniform sand case. Large economic gains can be realised where current conservatism prompts unnecessary remedial work, and risk will be reduced by identifying unrecognised liquefaction hazard. The research program will create an interactive environment where undergraduate, master's and PhD students performing advanced laboratory tests and innovative numerical analyses will collaborate with each other, and with leading national and international experts from both industry and academia.