Physical Modeling of Soil Liquefaction: Repeatability of Centrifuge Experimentation at RPI

Abstract

Numerical models are commonly used in practice to predict the occurrence and propagation of liquefaction and lateral spreading. Nevertheless, there are still only a few good-quality data sets to validate these models. An international collaboration between physical and numerical modelers has been initiated through the Liquefaction Experiments and Analysis Projects (LEAP) to establish a standard method for validation of numerical models. One of the main goals of LEAP is to provide numerical modelers with high-quality experimental data that can be used in a series of prediction exercises. In this framework, a small-scale centrifuge test of a sloping deposit was conducted at Rensselaer Polytechnic Institute (RPI) in January 2015 and provided comprehensive data sets for evaluation of soil liquefaction numerical tools. The same experiment was repeated in September 2015 to assess the robustness and repeatability of the physical modeling methods at RPI. This paper serves a double objective: (1) to study the response and liquefaction potential of sloped deposits experimentally and (2) to establish the repeatability of experimental data to be used in numerical simulations.