Toward Model-Based Simulation of Earthquake-Induced Soil Liquefaction Effects

将开发一个新的框架，用于可靠地计算模拟地震引起的土壤液化对土壤和土壤-结构系统的影响。这是一个具有重大经济意义的重要问题，由于其复杂性，在很大程度上仍未得到解决。饱和水砂和其他液化的土壤的特征是它们接触处的颗粒相互作用，以及与孔隙水和结构表面的相互作用。孔隙水随着压力的变化而变化，在地震期间和地震之后在土体内部和外部流动。当沙子液化时，颗粒分离，物理性质发生变化。其他物理变化包括土壤和土壤-结构边界处水夹层的形成、非层流的实例、剪切带的形成、土壤和土壤-结构边界处的分离和其他局部现象。在目前最先进的技术中，典型的均质本构关系被用于biot型相耦合公式中。提出的新模拟框架将是多物理场、多尺度和自适应的，土壤采用两个耦合尺度。其中一个尺度将保持均匀化，而第二个尺度将明确考虑关键区域（如剪切带和边界层）的颗粒尺度。调查将包括强大的实验、数学和计算组件，通过系统使用系统识别技术、自适应分析技术、反馈回路以及计算和物理模拟的比较可视化联系起来，并在一般框架内综合。数学和计算组件将包括均质化和微观尺度模型开发和模拟所需的微力学研究，包括孔隙水的处理及其与土壤颗粒的相互作用。本研究将基于地震诱发土壤液化及其影响研究的第一性原理，启动可操作的自适应多尺度多物理场计算模拟框架的开发。

A new framework will be developed for reliable computational simulations of earthquake-induced soil liquefaction effects on soil and soil-structure systems. This is an important problem of great economic implications, which remains largely unsolved due to its complexity. Water-saturated sands and other soils that liquefy are characterized by particle interactions at their contacts as well as with pore water and with structural surfaces. The pore water changes in pressure and moves within the soil mass and out of it during and after earthquake shaking. When the sand liquefies the particles separate and the physics change. Other changes of physics include the formation of water interlayers at soil and soil-structure boundaries, instances of nonlaminar flow, and formation of shear bands, separation and other localized phenomena in the soil and at soil-structure boundaries. In the current state-of-the-art, typically homogenized constitutive relations are used in a Biot-type phase-coupled formulation. The proposed new simulation framework will be multiphysics, multiscale and adaptive, with two coupled scales used for the soil. One of these scales will remain homogenized, while the second scale will explicitly consider the particle scale in the critical regions, such as shear bands and boundary layers. The investigation will include strong experimental, mathematical and computational components, linked by a systematic use of system identification techniques, adaptive analysis techniques, and feedback loops as well as comparative visualizations of computational and physical simulations, and synthesized within a general framework. Both mathematical and computational components will include micromechanical studies needed for both homogenized and micro-scale model development and simulation, including treatment of pore water and its interaction with the soil particles. This study will initiate the development of an operational adaptive multiscale multiphysics computational simulation framework based on first principles for the study of earthquake-induced soil liquefaction and its effects.