Characterizing granular material constitutive behavior using SelfSim with boundary load-displacement measurements

摘要

SelfSim, self-learning simulations, is a novel method that extracts the material constitutive behavior using boundary measurements of load and displacement. The method allows the numerical model to continuously learn from laboratory and field observations, and to characterize any material constitutive behavior. SelfSim has been extensively verified and applied to various problems using simulated and physical boundary measurements. The applications include: structural tests on metals, bridges, deep excavations and laboratory tests on granular material. This paper presents a brief overview of SelfSim method and its applications on the characterization of granular materials. For extra-terrestrial (lunar and Martian) environments it is possible to apply SelfSim in novel ways to extract relevant regolith constitutive properties using measurements of forces and displacements on instrumented rovers or other surface equipments. Specially designed probes can also be utilized to extract constitutive behavior at depth. The paper will describe conceptually how such probes and measurements can be made without the need to bring back samples for terrestrial laboratory testing.