An inverse algorithm for the identification and the sensitivity analysis of the parameters governing micropolar elasto-plastic granular material

Abstract

The article concerns the complex determination process of the material parameters governing micropolar granular material with elasto-plastic material properties. Proceeding from a gradient-based method, we split the total set of parameters and the overall identification procedure into two major categories. These are, firstly, the identification of the parameters of a standard non-polar elasto-plastic continuum, and, secondly, the determination of the remaining parameters governing the micropolar part of the constitutive model. While the first set of parameters can be obtained from homogeneous triaxial tests on, e. g., granular, cohesive-frictional materials like sand, the second set can only be determined from inhomogeneous tests, such as biaxial tests including the onset and the development of shear bands. Following this, one can obtain the first part of the identification process from a simple inverse algorithm applied to the elasto-plastic material model of non-polar solids, while the second part requires a fully inverse computation in the sense of a back analysis of the underlying boundary-value problem. In the present article, this procedure is carried out by use of the semi-discrete sensitivity analysis. Finally, the whole model is applied to the data of Hostun sand taken at the universities of Grenoble and Stuttgart.