Phase distribution and properties identification of heterogeneous materials: A data-driven approach

This paper presents a new methodology to extend the Data-Driven Identification (DDI) heterogeneous samples made of multiple elastic materials. By using Correspondence Analysis (CA) technique post-process DDI, we are able identify material databases representative behavior each phase. Simultaneously, localize different phases (matrix and inclusions) in sample. For contrasts between phases, method is tested on synthetically generated data parametric study performed. Furthermore, show that it possible iterate DDI CA order improve method’s predictions when limited by scarce input data. In all cases studied, proves be effective for estimating stresses, as well identifying