Data-Driven Design Optimization for Composite Material Characterization

Abstract

The main goal of the present paper is to demonstrate the value of design optimization beyond its use for structural shape determination in the realm of the constitutive characterization of anisotropic material systems such as polymer matrix composites with or without damage. The approaches discussed are based on the availability of massive experimental data representing the excitation and response behavior of specimens tested by automated mechatronic material testing systems capable of applying multiaxial loading. Material constitutive characterization is achieved by minimizing the difference between experimentally measured and analytically computed system responses as described by surface strain and strain energy density fields. Small and large strain formulations based on additive strain energy density decompositions are introduced and utilized for constructing the necessary objective functions and their subsequent minimization. Numerical examples based on both synthetic (for one-dimensional systems) and actual data (for realistic 3D material systems) demonstrate the successful application of design optimization for constitutive characterization.