High performance data-driven multiscale inverse constitutive characterization of composites

Abstract

In this chapter, we present recent advances in the multi-scale data-driven inverse constitutive characterization of composites via multi-axial robotic testing, high-performance full-field strain measurement methods and surrogate models. Emphasis is given first on the motivating aspects for multi-scale model inversion for the case of laminated composites, followed by a description of a data-driven inverse constitutive characterization of the bulk lamina of composites via multi-axial loading of laminate-level coupons. A description of a surrogate modeling approach is presented in order to address the computational cost involved in the forward model evaluations associated with the optimization required for material characterization. Finally, we present progress in identifying material properties at the lower scales by utilizing experiments in the laminate-level macro-scale.

Citation

Michopolous, John; Iliopoulos, Athanasios; Hermanson, John; Steuben, John; Komnineli, Foteini. 2017. High performance data-driven multiscale inverse constitutive characterization of composites. In: Brancherie, D.; Feissel, P.; Bouvier, S.; Ibrahimbegovic, A., eds. Mechanical engineering and solid mechanics series: from microstructure investigations to multiscale modeling: bridging the gap. London, UK: ISTE Ltd; Hoboken, NJ: John Wiley & Sons, Inc.: 197-238. Chapter 8.