A new performance indicator for heterogeneous elastoplastic material test specimens

A wide range of heterogeneous specimen designs has been proposed for the inverse calibration of anisotropic elastoplastic models, yet a consistent framework to quantitatively evaluate and compare their performance remains lacking. This study presents a performance indicator for the systematic evaluation and design of heterogeneous mechanical tests aimed at the inverse calibration of anisotropic metal plasticity models. The indicator integrates two fundamental criteria: load state diversity, which quantifies the richness of the strain space via convex hull analysis; and measurement efficiency, which assesses the spatial compactness of high-value plastic regions relevant to Digital Image Correlation. The formulation operates on strain data in a global coordinate system and is applicable to both two-dimensional and three-dimensional specimens. It can also account for material symmetry, tailoring its application to any form of anisotropy. Application of the indicator to ten 2D geometries revealed a clear ranking of designs according to their strain diversity and experimental practicality, while its use in optimization led to the development of a 3D curved-beam specimen exhibiting several orders of magnitude higher load state diversity and twice the efficiency of planar counterparts. Experimental validation using a multi-view stereo-DIC system is performed to confirm the generation of rich heterogeneous strain fields that can facilitate inverse calibration of constitutive models. The framework provides a unified and practical methodology for integrating strain-space analysis, shape optimization, and experimental testing in the context of anisotropic elastoplastic constitutive model calibration.