A data-driven procedure for the analysis of high strain rate tensile tests via visible and infrared image processing

Abstract

In the present work, a previously proposed procedure for analyzing quasi-static tensile tests during the post-necking phase is extended to tests at high strain rates. The method utilizes a database built from numerical simulation which correlates the relationship between the equivalent stress and the equivalent plastic strain with the shape of the necking profile and with the engineering stress applied to the specimen. Therefore, such database can be used for characterizing material hardening behavior through appropriate processing of the information collected in the database itself. This significantly reduces the computational effort compared to FE-based inverse methods. More specifically, this paper demonstrates how the proposed method can be used to analyze tensile tests conducted on axisymmetric specimens made of isotropic metallic materials whose plastic behavior depends on both the strain rate and the temperature. Regarding the temperature, the authors focused on the temperature rise caused by material self-heating in dynamic tests, whether under adiabatic conditions or not. From an experimental perspective, both visible and infrared cameras were employed to acquire all the data necessary for analyzing, using the proposed method, the material behavior under dynamic conditions. The proposed approach entailed recording the test with proper spatial and time resolution. The significant advantage is that Digital Image Correlation measurements and evaluation of strains are not necessary, as it is sufficient to extract the external contour of the sample.