A computational mechanics engineering framework for predicting the axial crush response of Aluminum extrusions

Abstract

Automakers are developing new lightweight aluminum alloys for automotive structures to reduce vehicle weight. However, these alloys require extensive mechanical characterization for accurate calibration of a numerical model, which is often a painstaking task. Automakers are exploring alternative strategies to reduce the number of experiments required for characterizing these new alloys without sacrificing accuracy in predicting performance. The method of virtual experimentation through computational mechanics engineering (CME) with crystal plasticity is showing promise in satisfying this need. This work presents a CME framework for predicting the axial crush behavior of an aluminum alloy AA6060-T6 extrusion using only a single uniaxial stress-strain response and 2D electron backscatter diffraction (EBSD) scans. An anisotropic phenomenological model is generated using a 3D reconstructed microstructure and a calibrated crystal plasticity model. Additional mechanical characterization is performed to qualify the proposed CME framework. Finite element simulations that employ the CME framework are performed to evaluate the suitability of this methodology in quasi-static axial crush applications. Quasi-static axial crush experiments of the extrusion are performed to validate the finite element simulations and the CME framework. Simulations using the CME framework were capable of predicting the experimental crush response with 3–4%. The proposed CME framework can help automakers reduce the number of experiments needed for the development of components in large deformation, such as crush.