Identification of Plasticity and Fracture Models for Automotive Extruded Aluminum Parts Using Finite Element Model Updating Algorithm

• Published in: JOM (1989) Vol. 75; no. 12; pp. 5479 - 5493
• Main Authors: Won, Jung Yun, Hong, Seojun, Nam, Byunggun, Jung, Jaebong, Kim, Youngwan, Lee, Myoung-Gyu
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.12.2023; Springer Nature B.V
• Subjects: Algorithms; Alloys; Aluminum alloys; Aluminum base alloys; Automobile industry; Chemistry/Food Science; Constitutive models; Deformation; Ductile fracture; Earth Sciences; Engineering; Environment; Experiments; Extrusion; Finite element method; Geometry; Heat treating; Identification; Identification of Anisotropic Constitutive Models for Complex Loading Paths; Load; Mathematical models; Mechanical properties; Mechanical tests; Mechanics; Model updating; Partial differential equations; Physics; Plastic properties; Weight reduction
• ISSN: 1047-4838; 1543-1851
• DOI: 10.1007/s11837-023-06164-z

For extruded aluminum alloys, the identification of mechanical properties and the corresponding constitutive modeling are very challenging, due to the complex structured part geometry, microstructure variance, difficulty in performing standard testing, and discrepancy between material- and structural-scale deformation. With these challenges, in this study, the plasticity and ductile fracture models for an aluminum extrusion part having a complex cross-sectional shape are identified based on an inverse experimental–numerical approach. In particular, bending experiments in part-scale are employed as alternatives to standard mechanical tests. A single and double finite element model updating scheme are newly suggested and performed to predict plastic hardening behavior and ductile fracture criterion from measured load–displacement curves. To overcome the limited deformation history available at various stress states, a virtual (deformation) path generation method is proposed for calibrating the fracture model. The feasibility of the optimized constitutive models is evaluated through a number of trials with modifications in the optimization process, which are successfully validated through comparison with experiments on load–displacement curves and fracture initiations. Finally, it is confirmed that the proposed inverse identification approach can offer a computationally efficient method for constitutive modeling, with potential applications in various engineering fields.