[PO-C2] Poster Session 2

[P2-18] Prediction of Biaxial Tensile Deformation Behavior of Aluminum Alloy Sheets using Crystal Plasticity Finite Element Method and Machine Learning

^○Kota Koenuma¹, Akinori Yamanaka¹, Ikumu Watanabe², Toshihiko Kuwabara¹ (1.Tokyo University of Agriculture and Technology, Japan, 2.National Institute of Materials Science, Japan)

Aluminum alloy sheets are used for automotive body parts to reduce the weight of automotive. However, because the formability of aluminum alloy sheets is much lower than that of steel sheets, the plastic deformation behavior of aluminum alloy sheets during plastic forming processes has been predicted by sheet metal forming simulations based on the finite element method. In order to predict the plastic deformation behavior of an aluminum alloy sheet subjected to a biaxial stress state in a sheet metal forming process, we proposed the numerical biaxial tensile test (NBT) methodology using the crystal plasticity finite element method (CPFEM) and the mathematical homogenization method. The NBT method has been already applied to predict the contour of equal plastic work of a 5000-series aluminum alloy sheet. The results demonstrated that the contour of equal plastic work (the yield locus) calculated by the NBT method shows a good agreement with that measured by the experimental biaxial tensile test using a cruciform specimen. Unfortunately, because we need to perform multiple NBTs to calculate the contour of equal plastic work, the disadvantage of the NBT method is its high computational cost. In this study, in order to predict the biaxial tensile deformation behavior of aluminum alloy sheets more efficiently, we propose a new numerical material testing methodology by combining the NBT method with machine learning methods. The deformation behavior of aluminum alloy sheets predicted by the proposed method is verified by experimental multi-axitial material tests.