|About this Abstract
||2022 TMS Annual Meeting & Exhibition
||Defects and Properties of Cast Metals IV
||Reduced-Order Multiscale Modeling of Elasto-Plastic Cast Alloys with Process-Induced Porosity
||Shiguang Deng, Carl Soderhjelm, Diran Apelian, Ramin Bostanabad
|On-Site Speaker (Planned)
Cast aluminum alloys often contain heterogeneously distributed pores of complex morphologies that significantly affect material behaviors. In this talk, we will introduce a computationally efficient reduced-order multiscale framework to simulate the behavior of metallic components with process-induced porosity under irreversible nonlinear deformations. Major components of our approach include: (1) a data compression stage which dramatically reduces the number of degrees of freedom by agglomerating neighboring finite element nodes; (2) a novel reduced-order method which projects displacement vectors into a lower dimensional space where the material’s elasto-plastic behaviors are approximated; and (3) a porosity oriented microstructure reconstruction algorithm which mimics the material’s local heterogeneity with reconstructed pores from tomography characterization. We will illustrate the effects of spatially varying microstructures on cast alloy behaviors and compare our approach against the results from direct numerical simulations to demonstrate its performance and versatility.
||Computational Materials Science & Engineering, Modeling and Simulation, Aluminum