A Computational-Experimental Program for Multiple Scale Fracture Simulation and Design of Metal Forming Processes
Sponsor: National Science Foundation GOALI Research Grant
Co-Sponsor: OSU CAMMAC-TREP
Industry Partner and co-Sponsor: Alcoa Technical Center (Dr. M. Li)
PI: S. Ghosh
Keywords: FEM simulations of Shearing and Cold Rolling Processes, Genetic Algorithms for Design,
Abstract:
(1) This work identifies various modeling issues that are necessary for successful simulation of the cold rolling process by comparing it with experiments on aluminum alloys. It combines considerable experimental studies with finite element simulations using the ABAQUS/Explicit commercial finite element code to identify and evaluate modeling parameters, such as the material properties and friction laws. Damage models are incorporated in the numerical simulations by using plasticity with damage variables e.g. the Gurson model with evolving porosity and Cockcroft-Latham with damage in terms of plastic work. The 3D model predictions are compared with predictions from 2D models to understand the limitations of 2D simulations in predicting the stresses, strains and evolving damage in the rolled strip.
(2) This work combines experimental studies with finite element simulations to develop a reliable numerical model for the simulation of shearing process in Aluminum alloys. The critical concern with respect to product quality in this important process is burr formation. Numerical simulations are aimed at understanding the role of process variables on burr formation and for recommending process design parameters. The commercial code ABAQUS-Explicit with the arbitrary Lagrangian-Eulerian kinematic description is used in this study for numerical simulations. An elastic-plastic constitutive model with experimentally validated damage models are incorporated through the user subroutine VUMAT in ABAQUS, for modeling deformation and ductile fracture in the material. Macroscopic experiments with microscopic observations are conducted to characterize the material and to calibrate the constitutive and damage models. Parametric study is done to probe the effect of process parameters and finally, a Genetic Algorithm (GA) based design method is used to determine process parameters for minimum burr formation.
OSU Research Team
Professor: S. GhoshGraduate Student: D. Gardiner, A. Khadke