Calibrating material parameters to model the thin-walled components made of die cast AM60B magnesium alloy

Abstract

As a novel lightweight metal, die cast AM60B magnesium alloy continues to be considered as a potential replacement for steel in certain automotive components. However, proper numerical model representing this alloy has not yet been fully explored. Thus, an optimisation methodology was developed to calibrate the material parameters needed for four available material laws, namely, MAT_99, MAT_81, MAT_24 and MAT_107 in LS-DYNA explicit FEA package (LSTC, Livermore, CA). The basic idea was to combine computations with a set of optimisation procedures to systematically adjust various material parameters until the calculated mechanical responses optimally match those measured experimentally. The optimisation was based on uni-axial tensile coupon tests at different nominal strain rates, ranging from quasi-static to 800 s−1. The optimisation results were evaluated using component experimental data from both slow- and high-speed axial crushing and quasi-static four-point bending tests of a thin-walled top-hat structure, and then quantified using a gross correlation index (GCI). Calculated results indicate that using the material constants generated from this set of procedures, all of the four material models replicated experimentally obtained stress–strain curves well, and yielded similar values of GCI (The difference was less than 4%). But only MAT_99 could accurately capture the damage patterns as observed in structural component experimental tests, thus being considered the best among the four material types studied. It is also concluded that the damage behaviour of AM60B alloy is sensitive to the strain hardening model and failure criterion selected.