A Methodology for Evaluating the Performance of Material Models Under Biaxial Loads

Abstract: The objective of this thesis is to develop a methodology for evaluating the performance of material models during biaxial loads. This is done by executing identical punch tests, one in a simulation and one in a physical experiment, and then comparing the results from the two. The material model that is to be evaluated is used in the simulation and a specimen of the corresponding material in the physical experiment. The more the results resemble each other, the more accurate the material model is considered to be. A physical test jig is designed using a concept development process. Results from the experiment are acquired using digital image correlation and then compared to data from the simulation. Three comparisons are then made. First, the global force-displacement curves are compared. Second, a field analysis where local displacements and in-plane strains are compared. Third, the distribution of major vs. minor strains is compared. The mentioned comparisons result in a number of graphs and figures where the differences between the simulation and the experiment are represented both graphically and numerically. The results from using this methodology can be used to calibrate and compare different material models. In addition, the material behavior can be analyzed qualitatively. The results can also reveal in which deformation range the model performs best.