Abstract
The anomalous behavior of aluminum alloys is modeled with quadratic Hill48 yield criterion in this study. An identification method based on minimization of the error function is applied and the effect of the number of experimental input and weight coefficients used in the identification are investigated. Two highly anisotropic aluminum alloys (AA2090-T3 and AA5182-O) are selected in the study. Firstly, Hill48 parameters are determined with four different experimental data set, then the effect of the weight coefficients for each set is investigated. In-plane variations of plastic properties and yield surfaces of the materials are predicted with determined Hill48 parameters and the most appropriate pair (experimental data set and weight coefficient) are selected by comparison of the predicted results with experiment.