Optimization of the Phenomenological Constitutive Models Parameters Using Genetic Algorithms

The lack of accuracy of numerical results is still nowadays one of the main drawbacks of sheet metal forming process simulation. One of the main reasons for such a lack of accuracy is the constitutive models used to describe the real material’s mechanical behavior. The most widely used phenomenological constitutive model is based on the classical Hill 1948 yield criterion. In the last decade several new yield criteria have been proposed, with the constraint that a parameter identification procedure has not always been clearly set. This study presents a general approach to optimize anisotropic plastic description. A weight-based optimization procedure is presented in order to perform the optimization of several constitutive models based on experimental results. Using this procedure is expected to improve the plastic description with a global optimum solution.