Automated identification and calculation of the parameters of nonlinear material models

Material models can be defined as a set of ordinary differential equations or as an analytical function. Usually, material models include model parameters. These model parameters are adjusted with respect to experimental data. A numerical tool, called FitIt [1], for an automated optimisation of model parameters is described. The underlying methods are the Levenberg-Marquardt method and the simple Gradient-method. First and second order derivatives of a quality function with respect to the model parameters are used. The derivatives are calculated numerically; an empirical method for the calculation of these derivatives is established.