Author(s)

M.G.D. Geers, F.P.T Baaijens & C.W.J. Oomens

Abstract

The identifiability problem that rises in mixed numerical-experimental problems is one of the most important issues if an experiment has to be designed for the quantification of material parameters. This paper focuses on the relation between various aspects of an experiment and the condition number of the information matrix used in gradient algorithms for the parameter estimation. The initial estimate of the parameter sets permits to indicate whether or not the designed experiment provides the required information. A comparison is made for a tensile specimen and a bend specimen, where different types of measurements are simulated. Various aspects influence the optimal relation

Keywords