Hi all,
I wanna optimize a geometrical parameter and a parameter as factor for a material property to fit my 3d FEM model (displacement on an edge) against real life measured data (inverse methode) resp. an analytical representation (linear function). The mechanical problem I´m getting solved incl. an auxiliary sweep in a stationary study, but if I implement the optimization part, the model does´nt converge anymore. Do you have any recommendations on which opt. nodes are sufficient and what are the setup basics for the opt. solver. Currently I tried the squared error node and the global opt. node in combination with all solver algorythms. Additionaly I´m struggeling with defining an objective function regarding the measured and sim. displacement.
Any comment is welcom,
thanks in advance
Sven