Robert Mehlig

Thank you! I will try this.

Generally, I want a as simple model as possible, any complexity might only result in more sources of errors. Trying to model the piezos as seperate 2D or 3D objects would mean I'd have to know their material properties well enough and define the contact region aswell. Since they are glued on in our experiments. I have noticed, this kind of work takes trial and error. 

I will move on and define the excitation by pressure on a circular surface with the radius of the piezo, later maybe on a combination of force/pressure/shear...

The current problem is that the shell element, layered section model takes to long to solve, without yet converging, i.e. the results are useless. Reducing the number of shell integration points to 1 did however reduce computing time by 1/6 and producing ~the same result. Now I can reduce the element- & timestep size to see if it converges.

Thanks again,