Discussion Forum

Hello everyone,

I am trying to solve an optimization problem with a fluid-structure interaction.

For the FSI problem, I took the existing COMSOL application of a vibrating beam in a flow: www.comsol.com/model/vibrating-beam-in-fluid-flow-9408.

The FSI has been solved with E=5.6MPa in the beam and calculates transient drag and lift coefficients.

For the Optimization module, I am following the explanations on this thermal conductivity estimation problem: www.comsol.com/video/using-the-optimization-module-for-parameter-estimation

Here is what I did:
I exported the lift values between 5 and 6 seconds in a .csv file, and I want to solve an optimization problem, where E is the control variable between 4 and 7, and the objective function is to minimize the global least square error between computed lift and the lift given in the .csv table. So I expect the optimization solver to deliver a value E close to 5.6.

I tried several things but I always end up with a list of :
Warning: Inverted mesh element near coordinates (0.597038, 0.193043).
Warning: Inverted mesh element near coordinates (0.597038, 0.193043).
Warning: Inverted mesh element near coordinates (0.606098, 0.192952).
....

and a final error :
Undefined function at the first feasible point.
Returned solution is not converged.

I don't think the mesh is wrong or the initial and bound values of E are wrong, because I applied a finer mesh, but I obtained the same error. And I also solved the direct FSI problem with the same mesh with E=4 and E=7 (manual parameter sweep =) ) and it worked fine.

Do you have an idea what I can try to solve this problem? Is it a problem between the optimization module and moving meshes of FSI?

Thank you for your support.
Best regards,
N. Alia