Computer Aided Engineering Journal & Forums

how do I solve this issue?
i am not sure how to address this? anyone has a clue? thanks

Quote:Numerical accuracy is lost after 15 eigenvectors have been calculated. The number of eigenvectors in the iteration is reduced accordingly.

Numerical accuracy is lost after 14 eigenvectors have been calculated. The number of eigenvectors in the iteration is reduced accordingly.

Numerical accuracy is lost after 13 eigenvectors have been calculated. The number of eigenvectors in the iteration is reduced accordingly.

Numerical accuracy is lost after 12 eigenvectors have been calculated. The number of eigenvectors in the iteration is reduced accordingly.

Numerical accuracy is lost after 11 eigenvectors have been calculated. The number of eigenvectors in the iteration is reduced accordingly.

The numerical method is only valid for low frequency so the calculation after 15 eigenvectors is not precision. Just use the frequncies less than 15 eigenvectors .

Thanks a lot for the info

Hi Mandarin , did you use subspace method in your model.
I think you can try Lanczos method instead.