Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.

Unexpected Complex Frequencies in Acoustic Eigenfrequency Solver

Please login with a confirmed email address before reporting spam

I have a simple 3D model, with which I'm analyzing the phononic band structure, similar to this blog post: https://www.comsol.com/blogs/modeling-phononic-band-gap-materials-and-structures/. I'm seeing some odd behavior near zero frequency solutions, and I believe the problem can reduced to a simpler problem, stated below.

In my 3D model, I impose periodic boundary conditions on each opposing side. When solving the pressure acoustics equation, there should clearly be a trivial constant pressure solution, with eigenfrequency zero. When solving for the eigenfrequency, however, I get an eigenfrequency of around 0.7 + 100i Hz. While I'd normally brush this off as an unimportent numerical error, the abnormaly-large imaginary part persists in analyzing the band structure for finite wavevectors, and gives me non-physical solutions when I try looking at the bandstructure at small frequencies. What could be causing these complex frequencies, and how can I address them?


1 Reply Last Post 5 nov 2018, 08:21 GMT-5
Mads Herring Jensen COMSOL Employee

Please login with a confirmed email address before reporting spam

Posted: 6 years ago 5 nov 2018, 08:21 GMT-5

Hi Henry

I think there can be several reasons for this behavior. We would need to see the model to figure this out and I suggest you send it to support. Here are a couple of suggestions:

1) Make sure you use the "Eigenfrequency search method around shift" in a good way, possible using the search for "Larger real part" option. 2) Another problem can be the linearization point in the model. If you have losses in the model then the eigenfrequencies (eigenvalues) may enter in a nonlinear fassion in the equations. If this is the case you should set the linearization point appropriately. In the documnetaiton for the acoustics module this is discussed in the chapter about the different study types.

Best regards Mads

Hi Henry I think there can be several reasons for this behavior. We would need to see the model to figure this out and I suggest you send it to support. Here are a couple of suggestions: 1) Make sure you use the "Eigenfrequency search method around shift" in a good way, possible using the search for "Larger real part" option. 2) Another problem can be the linearization point in the model. If you have losses in the model then the eigenfrequencies (eigenvalues) may enter in a nonlinear fassion in the equations. If this is the case you should set the linearization point appropriately. In the documnetaiton for the acoustics module this is discussed in the chapter about the different study types. Best regards Mads

Note that while COMSOL employees may participate in the discussion forum, COMSOL® software users who are on-subscription should submit their questions via the Support Center for a more comprehensive response from the Technical Support team.