I found an interesting result from COMSOL 4.4 Acoustic-Shell Interaction. Let me describe it here:
I have a 1 square meter cubic whose all 6 walls were set to Sound Hard Boundary except a A2 size window in the front wall where a 2.5 mm thick Aluminum Plate was mounted with 4 edges Fixed Constrained. Its a simple set-up. Then I just add the Acoustic-Shell Interaction physics and set the plate as the Exterior Shell. And I do a Eigenfrequency analysis.
Then comes the interesting thing. If I set the mesh size to "Normal", COMSOL will give out a reasonable result with the first coupling mode caused by the plate around 95Hz. But if I set the mesh size to "fine", there will be several local modes under 95Hz, which should not be real.
Why should this happen? How can I get a more accurate result for the acoustic-structure interaction analysis in COMSOL?
I have a 1 square meter cubic whose all 6 walls were set to Sound Hard Boundary except a A2 size window in the front wall where a 2.5 mm thick Aluminum Plate was mounted with 4 edges Fixed Constrained. Its a simple set-up. Then I just add the Acoustic-Shell Interaction physics and set the plate as the Exterior Shell. And I do a Eigenfrequency analysis.
Then comes the interesting thing. If I set the mesh size to "Normal", COMSOL will give out a reasonable result with the first coupling mode caused by the plate around 95Hz. But if I set the mesh size to "fine", there will be several local modes under 95Hz, which should not be real.
Why should this happen? How can I get a more accurate result for the acoustic-structure interaction analysis in COMSOL?