I'm trying to get the frequency response of a small box, filled with air and having a hole in the middle (kind of like the soundboard of a guitar). I have tried a few approaches, but none of them seem to give me what I need. I'm new to Abaqus, and mainly use Abaqus/CAE (haven't used the keywords that much yet).

Here is my model:

The model consists of a hollow cylinder shell, with a hole on the top plate. The cylinder is filled with air, modeled as an acoustic solid. The pertruding cylinder is a column of air with non-reflecting walls, to model the behaviour of air in the hole.

I'm trying to get the frequency response of this system, measured from a node inside the shell, from the acoustic solid (POR).

Here is what I've tried, and some questions:

1) Lanczos eigensolver for the eigenmodes of the coupled system.

This seems to work fine. But it's not detailed enough.

2) SSD-mode based excitation on a specific point

I think the resonance frequencies are about right, but I cannot understand how I get the peaks to be correct (see the image below). The response nears it's peak when the frequency is increased, but at the peak the value is suddenly "reversed".

Any pointers and help to this problem would be greatly appreciated!