This post has NOT been accepted by the mailing list yet.
I just started a PHD for a car company, and I work with acoustic transmission through panels. I developed a simple planar model of a glass panel exited by a diffuse acoustic field using abaqus acoustics. On the non-exited side of this panel I created an air mesh, with a non-reflecting condition on its outer surface. I output POR pressure (p) and velocity (v) on the solid-acoustic interface.
What I expected was a dip on the Transmission Loss curve, around the coincidence frequency (4000Hz-4100Hz), as we would expect in all diffuse excited panel transmission loss curves. Instead I get a curve with a 1st resonance dip and other resonances, but no apparent coincidence dip.
I then developed a simpler model. The plate was excited by a punctual pressure (instead of a diffuse field) and so it became a radiation model. In this case theoretically, for frequencies under the coincidence frequency the product of p x v* on the receiving surface (v* being conjugated NORMAL velocity) should be purely imaginary (because of the different dispersion of air and glass, wavelengths don't match and the transmitted wave is evanescent and cannot travel further into the air part), real(p x v*)=0, so no radiation. Instead real( p x v*) when simulated on my model is completely unaffected by coincidence and shows peaks on the resonant frequencies (80-...Hz) un et d of being 0. As if the received waves had no problem traveling into
I am under the impression that either my structure-fluid coupling isn't correct (I used *TIE), or my non-radiation conditions are not well applied. Here first and second case input files( I hope, I am not sure that I know how to do this)