I am trying to simulate a rectangular elbow which filled with acoustic glass fiber by 3D models, but the pressure drop i got is about 50% of tested data. Your help is appreciated. Rectangular elbow inlet and outlet 25" width, inner wall R11", outer wall R37", upstream and downstream ducts 40" width, elbow and ducts 24" thick. I tried standard k-e, realizable k-e, RNG k-e, RSM models, Y+=110~200, standard wall function. The pressure drops I got by these models are quite close but far away to tested data. Your comments are highly appreciated.

Tony

how did you model the glass fiber zone?

I do nothing to Fiber zone.The fiber is covered by two perforated sheet metal plates and I just want the pressure drop at air side. Howerver, the perforated walls are pretty rough, I give a roughness and just a few percent of pressure drop increased. I am not comfortable with the ROUGHNESS of the covers, they have 1/8" dia. holes @ 3/16" centers. I don't know it's good enough to assume a roughness for them, I hope it's not the reason why the result is off. Because I simulated a rectangular straight silencer, the pressure drop looks good. I am wondering if the flows are too BAD to be simulated? Thanks