Acoustic simulation across a pipe with leakage.
 

I am doing a project on acoustic simulation across a pipe with leak. I am to study the SPL at the inlet and outlet of the pipe when there is a leakage. I know that fluent acoustic model should not be used for a duct flow but that is the only software I have for now. I wonder if my solution cannot converge mainly because of this.

My input is as followed.

Models
Energy
Viscious - LES WALES
Acoustic - FW-H

The fluid is compressible air
Density - ideal gas

Boundary conditions
Inlet - mass flow rate/pressure inlet/velocity inlet (all failed)
Pipe and leak outlet - pressure outlet
Pipe wall - no slip

Timestep - 1*10^-6

The solution cannot converge and error such as courant number is reduced to 5e-6 and temperature change is too large. Is there anyway I can fix this? Thank you.

You want to master the numerical setup before jumping into LES.

Try running a simple steady case of just simple flow in a pipe with a leak. Use laminar. It will also probably fail because there's something wrong with your setup. Check all the boundary conditions and check your mesh.

Are you sure you even need the acoustics model? You're doing LES so the sound generated by the leak should already be captured.

This isn't related to your convergence but make sure you use the non reflecting boundaries for the inlets and outlets so that your pressure doesn't continue to rise with time.

Thank you very much. I fixed the parameters and it is running smoothly now. Can I ask if you know of any past research and papers at involved in similar project? I can't seem to find anyone who did such work before.

I am also working on this project. And there are few reliable references.
Moreover, the meshing is difficult when the ratio of leak and pipe diameter is very small. How did you handle the high aspect ratio problem?

Hello,

1- Are the acoustic models in Fluent applicable to predict noise in wall-bounded domains? I want to simulate the noise of flow in a pressure relief valve and a silencer that installed at its outlet.


2-Is there any BC in acoustic models that can help me to import magnitude and frequency distribution of noise at the inlet of a duct(where the air is flowing) and predict the noise attenuation at the outlet? (if possible I don't want to simulate valve and I want to use empirical relations for valve noise-producing)


Thanks,