[jimteb]

Hi Foamers,

I am new to openFoam and I am trying to use it to model the effects of a pressure wave exiting a tube into the open environment. At the moment this has been solved as an axisymmetric model.

I am using the timeVaryingUniformFixedValue boundary condition at the pipe inlet (with clamp for out of bounds) to create the pressure pulse which then propagates out of the tube.

There are a couple of problems I am having however, mainly concerning reflection of pressure waves:

1) I need to use a non-reflective BC at the inlet, however I am already using the timeVaryingUniformFixedValue BC here. Is there any BC which allows waveTransmissive to be used in conjunction with timeVaryingFixedValue, or something else that might be suitable?

2) Would rhoCentralFoam or rhoPimpleFoam be more suitable for this problem? rhoCentralFoam seems to be more stable and gives more realistic answers, but the flow is subsonic and it is a fairly weak pressure wave.

3) Implementation of waveTransmissive in the farfield gives the error "keyword inletOutlet is undefined in dictionary" for rhoCentralFoam. Any ideas why this might be? On top of this what would you use when solving with rhoPimpleFoam in place of waveTransmissive as I don't believe you can use a NRBC with a pressure based solver?

Any help would be greatly appreciated.

Thanks,

James

[alessandroB]

Hi,
Maybe it's a little late to ask you about the wavetransmissive time dependent boundary condition,but ... Did you solve the problem???
Best regards.

[jimteb]

Hi,

I did manage to fix this problem, in a work-around manner:

I created an initial pressure and velocity field, representative of the time varying pressure/velocity waveform, with ambient conditions everywhere else. This required me to extend my domain a bit, in order to factor in the initial field region at the inlet. So the downside is that I needed more cells than is necessary, compared to say a mixed waveTransmissive-timeVaryingFixedValue BC. Then I just applied a normal waveTransmissive BC at the inlet boundary. In order to get only a forward propagating acoustic wave, you need to set U = p/(rho*c).

This seems to work well, but a mixed waveTransmissive-timeVaryingFixedValue BC would possibly be better.

Regards,

James