[ankgupta8um]

Hello,

I am using coodles solver to simulate the expansion of annular jet into a cylinder. The geometric configuration consists of an annular pipe attached to a cylinder.
I created the movies for pressure and axial velocity, showing their transient behavior. The movie for pressure shows that there is a spurious pressure wave bouncing between the inlet and the junction where annular pipe meets the cylinder. The movie for axial velocity shows that the flow in the annular pipe is pulsating, which I think is due to the pressure waves. Due to this pulsation, the computed mean axial velocity in OpenFOAM is far off from the experimental data.

Movies are attached here:
Pressure:
pres.avi
Axial velocity:
vel.avi

Could any body at the forum please provide me with some pointers on how to get rid of the pressure waves ? Any information would be greatly appreciated.

The BCs I used are fixed velocity, zero pressure gradient, and fixed temperature at the inlet. At the outlet, it is inletOutlet for velocity and temperature, and pressure transmissive BC for pressure.

I would me happy to provide more information on my case set-up.

Thanks!!
Regards,
Ankur

[ankgupta8um]

Hello,

Sorry.. I couldn't attach the movies as it violated the file size restrictions on the board.
Please find the movies at these links:
www.personal.psu.edu/axg330/pres.avi
www.personal.psu.edu/axg330/vel.avi

Thanks!!
Regards,
Ankur

I haven't tried it and have yet to get a good coodles case to work, but maybe if you split the outlet patch in two and placed them at an angle to each other it would dampen the reflected wave.

[ankgupta8um]

Hello Conn,

Thanks for the reply. The problem that I am seeing is the reflection of the pressure wave from an internal plane in the geometric configuration. The wave reflects back from the location where the annular portion meets the cylinder.
If you play the pressure movie real slow, it is clearly visible that the pressure wave moves from inlet to the location where annular pipe meets the cylinder, and then it reflects back from there.

I am getting a feeling that this pressure wave is either due to the specified BCs at the inlet, or due to the wrong choice of numerics. I did try some other BC specifications at the inlet, but that didn't help.

I would really appreciate any ideas/comments on resolving this issue of pressure waves !!

Thanks!!
Regards,
Ankur

[eugene]

I have had this problem in the past, but unfortunately have not had the opportunity to do something constructive about it. Your problem is the zeroGradient pressure boundary at the inlet. This boundary is reflective with respect to pressure waves. The interface between the annular and open sections of the domain is also partially reflective because of the large disparity in scale, so you end up with a standing wave in the annular section.

The solution: formulate and implement a non-reflecting inflow boundary. This is not as hard as it seems, as all the building blocks are available in the waveTransmissive outlet code.