Cyclic mesh for fully-developed profile
I am trying to find the fully developed velocity profile in a straight duct.
Rather than compute a 3D duct, and look at its profile towards the end, I would like to apply periodic boundary conditions to make it more efficient. I have never done this on openFOAM and would like some help.
I have a inlet and an outlet boundary. So:
(0 4 7 3)
(2 6 5 1)
However, now I am wondering what kind of boundary field definitions I need.
Previously I had zeroGradient or fixedValue fields (depends on variable) for both the inlet and outlet. What do I need to set for the periodic case?
Thank you very much.
Take a look at the channelFoam tutorial. This will help you I think.
Please keep in mind that the pressure field is never periodic. Therefore you will have to split the pressure field in a periodic part and a pressure drop. This can be seen in the channelFoam solver.
So, it means that I need to modify my solver?
Looking at channelFoam.c, I would need to slightly change the first part before the PISO loop and add the section correcting the driving force for a constant mass flow rate. I would not need to change any field definitions, apart from creating createGradP.H, is that correct?
In the first section, what is "sgsModel->correct(); " ?
That you can do indeed, or you can use a jumpcyclic. This has different names for different OF versions I think. Browse around the OF source to see the difference.
src » finiteVolume » fields » fvPatchFields
The same holds for sgsModel->correct(); Browse around the turbulence models to see what it does.
src » turbulenceModels
Hmm, I don't quite understand how jumpCyclic works (especially trying to read the source code).. Is there a tutorial on how to use it?
There is a fan tutorial in OF that apparently uses it but I'm not sure how that is relevant to what I'm trying to do...
I also met the same problem that I want to set zeroGradient or fixedValue fields (depends on variable) for both the cyclic inlet and outlet. Did you find a way to achieve it?
My best regards,
|All times are GMT -4. The time now is 21:51.|