[AlmostSurelyRob]

Dear Foamers,

I am seeking some advice on interFoam, pressure reference cells and cyclic boundaries conditions. I am trying to modify interFoam to run cyclic boundary conditions in a similar manner to channelFoam and then apply this model to simulate stratified flow in a pipeline.

I set up pipeline geometry, I put cyclic patches, I write velocity and pressure corrections to account for specified flow rates for each of the two phases.

When I was preparing my case I removed all pressure boundaries so eventually the only patches I have are wall and cyclic. interFoam would protest when I run it so I had to add

pRefCell 0
pRefValue 0

to system/fvSolution. My simulations seem to converge but after several hundred iterations they begin to oscillate.

So I modified the code to output cell residuals for each quantity. The most exciting action happens at the interface between phases (expected!) and at pressure reference cell (surprising?). For some reason velocity residuals are always very high at pressure reference cell. Please have a look at the attachment.

It would be great if you had the time to lend me some assistance on this matter. I can send/attach some case files if think it necessary.

[AlmostSurelyRob]

Something curious, which seems to be related to this problem, are increased values of phase fraction. In the region around the pressure reference cell I get values of alpha1 greater than 1. What is the significance of it?

[nimasam]

hi friend
i had the same problem in cyclic boundary in interFoam, i dont know its for pressure refrence or not but it cant be solved with finner mesh or finner delta time step or alpha correction or ...... alpha residual is up rising !!!!!!!!

[AlmostSurelyRob]

Thanks for your interest. I don't know what type of multiphase case you're doing. I am trying a stratified flow in a simple horizontal pipeline. It's quite possible that I have chosen inputs that do not give steady solutions so I am trying to find some experimental/theoretical cases for which I am sure that the interface is smooth.


I was wrong about the pressure cell. The curious behaviour of high alpha values occurs somewhere else. You can see my geometry, pressure reference cell and high alpha values in the attached picture. Also the Maximum value of alpha is increasing up to 1.14 and then decreasing again.

[Ingenierias2003]

Dear Robert,

I am working too in two-phase flow regarding periodic flow. I am trying to modify the interFoam solver in according to periodic flow (channelFoam).
I was not sure if you got to do it. I trying to do that my be could you help with any doubts.

My kind regards.

Respectfully
Javier Basurco

[AlmostSurelyRob]

Originally Posted by Ingenierias2003

I am working too in two-phase flow regarding periodic flow. I am trying to modify the interFoam solver in according to periodic flow (channelFoam).

Well, in this case we are in the same boat or maybe pipeline, channel... or other conduit which is relevant to your application. ;-)

For now, I can recommend some literature and other posts on this forum.

Single phase:
1. J.Y.Murthy, S.Mathur "Periodic flow and heat transfer using unstructured meshes", Int. Journal for Numerical Method in Fluid, Vol 25,659-677(1997)

2. S.V.Patankar and D.B.Spalding "A calculation procedure for ehat, mass and momentum transfer in three dimensional parabolic flows" Int. J. of Heat Mass Transfer, Vol 15, pp. 1787-1806, 1972

There is also a splendid book which explains the basics of SIMPLE method:
"Numerical Heat Transfer and Fluid Flow" by S.V. Patankar, 1980

Two phase
Level set method:
3. G.Son "A numerical method for incompressible two-phase flow with open or periodic boundaries.", Numerical Heat Trasnfer Part P, 2001


There is also this post which you might find helpful
http://www.cfd-online.com/Forums/ope...interfoam.html

We could have a chat about it to exchange some ideas, but this will be a bit off topic. Tell me what you think.

[Ingenierias2003]

Dear Robert,

Thanks for your quick reply. Definitively, I would like to exchange ideas with you. My e-mail is javierbas82@hotmail.com or javier.basurco@gmail.com. If you could add me to msn, I think is the best way to talk about that or maybe do you have a skype account.

I a look a forward to hearing with you

My deep regards

Respectfully
Javier Basurco

[nimasam]

i use a cyclic boundary condition, you can find my case here! i simulate a bubble train flow!
http://www.cfd-online.com/Forums/ope...interfoam.html