[srinath]

Hello

I find the following quite puzzling.
In icoFoam.C, i find the foll. lines

solve(UEqn == -fvc::grad(p));

// --- PISO loop

for (int corr=0; corr<nCorr; corr++)
{
volScalarField rUA = 1.0/UEqn.A();

U = rUA*UEqn.H();
......
}

Solving UEqn presumably changes the value of U. But what is the point if couple of lines later, we store rUA*Ueqn.H() in U

If solving UEqn is a predictor for a U value, what is the point if we never use the value and overwrite it?

I would appreciate it if someone could clarify this as i find it quite puzzling.

Regards
Srinath
p.s. commenting out the solve UEqn line, seems to give visually acceptable results for the cavity case, but diff says the results are different. So clearly it is not a bug, but is doing something quite subtle

[fabianpk]

Hello,
Thank you for your question. This gives me an excellent opportunity to advertise my thesis. This exact question is covered in Appendix A.
The thesis main topic is diesel spray simulations using OpenFOAM, as well as simulation of cavitating flow in a model injector. However, since I have pondered on questions similar to the one you put out during my PhD studies, I put it as an appendix for myself to read, should I ever forget what happens in icoFoam.
A free version is available at:

http://powerlab.fsb.hr/ped/kturbo/Op...olmPhD2008.pdf

If you want a paper copy, contact me by email.

/Fabian

[srinath]

Thanks Fabian

Appendix A in your thesis was very useful.

Am also reading the rest of it as it is very interesting.

Cheers
Srinath

[kumar]

Hello, Fabian
My name is Suresh. I have just stârted using OpenFOAM a few weeks back. I saw your thesis, your work seems to be quiet interesting. I would like to let you know that My P.hD research work is also similar to your topic. I amalso using OpenFOAM to study the break up of a liquid sheet from a nozzle. But a major part of my research work will include Nonlinear instability analysis of the liquid sheet.
I am using lesInterFOAM of OpenFOAM v1.5 to do these simulations and I have already performed one simulation on a simple geometry.
Since I have just started using OpenFOAM, i have a few basic questions and I hope you wouldnt mind answering them:
1) I found that the icoFOAM, interFOAM, lesInterfOAM are segregated solvers based on the PISO algorithm, just correct me if i am wrong in understanding .(with VOF for surface tracking and LES for turbulence modeling)

2) But I found in your thesis that you are including the compressibility effects, isnt it advicable to solve the equations in a coupled manner like in a coupled solver to solve a compressible fluid problem.
3) I also have a very basic question, is there any coupled solver in OpenFOAM?
I have some experience in developing a coupled solver, so i understand the structure of a coupled solver. I also have some idea about the basic methodology of a segeregated solver as explained in basic textbooks by S.V.Patankar and Malalkasera.
4) For my simulations in OpenFOAM I have a setup like injecting 200bar pressure fluid into the chamber with 10bar pressure. This is just a test case given to me by Delphi.

bye
thanks for your reply inadvance
with regards
K.Suresh kumar

[fabianpk]

Hello,

1) Yes, to find an example of a solver that's not PISO, look at simpleFoam.
2) Perhaps it might be suitable, and there are quite a few coupled solver codes out there now, but PISO works fine for compressible flow (in my experience).
3) Not as far as I know, no. I think to develop a coupled solver in OpenFOAM you would have to dig into the source code much further than the top-level code.

/Fabian

[kumar]

Hello Fabian,
Thankyou very much for your reply. I will have a look at simpleFoam.
But for my problem I dont have to worry about the coupled solver as my problem corresponds to an incompressible fluid.
I was just curious to know more about the solver.
I just have one more question can you provide me some reference on the algorithm and some details about the lesInterFOAM solver, if you have any?
thankyou very much for reply
bye
K.Suresh kumar