PimpleFoam problem
 

Hi all!

Maybe a dumb question, but I cannot run pimpleFoam. Seems to be something general.
I have done simpleFoam with SST, everything was fine. now I want to run pimpleFoam with same model, but...

[3] #0 Foam::error::printStack(Foam::Ostream&) in "/home/geza/OpenFOAM/OpenFOAM-1.6/lib/linux64GccDPOpt/libOpenFOAM.so"
[3] #1 Foam::sigFpe::sigFpeHandler(int) in "/home/geza/OpenFOAM/OpenFOAM-1.6/lib/linux64GccDPOpt/libOpenFOAM.so"
[3] #2 __restore_rt in "/lib64/tls/libc.so.6"
[3] #3 Foam::DILUPreconditioner::calcReciprocalD(Foam::Fi eld<double>&, Foam::lduMatrix const&) in "/home/geza/OpenFOAM/OpenFOAM-1.6/lib/linux64GccDPOpt/libOpenFOAM.so"
[3] #4 Foam::DILUPreconditioner::DILUPreconditioner(Foam: :lduMatrix::solver const&, Foam::dictionary const&) in "/home/geza/OpenFOAM/OpenFOAM-1.6/lib/linux64GccDPOpt/libOpenFOAM.so"
[3] #5 Foam::lduMatrix::preconditioner::addasymMatrixCons tructorToTable<Foam::DILUPreconditioner>::New(Foam ::lduMatrix::solver const&, Foam::dictionary const&) in "/home/geza/OpenFOAM/OpenFOAM-1.6/lib/linux64GccDPOpt/libOpenFOAM.so"
[3] #6 Foam::lduMatrix::preconditioner::New(Foam::lduMatr ix::solver const&, Foam::dictionary const&) in "/home/geza/OpenFOAM/OpenFOAM-1.6/lib/linux64GccDPOpt/libOpenFOAM.so"
[3] #7 Foam::PBiCG::solve(Foam::Field<double>&, Foam::Field<double> const&, unsigned char) const in "/home/geza/OpenFOAM/OpenFOAM-1.6/lib/linux64GccDPOpt/libOpenFOAM.so"

Have tried different BC's for k and omega. Have also changed schemes. Nothing helps. Every time I start pimple, it goes to large Co and crashes.
My geometry actually is just a domain with walls, inlet and outlet. If I take simple tube with same BC's, everything is fine.
Have also tried different meshes - all of them works fine with simple, but none of them with pimple.
Have also done LES with piso, and it was also fine. When I use mesh and timestep from LES in pimple with SST, crash again.
Please give me some advice!


Thanks in advance!

Dear Vadims,

I am having a similar problem like you had.

How did you resolve it?

Kind Regards,
Robert

Dito. I do also have this problems.

Did you change your ddtSchemes from steadyState (used with simpleFoam) to Euler ?

Yes, I tried different ddtSchemes with pimple.

Well I get some results, but this results are completly different to the results out of simpleFoam...

@krene

If you are getting results from a transient solver (pimpleFoam) that differ from your steady state solution from simpleFoam, it might be that you are in a transitional state and not fully turbulent. Does your turbulence dissipate? Is it dissipating from your estimated inlet values of your turbulent parameters? I will add that I'm not an LES expert, but how are you comparing the RANS results from simpleFOAM with the LES of your simulations? Also, RANS simulations are known to be dissipative with respect to turbulence (I have no source on this one, only from my notes at the AIChE conference from the turbulence sessions). Just my two cents. Good luck.

Thanks to Gregor and Dan for your responses.

First I'll introduce you my case, how my case looks like, and afterwards where my problems are.

My case is a cylinder with a diameter of D = 300mm and a length L = 2000mm. There is a inlet 160mm from the top of the tank with a diameter of Di = 3mm. The velocity is assumed to be 1.67m/s which leads to a Reynoldsnumber of around 5000. The fluid is water. At the bottom there is an outlet with a diameter of Do = 25mm. While running the experiment, the pressure at the outlet is constant.

I'm interested to simulate the jet at the inlet. The goal is to estimate the entrainment rate.

I'd like to use a transient solver like PISO or PIMPLE to simulate the transient case. I could simulate the case with SIMPLE, but I'm not so happy with the results.
As BC I do have for U:
Inlet: fixedValue
Outlet: zeroGradient
Wall: fixedValue (0 0 0) ==> with zeroGradient, SIMPLE will work for 400s

BC for p:
Inlet: zeroGradient
Outlet: fixedValue, 0
Wall: fixedValue, 0

With this BC, the results looks OK, but there is also the unlogical behaviour, that there is no backflow near the walls.

If I take this BC to the PIMPLE solver, the jet does not look like a jet, it's more a pencil, which goes through a tart. There is no entrainment and there is also no interaction at the boarder between the turbulent and the laminar layer. The same behaviour can be seen by increasing the velocity to 3.33m/s (Re = 10'000).

The results are compared visually, the errors are obviously. Well, I'm not an expert in modelling with LES and RANS. It's the first time I'm working with OpenFOAM, so I do not have to much experiences.

If anyone has time to have a look at my case, I could also upload it here. Thanks a lot.

At the walls, BCs for velocity and pressure should be:

U--->fixedValue (0 0 0)

p--->zeroGradient (not fixedValue!!!)

Regards

V.

Hi Vesselin,
thanks for your replay.
You're right, p should be zeroGradient.

After setup the first model, I got some bad results, so I adjusted the BC so, that I got the wanted results. And the results were good with fixedValue for p at the wall.

I made a run with SIMPLE, PISO and PIMPLE.

The results out of SIMPLE (RANS) are reasonable, also there is the expected backflow in the tank.

With PISO and PIMPLE (LES), the results are still not useable. Same behaviour as before, there is no jet.

Thank you guys so far for your help! :)