[shanvach]

Hi all,

I am trying to add local time stepping to twoPhaseEulerFoam. I am using reactingTwoPhaseEulerFoam as a reference. I have made all the necessary changes to twoPhaseEulerFoam.C, twoPhaseSystem.C and createFields.H. However I am getting the following error

--> FOAM FATAL ERROR:

    request for volScalarField rSubDeltaT from objectRegistry region0 failed
    available objects of type volScalarField are

25
(
thermo:alpha.air
alpha.air_0
thermo:alpha.water
psi
alpha.water
T.air
thermo:psi.water
T.water
p_rgh
dpdt
gh
thermo:psi.air
rDeltaT
thermo:rho.air
alpha.air_0_
p
K.air
thermo:rho.water
K.water
alpha.air
thermo:mu.water
e.air
thermo:mu.air
dgdt
e.water
)


    From function const Type& Foam::objectRegistry::lookupObject(const Foam::word&) const [with Type = Foam::GeometricField<double, Foam::fvPatchField, Foam::volMesh>]
    in file /opt/CFDSupportFOAM4.0/beta/OpenFOAM-dev/src/OpenFOAM/lnInclude/objectRegistryTemplates.C at line 193.

FOAM aborting

I think I have missed including some file in my twoPhaseEulerFoam.C .
Can anyone help me out with this problem?

Your help in this matter is greatly appreciated.

Thanks and Regards,
Shantanu Vachhani

[wyldckat]

Quick answers: I received your PM but I was only able to try and look into this right now.

Without being able to see the modified code, it's fairly hard to diagnose the actual issue, specially since I've never done this myself. The best I can do is point you to commits where this change has been done:

• Recent change in pimpleFoam: https://github.com/OpenFOAM/OpenFOAM...84e00a2ea34ec5
• reactingParcelFilmFoam: 9d928c3a98c89fb4ea9045a9c627addaa2c3ba86

These are just 2 complete-looking examples. You can get more examples if you git clone the OpenFOAM-dev repository then use gitk to look for commits with "LTS" in their descriptions...


From the PM you sent me:

Originally Posted by shanvach

PS:- This might be a bit of stretch but I would like to have your advice. Basically I am simulating blood flow using twoPhaseEulerFoam where I want to arrive at a steady state solution. I tried to change the solver for steady state but without any success. So right now I am running a transient solution and hence trying to implement LTS to speed up the solution. Do you have any suggestions to arrive at the steady state solution faster?

I don't know is this solver works with larger time steps, but if PIMPLE works as intended with it, see this wiki page: https://openfoamwiki.net/index.php/O...hm_in_OpenFOAM

The idea is that you can increase the "maxCo" value in "controlDict", as long as you also increase the "nOuterCorrectors" accordingly... but you try in a separate run, to see how it affects the simulation and how to calibrate it to your specific case. From my experience, sometimes it works with fairly high Courant Numbers (have used maxCo 40 and some 6-8 nOuterCorrectors with it), but the results weren't exactly trustworthy... the time diffusion was too high and there were instances where it generated artificial mass to compensate for an instability.

Beyond this, I would suggest a coarser mesh, so that the time steps could be higher... if possible.