Bubble Foam momentum equation
I want to use the standard momentum equation in bubble Foam rather than the phase intensive one as suggested by Henrik Rusche in his PhD thesis... I got the following problems :
Code:
UEqns.H Code:
/*--------------------------------*- C++ -*----------------------------------*\ Code:
gradientInternalCoeffs cannot be called for a calculatedFvPatchField |
Are you sure about doing this? There are some very good reasons for dividing the momentum equation by the phase fraction! Alberto Passalacqua was working on this, have a read through this thread:
Conservative Two Phase Euler The error is probably due to the boundary condition not being explicitly stated. You can either setup the Uamod field in the 0/folder, use the == operator instead of = (not sure if that works... try it) or Code:
volVectorField Uamod |
thanks for the link ....
|
I read the same arguments in Dr. Rusche's thesis .... The doubt I had was what is exactly meant by equations becoming singular ... I didnt quite understand the effect it would have on the flow ... For the same purpose , I decided to implement momentum eqn in the conservative form and check out how it affected the solution .....
Another doubt I face is : How is the non conservative form of the momentum equation derived ? i.e. how is it possible to divide by alpha when alpha is a variable in itself .... |
Presently I am getting the following error on running the program :
Code:
FOAM FATAL ERROR: |
Another link in case you haven't spotted it is the bubbleFoam wiki article. Again, thanks to Alberto.
I haven't got first hand experience of the difficulties you'll encounter, but effectively in areas where the phase fraction is zero, your momentum becomes zero and the velocity becomes undefined. The matrix probably becomes ill conditioned. I have a full derivation. If you want to see how to get it for yourself, just expand the derivatives in the momentum equation and then divide the whole equation by its phase fraction. |
I just spotted that your UaEqn/UbEqn should have Uamod and Ubmod as the variable to solve for, not Ua and Ub. There may be other mistakes.
|
can u send me a link to the derivation .... thanks ...
|
got it!!! derived it myself ... but are there cases when using a non conservative form for multiphase flow may result in spurious results ...
|
Have you shared (made it open source ) the PEA source code ???
|
Not yet, maybe soon!
|
I have started working on the same ..... right now working on two things ...
A spalding approach to calculate the volume fractions and the second one being the PEA ... Have completed the first one ... and will start the second part soon .... How are you programming the PEA ? As a general elimination algorithm or hard coding it ... |
The PEA is Spalding's approach, no? That's the method where you substitute one momentum equation into the other and rearrange.
I created a multiphase PISO abstract class and have semiImplicit/partialElimination/multiphasePartialElimination derived classes. I don't think you need the PEA for bubbles. It's only necessary when the drag force is large, which is true for small particles in fluidised beds. |
I meant the spalding approach to solving the volume fraction ... discretise the volume fraction eqns and then calculate alpha = alphadis/(alphadis + betadis) where alphadis and betadis are the discretized volume fraction eqns ...
|
I have come up with an implementation of the PEA .. .....
The equations i have used are on this page ... https://sites.google.com/site/balkrishnanitt/eq Can anyone tell me whether the implementation is correct .... Code:
fvVectorMatrix UaEqn(Ua, Ua.dimensions()*dimVol/dimTime); |
The equations are not in that link. Best to check those before code.
You're not solving the equations simultaneously, you're eliminating the other phase's velocity from each velocity equation. |
I have updated the link ....
|
I am getting the following error ... How can it be resolved ??
<code> --> FOAM FATAL ERROR: incompatible fields for operation [Ubmod] == [Ub] From function checkMethod(const fvMatrix<Type>&, const fvMatrix<Type>&) in file /home/ifmg/OpenFOAM/OpenFOAM-1.7.0/src/finiteVolume/lnInclude/fvMatrix.C at line 1181. FOAM aborting #0 Foam::error::printStack(Foam::Ostream&) in "/home/ifmg/OpenFOAM/OpenFOAM-1.7.0/lib/linuxGccDPOpt/libOpenFOAM.so" #1 Foam::error::abort() in "/home/ifmg/OpenFOAM/OpenFOAM-1.7.0/lib/linuxGccDPOpt/libOpenFOAM.so" #2 #3 #4 __libc_start_main in "/lib/tls/i686/cmov/libc.so.6" #5 Aborted </code> When is the function checkMethod executed ?? |
Not very relevant to the original poster's problem, but I'd like to have make sure about something in bubblefoam momentum equation.
Code:
surfaceScalarField phiRa = - fvc::interpolate(nuEffa) *mesh.magSf()*fvc::snGrad(alpha)/fvc::interpolate(alpha + scalar(0.001)); If yes, then why not VSMALL but a random scalar number? |
Not very relevant to the original poster's problem, but I'd like to have make sure about something in bubblefoam momentum equation.
Code:
surfaceScalarField phiRa = If yes, then why not VSMALL but a random scalar number? |
I guess that if you'd use VSMALL, you'd end up with a huge number if alpha1 is indeed zero. To avoid problems the authors chose a small, but not too small value.
|
But what if the alpha values that I am working with are also small? smaller than 0.001? then this hard-coded value will affect my results?
should I decrease this scalar value only few order of magnitudes then? |
help
I want to use bubblefoam solver but it prevents on my work, because of bubblefoam limitations.
BubbleFoam limitations The diameter of the particles constituting the dispersed phase is assumed to be consistent. Aggregation, breakage and coalescence phenomena are not accounted for. What is your suggestion? DO you know any solver uses Euler mixture model ? thank you |
I want to use bubblefoam solver but it prevents on my work, because of bubblefoam limitations.
BubbleFoam limitations The diameter of the particles constituting the dispersed phase is assumed to be consistent. Aggregation, breakage and coalescence phenomena are not accounted for. What is your suggestion? DO you know any solver uses Euler mixture model ? thank you |
All times are GMT -4. The time now is 04:50. |