[gaza]

Hi Kanarya,
Nobody gave respond. I do not know how to properly implement MULES in this solver.

[Kanarya]

I just changed the sign of source term of the mass transfer in order to get condensations and it gives wrong results...some places I have minus condensation rate...do you know why?

Originally Posted by gaza

Hi Kanarya,
Nobody gave respond. I do not know how to properly implement MULES in this solver.

[wyldckat]

Greetings to all!

Kanarya sent me a PM this week asking me to take a look into this thread. The first detail I would like to point out is that I barely know anything about MULES itself, therefore I won't be able to help very much with the algorithm itself.

Nonetheless, I might be able to help with compilation problems and with interpreting the case set-up and physics in play here. Therefore, my requests here are as follows:

1. There were some additional improvements to multiphase modelling in OpenFOAM 3.0 that might be useful for you two, therefore keep in mind to take a look into it in the future: http://www.openfoam.org/version3.0.0/
2. But for right now, I prefer to first isolate the first problem: understanding why the flow profiles with interFoam and the other multiphase solver give such different results. My guess is that the boundary conditions are responsible for this, because if the pressure field is incorrectly defined, this could mean that X percent of the water might evaporate in the first few milliseconds due to high pressure gradients, just because the pressure field was incorrectly defined. On this topic, check this thread: http://www.cfd-online.com/Forums/ope...ity-field.html
   • My request here is that you first try and properly set-up a case using interFoam and interPhaseFoam so that they give similar results. In other words, use OpenFOAM's own solvers and make sure the boundary conditions and initial fields are all well defined. Only then we can evolve to the next step.
3. Next is regarding the new source code. I got confused with the previous posts and I didn't figure out which OpenFOAM versions the new custom code is meant to work with. In addition, I didn't understand if there are still compilation problems or not.
   • Therefore, I need a good description of what is the current status on this part and what versions of OpenFOAM are being used.

Best regards,
Bruno

[Kanarya]

Dear Bruno,

Thanks for the detailed answer!
I am using OF 2.3.1.
I have also compilation problems but lets start first with more basics for example BC for pressure p_rgh:

internalField   uniform 1e5;

boundaryField
{
    hot_wall
{
 type            fixedFluxPressure;
adjoint              0;
     value           uniform 0;

}
    cold_wall
{
 type            fixedFluxPressure;
adjoint              0;
        value           uniform 0;
}
ocold_wall
{
 type            fixedFluxPressure;
adjoint              0;
      value           uniform 0;
}
icold_wall
{
 type            fixedFluxPressure;
adjoint              0;
       value           uniform 0; 
}

outlet
{

type           totalPressure;
        p0              uniform 1e5;
        U               U;
        phi             phi;
        rho             rho;
        psi             none;
        gamma           1;
        value           uniform 1e5;
}
inlet
{
  type            fixedFluxPressure;
adjoint              0;
      value           uniform 0;
        
} 
}

Question 1:
in version OF 2.3.1 fixedFluxPressure is used instead of bouyantPressure, is it correct?
Question 2:
why we need to "define adjoint 0;" as well, because when I am not putting adjoint together with fixedFluxPressure, it complains! do you know why?
Question 3:
My BC for p_rgh make sense?

Thanks in advance!
Best regards,

Kanarya

Originally Posted by wyldckat

Greetings to all!

Kanarya sent me a PM this week asking me to take a look into this thread. The first detail I would like to point out is that I barely know anything about MULES itself, therefore I won't be able to help very much with the algorithm itself.

Nonetheless, I might be able to help with compilation problems and with interpreting the case set-up and physics in play here. Therefore, my requests here are as follows:

1. There were some additional improvements to multiphase modelling in OpenFOAM 3.0 that might be useful for you two, therefore keep in mind to take a look into it in the future: http://www.openfoam.org/version3.0.0/
2. But for right now, I prefer to first isolate the first problem: understanding why the flow profiles with interFoam and the other multiphase solver give such different results. My guess is that the boundary conditions are responsible for this, because if the pressure field is incorrectly defined, this could mean that X percent of the water might evaporate in the first few milliseconds due to high pressure gradients, just because the pressure field was incorrectly defined. On this topic, check this thread: http://www.cfd-online.com/Forums/ope...ity-field.html
   • My request here is that you first try and properly set-up a case using interFoam and interPhaseFoam so that they give similar results. In other words, use OpenFOAM's own solvers and make sure the boundary conditions and initial fields are all well defined. Only then we can evolve to the next step.
3. Next is regarding the new source code. I got confused with the previous posts and I didn't figure out which OpenFOAM versions the new custom code is meant to work with. In addition, I didn't understand if there are still compilation problems or not.
   • Therefore, I need a good description of what is the current status on this part and what versions of OpenFOAM are being used.

Best regards,
Bruno

[wyldckat]

Hi Kanarya,

Originally Posted by Kanarya

Question 1:
in version OF 2.3.1 fixedFluxPressure is used instead of bouyantPressure, is it correct?

Yes. This bug report doesn't give much information about it, but it's something: http://www.openfoam.org/mantisbt/view.php?id=1292
Quoting the User Guide: http://cfd.direct/openfoam/user-guide/damBreak/

At all wall boundaries, the fixedFluxPressure boundary condition is applied to the pressure field, which adjusts the pressure gradient so that the boundary flux matches the velocity boundary condition.

Originally Posted by Kanarya

Question 2:
why we need to "define adjoint 0;" as well, because when I am not putting adjoint together with fixedFluxPressure, it complains! do you know why?

That can't be right! "adjoint" shouldn't be necessary, as detailed in the last comment here: http://www.openfoam.org/mantisbt/view.php?id=1323
Are you 100% certain you are using OpenFOAM 2.3.1?

Are you perhaps using older code built along with OpenFOAM 2.3.1?

Originally Posted by Kanarya

Question 3:
My BC for p_rgh make sense?

Without additional context, seems to make sense.

Best regards,
Bruno

[coccone]

Howdy!

I've been working on a solar heat-pipe application and wanted to start simulating its functioning. I realized that no solver is suitable "as is" not even for a very simplified model of the phenomenon. Either Teqn solution missing or two phase but without phase-change modelling and so on. Given my still rookie skills with OF and C++, I am afraid I don't really have the necessary understanding of solver structures to modify one by my own so I compiled yours on 2.3.1 and i am playing around with it.

At what point are you with the development, are you still working on it? If I got it right, condensation mass flux is set to 0 and there still are some issues with MULES. Could I be of any help in developing/testing? Of course I'm interested in speeding up the whole process, so just let me know...



Ciao

[lowlow]

Originally Posted by gaza

Hi Kanarya,
Nobody gave respond. I do not know how to properly implement MULES in this solver.

Hi Przemek,

I tried the solver that you uploaded, something weird happened when setting up a case to test the

interevapcondphasechangefoam. A water vapor bubble on heating wall with water flowing from left

to right (Re=100). After several time steps, ALPHA is over 1 (look at the fig). ??? is there

anything wrong with my case. or the problem is in the solver?



Besides, The bubble interface temperature was set to saturation temperature, and the T field look not bad. But

I'm not sure with the mass transfer.

Are you still working on it? do you have any progress? waiting for ur reply

regards

L

[gaza]

Hi Lowlow,
I didn't validate solver for more sophisticated case.
Maybe decrease Courant number and increase mesh density.
best regards
Przemek

[gaza]

also I think that MULES is not well implemented in the interEvapCondPhaseChangeFoam
and it could be the reason why alpha exceeds 1.0 value
probably only Henry Wheller knows how to properly implement MULES for this case

[lowlow]

Originally Posted by gaza

also I think that MULES is not well implemented in the interEvapCondPhaseChangeFoam
and it could be the reason why alpha exceeds 1.0 value
probably only Henry Wheller knows how to properly implement MULES for this case

cannot agree more.

[olivierG]

hello,

You should take a look at this: www.cfd-online.com/Forums/openfoam-solving/150535-evapvofhardt-discussion-come-join.html and https://github.com/wyldckat/evapVOFHardt

Work better than the "interPhaseChangeFoam" approach.

regards,
olivier

[wanghongjie]

Originally Posted by gaza

Hi Foamers,
I created solver for evaporation/condensation modeling based on interPhaseChangeFoam. My solver is called interEvapCondPhaseChangeFoam
and is a result of information I read from similar threads on cfd online.

Unfortunately solver does not work well. I run my solver on damBreak case and compared with interFoam but results are different. In attachement there are figures for 0.15 s for both solvers.

I implemented Ganapathy model for mass transfer according to

http://www.sciencedirect.com/science...7931013004341#

I think that something is wrong in my MULES implementation.
Here are all the necessary files

http://fluid.itcmp.pwr.wroc.pl/~pblasiak/download.html

Here are the commands to type in terminal to compile my solver and run damBreak test case for OpenFOAM 2.3.0.:

unzip interEvapCondPhaseChangeFoam_allfiles.zip
tar xzvf incompressible.tar.gz
tar xzvf interEvapCondPhaseChangeFoam.tar.gz
tar xzvf damBreak_interEvapCondPhaseChangeFoam.tar.gz
cp -r incompressible $WM_PROJECT_USER_DIR/src/transportModels/
cp -r interEvapCondPhaseChangeFoam $WM_PROJECT_USER_DIR/applications/solvers/multiphase/
cp -r damBreak $FOAM_RUN/tutorials/multiphase/
cd $WM_PROJECT_USER_DIR/src/transportModels/incompressible
wclean
wmake libso
cd $WM_PROJECT_USER_DIR/applications/solvers/multiphase/interEvapCondPhaseChangeFoam
./Allwclean
./Allwmake
cd $FOAM_RUN/tutorials/multiphase/interEvapCondPhaseChangeFoam/damBreak
./Allrun

As I said the solver is a result of available knowledge on cfd online but
it still does not give correct results. I have also test case for one dimensional Stefan problem and results are close to analitycal but error is too large. If anybody interested I can upload this test case too.

Can anybody help to fix this solver?

I can not find any solver in this link, could you send me again, thanks.
http://fluid.itcmp.pwr.wroc.pl/~pblasiak/download.html

[rocco96]

Hi, is this solver, still avaible? This link doesn't work.
Thank you!

[gaza]

Originally Posted by rocco96

Hi, is this solver, still avaible? This link doesn't work.
Thank you!

No it is not available.