[towanda]

Dear community,
I am a beginer using the solver rhoCentralFoam. I am facing a problem and I run out of ideas. I will be glad if someone could give a hand. Here is the problem:

I am using OF2.4.0 on Debian Jessie. I am simulating the flow through a laval nozzle using rhoCentralFoam. The fluid is air. The flow pass from subsonic regime to supersonic regime and agrees with theoretical prediction from the book. Without any problem the simulation reaches the steady state. In the stationary regime, the fields rho,T,p,U calculated agree with the results obtained with other tool (fluent).

The problem is that the mass conservation is not acomplished between the inlet and the outlet. On the outlet I have more mass flow than in the inlet. I think that this could be due to numerical diffusion. I attach the schemes that I am using to solve the problem(fvSchemes).

In particular I was trying to modify how the convective term is solved but the only scheme that I am able to use is:

div(tauMC) Gauss linear;

¿Does someone has a similar problem with the mass conservation? ¿How can I use high order schemes to solve the convective term? Any suggestion to face the problem will be apreciated.

Thank you in advance

[julien91]

Hi Towanda,
Did you manage to understand your problem? I am working with the same problem using rhoCentralFoam, and I get a non constant mass flow rate inside the nozzle.
Thanks for your help

[towanda]

Hello julien91,
I managed to make a little step towards a better understanding of the problem. I have used the run-time function flowRatePatch of OF and I could calculate the mass flow on the inlet patch and on the outlet patch. To use it you should:

1)Add in your controlDict

functions
{
#include "flowRatePatch"
}

2)In the /system/ folder include the file flowRatePatch with the following:

/*--------------------------------*- C++ -*----------------------------------*\
| ========= | |
| \\ / F ield | OpenFOAM: The Open Source CFD Toolbox |
| \\ / O peration | Version: 3.0.0 |
| \\ / A nd | Web: www.OpenFOAM.org |
| \\/ M anipulation | |
\*---------------------------------------------------------------------------*/
FoamFile
{
version 2.0;
format ascii;
class dictionary;
object flowRatePatch;
}
// * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * //

// This packaged function object sums the flux phi on patch faces so the
// calculated volume flow rate for solvers where phi = volumetric flux, and
// mass flow rate for solvers where phi = mass flux.

flowRatePatch
{
patch <name_patch>;

#includeEtc "caseDicts/postProcessing/flowRate/flowRatePatch.cfg"
}

//***********************************************


By using this run-time function , during the running a /postProcessing/ folder will be created and it will contain a file with the mass flux as function of time.
It works well for my transient simulation using rhoCentralFoam solver. During the transient solution may exist compression or expansion of the fluid so the mass flow at the inlet is not equal than the mass flow at the outlet. But when the steady state is reached, the inlet mass flow equals the outlet mass flow (almost equal , it may exist some numerical diffusion) .

I hope this will help

[julien91]

Thanks a lot for your help Towanda, I will try it and I will keep you posted

[lenzo]

Hello there!

Any of you have the clue on how to set the write interval on this post-process?

Because This flowratepatch creates a huge file, and I don't need that many data.

Thank you!

[lenzo]

Originally Posted by lenzo

Hello there!

Any of you have the clue on how to set the write interval on this post-process?

Because This flowratepatch creates a huge file, and I don't need that many data.

Thank you!

Any one?? help

[towanda]

Hello Lenzo,
I am sorry the delay in the reply. I do not know if there is a setting for output times in the postProcess flowRatePatch. In a quick search using gdb I could not find it but I think that debugging the code will allow you to see where the postProcess is solved and perhaps you can find a setting or add one. I am sorry I can not be helpful.

Best Regards

[lenzo]

Dear Towanda,

Lately I found how to reduce the data captured by suing flowRatePatch:

In the controlDict, where it is added the flowRateInlet, you can also add function objects. See the example:



flowRateInlet
{
#include "flowRatePatchInlet";

//other function objects here
//p.e.

writeControl adjustableRunTime;
writeInterval 5e-04;

}