[sippanspojk]

Hi @mkraposhin,

I am facing a similar challange as openfoammaofnepo did however, mine is a bit different.

I am writing a functionObject that I use for parallel computing. I define one or a few patches (let's say I select Building from Inlet, Outlet, Tunnel & Building) and I read the face pressure values on that patch. Then I do some computations and lastly write that to a custom made field.

My problem is that I don't know how to refer the pressure value of a face on one processor with the global face ID. Meaning I could read the p-value on face A, do my computations and then write my new value to face C.

I found this thread and it seams to be more or less my answer however, I do have some quesitions:

1) How would I implement a patch selection? I have a dict file where I define some patch names and that files is read into the list patchSet, which is a list of integers.

2) Can I implement your example code directly in my functionObject or do I need to do that in a solver?

I would be very gratefull for your help on this.

Cheers,
David

[sippanspojk]

Hi @mkraposhin,

I am facing a similar challange as openfoammaofnepo did however, mine is a bit different.

I am writing a functionObject that I use for parallel computing. I define one or a few patches (let's say I select Building from Inlet, Outlet, Tunnel & Building) and I read the face pressure values on that patch. Then I do some computations and lastly write that to a custom made field.

My problem is that I don't know how to refer the pressure value of a face on one processor with the global face ID. Meaning I could read the p-value on face A, do my computations and then write my new value to face C.

I found this thread and it seams to be more or less my answer however, I do have some quesitions:

1) How would I implement a patch selection? I have a dict file where I define some patch names and that files is read into the list patchSet, which is a list of integers.

2) Can I implement your example code directly in my functionObject or do I need to do that in a solver?

I would be very gratefull for your help on this.

Cheers,
David

[bany]

Originally Posted by openfoammaofnepo

Dear Matvey,

How did you define the "localmesh" in the following:

    //read local cell addressing
    labelIOList localCellProcAddr
    (
        IOobject
        (
            "cellProcAddressing",
            localMesh.facesInstance(),
            localMesh.meshSubDir,
            localMesh,
            IOobject::MUST_READ,
            IOobject::NO_WRITE
         )
    );

Thank you. When I compiled, I have the error information:

CFD_Cell_Local2Glob.H(27): error: identifier "localMesh" is undefined
            localMesh.facesInstance(),

The error can be solved by :

const fvMesh& localMesh(mesh);

In addition, the code can be simplfied by using

gatherList

List<List<label>> processCellToGlobalAddr_;
List<label> globalCellToProcessAddr_;

if (Pstream::parRun())
{
	processCellToGlobalAddr_.resize(Pstream::nProcs());
	 const fvMesh& localMesh(mesh);

	//read local cell addressing
	labelIOList localCellProcAddr
	(
	    IOobject
	    (
			"cellProcAddressing",
			localMesh.facesInstance(),
			localMesh.meshSubDir,
			localMesh,
			IOobject::MUST_READ,
			IOobject::NO_WRITE
	    )
	);
	
	processCellToGlobalAddr_[Pstream::myProcNo()] = localCellProcAddr;
	Pstream::gatherList(processCellToGlobalAddr_);
}

If the usage of "gatherList" has errors, please let me know.