[EricS]

Hello everyone!

I'm trying to solve a problem with a gradient but I can't figure out how to get the value of a field, in this case Temperature, adjacent to a face.
So far I wrote this:

const scalarField& Tp = 
          patch().lookupPatchField<volScalarField, scalar>("T"); 
const scalarField& Tc = 
          patch().lookupPatchField<volScalarField, scalar>("T").internalField();
const scalarField& deltaInv = patch().deltaCoeffs(); 
const scalarField gradT_ = - (Tp-Tc)*deltaInv;

But if I write the size of Tp and Tc I get different values (namely 2000 and 80000).


I'm sure I'm doing something wrong with the "internalField()" member function, but ... what? Do you have an idea?

Thank you!

[olesen]

Originally Posted by EricS

Hello everyone!

I'm trying to solve a problem with a gradient but I can't figure out how to get the value of a field, in this case Temperature, adjacent to a face.
So far I wrote this:

const scalarField& Tp = 
          patch().lookupPatchField<volScalarField, scalar>("T"); 
const scalarField& Tc = 
          patch().lookupPatchField<volScalarField, scalar>("T").internalField();
const scalarField& deltaInv = patch().deltaCoeffs(); 
const scalarField gradT_ = - (Tp-Tc)*deltaInv;

But if I write the size of Tp and Tc I get different values (namely 2000 and 80000).


I'm sure I'm doing something wrong with the "internalField()" member function, but ... what? Do you have an idea?

Thank you!

I'll take a not-so-wild guess here and bet that your mesh has 8000 cells in it. In which case, the value returned by internalField() is correct. If you actually wanted the values of the internal field immediately next to the patch, that would be a different method call. Instead of providing a quick answer, I'd suggest you take a look at fvPatch.H to see what the first lookupPatchField is doing. After that, you'll probably (hopefully) be motivated to look at fvPatchFIeld.H for which method you actually wanted to use. This is the best way to understand which code to write. I am confident that you will find the answer - please post it when you find it.

[EricS]

Hi Mark, thank you for your reply and yes, I think I found a not very efficient solution to the problem, but it works! The code I wrote is:

    // store in "Tp" the field of scalar temperature on the patch
      const scalarField& Tp(patch().lookupPatchField<volScalarField, scalar>("T")); 
                
    // store in "Tc" the field of scalar temperature in the computational domain
      const scalarField& Tc(patch().lookupPatchField<volScalarField, scalar>("T").internalField());         
    //Info << "Size of Tc: " << Tc.size() << endl;
    
    // initialize cell's temperature variable
    scalarField Tcell(Tp.size(), Zero);
        
    // distance from the face center to the cell center on patch's cells
    const scalarField& deltaInv(patch().deltaCoeffs()); 
    
    // separately calculate the homogenization coefficient "lambda_theta"        
    const scalarField Coeff(kF_/kS_*solidThickness_*deltaInv);

    // store in "cells" the labels of the cells adjacent to the face
    const labelList cells(patch().faceCells());
    
    // initialize the counter
    label cellN(0);
    forAll(cells, patchCell) {
        cellN = cells[patchCell];
        Tcell[patchCell] = Tc[cellN];
    }

Its purpose is to work with temperatures on the cells adjacent to a patch. The important part is:

    
label cellN(0);
    forAll(cells, patchCell) {
        cellN = cells[patchCell];
        Tcell[patchCell] = Tc[cellN];
    }

If anyone has ideas on how to make it more efficient please feel free to tell me how to improve it since now it is quite slow!

[olesen]

If you look at polyPatch.H, you'll see there is a patchInternalList method that returns the values you are looking for.