[samwong]

Hi All,

I want to set the following condition for T for my problem.
i.e. some Flux due to Temperature and Pressure to be equal.



I have tried to do this by calling lookupPatchField, then using Tp.snGrad().
Then I read Eugene's post for setting



by calling lookupObject then using fvc::grad.

I think the difference between fvc::grad and snGrad() is that,
fvc::grad use both internalField and patchField,
but snGrad() only use the patchField (boundary face and its cell's cell-center).
I wonder if using fvc::grad is better(more rigorous) then snGrad().

Eugene's post:
Second Derivative Zero - Boundary Condition

His code:

void extrapolatedGradientFvPatchScalarField::updateCoeffs()
{
    if (updated())
    {
        return;
    }
    
    word field = this->dimensionedInternalField().name();
    
    vectorField nf = patch().nf();
    
    volVectorField gradField 
        = fvc::grad(db().lookupObject<volScalarField>(field));

    gradient() 
        = (nf & gradField.boundaryField()[patch().index()]
        .patchInternalField()());
    
    fixedGradientFvPatchScalarField::updateCoeffs();
}

Here is my code

void equalGradientFvPatchScalarField::updateCoeffs()
{
    if (updated())
    {
        return;
    }
    const fvPatchField<scalar>& Pp =                                                                                 
           patch().lookupPatchField<volScalarField, scalar>(PName_);   
    gradient() =  - Pp.snGrad();
    
    fixedGradientFvPatchScalarField::updateCoeffs();
}

Thanks,
Sam