[MafioTia]

Hi everyone.

I'm trying to implement a fluent-type wall function sensitized to pressure gradient, i.e. "non-equilibrium WF". There is no problem in epsilon/omega WF when I get the pressure gradient, needed for my calculation, in this way:

const volScalarField& P = db().lookupObject<volScalarField>("p");
volVectorField grdntP = fvc::grad(P);

If I apply the same procedure in nutSpalartAllmarasWF ( that follows the general Spalding law-of-the-wall but it has to be sentivized to pressure gradient, too!) the solution explode in an unexplainable way. I've already tried to debug my code on every line that I added and I've concluded that the problematic line is the second I wrote before.
My idea is that this procedure applied to nut-WF result in an unintentional modify of the pressure field, but I thin it's a weak idea, if I had understood OF works.

Any idea?

[maolongliu]

Hi, have you solved this problem?
I am now also trying to write a new wall function including pressure gradient.

Originally Posted by MafioTia

Hi everyone.

I'm trying to implement a fluent-type wall function sensitized to pressure gradient, i.e. "non-equilibrium WF". There is no problem in epsilon/omega WF when I get the pressure gradient, needed for my calculation, in this way:

const volScalarField& P = db().lookupObject<volScalarField>("p");
volVectorField grdntP = fvc::grad(P);

If I apply the same procedure in nutSpalartAllmarasWF ( that follows the general Spalding law-of-the-wall but it has to be sentivized to pressure gradient, too!) the solution explode in an unexplainable way. I've already tried to debug my code on every line that I added and I've concluded that the problematic line is the second I wrote before.
My idea is that this procedure applied to nut-WF result in an unintentional modify of the pressure field, but I thin it's a weak idea, if I had understood OF works.

Any idea?