CFD Online Discussion Forums

CFD Online Discussion Forums (
-   OpenFOAM (
-   -   fvm::laplacian appropriate for d/dx(d/dy) (

sven August 11, 2009 21:40

fvm::laplacian appropriate for d/dx(d/dy)

I am currently working on understanding the source code of the LaunderGibsonRSTM model.

In this model the Reynolds Stress equation reads:


tmp<fvSymmTensorMatrix> REqn
      + fvm::div(phi_, R_)
      - fvm::Sp(fvc::div(phi_), R_)
    //- fvm::laplacian(Cs*(k_/epsilon_)*R_, R_)
      - fvm::laplacian(DREff(), R_)
      + fvm::Sp(Clg1_*epsilon_/k_, R_)
      + (2.0/3.0*(Clg1_ - 1)*I)*epsilon_
      - Clg2_*dev(P)

        // wall reflection terms
      + symm
            I*((yr_.n() & reflect) & yr_.n())
          - 1.5*(yr_.n()*(reflect & yr_.n())
          + (yr_.n() & reflect)*yr_.n())
        )*pow(Cmu_, 0.75)*pow(k_, 1.5)/(kappa_*yr_*epsilon_)

My question now concerns the forth term from the top:


//- fvm::laplacian(Cs*(k_/epsilon_)*R_, R_)
This term seems to represent a part of the diffusion of the turbulent reynolds stresses. However, in all books on turbulence this equation looks like d/dx(Cs*k/epsilon*R*dR/dy), where d/dx and d/dy are derivatives in different directions.

Is it possible to translate this term into OpenFOAM source Code by using the fvm::laplacian scheme, even though the two derivatives (d/dx,d/dy) have different directions?

chai December 10, 2009 13:24

Hi, Sven
I am facing a similar problem. Did you find a solution to your problem?

All times are GMT -4. The time now is 20:48.