[haoyuan]

Hi Foamers,
I recently study the compressible solver rhoCentralFoam in OpenFOAM V4.1. I need to add some non-linear turbulence model in this solver, the viscous term in the present solver is not capable of the non-linear shear stress.

I replace " - fvm::laplacian(muEff, U) - fvc::div(tauMC)' in the momentum equation with turbulence->divDevRhoReff, and it seems to work fine.

However, I replace the "sigmaDotU" in the energy equation with (fvc::dotInterpolate(mesh.Sf(), -devRhoReff))& (a_pos*U_pos + a_neg*U_neg), it goes wrong.

In my opinion, "(muEff)*fvc::grad(U) + tauMC" is the same as "-devRhoReff", which is "((this->alpha_*this->rho_*this->nuEff()))*dev(twoSymm(fvc::grad(this->U_)))" .

Please note that there is a "-" in the functions of "devRhoReff".

Am I right? Does anyone have a similar experience about this?

Regards,

Haoyuan