[Gerry Kan]

Howdy Folks:

I am wondering if others experience a similar issue:

I have a case using rhoPimpleFoam that works on (single and multiple) processes on a single machine. But the same job would choke on the following error:

--> FOAM FATAL IO ERROR:
keyword div((nuEff*dev2(T(grad(U))))) is undefined in dictionary "IOstream.divSchemes"
file: IOstream.divSchemes from line 0[10]

Here is the divSchemes block in my system/fvSchemes file:

divSchemes  {
    default             none;
    div(phi,U)          Gauss limitedLinearV 1;
    div(phi,h)          Gauss limitedLinear 1;
    div(phi,K)          Gauss limitedLinear 1;
    div(phid,p)         Gauss limitedLinear 1;
    div(phi,epsilon)    Gauss limitedLinear 1;
    div(((rho*nuEff)*dev2(T(grad(U)))))     Gauss linear;
}

Any ideas what might have caused this?

Thanks in advance, Gerry.

[pete20r2]

Maybe the excessive bracketing?

[Santiago]

Originally Posted by Gerry Kan

Howdy Folks:

I am wondering if others experience a similar issue:

I have a case using rhoPimpleFoam that works on (single and multiple) processes on a single machine. But the same job would choke on the following error:

--> FOAM FATAL IO ERROR:
keyword div((nuEff*dev2(T(grad(U))))) is undefined in dictionary "IOstream.divSchemes"
file: IOstream.divSchemes from line 0[10]

Here is the divSchemes block in my system/fvSchemes file:

divSchemes  {
    default             none;
    div(phi,U)          Gauss limitedLinearV 1;
    div(phi,h)          Gauss limitedLinear 1;
    div(phi,K)          Gauss limitedLinear 1;
    div(phid,p)         Gauss limitedLinear 1;
    div(phi,epsilon)    Gauss limitedLinear 1;
    div(((rho*nuEff)*dev2(T(grad(U)))))     Gauss linear;
}

Any ideas what might have caused this?

Thanks in advance, Gerry.

Ehhh.... you have a rho that shouldnt be there...

[simrego]

I think you are using a different solver. Not what you need.
rhoPimpleFoam need div(((rho*nuEff)*dev2(T(grad(U))))) what you got in fvSchemes.
But the solver what you are running need div((nuEff*dev2(T(grad(U))))).
I think this is an incompressible solver. Can you check the beginning of the log?

[Gerry Kan]

Folks:

I think I know what happened. I have a divergence definition for rho*nuEff, but I also need one for rho, so that OpenFOAM knows what to do with rho *and* nuEff. So my divSchemes looks like

divSchemes  {
    default             none;
    div(phi,U)          Gauss limitedLinearV 1;
    div(phi,h)          Gauss limitedLinear 1;
    div(phi,K)          Gauss limitedLinear 1;
    div(phi,rho)        Gauss linear;
    div(phid,p)         Gauss limitedLinear 1;
    div(phi,epsilon)    Gauss limitedLinear 1;
    div(((rho*nuEff)*dev2(T(grad(U)))))  Gauss linear;
}

Now rhoPimpleFoam runs, although it crashes right on 1st iteration. This is some other problem.

Thanks for the ideas and at least this problem is solved, Gerry.

[moz455]

Hi,
I want to use the "Gauss limitedLinear 1" scheme for div((nuEff*dev2(T(grad(U))))) but I get an error in simpleFoam. Whay does this scheme does not work for this term?

Originally Posted by simrego

I think you are using a different solver. Not what you need.
rhoPimpleFoam need div(((rho*nuEff)*dev2(T(grad(U))))) what you got in fvSchemes.
But the solver what you are running need div((nuEff*dev2(T(grad(U))))).
I think this is an incompressible solver. Can you check the beginning of the log?