[wernsen]

Dear Foamers,

I am having trouble with a rhoSimpleFoam case which is running under 2.0.x without a problem. If I try to run this case under 2.1.x I am gettin the Fatal Error:

--> FOAM FATAL IO ERROR: 
keyword div(phi,K) is undefined in dictionary

What is the capital K and why do I need it since in 2.0.x it is not necessary?

thx
martin

[doubtsincfd]

something is missing in fvSchemes

It might also be a missing comma or a {

Try giving a default scheme for div

[wernsen]

my fvSchemes::divSchemes looks like that:

divSchemes
{
   default            none;
   div(phi,U)                     Gauss upwind;
   div(phi,h)                     Gauss upwind;
   div(phi,k)                     Gauss upwind;
   div(phi,epsilon)               Gauss upwind;
   div(phi,omega)                 Gauss upwind;
   div(U,p)                     Gauss linear;
   dic(phi)              Gauss linear;
   div((muEff*dev2(T(grad(U)))))  Gauss linear;
}

so I already have a scheme for the div(phi,k) but not for div(phi,K) which confuses me. So what is the K?

[maHein]

The formulation of the energy equation has changed from 2.0.x to 2.1x. In the later one, the divergence on the kinetic energy K is part of it. That's why you need div(phi,K) in your fvSchemes. See line 8 of the hEqn.H file:

- fvc::div(phi, 0.5*magSqr(U), "div(phi,K)")

By the way, I noticed considerable differences between results from both versions. I tested both for turbomachinery simulation and the 2.1.x resulted in negative isentropic efficiencies. So make sure to compare both results and see if they are reasonable.

[wernsen]

Thank You very much "maHein"!

[estang]

Originally Posted by maHein

The formulation of the energy equation has changed from 2.0.x to 2.1x. In the later one, the divergence on the kinetic energy K is part of it. That's why you need div(phi,K) in your fvSchemes. See line 8 of the hEqn.H file:

- fvc::div(phi, 0.5*magSqr(U), "div(phi,K)")

By the way, I noticed considerable differences between results from both versions. I tested both for turbomachinery simulation and the 2.1.x resulted in negative isentropic efficiencies. So make sure to compare both results and see if they are reasonable.

I use Version 2.3.0 and i do not understand why the angledDuctExplicitFixedCoeff tutorial for rhoSimpleFoam runs with

div(phi,k)      bounded Gauss upwind;

and the same fvSchemes file doesn't work with my case.

I get the following errormessage:

--> FOAM FATAL IO ERROR:
keyword div(phi,K) is undefined in dictionary "/home/estang/OpenFOAM/estang-2.3.0/run/2_fluid_Testcase_comp_heat_transfer/system/fvSchemes.divSchemes"

file: /home/estang/OpenFOAM/estang-2.3.0/run/2_fluid_Testcase_comp_heat_transfer/system/fvSchemes.divSchemes from line 30 to line 40.

From function dictionary::lookupEntry(const word&, bool, bool) const
in file db/dictionary/dictionary.C at line 437.

FOAM exiting

i just want to understand

[maHein]

It is a capital "k" in your warning: div(phi, K).

[estang]

Originally Posted by maHein

It is a capital "k" in your warning: div(phi, K).

That's why i'm asking here in this thread. I copied the fvSchemes without any edit from the tutorial. I can run the tutorial but when i try to run my own case with the same fvSchemes the solver complains. Shouldn't the same message appear with running the tutorial case?

[maHein]

You have to check whether you use sensibleEnthalpy or sensibleInternalEnergy in your thermodynamicProperties. This changes the form of the energy equation and thus introduces different terms such as div(phi,K).

[estang]

Thank you very much.

Bad that i didn't see the difference in my thermophysical Properties before.

I should have a look to the different formulations.