[superkelle]

** Prandtl : )

I worked me throug the source code of kOmegaSST and notices that in the model Coefficients no turbulent Prantl number is set. So I went to the calulation of alphat in



/src/TurbulenceModels/compressible/EddyDiffusivity


and there Prt is just set to 1.0 when not defined in your simulation constant/turbulenceProperties file.


I think normally Prt is assumed to be 0.8 to 0.9like in:
"ELEMENTS OF INDUSTRIAL HEAT TRANSFER PREDICTIONS " by Menter



So do you also change Ptr or does everyone go with Ptr=1.0 because it is maybe not big of a deal?


I changed it

simulationType  RAS;

RAS
{
    RASModel        kOmegaSST;

    turbulence      on;

    printCoeffs     on;

kOmegaSSTCoeffs
    {
        Prt             0.85;
    }
}

If I do so I also get Prt printed in the list of coefficients. Otherwise Prt is not even mentioned.

[agustinvo]

Hi superkelle,


the turbulent Prandtl number is not an obvious question. It depends on the flow conditions rather than the fluid properties.


It is usually set to 0.85-0.9-1 for Prandtl numbers close to 1 or above. At that stage this approach is a fair estimation.


If you go for low Prandtl values, then this appraoch is not as valid as before. Using a larger Prt value (around 2) is the usual approach, but the turbulence modeling must be reconsidered.


I hope it serves you for setting your Prt value!

[superkelle]

Originally Posted by agustinvo

Hi superkelle,
the turbulent Prandtl number is not an obvious question. It depends on the flow conditions rather than the fluid properties.

Thank you for your response. Is Prt a pure experimental to determin? I looked into different textbooks and they were all very quiet about Prt I only found the statement that it should be around 0.9 for air water like fluids , and for liquid metals at around 3. And also in OF tutorials Prt is very rarely set directly in RAS properties.

[agustinvo]

Hi,


it is not present in some RANS properties as the few incompressible solvers that use it they define it at transportProperties. For compressible flows, as you're forced to solve the energy equation you must provide it at the RANS properties.


The value you set depends on your application and working Pr value. You can have a guess with Prt=0.85 and then modify it if needed. You can check similar studies to yours to see what they propose for Prt.