[gu1]

Hi,

I would like help to understand the reason for having a function intended for epsilon() in the KEqn model, which is an one SGS equation model and I did not see where the value obtained in the function is used.

template<class BasicMomentumTransportModel>
tmp<volScalarField> kEqn<BasicMomentumTransportModel>::epsilon() const
{
    return volScalarField::New
    (
        this->groupName("epsilon"),
        this->Ce_*k()*sqrt(k())/this->delta()
    );
}

Thanks

[Tobermory]

It's just a utility function, to allow you to calculate epsilon if you need it.

[gu1]

Originally Posted by Tobermory

It's just a utility function, to allow you to calculate epsilon if you need it.

...and how I can calc? How to invoke the utility function?

Thanks.

[Tobermory]

You first need to get a reference to the instantiated kEqn object, let's call that turbulence, then you just write turbulence.epsilon().

Here's an analogous situation, from a quick "grep -r" on "epsilon()" in the src folder:

     const compressibleMomentumTransportModel& turbulence =
         combModel_.turbulence();
 
     // Total strain
     const volScalarField sigmaTotal
     (
         sigma + alpha_*turbulence.epsilon()/(turbulence.k() + kMin)
     );

which shows both the epsilon() and k() access functions in action.

[gu1]

Originally Posted by Tobermory

You first need to get a reference to the instantiated kEqn object, let's call that turbulence, then you just write turbulence.epsilon().

I understood.
I believe I'm even in the wrong forum, the right thing to do would be to question it in post-processing.

What I want to learn is, once I have finished my analysis using the KEqn model, how do I calculate the epsilon for the last time of the simulation? Can I access it from outside, something like: 'postProcess -func calc epsilon -latestTime'?

Thanks

[Tobermory]

Aaah - okay. There's no direct way, I think, unless you can get the cacheing to work, as per my response on an earlier post - Production and Destruction terms.

Probably the best way to do this would be to write it your own functionObject in which you generate the reference to the kEqn object, call the epsilon access function and then write the resulting volScalar field. You could do this as a coded FO if it's a one-off, or implement as your own full FO if you plan to use it regularly.

[gu1]

Ok. Thanks.

I even made a post about 'R', present in the multiphase model. I know he's not there for nothing, I just need to figure out what to call him.