[motaz.medhat]

I use realizable k-e turbulence mخdel of fluent and I want to change the value of c1e which equal 1.44 to anthor value. I only found that I can edit values of c2e, tke prandtle,tdr prandtle....etc
Please advice

[LuckyTran]

What software....?

[sbaffini]

You are obviously using a cfd software. Now, there are three options:

1) It has a dedicated forum here. You should go there if you want to keep this information secret.

2) It doesn't have a dedicated forum here. We really have no chance to help if you don't tell us which software you are using.

3) Independently from which one of the two above applies, it is actually possible to change it by simply looking better.

[motaz.medhat]

fluent cfd

[sbaffini]

My apologies, I haven't looked at the turbulence models in a while.

The reason for C1e control missing is just because the realizable k-epsilon model doesn't actually have a C1e.

Basically, the epsilon equation in the realizable model is different from the usual one and the production term, the one multiplied by C1e in the standard k-e model, is one of those terms that are different.

See, for example, here:

https://www.afs.enea.it/project/nept.../th/node60.htm

And here (pages 9 and 10 and eqs. 23 and 25):

https://www.researchgate.net/publica...and_Validation

[sbaffini]

What the Fluent manual cites as C1e in the realizable model, at least the one I have access to at the moment, is actually a constant in the buoyancy term, which appears to have also C3e:

Pb = C1e * e/k * C3e * Gb

If this is what you are referring to, then maybe try to activate buoyancy (both in general and in the k-e model if I recall correctly) and see if you have both C1e and C3e... then let us know

[LuckyTran]

Fluent has some limitations on the buoyancy term. Normally the buoyancy term in the epsilon equation is disabled (Gb is 0) and the C1e would never come into effect. If you enable the full buoyancy effects, then Gb is included. The way C3e is implemented in Fluent is also a bit special (following the work of Henkes et al.) and there is not user-definable C3e but instead involving a tanh. I don't have this paper right now to verify what happens to the C1e constant in this approach, it might not even be applicable. Regardless, there's no option to tune either C1e or C3e in the realizable k-epsilon model in Fluent.


Please make sure you actually care about this term.

I was able to verify that there is no C1e specific to the rke model (there's no parameter for it). If it did exist, it would have to be hard-coded or reference by passing another variable (e.g. the C1e from the standard k-epsilon model). I don't think any developer would do this second option. If it did reference the standard model constant, then you could just change the constant in standard k-epsilon and then switch to rke. But again, that would be stupid. I would first read through Henkes' paper first to understand it better.

[vinerm]

used in all models is same. It does not matter whether it is standard, RNG, or realizable. The only difference is in their values. For standard and realizable, value is set to 1.44 while for RNG, it is set to 1.42. It is not really a good idea to change these values, however, if you really want to see the effect, you can modify it using

(rpsetvar 'kec1 1.4)

Default is 1.44, the command above changes it to 1.4

[sbaffini]

Originally Posted by vinerm

used in all models is same. It does not matter whether it is standard, RNG, or realizable. The only difference is in their values. For standard and realizable, value is set to 1.44 while for RNG, it is set to 1.42. It is not really a good idea to change these values, however, if you really want to see the effect, you can modify it using

(rpsetvar 'kec1 1.4)

Default is 1.44, the command above changes it to 1.4

Well, the point now is to know which term this C1e actually multiplies in the realizable epsilon equation: regular production as in the other k-epsilon models (unlikely and in contrast to what the manual says) or just the buoyancy term (more likely at this point)? And more importantly, what the original poster actually wants to achieve?

[vinerm]

One of the major differences in Realizable mode is the equation. It does not contain production term for turbulent kinetic energy per se; only the mean rate of strain tensor is included.

[SergioChachoBatista]

Dear Community

By using Fluent, I want to use rpsetvar to evaluate the effects of changing Swirl Factor on RNG k-e Model.

I am having troubles to find the exact 'var_name for Swirl Factor on (rpsetvar 'var_name value).

Can you please help me?