selecting turbulence model and its boundary -- CFD Online Discussion Forums

June 22, 2016, 07:13

selecting turbulence model and its boundary

BakedAlmonds

Member

Almond Wong

Join Date: May 2016

Posts: 68

Rep Power: 9

Hi,

I am looking into the different turbulence model and would like to know more about it.
At the moment, I am looking into k-epsilon, k-omega and RSM model.

k-epsilon and k-omega
These are common RANS model used in OpenFOAM, aka RAS. Depending on which model is chosen, either epsilon or omega has to be defined.
Should a wallfunction (ie { type kqRWallFunction; value uniform 0.0; } ) or intensityfunction { type turbulentIntensityKineticEnergyInlet; intensity 0.05; value uniform 1; } be a better choice as BC for the two models?

Or are the two BC actually the same since the intensity, k and e are actually all related through: (ref: http://www.cfd-online.com/Wiki/Turbu...ary_conditions)

Quote:

Originally Posted by l_r_mcglashan

$k = \frac{3}{2} \left(\bar{U} I \right)^2$
$\varepsilon = C_{\mu}^{3/4} \frac{k^{3/2}}{l}$
$\omega = \frac{\sqrt{k}}{l}$

Where I = turbulent intensity and l = mixing length.

This bring me to another question. How can the turbelence mixing length be defined for an open flow (ie over a flat plate or terrian) where it is defined by % of enclosed height (ref: http://www.cfd-online.com/Wiki/Turbulent_length_scale)

RSM
For this model, I have yet to understand it much. Am I right to say that for OpenFOAM RSM is operated through an LRR RASModel?

It will be great if more info could be provided for this model.

Regards