ICE simulation using RNG k-eps model - adjust turbulent dissipation constant
 

Hi all,


I have some questions about adjusting turbulent dissipation constant in RNG k-epsilon model. In Converge, this constant's default value is -1. However, some folks who use Star-CCM before suggest that this constant should be +0.33 for more realistic. What they were told is that, the story of this value changed was originally from Toyota ICE R&D center in Japan. I think we may need something more convincible to explain it.


My questions are:
1. Why this constant is default as -1 in RNG k-epsilon model?

2. Are there any suggested values for this constant while simulating ICE transient process and steady-flow process as well?


Thank you very much. You are more than welcome to share your thought on this thread.

SC

Hey Shanghua chen,

This is Basu Parmar from CONVERGE support. I cannot comment on Star-CCM, since different solvers can have different implementations of the same models.


In CONVERGE, the Ceps3 term relates to dilatation in the dissipation equation, i.e. Ceps3*density*(divergence(velocity))*epsilon.


1. For incompressible flows, this term goes to zero so the value of Ceps3 is irrelevant.
2. Its simply -1 in standard RNG model and even in standard k-eps model for compressible flows
3. In rapid distortion limits, use the Rapid distortion RNG model available in CONVERGE, which autocalculates this coefficient. The same goes for Generalized RNG model.


You can look at the equations for these models in CONVERGE manual. Please take a look at this article which describes these models in detail and what model to use in different scenarios:

Wang, Bao-Lin, Paul C. Miles, Rolf D. Reitz, Zhiyu Han, and Benjamin Petersen. Assessment of RNG turbulence modeling and the development of a generalized RNG closure model. No. 2011-01-0829. SAE Technical Paper, 2011.

Thanks.

Thank you Basu. This is very helpful.