turbulenc models
 

i am working on the time cost of the RANS models in comparison together in 2D in a ribbed square duct, hence i got an interesting point: the time cost of RNG based on K-e model is 50% (half) of the RM based K-e model.
does any one know the reasons of this result?