We have implemented the 1998 Wilcox k-w model in CFX. We have very good comparison with backward facing step data on velocity, kinetic energy profiles, friction coefficient and heat transfer coefficient. In a comparison we have an almost exact fit with DNS data from stanford. Work is to be continued for more complicated flows.

(1). Where is the source reference of the 1998 Wilcox k-w model?

Turbulence Modeling for CFD Second Edition by David C. Wilcox http://www.dcwindustries.com/

(1). Thank you very much. I had the book for a while. It is the second edition, 1998, first printing July,1998, "Turbulence Modeling for CFD" by David C. Wilcox. (2). In Chapter 4, One-equation and Two-equation Models, Section 4.3.1 The k-w (omega) Model, page-122, it says " To avoid confusion, from this point on, we will refer to Equations(4.36) through (4.42) as the Wilcox(1998) k-w model." (3). The model is a version with improved accuracy, based on the previously published version Wilcox(1988)in AIAA Journal, Vol26,No.11, pp.1299-1310. (4). I think, the new version did not appear in technical journals when the book was published in 1998. In the book reference section, the most recent paper by Wilcox was published in 1995. (5). So, I think, it is important to know that the new version in his 1998 book was not previously published, and is an improved version of his 1988 version, published in AIAA.

For information:

The standard (i.e. not transition) 1998 k-w model (new one!) makes the closure coefficients bounded functions of (a) a vortex stretching parameter similar to the Pope (1978) correction for the round jet plane jet anomaly and (b) a cross-diffusion parameter - what is "left out" of the k-w model when it is converted back into a k-e model. This is meant to improve predictions in free shear flows without effecting the 1988 model's good prediction of boundary layer flows. The model is integrated to the wall and does not require near wall damping functions.

Rick