ON THE RNG THEORY OF TURBULENCE

The Yakhot and Orszag [J. Sci. Comput. 1, 3 (1986) ] RNG theory of turbulence has generated a number of scaling law constants in reasonable quantitative agreement with experiments. The theory itself is highly mathematical, and its assumptions and approximations are not easily appreciated, particularly since several rather unconventional artifices are employed. The present paper reviews the RNG theory and recasts it in more conventional terms using a distinctly different viewpoint. A new formulation based on an alternative interpretation of the origin of the random force is presented, showing that the artificially introduced epsilon in the original theory is an adjustable parameter, thus offering a plausible explanation for the remarkable record of quantitative success of the so-called epsilon-expansion procedure.