Computational Modeling of Turbulent Flows

This chapter discusses that in many situations of practical importance, “second-order modeling” technique makes possible computations that often agree with what data is available. Inevitably, the technique is also applied in many situations in which data does not exist, which must be regarded as a dangerous practice as the limitations of the technique are not known with any precision. It is primarily the possibility of practical computation that has been responsible for the great interest in this method. Even in its most stripped-down form, it results in general in the simultaneous solution of four partial differential equations in the domain of interest; more elaborate models in a three-dimensional situation might require the simultaneous solution of as many as 36 partial differential equations to obtain the mechanical field only. This is within the capabilities of present computers at a reasonable price, which cannot be said of any other technique. The chapter explores that direct simulation is not an alternative for practical computation. The various sophisticated closures suffer from essentially the same problems as the direct simulations and therefore, are limited to homogeneous situations. Thus, the second-order modeling is the only possibility for practical computation. © 1978, Academic Press Inc.