Turbulent boundary layers. II: Further developments (Q4215078)

We use asymptotic analysis to consider various aspects of turbulent boundary layers in the limit of large Reynolds number. First, some classical results for two-dimensional incompressible flows are reviewed and extended; special emphasis is placed on the numerical algorithms that are required for the prediction of such flows. Then we describe a model for the mean flow profile in the near-wall region of the boundary layer; this model is based on the observed coherent structure of the wall-layer flow and provides a simple alternative to conventional mixing-length formulations. The case of incompressible two-dimensional flow with heat transfer at the wall is addressed in the next section; the asymptotic theory constraints the types of models that can be used in the energy equation and provides an effective way to determine the heat transfer at the surface in a prediction method. The dynamics of the time-dependent turbulent flow considered, and it is argued that the main features of the turbulent boundary layer, including regeneration, can be explained in terms of how hairpin vortices interact with one another and their environment. Attention is returned to the mean flow problem through the consideration of pressure-driven three-dimensional flows. It is shown via asymptotic theories how the streamwise and cross-stream velocities must behave in both the outer part and the wall layer of the boundary layer. We describe embedded function algorithms wherein accurate numerical predictions can be produced by matching an outer-layer numerical solution with a set of embedded functions in the wall layer. Lastly, the compressibility is taken into account at mainstream Mach numbers approaching and throughout the supersonic range. Modifications of existing turbulence models for compressibility are suggested, and a self-consistent general formulation is given in terms of the Howarth-Dorodnitsyn variable. We present a set of self-similar profiles and compare them with experimental data. An extension of the embedded function methodology to the prediction of high-speed compressible flow concludes this work. For the entire collection see [Zbl 0896.00016].