Turbulence at the cross roads: The utility and drawbacks of13; traditional approaches

This survey argues that traditional approaches to turbulence have led to some notable but incomplete successes, such as the establishment of a plausible connection between diffusion and correlation, and the13; similarity and matchability arguments of Millikan and Kolmogorov.Experience with shear flows may be summarized in the form of some basic working rules, although their validity cannot be considered beyond doubt. On the other hand, traditional approaches seem to ignore the wide prevalence of exponential distributions (rather than13; Gaussians), and have not led to any rational models for turbulent flow yet. For many decades now the scientific initiative has remained with experiment, and extensive work on coherent structures in a variety of flows has suggested novel methods of turbulence management and13; possible new candidates for molecules of turbulence that may enable progress in theory. While the initiative may now be passing to computers at low Reynolds numbers, a reconciliation between the notions of ordered motion and statistical theory is considered likely in the near future, and will probably emerge from greater13; understanding of the apparently random occurrence of coherent events through a dynamical systems approach