The boundary layer of swimming fish.

Tangential and normal velocity profiles of the boundary layer surrounding live swimming fish were determined by digital particle tracking velocimetry, DPTV. Two species were examined: the scup Stenotomus chrysops, a carangiform swimmer, and the smooth dogfish Mustelus canis, an anguilliform swimmer. Measurements were taken at several locations over the surfaces of the fish and throughout complete undulatory cycles of their propulsive motions. The Reynolds number based on length, Re, ranged from 3x10(3) to 3x10(5). In general, boundary layer profiles were found to match known laminar and turbulent profiles including those of Blasius, Falkner and Skan and the law of the wall. In still water, boundary layer profile shape always suggested laminar flow. In flowing water, boundary layer profile shape suggested laminar flow at lower Reynolds numbers and turbulent flow at the highest Reynolds numbers. In some cases, oscillation between laminar and turbulent profile shapes with body phase was observed. Local friction coefficients, boundary layer thickness and fluid velocities at the edge of the boundary layer were suggestive of local oscillatory and mean streamwise acceleration of the boundary layer. The behavior of these variables differed significantly in the boundary layer over a rigid fish. Total skin friction was determined. Swimming fish were found to experience greater friction drag than the same fish stretched straight in the flow. Nevertheless, the power necessary to overcome friction drag was determined to be within previous experimentally measured power outputs. No separation of the boundary layer was observed around swimming fish, suggesting negligible form drag. Inflected boundary layers, suggestive of incipient separation, were observed sporadically, but appeared to be stabilized at later phases of the undulatory cycle. These phenomena may be evidence of hydrodynamic sensing and response towards the optimization of swimming performance.