The structure of swirling and accelerating turbulent curved shear flow

This thesis describes the effects of plane mean flow curvature and acceleration on the structure of uniformly sheared turbulence. In one experiment, the plane of flow curvature is normal to the plane of the mean shear, and the flow develops a swirling motion in which the mean shear rotates about the streamwise direction. In another set of experiments the flow curvature and acceleration are applied in the same plane as the mean shear. Here the acceleration is applied independently so that its application and removal from curved shear flow can be studied. In all experiments, the mean shear rate was approximately 25 s-1 and the shear flow was allowed to develop a self-preserving structure in a straight tunnel before entering a curved tunnel with a centerline radius of 3 m.;For the swirling flow, the turbulence structure was only weakly affected until the shear had rotated sufficiently so that a stabilized case of two-dimensional curved shear flow was approached. Throughout the development the direction of the turbulence shear stress was found to lag the direction of the mean shear as it rotated around the streamwise direction.;Several accelerating shear flow experiments were performed. These included the simultaneous application of positive acceleration and positive curvature (stabilizing) to shear flow and the application of positive acceleration to well developed curved shear flow. The acceleration was found to have a very strong effect on the turbulence structure and, in particular, it substantially increased the ratio of the turbulence shear stress to the kinetic energy when compared to a nonaccelerating flow of equal curvature. Removal of the acceleration caused a relaxation of the flow structure towards that of a curved shear flow.;Some novel models of the effects of mean flow curvature and acceleration on the structure of turbulence are introduced. The predictions of the model are found to be in qualitative agreement with the experimental data.