Interaction of a longitudinal vortex with a three-dimensional, turbulent boundary layer

Abstract

THE distortion of a pressure-driven, three-dimensional boundary layer by an embedded longitudinal vortex was studied experimentally. The peak vorticity and secondary velocities were found to decay much more rapidly in a threedimensional boundary layer than in a similar two-dimensional layer. The distortion of the boundary layer was strongly dependent on the sign of the vortex. Contents Longitudinal vortices are used frequently to manage separation on aircraft surfaces and in internal flow passages. Such vortices when embedded in a turbulent boundary layer cause large distortion in both the mean velocity and turbulent stress fields. There have been extensive experimental investigations1'2 examining vortex/boundary-layer interactions in support of turbulence model development for these complex but important flows. The previous detailed studies have all been conducted in flows where a longitudinal vortex interacts with an otherwise two-dimensional boundary layer. However, in virtually all of the applications, the boundary layer of interest is three-dimensional, the flow on a swept wing with attached vortex generators being the classic example. Three-dimensional boundary layers have distributed mean flow longitudinal vorticity which may interact with the rolled up longitudinal vortex. It will be shown here that this interaction leads to substantial differences in the development of the vortex and the boundary layer mean velocity field. Full details of the experiment and results are outlined in Ref. 3. The experiments were conducted in the attached threedimensional boundary-layer apparatus used by Anderson and Eaton.4 A two-dimensional boundary layer with a momentum thickness Reynolds number of 3750 at a freestream velocity of 16.3 m/s approached a 60-deg included angle wedge which split the boundary layer and deflected it in the spanwise direction. The flow near the wall turned through a larger angle than the freestream creating strong skewing across the boundary layer. The resulting three-dimensional boundary layer contained distributed longitudinal vorticity of negative sign in the outer layer and concentrated positive vorticity in a thin layer adjacent to the wall. A longitudinal vortex was introduced at the end of the two-dimensional development section using a 2-cm high by 4.7-cm long half-delta vortex generator mounted to the wall. Both signs of the longitudinal vortex were investigated; case 1 was a vortex with negative circulation, and