Numerical and experimental investigations of a turbulent junction flow with upstream ribbed surface

Junction flow occurs when a boundary layer encounters an obstacle attached to the same surface. The presence of adverse pressure gradients upstream of the obstacle causes flow separation and generation of horseshoe vortices that wrap around the obstacle. These vortices are highly unsteady and are responsible for high turbulence intensities, high surface shear stress, and larger drag and heat transfer rate.; The objective of the investigations is to provide a method for reducing the strength of the horseshoe vortex in a junction flow which should lead to better design of control surfaces on aircraft, submarines and other vehicles as well as turbomachinery.; This dissertation involves detailed experimental and numerical investigations to study the effects of placing riblets of specific configuration upstream of a NACA 0012 airfoil wing on the flat plate surface to control the strength of the horseshoe vortex. Detailed distributions of the mean and turbulence flow components were obtained at four different streamwise planes. The flow domain was also modeled by a computational Fluid Dynamic program. The numerical solution included solving the Reynolds-Averaged Navier-Stokes equations using various turbulence models. The computational results were validated by comparison with the experimental results.; The experimental results indicate that the riblets reduce the strength of the secondary flow, in particular at the corner and at the immediate downstream interaction regions of the wing. In the downstream wake regions the effects of the riblets in reducing the horseshoe vortices are negligible. The measured drag and shear stress results show overall reduced drag, including reduced skin friction, thus less mixing and heat transfer.; There were obvious qualitative differences in the shapes of the experimental and numerical results, but these do not represent large quantitative differences at most planes. Most of the larger discrepancies were seen in the near-wall and the junction regions. Of the turbulence models evaluated none captured the vortex structure motions, in particular at the wall surface. The Cebeci-Smith model showed better qualitative agreement with the experimental data than the κ-&egr; model.