Local heat/mass transfer distributions on the bottom surface of a cavity exposed to an approaching turbulent boundary layer

Abstract

Heat/mass transfer on the bottom surface of a rectangular cavity in an incompressible turbulent boundary layer flow is investigated, with emphasis on the effects of cavity width (W) to depth (d) ratio, for a cavity with aspect ratio L/d (length L to depth) of 6. For a wide cavity, this value of L/d is known to establish a flow structure in which the streamwise flow does not reattach on the bottom surface of the cavity, and instead shears past the entire open surface of the cavity, impinging on the downstream wall. A naphthalene sublimation mass transfer technique is used to evaluate local (mass) Stanton numbers at the cavity bottom surface for W/d ranging from 0.51 to 10, and two Reynolds numbers of 8100 and 12,800 defined using the cavity depth and freestream velocity. For W/d>2, the area-averaged mass transfer Stanton number asymptotes to a constant value. Despite the nominal two-dimensional behavior implied by the area-averaged Stanton number for W/d > 2, the details of the local mass transfer distributions show the effects of a system of multiple vortices that remain influential beyond W/d>2 and cause strong spatial variations in transport on the bottom surface. For W/d<2, the area-averaged Stanton number decreases as width is reduced, scaling as a power law with W/d, and appears to be independent of Reynolds number. Computations using the SST k−ω RANS model are able to predict the measured heat/mass transfer at the bottom surface well for W/d>3. For 0.8<W/d<3, the computations are able to capture the trends only qualitatively. For W/d<0.8, computations fail to predict the observed distributions and significantly under-perform.