Coupling of U-Tube and shear layer oscillations in an interconnected flow compartment

Results of experiments conducted in a 2m high flume at large Reynolds numbers are reported in this paper. The flume was partitioned into two compartments. Flow entered the bottom of the upstream test compartment as a wall jet, at jet Reynolds number ranging from 11,000 to 170,000. Periodic oscillations of the free surface in the two compartments resembling the oscillatory flow in a liquid-filled U-tube, and large coherent structures formed above the potential core of the wall jet were observed. Coupling of the U-tube oscillations and vortex shedding is attributed to fluid-dynamic and fluid-resonant feedback processes. For test compartment length, Lc=0.8m , fluid-resonant feedback was found to be dominant, and the shear layer was observed to oscillate at the natural frequency of the two-compartment, U-tube system. The observed U-tube oscillations are initiated by the oscillations of the shear layer at a frequency equal to the subharmonic component for the U-tube. The flow oscillations were generally weaker for Lc=1.2 and 2.0m with oscillation frequencies governed by fluid-dynamic feedback, verified from a comparison with the results from a previously reported study.

A non-darcian numerical modeling in domed enclosures filled with heat-generating porous media

Numerical study of the natural-convection flow and heat transfer in a dome-shaped, heat-generating, porous enclosure is considered. The general conic equation for the top dome is used to consider various conical top sections such as circular, elliptical, parabolic, and hyperbolic. The individual effect of fluid Rayleigh, Darcy, and heat-generating parameters on flow patterns and heat transfer rates are analyzed and presented. The predicted results show that the heat-generating parameter has the most significant contribution toward the growth of bicellular core flow. Moreover, there is significant change in temperature distribution in comparison to rectangular enclosures, due to the existence of the domed-shape top adiabatic cover. The results also show that, regardless of Darcy and Rayleigh values, a flat adiabatic top cover tends to yield the highest value of Nusselt number, followed by circular, elliptical, parabolic, and hyperbolic top covers, respectively.