The effect of rib installation on heat transfer enhancement of two-dimensional impingement jet has been studied in the wall jet region by mass transfer experiments using a naphthalene sublimation method and numerical simulations with Reynolds-Averaged Navier-Stokes (RANS) models. The effect of rib-height, rib-location, and distance between the ribs on the heat transfer enhancement was systematically examined. As a result, the rib installation was found to induce flow separation and reattachment behind the rib and to promote heat transfer in the wall jet region. The local peak Nusselt number at the reattachment point attained the highest value when the rib height was as tall as the corresponding boundary layer thickness of a no-rib case at the rib position. There is an optimal rib-height to maximize the area-averaged Nusselt number, which is determined by a trade-off between the peak Nusselt number and heat transfer area enhanced by the rib. From numerical results, it appeared that the Abe-Kondoh-Nagano (AKN) model was found to predict reattachment length best among three RANS models, which then resulted in good agreement of the local Nusselt number distributions between experiments and simulations except the stagnation and reattachment regions, where over-and under-estimation occurred.

抄録全体を表示

  Large Eddy Simulation (LES) was conducted to investigate two-dimensional impingement jet heat transfer enhanced by a square-rib turbulator placed in the wall jet region, where rapid decay of local heat transfer rate occurs in the case of impinging jet on a flat target wall. The effect of rib height on time-mean local Nusselt number upstream and downstream the square-rib was systematically examined by changing the height from 0.2 to 0.8 mm at an interval of 0.1 mm. In order to validate current LES, predicted time-mean local Nusselt number distribution was compared with the measured data by naphthalene sublimation method and found to be in good agreement with the experimental data except the jet-stagnation and flow-reattachment regions, where over- and under -estimation occurs. LES has successfully reproduced the decaying tendency of local Nusselt number downstream the flow reattachment point, while a RANS model failed to predict it. Concurrently performed LES on each rib-height realized the direct comparison of instantaneous Nusselt number among different ribs, and it was found that turbulent structures around the rib turbulator are not affected by the rib-height in the upstream region, but much affected by the rib-height in the downstream region.

抄録全体を表示