Prediction of Heat Transfer & Flow in Rotating Two-Pass Channels Connected by Holes
A computational study of the effect of rotation on the velocity and thermal field for a two-pass channel connected by rows of holes on the divider wall has been conducted. Detailed velocity and Nusselt number distributions are presented inside the rotating two-pass coolant channel. The enhanced cooling in this passage design is achieved by a combination of impingement and crossflow-induced swirl. The cross flow is generated from one coolant passage to the adjoining coolant passage through a series holes placed along the dividing wall. The holes deliver the flow from one passage to another typically achieved in a conventional design by an 180o U-bend. The holes direct flow perpendicular to the axial direction. The impingement and swirl produces significantly high heat transfer enhancement over conventional heat transfer enhancement mechanisms such a rib turbulators, pin fins, etc. Commercial software, FLUENT, is used for predicting the flow using the standard k-e turbulence model. The results are primarily presented at a channel flow Reynolds number of 25000. The effect of rotational speed is achieved by varying the rotation number from 0, 0.1, and 0.2. The effect of coolant-to-wall density ratio is also varied from 0.05, 0.15, and 0.25. Results show that the impingement and swirl flow are affected by rotation induced Coriolis and centrifugal forces. There appears to be little effect of buoyancy for this geometry as velocity profiles are seem to be unaffected by the wall temperature changes. In the absence of adequate experimental data for rotational cases, the detailed heat transfer distributions for some stationary cases obtained using the liquid crystal technique were compared. The detailed flow field predictions effectively explain the experimentally obtained detailed surface heat transfer distributions. The pressure distribution and Nusselt number distribution from the predictions are in good agreement with the experimental data for such stationary cases and raises the confidence in predicting the same for rotating channels.
Advisor:Michael M. Khonsari; Yitshak M. Ram; Srinath V. Ekkad
School:Louisiana State University in Shreveport
School Location:USA - Louisiana
Source Type:Master's Thesis
Date of Publication:07/12/2002