Conjugate analysis of asymmetric heating of supercritical fluids in rectangular channels

Hogirl Jung, Purdue University

Abstract

The conjugate problem of the heat transfer to a supercritical fluid in an asymmetrically heated high aspect ratio (AR) channels was analyzed by a computational approach. The domain included both the fluid and the solid regions. The Navier-Stokes equations along with the continuity and the energy equations were solved in the fluid region and the energy equation only was solved in the solid. The fluid and solid regions were coupled through the interface condition requiring a balance of the heat flux across the fluid-solid interface. An adaptive Cartesian look-up table provided the property information for the supercritical fluid. A modified two-equation model of turbulence was implemented to give turbulence characteristics with secondary motions in the fluid. Because of the high Reynolds of interest, the effect of surface roughness was considered with the equivalent roughness set as ks = 3krms. Two dimensional solutions are used to verify the roughness model and the conjugate heat transfer model. The focus of the results then lies on three-dimensional solutions. From these results, series of cross plane views for channel aspect ratios of 4 and 8 cases are used to show details of the streamlines and velocity vectors in the cross plane along with contours plots of the velocity and temperature profiles. These help to improve the understanding of heat transfer in conjugate, high Reynolds flows. The secondary flow in the channel corners increases the heat transfer and decreases the pressure drop while surface roughness augments the heat transfer and increases the pressure drop. Comparisons of the pressure drops along the channel and the temperature at the outer surface of the solid between the computation and the measured data show qualitatively and quantitatively reasonable agreement. Comparison between the AR = 4 and the AR = 8 channel showed that the latter had a larger heat flux than the AR = 4 channel under the same wall temperature conditions. Under same heat flux condition, the maximum wall temperature for the AR = 8 channel was 37% lower than that for the AR = 4 channel. Based on these computational results, a Nusselt number correlation was suggested that gives predictions within ± 20% of the computation.

Degree

Ph.D.

Advisors

Merkle, Purdue University.

Subject Area

Aerospace engineering

COinS