Doctoral Dissertations

Li Chen

Date of Award

8-1989

Degree Type

Dissertation

Degree Name

Doctor of Philosophy

Major

Mechanical Engineering

Major Professor

D. R. Pitts

Committee Members

M. Keyhani, R. V. Arimilli, J. J. Perona

Abstract

The first part of this dissertation shows an experimental study of natural convection heat transfer in a rectangular enclosure with ten protruding heaters from one vertical wall. This experimental study is termed Experiment I for the sake of brevity and clarity. The top surface of the enclosure was a heat exchanger maintained isothermally as the heat sink. All other surfaces, except at the heater locations in the vertical wall, were unheated. The enclosure had a height-to-width ratio of 7,26 and a depth-to-width ratio of 8.42. The ten heaters were identical each having a height to enclosure width of 0.35 and a horizontal protuberance to enclosure width ratio of 0.48. The vertical spacing between heaters was equal to the heater height.

Distilled water and ethylene glycol were used as working fluids. Experimental results show that the bottom heater (heater 1), except for high Rayleigh number runs, has the highest heat transfer coefficient. The heat transfer coefficients at heaters 7, 8, and 9 are nearly the same and present the lowest values among all the heaters. It is also shown that the heat transfer coefficients decrease up to heater 7. At high Rayleigh numbers, the top heater (number 10) has the highest heat transfer coefficient. For each fluid a single correlation of Nuy verses Ray*, represented the heat transfer coefficient for all the heaters.

Flow visualization experiments have also been performed. Photographs of the flow patterns under several power inputs with glycol as the convection medium indicate a core flow within the enclosure and a recirculating cell in the gap between heaters.

The second part of the dissertation covers the experimental investigation of the effect of enclosure width on natural convection heat transfer in a rectangular enclosure with protruding heat sources. Again for the sake of brevity and clarity, this experimental study is termed Experiment II. Five protruding heaters were uniformly mounted on one vertical wall. The vertical wall opposite to the heated wall was movable so that the enclosure width could be adjusted to the desired value. The top plate was a heat sink, and all other surfaces except the two end vertical surfaces were insulated. The five identical heaters had a protuberance of 9mm and a vertical height of 15mm. The vertical spacing between heaters was also 15mm. The enclosure width was varied in experiments from 13.5mm to 45mm.

The convection medium was glycol. Experimental results show that the starting point of core flow directly effects the local heat transfer behavior of the bottom heater, and the secondary flow cell between the top heater and the top sink influences the heat transfer process at heater 5. Width variation influences heat transfer mainly by the variation of flow pattern. This influence is weak when the ratio of width over protruding height is greater than 4.0 and negligible at the ratio of 5.0 or higher. Furthermore, such influence depends on heater location and power input. Photographs from visualization experiments under various widths and power inputs lead to the above conclusion.

The correlation of Nusselt number and modified Rayleigh number based on heater local height y is independent of the heater number, protruding height and vertical height of each heater under the following conditions: protruding heater height from 9 to 11 mm; vertical height of heater from 8 to 15 mm and heater number from 5 to 10. However, the heat transfer coefficient based on the top sink area at the smallest width is the highest iv value and this decreases with increase of heat sink area. With a protruding height of 9 mm, the heat transfer coefficient is lower than that of the flush case for a given power input, but the difference in heat transfer coefficient between these two cases changes with variation of enclosure width, power input and heater location.

Download
Files over 3MB may be slow to open. For best results, right-click and select "save as..."

Share

COinS