Date of Award


Degree Name

M.S. in Engineering


An axially loaded heat pipe, with a length to diameter ratio of unity, has been theoretically and experimentally investigated. Water and methanol have been used as the working fluids and the container material is oxygen-free hard copper. The limits used to define the operation of a conventional fixed-conductance heat pipe have been modified so that the operation of the axially loaded heat pipe could be predicted. The results of this analysis conclude that the axially loaded geometry will reach a boiling limit far before reaching the capillary limit. The axially loaded heat pipe was processed with 3.1 grams of doubly distilled deionized water. Testing of the water filled heat pipe included characterization of the steady-state performance for various heat inputs and inlet coolant conditions. The results of these tests confirm that this geometry operates in a linear manner with the vapor temperature being directly influenced by the source and sink boundary conditions. High power water testing revealed a unique phenomenon which is speculated to be due to some type of internal oscillatory flow. Continued investigations of this phenomenon verified that the magnitude and frequency of the oscillations were directly effected by the input power. Recognizing this as a potential flow instability, additional investigations revealed that positive inclinations, gravity assisted modes of operation, tended to dampen out the observed phenomenon. To this author's knowledge this phenomenon has not been previously observed by other heat pipe researchers. The data and trends observed in the high power water testing are discussed in Appendix B. The experimental set-up was power-limited for the water heat pipe such that boiling conditions could not be reached. In order to approach the theoretical boiling limit, methanol was used as the working fluid. Methanol heat pipe testing was initiated at 47.2 W and progressed to a maximum of 151.8 W or a surface heat flux of 13.3 W/cm2. As the input power was increased, evaporator temperature fluctuations showed that nucleate boiling was achieved. Continued operation in the boiling mode demonstrated over 700 seconds of steady-state data could be obtained with no significant change in the heat pipe's performance. The significance of this test reveals that, for this particular geometry, the initiation of boiling does not automatically lead to a dryout condition. The boiling limit, as it currently exists, should be revisited so as to account for such things as nucleation rates, bubble departure diameters, surface roughness, and capillary radius ratios for composite wick structure to minimize the potential of vapor lock.


Heat pipes, Heat Transmission

Rights Statement

Copyright © 1993, author