Pool boiling experiment with Novec-649 in microchannels for heat flux prediction

• Published in: Experimental thermal and fluid science 2023-02, Vol.141, p.110802, Article 110802
• Main Authors: Kaniowski, Robert, Pastuszko, Robert
• Format: Article
• Language: English
• Subjects: Bubble departure diameter; Heat transfer coefficient; Mechanistic model; Microchannel; Pool boiling
• ISSN: 0894-1777; 1879-2286
• DOI: 10.1016/j.expthermflusci.2022.110802

•Object of analysis: plain smooth surface, microchannel surfaces.•Boiling curves for Novec-649.•Comparison of the results.•Visualization studies.•Simplified mechanistic model for heat flux determination at pool boiling. Boiling dissipates significant amounts of heat as a result of small temperature differences between a wall and a fluid. For this reason, the heat transfer process is widely used across many industries, e.g. for cooling electronic devices, power sources, etc. In practice, heat dissipation from an electronic device occurs mainly during the controlled heat flux loss through the wall to the coolant. This mechanism was studied for boiling Novec-649 (GWP = 1) at atmospheric pressure on copper surfaces with 0.2 to 0.5 mm deep grooves milled in parallel. The resultant microchannels and the spaces between them were 0.2 mm, 0.3 mm, and 0.4 mm wide. The maximum heat flux for the surface with 0.4 mm deep and 0.3 mm wide microchannels was 274 kW/m2. The highest heat transfer coefficient, 22.3 kW/m2K, was obtained for the surface with a 0.5 mm deep and 0.3 mm wide microchannel. A twofold increase in the maximum heat flux and a fivefold increase in heat transfer coefficient were obtained compared to the smooth surface. The effect of geometric parameters on the heat exchange process was investigated for the heat flux density range of 6.6–274 kW/m2. Diameters and frequencies of departing vapor bubbles were determined experimentally with the use of a high-speed camera. A simplified model was proposed to determine the diameters, frequencies, and heat fluxes at different superheats. The model provided a heat flux prediction with a maximum mean deviation of 35 %.