Date of Award

Spring 2021

Document Type

Open Access Dissertation


Mechanical Engineering

First Advisor

Jamil A. Khan


Due to shrinking component sizes and faster switching speeds, the volumetric heat generation rates and surface heat flux in many advanced power electronic devices have increased exponentially, creating an urgent need for efficient cooling technologies. It has become particularly challenging to maintain an optimal temperature at the device level. Among advanced cooling techniques, single-phase and two-phase microchannel heat sinks show considerable promise in removing a significant amount of heat from a small surface area. Considered one of the most efficient heat dissipaters, the two-phase microchannel heat sink can handle very high heat flux at a small footprint and requires relatively low coolant flow rates. However, ever-increasing high-heat-flux cooling requirements have pushed the single-phase microchannels to their limit. Therefore, further enhancement of the Heat Transfer Coefficient (HTC) in microchannel is needed. Although two-phase microchannels provide higher heat transfer rates, they suffer from flow instability issues, which cause significant oscillations in the flow rate, temperature, and pressure, as well as deterioration of the HTC and an early occurrence of Critical Heat Flux (CHF).

The objectives of this research were to: (1) enhance the performance of the single-phase microchannel heat transfer, and (2) suppress the two-phase microchannel flow instabilities to increase CHF. The surface modification techniques knurling and sandblasting, two methods which are conventionally used in macro duct flow systems, were experimentally investigated for the first time to improve the single-phase microchannel heat transfer performance.

A micro-knurled surface technology was found to be an effective approach for improving the single-phase microchannel heat transfer performance, resulting in a maximum HTC enhancement of 255% compared to that of the smooth microchannel. In contrast, fully sandblasting the bottom surface of the microchannel only slightly improved the heat transfer performance; however, modifying the surface with a hybrid micro-sandblasting of elliptical patterns enhanced the heat transfer performance substantially.

A knurled surface was also used to study its effects on the thermal performance in flow boiling microchannels. A diamond-patterned knurling was fabricated on the bottom surface of the microchannel, with two different knurling heights R-1 (0.25 mm) and R-2 (0.17 mm). The CHF for both the R-1 and R-2 cases were all improved, at mass flux G = 66.48 kg/m2s and 172.87 kg/m2s, respectively. Specifically, at mass flux G = 172.87 kg/m2s, the improvements of the CHF for the roughened surfaces were around 30% higher than the CHF for the smooth surface under similar conditions tested. In addition, the results show that the knurling does not significantly influence the pressure drop. Moreover, the optimal design of the IRs was achieved when combined with knurling to further enhance the heat transfer performance of the two-phase microchannels. The hybrid design decreased the amplitudes of the pressure-drop oscillations up to 60%.