Date

2013

Document Type

Dissertation

Degree

Doctor of Philosophy

Department

Chemical Engineering

First Adviser

Chaudhury, Manoj K.

Other advisers/committee members

Tuzla, Kemal; Mittal, Jeetain; Kothare, Mayuresh; Zuo, Zhijun J.

Abstract

For most of the twentieth century, advanced heat transfer research was geared towards solving the two-phase flow and thermal problems of the nuclear, chemical, and materials processing industries that emerged near the start and middle of the century. These applications demanded relatively large scale systems, with scale up being the challenge of the day. At these power levels and length scales, understanding the interaction between the respective shear, momentum, and body forces of the vapor and liquid were of critical importance. Fast forward to the end of the twentieth century and start of the twenty-first century, and advanced thermal research has shifted towards solving the "smaller" problems of the electronics cooling industry with scale down being the challenge of the day. In large, many of the benefits associated with two-phase heat transfer at these smaller length scales are consistent with the high convection coefficients and lower flow rate requirements observed at larger length scales. Understanding two-phase heat transfer at these smaller length scales requires the understanding of the same macro-scale two-phase heat transfer physics along with additional physics such as capillary forces and wetting phenomenon. Perhaps this is most evident in heat pipes, where two-phase heat transfer in combination with capillary wick structures have emerged as a staple in the electronics cooling community. Increasing power density requirements have pushed the use of the monolithic hydrophilic wick structures used in commodity heat pipes to their limits, including their power transport capabilities and their convection thermal resistances. One way to improve upon traditional wick structures is to manipulate large droplets on hydrophobic surfaces in the condenser portion of heat pipes. Fluid can be circulated more readily as a droplet on a surface compared with fluid through small pores. Further, the convection resistance associated with dropwise condensation is an order of magnitude better than condensation on a hydrophilic wick structure. Surfaces with graded hydrophobicity have shown to be capable of passively moving condensing droplets while exhibiting the characteristic high heat transfer coefficient of dropwise condensation on hydrophobic surfaces. Vapor-liquid two-phase heat transfer alone cannot solve all modern thermal problems. The electronics community has pushed transistor sizes to the nano-scale feature size levels, where more complicated physics such as phonon-phonon and electron-phonon interactions are required to understand heat transfer physics and engineer solutions at these levels. For communications applications where the chips operate in pulsed modes, thermal storage through freeze-thaw cycles of a solid-liquid phase change material is an attractive solution. However, the thermal storage effect at high pulse frequencies is only warranted if the phase change material is located in close proximity, within a few microns, of the active transistors where electron-phonon and phonon-phonon interactions dominate conventional Fourier conduction thermal resistances. In high power density systems operating at 1,000's of W/cm2 heat fluxes, such as high power laser diodes, compact micro-channel coolers using single phase coolants have emerged as a potentially viable cooling solution. The laminar flow and heat transfer physics involved in these devices are well understood. However, the compact nature of the micro-channel coolers along with the high flow rates needed to move the required power have lead to unavoidably high velocities. Copper is the only engineering material with thermal conductivities high enough to meet thermal performance requirements, but suffers from erosion at the high velocities required. Meanwhile, ceramic thin films have the necessary properties to limit erosion under the high velocities required. Although conventional fabrication techniques cannot be used to apply ceramic coatings within the small micro-channel features, nano-scale coating methodologies such as Atomic Layer Deposition may hold the key to meeting the application demands.Thermal testing at smaller length scales poses measurement challenges that are under development. One such challenge is encountered by CPU manufacturers during their development process, where optical inspection of active transistors is required while operating. The CPU dissipates heat at a rate equivalent to normal operation in these tests. The optical view requirement removes the possibility of using traditional heat sinks to remove this waste heat. The only way to remove heat and meet the optical view requirement is to use impingement cooling. The small sizes of the so called "fireball" processing cores push the limits of the measurement technology used to validate cooling solutions.Interdisciplinary research is the key to meeting many emerging research challenges, thermal management included. The twenty-first century has brought a new set of thermal challenges, pushed largely by the demands of the electronic industry. This thesis addresses several of those challenges, where interdisciplinary research involving heat transfer, surface engineering, and small length scales are a common theme.

Share

COinS