Date of Award
Master of Science (MS)
High-powered electronics are extremely prevalent, and a vast majority of these devices use some form of heat sinks as either an active or passive cooling device. Although two-phase cooling technology has rapidly advanced to accommodate increasing heat loads on modern electronics, the added costs and complication inherent in two-phase cooling solutions makes effective air-cooled heat sinks potentially valuable.
A more effective heat sink can occupy a smaller footprint to dissipate the same heat flux, meaning material and packaging costs can be reduced. Aluminum heat sinks are produced through a variety of manufacturing processes, including machining, bonding, extrusion, casting, forging, 3D printing, and skiving. Of these processes, skiving stands out due to its relative simplicity. In addition, skived fin heat sinks can have very dense fin geometries which enables extremely high aspect ratios for increased heat transfer performance.
This work proposes a potential modification to the skiving process in order to manufacture more efficient strip-fin heat sinks. Many strip-fin heat sinks already include cross channels to increase heat transfer, but this work seeks to develop a method to offset the rows of fins along the direction of the cross channels. Because heat transfer is in general less efficient in a laminar flow, breaking up the channels with offset fins should induce turbulence, resulting in a modest increase in heat transfer.
The increase in thermal efficiency will be relatively small, and therefore it is imperative that the modification have a low implementation cost and very little impact on the existing manufacturing infrastructure. As such, this work explores the possibility of modifying existing skiving blades by using a grinding wheel to introduce serrations into the cutting edge of the blade. Due to the limitations in available equipment, a prototype was developed using multiple smaller high-speed steel blades mounted in a custom tool holder on a Trak DPM RX2 milling machine.
Tests are performed with a single small blade in order to determine the ideal cutting geometry for the blade tip, and then the effective mean friction coefficient between the tool and the work piece is determined to be approximately 1.488 with the cutting fluid used. Unfortunately, the combination of the prototype and the spindle carriage on the milling machine were unable to skive with multiple blades simultaneously due to the increased reaction load. However, it was determined that the process is very predictable once certain parameters are known. This is an important aspect of the viability of this proposed modification; a new offset geometry would require the grinding of a new blade, and as such predictability of fin formation is important for keeping costs low. While the prototype was unable to create a proof of concept, the issues responsible were not issues inherent with the concept itself. As such, this paper still lays the foundation for a potential innovation in heat sink manufacturing technology.
Damena Agonafer Jackson Potter Sanghoon Bae