Preparing a file for data collection
Author: Rafael Estrella | Major: Mechanical Engineering
My name is Rafael Estrella, and I am a senior mechanical engineering student and an aspiring researcher. At the end of the first semester of my sophomore year, I was invited to work in the lab of my Introduction to Material professor and current Honors mentor, Dr. David Huitink. Intrigued by the work of graduate student Reece Whitt on non-metallic heat spreaders for high-power electronics, I began analyzing the durability of the polymer materials used for these devices. During the 2021 fall grant term, I conducted an honors independent study on another part of the heat spreader project: improving the performance of a single jet nozzle.
High-voltage power electronics are largely used to convert battery storage into useable power for
electric motors and can suffer from deleterious heat concentrations during operation. One common thermal management solution, metallic cold plates, relies on conduction and convection via a liquid-cooled metallic block placed in direct contact with a power module to draw out excess heat. Concerns with this method are the inefficiencies with the metallic block cooling, the added weight, and the signal interference accentuated by the metal components. As such, this work is a sub-section of an effort to enable polymer AMHS (additively manufactured heat spreader) technology, which uses jet impingement nozzles to direct coolant to heat concentrations. This reduces the maximum device temperatures and allows fluid to come into direct contact with areas of concern, reducing thermal resistance.
The use of polymer materials for production means a lighter form factor and less signal interference. On the other hand, there are concerns about the durability of polymer AMHS devices at the temperatures and pressures of operation. Thus, part of my ongoing research is an erosion study on the degradation of the impingement nozzles at elevated temperatures and pressures. Preliminary data suggested that cavitation, or vapor bubbles caused by low pressure areas, can quickly enlarge the nozzle diameter. This is currently being studied alongside long-term effects of subtler forms of erosion. Due to the nature of the long duration studies and limited equipment availability, this portion of my work will proceed into the spring semester as planned.
A second aspect of my research is to augment the fluid chamber and jet arrangement within the heat spreader to maximize the convective heat transfer of non-metallic coolers. In this study, a single nozzle design is defined as having one jet impingement nozzle dedicated to each hot spot whereas dual nozzle designs provide a pair of nozzles targeted at a single heat concentration. This semester, the redesign of the device manifold resulted in four dual nozzle designs. This was accomplished in no small part through the mentorship I received from both my professor and the graduate student I worked with. Inexperienced in design work myself, my professor provided necessary guidance in this area. The graduate student working on the main heat spreader project, Reece, provided insight on many of the practical aspects of conducting research in this field, including robust methods for hermetic sealing and data collection and analysis.
In jet impingement, the major geometric parameters are nozzle diameter, jet-to-surface distance, and (in multi-jet contexts) the jet-to-jet spacing. Accounting for these variables, nozzle diameter was kept constant across all designs and each design was tested at a variety of jet-to-surface spacings. The jet-to-jet spacing was only altered in one dual jet design to change the point of convergence. Along with geometric variables, the flow rate of fluid coolant through the system was varied for each design at each geometric configuration to characterize the changes in pressure drop. The result was an array of parameters assessing the efficacy of each design. That said, this study provided an excellent example of how a few variations on multiple variables can quickly add up. Nearly 200 configurations were tested as a result. Designs were then compared via the heater temperature, heat transfer coefficient, and pressure drop.
Results of the study showed the single nozzle design provided the lowest heater temperatures and greatest heat transfer coefficients. This means that allocating a single nozzle per heat concentration should provide the heat spreader with the greatest efficacy compared to the dual nozzle designs tested. On the other hand, the dual nozzle designs provided much lower pressure drops. As the pressure drop is essentially the cost to the system to pump the coolant fluid, there are cases where these designs may provide an economic edge.
My research this semester provided an excellent opportunity for me to design an experiment in an academic setting and bring it to fruition. I am excited to continue my work on these projects into the spring semester. As I plan to pursue my master’s degree in mechanical engineering, I am looking forward to leveraging the experience I have gained this past semester in my career.