By Josh Perry, Editor
[email protected]

Archibald Amoako, a graduate student in the South Dakota State University (Brookings, S.D.) Department of Mechanical Engineering, recently presented research on using computer modeling to design devices to better remove heat from aerospace electronics at the American Institute of Aeronautics and Astronautics/American Society of Mechanical Engineers Joint Thermophysics and Heat Transfer Conference in Atlanta, Ga.

Using CFD, 12 different fin geometries were tested to identify which heat sink design is most efficient at dispersing heat from aerospace electronics. (South Dakota State University)

According to a report from the university, Amoako modeled an active heat sink that uses fans to pull the air from the device that needs to be cooled and into the fins of the heat sink, where it is dissipated. The key to the heat sink design is the fin geometry taking into consideration thermal resistance and pressure drop.

“Amoako tested fin configurations, such as rectangular and hexagonal pins, zigzag and arc plates and short plates, as well as pin-and-plate combinations,” the article explained. The research also incorporates additive manufacturing to produce new geometries that provide localized cooling where it is most needed.

The article continued, “The square zigzag plates had the lowest thermal resistance, while the separated short plates had the lowest pressure drop. However, Amoako found that no definite correlation exists between thermal resistance and pressure drop parameters when evaluating heat sink performance.”

Amoako uses STAR-CCM+ CFD software and the conjugate heat transfer process to simulate the flow of the air and the solid portions of the heat sink. The model was able to simulate the entire heat sink geometry.

His research advisor noted, “A major manufacturing advancement, additive manufacturing of 3D parts allows engineers to create exotic parts that could not be manufactured previously. We’re developing the tools to evaluate the potential of these new designs using computational fluid dynamics modeling.”