In electronics cooling with limited local convective or irradiative cooling, conduction path optimization is critical. Decreased junction temperatures (resulting in increased performance and reliability) can be achieved by maximizing the conductance through screw joints.
Thermal contact conductance across an Aluminum-Aluminum screw joint was experimentally measured which compared well with published data for uniformly distributed pressure, with increasing conductance versus the published data as the screw preload was applied. The geometry was modeled using a meshed surface application of an available model for thermal conductance of contacts with uniformly-applied pressure, resulting in a poor correlation due to limitations in the applied technique. Modifications to the model were explored, including capping pressure, imposing a non-zero pressure in the area not influenced by screw pressure, simplifying the pressure distribution, and scaling the model to match experimental results. Each resulted in somewhat better correlation between calculated and experimental results. Recommendations are made regarding the calculation of conductance in general, leaning toward more simplified models. Further work is suggested in the area of experimentally-verified modeling of other materials and surface conditions, geometry, and with an interstitial material other than air.