Experimental testing of thermal interface materials on non-planar surfaces

Abstract

As the power dissipated by electronic devices increases and package thermal performance improves, interfaces such as those between a package and heat sink become increasingly important components of the device thermal budget. As such, it is necessary to identify an interface material that will perform optimally under the conditions present in a given application. Unfortunately, standard test methods used to evaluate these materials usually idealize many of the test parameters, such as surface flatness, surface roughness, and test pressure, which characterize the application. Material performance can be significantly influenced by these parameters, necessitating testing in an application environment to identify an optimal interface. Since actual interfaces exhibit some distribution in flatness, roughness, etc., true application testing must involve a significant number of samples to capture expected results over the naturally occurring distributions. Enhancing standardized test methods to incorporate conditions which are reasonably expected to be found in most applications allows for a better comparison of materials. We propose additions to standardized test methods to capture deviations from idealized conditions and show experimental results which illustrate how relative material performance changes in response to surface flatness.