Nicholas R. Jankowski

Thermal interface materials (TIMs) have reached values approaching the measurement uncertainty of standard ASTM D5470 based testers of approximately ±1 × 10−6 m2 K/W. This paper presents a miniature ASTM-type steady-state tester that was developed to address the resolution limits of standard testers by reducing the heat meter bar thickness and using infrared (IR) thermography to measure the temperature gradient along the heat meter bar. Thermal interfacial resistance measurements on the order of 1 × 10−6 m2 K/W with an order of magnitude improvement in the uncertainty of ±1 × 10−7 m2 K/W are demonstrated. These measurements were made on several TIMs with a thermal resistance as low as 1.14 × 10−6 m2 K/W.

Defense Advanced Research Project Agency's (DARPA's) thermal ground plane (TGP) effort was aimed at combining the advantages of vapor chambers or two-dimensional (2D) heat pipes and solid conductors by building thin, high effective thermal conductivity, flat heat pipes out of materials with thermal expansion coefficients that match current electronic devices. In addition to the temperature uniformity and minimal load-driven temperature variations associated with such two phase systems, in their defined parametric space, flat heat pipes are particularly attractive for Department of Defense and commercial systems because they offer a passive, reliable, light-weight, and low-cost path for transferring heat away from high power dissipative components. However, the difference in thermal expansion coefficients between silicon or ceramic microelectronic components and metallic vapor chambers, as well as the need for a planar, chip-size attachment surface for these devices, has limited the use of commercial of the shelf flat heat pipes in this role. The primary TGP goal was to achieve extreme lateral thermal conductivity, in the range of 10 kW/mK–20 kW/mK or approximately 25–50 times higher than copper and 10 times higher than synthetic diamond, with a thickness of 1 mm or less.

A non-intrusive optical method for two-phase flow pattern identification was developed to validate flow regime maps for two-phase adiabatic flow in a small diameter tube. Empirical measurements of film thickness have been shown to provide objective identification of the dominant two-phase flow regimes, representing a significant improvement over the traditional use of exclusively visual and verbal descriptions. Use of this technique has shown the Taitel-Dukler, Ullmann-Brauner, and Wojtan et al. phenomenological flow regime mapping methodologies to be applicable, with varying accuracy, to small diameter two-phase flow.

The absence of phenomenological insights and accurate flow regime models makes it difficult to predict the improved effectiveness of internally-grooved tubes for two-phase heat transfer. A re-interpretation of available data and flow regime maps is used to suggest that performance improvement is a result of early transition to Annular flow. A modified flow regime map, with a newly-developed Stratified-Wavy to Annular transition criteria for internally-grooved tubes, is shown to increase regime prediction accuracy by 27% relative to the traditional, smooth tube flow regime prediction.

Ensuring adequate spreading of heat dissipated by high power density devices is a critical part of many electronics packaging designs. In many cases, passive wick-based heat spreaders can offer improved heat spreading performance relative to solid conductor alternatives. However, concerns related to performance degradation in high-inertial force environments frequently limit their use to static or near-static applications. In this work we investigate the performance of low coefficient of thermal expansion (CTE) wick-based heat spreaders cooling multiple high heat flux devices in static and high-g environments. Two high-power devices are simulated using custom-manufactured resistor-thermometer chips, enabling dissipation of die average heat fluxes in excess of 150W/cm². Comparative thermal performance is evaluated for wick-based heat spreaders and solid CuMo heat spreaders of equivalent CTE affixed with interface materials typical of those used when attaching a low CTE package to a high CTE cold plate (e.g., Al or Cu). Thermal performance is characterized as a function of heat input during exposure to increasing g-forces applied using a custom-built centrifuge. Experimental observations are interpreted through detailed modeling of fluid flow patterns within the wick structure of the passive heat spreader. Results from these experiments demonstrate that properly designed wick-based heat spreaders have utility in both static and dynamic environments, exhibiting effective conductivities in excess of that obtainable with competitive low-expansion composites.

Advancements in thermal interface materials have reached the resolution limit of typical ASTM D5470 based testers, which is around ±1×10⁻⁶ m²K/W. As the interfacial resistance is reduced, the temperature difference at the interface decreases and ultimately becomes difficult to measure. Standard ASTM testers utilize precise temperature sensors and knowledge of the thermal conductivity of the heat meter bar to resolve the temperature difference at the interface. It is difficult to resolve interface resistances on the order of 1×10⁻⁶ m²K/W, even when precision RTDs with a resolution of ±0.001°C are utilized, as the location uncertainty of the sensor can become important. Increasing the temperature difference across the interface is necessary for further improvement in the resolution. This work presents a miniature ASTM type tester that was developed to address the resolution limits of standard testers by reducing the heat meter bar thickness, using a chip resistor as the heater element, and using an IR camera to measure the temperature gradient along the meter bar. Reducing the length of the heat meter bars reduces the overall resistance, and increases the resistance of the interface relative to that of the meter bars. Because of the reduced size scale of the miniature tester, measurement of the temperature profile using the typical ASTM approach of embedding temperature probes along the length of the meter bar was not feasible but instead was achieved using a relatively inexpensive uncooled long wavelength infrared camera with a microscope attachment in order to focus down to ∼100 microns. Although the IR camera increases the uncertainty of the measured temperatures, this method is shown to measure a thermal interfacial resistance of 1.45×10⁻⁶ m²K/W with an uncertainty of ±1.1×10⁻⁷ m²K/W, where the sample interface was a 2 mil AuSn preform soldered interface.

Forced flow of fluids undergoing phase change in traditional single-phase cold plates is an effective way to manage waste heat removal of vehicle power electronics. Such cold plates come in a variety of standard styles, ranging from circular-tubed cold plates and flat-tube cold plates, to more exotic designs such as louvered and offset fin cold plates. The mechanisms of heat transfer in a two-phase system are different than single-phase, and as such, improvement can be made by custom designing cold plates for two-phase operation. Past research has focused on surface modifications that enhance nucleate boiling, convective vaporization, or both during two-phase operation. The method that has gained the most attention is the use of inner grooved tubes for performance improvement in refrigeration applications. Despite the popularity of inner grooved tubes, the flow mechanisms that deliver performance enhancement are not fully understood. The absence of phenomenological insights and physical models makes it difficult to transition inner grooved tube technology from conventional refrigeration equipment to cold plates. Therefore, understanding the physical mechanisms underpinning two-phase performance enhancement in inner grooved tubes is at the heart of this report.