Abstract
In the previous research, we prototyped the TIC in which a conventional TIM composed of silicone resin and filler was filled in pores of copper foam, and measured its thermal conductivity by a steady-state method. In addition, the effective thermal conductivity of TIC was predicted by Bhattacharya’s equation and Boomsma’s equation. As a result, it was reported that the experimental value and the predicted value match within 0.7 W/(m·K) by modifying the thermal conductivity of copper to 120 W/(m·K) in the Boomsma’s equation. The issue of that was to investigate the cause of the decrease in thermal conductivity of copper to 120 W/(m·K). In this paper, the effective thermal conductivity of TIC was predicted using the WP structure instead of the Kelvin structure, which is the basis of the Bhattacharya’s equation and Boomsma’s equation. As the result, it was clarified that the effective thermal conductivity predicted by the three-dimensional thermal conductivity calculation model based on the WP structure is more accurate than that predicted by the Kelvin model. And it was found that the experimental value and the predicted value match in the range of 0.4 W/(m·K) by considering the TIC surface structure without modifying the thermal conductivity of copper.
Recent advances in lattice thermal conductivity calculation using machine-learning interatomic potentials
