Bias-normal index: A new indicator of dense random packing for thermal polymer/ceramic composites

Abstract Thermal composites are widely used in thermal management. Usually, they consist of polymers and fillers. In pursuit of higher thermal conductivity, the fillers are mostly spherical particles for higher packing density. In previous work, the relationship between particle size distribution (PSD) and packing density was studied, especially the parameters such as polydispersity δ and the skewness S are defined to characterize the PSD. However, in our observation, for sieving process, which is very helpful in manipulating PSD and thus tuning thermal conductivity, δ and S are not good enough to explain most of the compositing results. Here, we suggest a new set of indices (bias-normal indices) to evaluate the sieving efficacy. Three kinds of ceramic powders (MgO, Al2O3, and AlN) and their single- and triple-order composites materials were prepared. The results show that the explanations based on bias-normal ratio index agree well with all the experiments. Thus, this new indicator is competent in guiding the recipe design of the thermal composite materials, and when well-designed recipes are employed, sieving is proven an efficient and simple way to promote the thermal performance, which is of great practical value.