Multiscale modeling of thermal conductivity of hierarchical CNT-polymer nanocomposite system with progressive agglomeration

Abstract

Agglomeration of carbon nanotubes (CNTs) is a key phenomenon that affects the effective thermal conductivity of CNT-polymer nanocomposites. This issue needs to be adequately studied especially in the practice of prediction of the effective thermal properties of CNT nanocomposites. In this study, we developed a multiscale model within the framework of effective medium theory to predict the effect of CNT agglomeration on the overall thermal conductivity of the hierarchical CNT-polymer nanocomposite system over a wide range of CNT loading. With an introduced Cauchy cumulative probability function, we developed a new model of CNT agglomeration wherein the CNT volume fraction progressively evolves with the total loading of the CNT fillers in the composite. This model is self-consistent and it leads to a very good agreement with experimental observations. The entire multiscale approach consists of the newly developed model of the progressive agglomeration as well as models that altogether take into account the effect of two types of CNT interfacial thermal resistance and the unique feature of the CNT nanostructure. We highlighted this multiscale approach with its implementation in the prediction of the overall thermal conductivity. Good agreement was confirmed between the prediction and the experimental data. The study shows that the existence of CNT agglomerates downgrades the thermal percolation networks and it can cause a decrease up to 50% in the overall thermal conductivity of the nanocomposite. Two important parameters for the progressive agglomeration model were identified and their influences to the predicted results were examined.