A semi-empirical model for thermal conductivity of polymer nanocomposites containing carbon nanotubes

Abstract

A new version of the semi-empirical Halpin–Tsai (H–T) model is presented to evaluate the effective thermal conductivity of general carbon nanotubes (CNTs)-reinforced polymer nanocomposites. The model captures the influences of the CNTs alignment, random orientation, aggregation, waviness, length, diameter and the CNT/polymer interfacial thermal resistance parameters. In order to verify the suitability of the new H–T model, the numerically calculated thermal conductivities are compared with existing experimentally measured ones. An excellent predictability is found of the modified H–T model over a wide range of the tests. The consideration of the CNT waviness and the interfacial thermal resistance parameters is seriously essential for a more realistic prediction in all conditions. For aligned CNT-reinforced polymer nanocomposites, considering the alignment factor seems to be very important. Moreover, in the case of well-dispersed CNTs into the matrix, it is necessary to incorporate the CNT random orientation parameter. Additionally, when CNTs are not well dispersed, the CNT aggregation and random orientation parameters must be incorporated in the analysis. The effects of the CNT volume fraction, length, diameter and non-straight shape on the nanocomposite thermal conducting behavior are estimated in details. The results clearly expose that it is needed to eliminate the aggregation, use the straight CNTs and form a strong interface if the full potential of CNT reinforcement is to be realized. Finally, the thermal conductivities of CNT-shape-memory polymer nanocomposites at different temperatures are obtained.