The focus of the dissertation is to find solutions to increase the through-thickness thermal conductivity of fiber-reinforced polymer matrix composites (PMC). The objective is to explore novel concepts and new approaches to improve the through-thickness thermal conductivity up to 30W/mK for PMCs. First, this research involves understanding the principles of thermal transport in composite and nanocomposite materials. Then the research proceeds to model and design high thermal conducting composites and develop fabrication processes and characterization methods for functioning prototype materials.

PMCs are advantageous for their light-weight, excellent strength and high modulus properties. However, due to insulation nature of polymer resin matrices, their bulk composites demonstrate poor through-thickness thermal conductivity making it unsuitable for applications that undergo thermal loads requiring a means for adequate heat dissipation. The research has carried out four technical approaches to achieve high through-thickness thermal conductivity. (1) Conductive Resins: Increasing the thermal conductivity of the matrix would increase the bulk through-thickness thermal conductivity. Experiments have been done using conductive fillers such as metallic nanoparticles and carbon nanotubes. Results have shown increase in the thermal conductivity but with the disadvantage of increased matrix viscosity making the fabrication process difficult. The thermal conductivity increases, however, is not adequate to achieve the objective solely. (2) Stitch Method: This method applies a continuous conductive path by stitching or inserting high conducting materials such as metal wires, high conducting carbon fiber or high conducting carbon yarns in the through-thickness direction of the composites. Experimentally, this method has proven to show a 27 fold increase in the through-thickness thermal conductivity at low volume fraction percentage of 5% with copper wire and 3.5 fold increase using K-1100 carbon yarn. (3) Long MWNT: Long MWNTs should create a conductive microstructure between fiber layers in composites. Providing conductive links improve the thermal transport of phonons, long MWNTs should more effectively provide thermal transport between fiber layers. However, the experimental results have yet to yield any improvements in the thermal properties of the composites. (4) Buckypaper: The use of thin film of dense nanotube networks or buckypapers is to improve the thermal connections between fiber layers as an interlayer material. If the buckypaper can make multiple connections between fiber layers, the nanotube network can be used to facilitate thermal transportation. However, the use of buckypaper has shown to have a reduced thermal conductivity value than that of a composite without buckypaper. Buckypaper in the experiment create resin rich areas between layers.

Modeling efforts were performed to understand thermal transport mechanism, find solutions and predictions to through-thickness thermal conductivity of the multiscale composites. Micromechanical models were used to predict thermal property values for conductive resins as well as nanoparticle/fiber multiscale composites. Results show that only a few models prove useful with close predictions to experimental data. On the other hand, finite element modeling (FEM), allows the exploration of the critical nanoparticle/fiber interactions and their effects on thermal properties of the resultant composites. The FEM results show that it is the interconnections between nanoparticle and fibers, rather than concentration of conductive fillers, significantly impact the through-thickness thermal conductivity in PMCs, where continuous thermal pathways were the most important for performance improvement. Discontinuous pathways of nanotubes and conducting materials showed very limited or no effects on thermal conductivity improvements. These results provide viable information for future design and fabrication of high through-thickness thermal conductivity composite materials for thermal management multifunctional applications.