Characterization and Modeling of Thermoelectric Properties of Fiber-Reinforced and Multiscale Hybrid Composites

Abstract

Thermoelectricity is one of the energy harvesting techniques that converts waste heat into electrical energy by using a thermoelectric device, which has been confined to metal semiconducting materials. Although inorganic conductors and semiconductors can serve as efficient thermoelectric materials, they have such drawbacks as high processing costs, scarcity and toxicity. For these reasons, organic electronic materials have received attention, and research on thermoelectric systems has been conducted utilizing either conjugated polymers or carbon nanotubes (CNT)/polymer composites to date. We have focused on composites consisting of reinforcing woven fibers as thermoelectric materials, in conjunction with functionalization of woven fabric composites, which can serve as structural materials. We have characterized the thermal and electrical conductivities and Seebeck coefficients of fiber-reinforced composites, which are the constituents of thermoelectric efficiency factor, and confirmed the feasibility of using composites for thermoelectrics. In addition, numerical modeling approaches were developed to predict the thermal and electrical conductivities of composites and to optimize the figure of merit, an index for thermoelectric efficiency, with respect to such material variables as fiber volume fraction, aspect ratio and orientation. Two types of composites, namely, carbon nanotube (CNT)/glass fiber (GF)/epoxy multiscale hybrid composites and carbon fiber (CF)/epoxy composites, were fabricated, and their thermoelectric properties were evaluated as n-/p-type thermoelectric materials. Experimental results showed that the electrical resistivity of the CNT/GF/epoxy composites decreased as CNT concentration increased. Inplane samples showed higher electrical and thermal conductivities due to partial alignment of CNTs in the multiscale composites and continuity of CFs in CF/epoxy composites. In general, CF/epoxy composites showed better electrical and thermal conductivities than multiscale composites. In the Seebeck coefficient measurement test, the multiscale composites showed n-type thermoelectric behavior, whereas the CF/epoxy composites showed p-type behavior. When temperature gradients were applied to closed circuits comprised of multiscale composites and CF/epoxy composites as n- and p-type materials, respectively, an electric current was successfully generated. In the process of optimizing the figure of merit, modeling approaches combining the methods for nanocomposites and woven fabric composites were proposed to predict the thermal and electrical conductivities of composites. The Mori-Tanaka method, thermal-electrical analogy and rule-of-mixtures were adopted as the modeling tools, and we validated the modeling approaches through comparison with experimental data. Thermal-electrical analogy and modified Mori-Tanaka method predictions, which take into account fiber volume fraction, conductivity, aspect ratio, continuity and undulation, agreed well with experimental results. This study covers evaluation of the thermoelectric properties of composites in which reinforcing woven fibers are incorporated, including manufacturing and characterization of materials, structure of constituting materials, and validity of the composites as thermoelectric materials. We proposed and validated the modeling methodologies to predict the thermal and electrical conductivities of composites, taking into account the influences of constituents’ structure, geometry and properties. The outcome of this study is expected to pave the way for a new method for energy harvesting where base materials are structural polymer composites that include reinforcing fibers and polymer matrix. Potential applications range from small devices like electronic and home appliances to large structures such as automotive, aerospace and civil structures. The uses of conductivity models developed can be expanded into other multifunctional applications, such as materials design for thermal management, electrostatic discharging, electromagnetic interference shielding, etc.