Methods to characterize and enhance the through-thickness thermal conductivity of three dimensional polymer composites

Yu, Hang
Journal Title
Journal ISSN
Volume Title
University of Delaware
Increased polymer composites are being used to replace traditional materials in modern structures and industrial applications due to their light weight and corrosion resistant features. However some of these devices and components are subjected to increasing thermal load and as polymer composites are poor conductors of heat, thermal management is critical. Thermal transport mechanisms are investigated and characterized when highly conductive fibers are embedded in three-dimensional polymer composites. Improved out-of-plane thermal conductivity of fiber reinforced polymer (FRP) composite materials will enable lightweight structures to integrate efficient thermal management. Different approaches have been explored to enhance out-of-plane thermal conductivity and address measurement issues associated with this class of composites, including Z-pin insertion and distribution, particle-fabric reinforcement and laminate stacking sequence design and some combinations of above factors. An approach in which a conductive coating is applied on the surface of composite samples to improve the through-thickness thermal conductivity of composites containing a small percentage of conductive fibers in the thickness direction is explored. A parametric study revealed that the thickness of the coating and the distribution of the conductive fibers play a crucial role in augmenting the heat transfer across the thickness of the polymer composite. For the enhancement of through-thickness thermal conductivity of nonwoven laminated composites, one can introduce conductive fibers and modify the form, direction and architecture of the fiber network. The role of the fiber tow orientation, volume fraction and stacking sequence is explored. A geometric scalar parameter is proposed to correlate laminate through-thickness thermal conductivity with fiber tow stacking sequence. For woven structures, it is found that Pilling's constitutive model for two phase system agrees well with the numerical predictions of the through-thickness thermal conductivity. The randomly distributed conductive particles can be loaded into the matrix to make a difference to the effective thermal conductivity of this three-phase system. A unified hybrid constitutive model is proposed for the through-thickness thermal conductivity prediction for the particle embedded woven fabric system, incorporating a combination of generalized rule of mixtures and Pilling model to address a wide range of particle volumes and particle and fiber thermal conductivities. For characterization of composite materials with heterogeneous surfaces, an experimental setup was designed, fabricated and validated to measure through-thickness thermal conductivity. A corresponding finite element model of the setup was developed to characterize and gain further understanding of the thermal field in the measurement cell. It is found that the temperature variation over the inhomogeneous surfaces cannot be captured by limited number of thermistors in the setup. To be able to record the temperature gradient over the entire surface, an approach combining an infrared camera temperature measurement system with a finite element analysis is developed to investigate the influence of natural convection on the effective thermal conductivity of such heterogeneous materials. Measurements of reference samples were conducted to validate the methodology. The conductive coating application in this scenario is also quantified for heat transfer enhancement. In summary, this dissertation developed various through-thickness thermal conductivity enhancement models for matrix, nonwoven laminate and woven structures and proposed an improved characterization technique with infrared thermography for heterogeneous materials.