Date of Award

2017

Document Type

Open Access Dissertation

Degree Name

Doctor of Philosophy in Mechanical Engineering-Engineering Mechanics (PhD)

Administrative Home Department

Department of Mechanical Engineering-Engineering Mechanics

Advisor 1

Gregory M. Odegard

Committee Member 1

Julia King

Committee Member 2

Ranjit Pati

Committee Member 3

Gowtham

Abstract

Carbon fiber (CF) composites are increasingly replacing metals used in major structural parts of aircraft, spacecraft, and automobiles. The current limitations of carbon fiber composites are addressed through computational material design by modeling the salient aerospace matrix materials. Molecular Dynamics (MD) models of epoxies with and without carbon nanotube (CNT) reinforcement and models of pure bismaleimides (BMIs) were developed to elucidate structure-property relationships for improved selection and tailoring of matrices.

The influence of monomer functionality on the mechanical properties of epoxies is studied using the Reax Force Field (ReaxFF). From deformation simulations, the Young’s modulus, yield point, and Poisson’s ratio are calculated and analyzed. The results demonstrate an increase in stiffness and yield strength with increasing resin functionality. Comparison between the network structures of distinct epoxies is further advanced by the Monomeric Degree Index (MDI). Experimental validation demonstrates the MD results correctly predict the relationship in Young’s moduli for all epoxies modeled. Therefore, the ReaxFF is confirmed to be a useful tool for studying the mechanical behavior of epoxies.

While epoxies have been well-studied using MD, there has been no concerted effort to model cured BMI polymers due to the complexity of the network-forming reactions. A novel, adaptable crosslinking framework is developed for implementing 5 distinct cure reactions of Matrimid-5292 (a BMI resin) and investigating the network structure using MD simulations. The influence of different cure reactions and extent of curing are analyzed on the several thermo-mechanical properties such as mass density, glass transition temperature, coefficient of thermal expansion, elastic moduli, and thermal conductivity. The developed crosslinked models correctly predict experimentally observed trends for various properties.

Finally, the epoxies modeled (di-, tri-, and tetra-functional resins) are simulated with embedded CNT to understand how the affinity to nanoparticles affects the mechanical response. Multiscale modeling techniques are then employed to translate the molecular phenomena observed to predict the behavior of realistic composites. The effective stiffness of hybrid composites are predicted for CNT/epoxy composites with randomly oriented CNTs, for CF/CNT/epoxy systems with aligned CFs and randomly oriented CNTs, and for woven CF/CNT/epoxy fabric with randomly oriented CNTs. The results indicate that in the CNT/epoxy systems the epoxy type has a significant influence on the elastic properties. For the CF/CNT/epoxy hybrid composites, the axial modulus is highly influenced by CF concentration, while the transverse modulus is primarily affected by the CNT weight fraction.