Date of Award


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

Advisor 2

Susanta Ghosh

Committee Member 1

Trisha Sain

Committee Member 2

Gowtham S.


The global efforts from major space agencies to transport humans to Mars will require a novel lightweight and ultra-high strength material for the spacecraft structure. Three decades of research with the carbon nanotubes (CNTs) have proved that the material can be an ideal candidate for the composite reinforcement if certain shortcomings are overcome. Also, the rapid development of the polymer resin industry has introduced a wide range of high-performance resins that show high compatibility with the graphitic surface of the CNTs. This research explores the computational design of these materials and evaluates their efficacy as the next generation of aerospace structural materials.

Process-induced residual stresses are a commonly observed phenomenon in composite structures during the manufacturing process. These are generated because of resin shrinkage and relative thermal contraction between the resin and reinforcement during the curing process. Experimental or computational characterization of these stresses can be a challenge due to their complex nature. Predictive models of the curing process require detailed knowledge of the resin thermo-mechanical property evolution during the cure. Molecular Dynamics (MD) is implemented to predict the resin properties of EPON 828-Jeffamine D230 as a function of the crosslink density at room temperature. The molecular models are developed using the Reactive Interface Forcefield (IFF-R). The physical, mechanical, and thermal properties are validated experimentally and using the literature data. The predicted progression of resin properties indicates that each property evolves distinctively.

The next generation of ultra-high strength composites for structural components of vehicles for crewed missions to deep space will incorporate flattened carbon nanotubes (flCNTs). With a wide range of high-performance polymers to choose from as the matrix component, efficient and accurate computational modeling can be used to efficiently down-select compatible resins, drive the design of these composites by predicting interface behavior, and provide critical physical insight into the flCNT/polymer interface. In this study, molecular dynamics simulation is used to predict the interaction energy, frictional sliding resistance, and mechanical binding of flCNT/polymer interfaces for a high-performance epoxy resin. The results, when compared to the sister studies, indicate that the BMI has stronger interfacial interaction and transverse tension binding with flCNT interfaces, while the benzoxazine demonstrates the strongest levels of interfacial friction resistance. Epoxy dwells in the “Goldilocks” zone with neither superior nor inferior properties. Comparison of these results indicate that BMI demonstrates the best overall compatibility with flCNTs for use in high-performance structural composites.

One critical factor limiting the potential of carbon-based composites in aerospace applications is the poor load transferability between the reinforcement and the polymer matrix, which arises due to low interfacial shear strength at molecular scale. Molecular dynamics (MD) simulations have been employed in several studies that investigate the interface, such simulations are computationally expensive. To efficiently explore and optimize the interfacial design space with the goal of improving the mechanical performance, it is important to develop a machine learning (ML) approach that can be used to assist in the identification of optimal combinations of interface variables. In this study, a MD-ML workflow is proposed to predict optimal functionalization strategies for a bismaleimide (BMI) and three-layer graphene nanoplatelet (GNP) nanocomposite with maximized interfacial shear strength. In turn, these predictions of pull-out force will be used to identify optimal surface functionalizations that maximize the pull-out force. The details on the MD modeling and training data generation for the ML model are discussed in this work.

Creative Commons License

Creative Commons Attribution 4.0 License
This work is licensed under a Creative Commons Attribution 4.0 License.