Manuscripts and Patents pending
until file(s) become available
ELECTROSPINNING OF NOVEL EPOXY-CNT NANOFIBERS: FABRICATION, CHARACTERIZATION AND MACHINE LEARNING BASED OPTIMIZATION
This investigation delineates the optimal synthesis and characterization of innovative epoxy-carbon nanotube (CNT) nanocomposite filaments via electrospinning. Electrospinning thermosetting materials such as epoxy resins presents significant challenges due to the polycationic behavior arising from intermolecular noncovalent interactions between epoxide and hydroxyl groups, resulting in a substantial increase in solution surface tension. In this study, electrospinning submicron epoxy filaments was achieved through partial curing of epoxy via a thermal treatment process in an organic polar solvent, circumventing the necessity for plasticizers or thermoplastic binders. The filament diameter can be modulated to as low as 100 nm by adjusting electrospinning parameters.
Integrating a minimal amount of CNT into the epoxy matrix yielded enhanced structural, electrical, and thermal stability. The CNTs were aligned within the epoxy filaments due to the electrostatic field present during electrospinning. The modulus of the epoxy and epoxy-CNT filaments were determined to be 3.24 and 4.84 GPa, respectively, resulting in a 49% improvement. Epoxy-CNT nanofibers were directly deposited onto carbon fiber reinforced polymer (CFRP) prepreg layers, yielding augmented adhesion, interfacial bonding, and significant mechanical property enhancements. The interlaminar shear strength (ILSS) and fatigue resistance demonstrated a 29% and 27% increase, respectively, under intense stress conditions. Up to 45% of the Barely Visible Impact Damage (BVID) energy absorption was increased. In addition, the strategic incorporation of CNT (multi-walled) networks between the layers of CFRP resulted in a significant increase in thermal and electrical conductivities.
This study also introduces a scalable fabrication procedure to address large volume processing, reproducibility, accuracy, and electrospinning safety. Electric fields of the experimental multi-nozzle setups were simulated to elucidate the induced surface charges responsible for the Taylor cone formation of the epoxy-CNT solution droplet on the nozzle tips. Electrospinning parameters were subsequently optimized for the multi-nozzle system and analyzed alongside simulated data to improve stability and synthesize fibers with smaller diameters.
Smaller diameter epoxy-CNT nanofibers proved critical as CNTs maintained alignment within the nanofibers when compared to larger diameter nanofibers. This research examines the impact of effective parameters on the diameter of electrospun epoxy-CNT nanofibers using artificial neural networks (ANNs). Consequently, employing a genetic algorithm (GA) and Bayesian optimization (BO) methods enable accurate prediction of epoxy-CNT nanofiber diameters prior to electrospinning. The presented models could aid researchers in fabricating electrospun thermosetting and thermoplastic scaffolds with specified fiber diameters, thereby tailoring these scaffolds for specific applications.
National Science Foundation (NSF) Small Business Technology Transfer (STTR) (#2036490)
- Doctor of Philosophy
- Mechanical Engineering
- West Lafayette