IN SITU PROCESS MONITORING OF CARBON-NANOMATERIAL-BASED MULTISCALE HYBRID COMPOSITES

Abstract

While fiber-reinforced polymer composites (FRPCs) exhibit excellent in-plane properties, dominated by microscopic fiber reinforcement, their out-of-plane performance, governed by the polymer matrix, is relatively poor, and their vulnerability to delamination is of great concern. Reinforcing the polymer matrix by adding nanofillers, such as carbon nanotubes (CNTs), has gained popularity as their extremely large surface area and exceptional properties allow high reinforcing efficiency. However, load transfer between nano-reinforcement and matrix relies on the capacity to disperse and distribute them in the host matrix and also to ensure good distribution on the fabric. When the nanofiller agglomerates are larger than the gaps in a porous fabric medium, “filtration” becomes an issue. The nanofiller aggregates tend to block these gaps leading to slow and incomplete resin impregnation. This inevitable phenomenon mainly occurs near the inlet, where most of filtration takes place, while the rest of the composite shows a “diluted” resin-rich phase even for well-dispersed nanofiller-resin mixture prior to infusion. It is important to effectively investigate the final properties of the composite prior to service to ensure a quasi-defect-free or safe composite when using liquid molding techniques. In a conventional vacuum-assisted resin transfer molding (VARTM) process, a filtration effect leads to an inhomogeneous microstructure of multiscale composites. For this reason, concerns over producing high-quality and multifunctional composites structures serve as the driving forces for the development of hierarchical composites via interface/interphase modification by means of coated nanoscale additives, and their applications in large-scale parts are therefore envisaged. However, to date, limited studies have been reported on monitoring and control of this novel materials processing in order to reduce part-to-part variability for reliable manufacturing. Here, we introduce novel methodologies for resin flow, cure and process monitoring of multiwall carbon nanotube (MWCNT)/glass fiber/unsaturated polyester composites using percolated CNT conductive networks. Two cases are considered where CNTs are: (1) premixed in a resin, or (2) spray-coated on fiber reinforcements prior to infusion. In the case of CNT-coated glass fibers, systematic experiments showed that resistance behavior of the as-deposited conductive network can be used to identify critical events that take place during the entire composite processing cycle, including the onset of crosslinking and gel point of resin, which are necessary to evaluate the part quality. A combination of short-length and low-density MWCNTs exhibited the best sensitivity. 3D real-time monitoring was successfully carried out with simultaneous in-plane and through-thickness resin flow monitoring. A case study was investigated, where the design and manufacturing guidelines applicable to spray-coated CNTs and CNT/exfoliated graphite nanoplatelets (xGnP) hybrids on fiber reinforcements were developed for electromagnetic (EMI) shielding. The research adopted a multi-phase approach that combined both numerical and experimental methods to help lay the groundwork for the fundamental understanding of the underlying physics governing the EMI shielding mechanisms of the composite materials. Simulation and experimental results showed that the contributions of reflection and absorption to EMI shielding is enhanced by sufficient impedance mismatching, while multiple reflections have a negative effect. For a given amount of carbon nanomaterial in the glass-fiber-reinforced composite, coating the outermost, instead of intermediate glass fiber plies with the MWCNT:xGnP weight ratio of 8:2 was found to maximize the conductivity and EMI shielding efficiency. In the case of infusing MWCNTs premixed in a resin, a study was performed to correlate the rheological behavior of MWCNT/unsaturated polyester mixture to the conditions experienced during flow. The resin containing well-dispersed MWCNTs resulted in a nominal increase in suspension viscosity with shear thinning behavior, and was within the desired range for VARTM. Trade-offs among sensitive parameters, such as viscosity, degree of dispersion, MWCNT type and concentration, was investigated to minimize CNT filtration. The degree of dispersion appeared as the dominant factor that dictates the degree of filtration. The filtration level was characterized in situ by monitoring the electrical resistance change in the glass fibers during infusion. As the CNT/resin mixture enters the initially insulating glass fiber layup, the electrodes placed in the fibers display decreases in electrical resistance as the mixture flow progresses. At the completion of infusion, the final resistance readings indicate the uniformity of CNT distribution affected by filtration. Concomitantly, the CNT/resin mixture flow behavior was monitored. Similarly, events that take place after the completion of infusion were indirectly monitored. We have demonstrated simple yet effective methods to monitor the manufacturing processes and predict the final part quality of multiscale hybrid composites, which can be integrated into the existing processes with minimal modifications.