Tailoring microstructure and morphology in printed carbon nanomaterials for strain sensing

Abstract

Composites offer significant advantages in the aerospace industry, including reduced structural weight, lower fuel consumption, and enhanced resistance to corrosion. However, unlike metals, which exhibit well-characterized mechanical behavior and predictable service life, composites display greater variability in mechanical performance. Their inherent heterogeneity makes them susceptible to various failure mechanisms, such as delamination, fiber fracture, and matrix microcracking. To address these challenges, Non-Destructive Testing (NDT) techniques have been developed to detect and monitor damage without compromising material integrity. Nevertheless, NDT methods share common limitations, including slow testing speeds, complex result interpretation, and limited detection of deep or small defects. While effective for periodic inspections, many NDT methods lack the capability for continuous real-time monitoring. As an alternative, Structural Health Monitoring (SHM) offers advanced early damage detection. Unlike NDT, which requires manual inspection and on-site evaluation and often leads to increased aircraft downtime, SHM enables continuous in-flight monitoring of structural components, reducing inspection time by up to 44%. This capability translates to an overall maintenance cost reduction of approximately 30%. Three main sensor technologies used in SHM systems are piezoelectric transducers, optical fiber sensors, and strain gauges. However, these technologies face challenges related to wiring and integration complexity, pre-surface treatment requirements, and noise cancellation. In the past decade, piezoresistive sensors have emerged as a promising solution for strain measurement in composite structures, offering benefits such as compatibility with composite materials, lightweight and tailored designs, and excellent sensing performance. Their integration into composites has proven both practical and effective. The sensing mechanism of these materials relies on dimensional changes that occur when the sensor is subjected to external stimuli, altering the distance between nanomaterials and leading to a change in resistance. In other words, the piezoresistive behavior depends on the physical dimensions of the nanomaterials. Low-aspect-ratio nanomaterials, such as zero-dimensional ones, result in higher resistance changes than one-dimensional particles. Chapter 2 provides a summary of sensor technologies commonly used in SHM, including piezoelectric sensors, optical fiber sensors, and metal strain gauges. A comprehensive literature review is presented on the state-of-the-art in carbon-based piezoresistive sensors. Various carbon nanomaterials reported in the literature, such as carbon nanotubes (CNTs), carbon nanofibers (CNFs), graphene, and carbon black (CB), are examined in detail. The review also discusses the effectiveness of different dispersants, including cellulose nanocrystals (CNCs), sodium dodecyl sulfate (SDS), and Triton X-100 (TX-100), in dispersing these nanomaterials in aqueous media. The chapter further explores the influence of sensor design parameters such as length, width, thickness, and geometry on electromechanical behavior. Various methods for integrating carbon-based materials with composite structures are reviewed, such as embedding the materials into the polymer matrix (self-sensing), coating fibers prior to composite fabrication, coating the surface of cured composites, and direct printing onto composite substrates. In addition, the multifunctional potential of carbon nanomaterial-based sensors is surveyed, with attention given to their ability to perform simultaneous sensing and other functionalities. Finally, antennas integrated with composite structures for use as strain or damage sensors are discussed, although this area is still at an early stage of development. Despite significant efforts, a comprehensive understanding of the electromechanical performance of carbon-based piezoresistive strain sensors compatible with composite structures remains lacking. Current carbon-based strain sensors are typically printed or engraved on an intermediate polymer substrate, such as Kapton film, which is then adhered to the composite surface. Although these solutions have demonstrated good compatibility with fiber-reinforced plastic (FRP) composites, their fabrication process is often multistep and requires careful attention to ensure proper adhesion. Chapter 3 introduces various types of CNTs, CNFs, graphene, and CB. Using these materials, both standalone and hybrid formulations were developed, and their dispersibility was assessed. The chapter also describes the fabrication of glass fiber-reinforced polymers (GFRPs) as composite substrates. Four unique strain sensor designs were developed: line, grid, o-line, and omnidirectional. The formulated inks were integrated using three methods: direct coating, screen printing, and spray printing. Chapter 4 discusses the printability and applicability of the developed inks based on the integration method. While direct ink application onto GFRPs was straightforward, it was limited by the deposited amount of ink and the lack of pattern precision. Screen printing and spray printing offered more precise ink deposition and better control over design. The sensor designs introduced in Chapter 3 were successfully demonstrated using these two printing methods. The screen printing method required careful adjustment and tailoring of the inks to improve wettability and enhance printing quality. This was achieved by controlling total solid content, solvent mixtures, and using different surfactant ratios. Notably, screen-printing of aqueous-based carbon inks was demonstrated using both high-viscosity inks and low-viscosity inks, the latter achieved by using a water/isopropyl alcohol (IPA) solvent mixture. Rheological tests were conducted to correlate printing quality with changes in rheological properties. Similarly, inks were prepared for spray printing, albeit with much lower viscosities compared to those used for screen printing. The spray printing method was shown to be easily adaptable for the aerospace industry. By using various carbon inks and adjusting their content, their printability was successfully demonstrated. In Chapter 5, the electromechanical performance of various sensors printed using different methods was investigated. CNT/CNC aqueous-based inks with various ratios were directly applied onto GFRP coupons. Following the ASTM D3039 testing standard, the host composites were subjected to tensile testing up to 1.3% strain. The sensitivity and linearity of the sensors depended on the CNT/CNC ratio. A modified fitting model was employed to differentiate between sensor performances. Further in Chapter 5, using the sensor patterns designed earlier and optimized ink formulations, including CNTs (from both research- and industrial-grade sources), CNF, CB, and various hybrids—sensors were printed onto GFRP coupons and subjected to 50-cycle cyclic loading at 0.6% strain. Sensor performance was quantified in terms of reliability, sensitivity, and linearity. Results showed that performance depended on pattern geometry. Sensitivity correlated with sensor length, with single-leg designs (line, o-line) showing higher sensitivity and more damage than multi-leg designs (grid, omnidirectional). Additionally in Chapter 5, electromechanical behavior of nanomaterials was further compared using the grid pattern. Resistance changes in 1D sensors (CNT, CNF) varied: research-grade CNT and CNF showed downward drift; industrial-grade CNT showed upward drift. All had sensitivity near 2, similar to metal strain gauges. CB (0D) showed slight upward drift and sensitivity up to 15× higher. Graphene (2D) also showed high sensitivity and downward drift. Hybrid sensors (CB with CNT or CNF) demonstrated significant improvements in sensitivity and reliability. All sensors showed good linearity. Continuing in Chapter 5, spray-printed sensors showed similar trends. However, finer lines resulted in less reliable but more sensitive behavior, especially in CB sensors, due to line width effects. Moreover in Chapter 5, complementary tests included higher strain loading (1.2%) on screen-printed CNT sensors, SEM imaging before and after cycling, and directional tests. Grid sensors' transverse sensitivity varied with filler type, indicating a need for design optimization. Omnidirectional CNT sensors confirmed multidirectional sensing. Thermal tests showed TCR depended on filler dimensionality, with 1D most sensitive. Impact tests revealed sensitivity decreased from 0D to hybrid to 1D. In Chapter 6, the investigation focused on patch antenna sensors printed onto GFRP substrates. The host composites were subjected to both tensile and bending forces to evaluate the sensors' mechanical and electromagnetic performance. Frequency response tests revealed that the electromagnetic properties of the antennas were sensitive to mechanical deformation, with measurable shifts in resonant frequency corresponding to the applied strain. These findings demonstrate the potential of printed patch antennas for use as strain sensors in composite structures.