Carbon Nanotube Yarns: Tailoring Their Piezoresistive Response Towards Sensing Application

Carbon nanotubes (CNTs) exhibit unique sensitivity to mechanical stress and strain, otherwise known as piezoresistivity. The piezoresistive effect causes a change in their electrical resistivity. By correlating this response to their relative change in electrical resistance, they can be used for strain sensing and damage detection. Due to their hollow structure, CNTs also possess low weight and high aspect ratio which is ideal for continuous sensing and real-time structural health monitoring. Additionally, they have tailorable geometry and outstanding mechanical properties that can be employed as a reinforcement in composites. By exploiting both their mechanical and electrical properties, it is possible to achieve distributed CNT-based self-sensing composite structures.;However, individual CNTs have length scale and size that limits their use. Thus, they need to be scaled up for the realization of various CNT-based piezoresistive smart materials and devices. One method to achieve this, is to grow and spin CNTs into macroscopic yarns. The challenge is in the production of a macroscopic CNT assembly that can reproduce the properties of individual CNTs. Properties of CNT yarns are reduced by imperfections in their structures like defects, discontinuities, entanglements, misalignments, orientation and interfacial properties with other materials. Consequently, the strain sensitivity of CNT yarns is about two orders of magnitude lower than that of individual CNTs. Moreover, CNTs have properties that depends on the type of CNT (i.e. the number of walls), the chirality, purity etc. It is pertinent to identify the properties of CNT yarns that are inherent and those that depend on the raw materials, processing and usage-induced structure to realize their full potential in sensing applications.;This study aims to understand the effect of material and physical parameters on the piezoresistivity of CNT yarns. Parametric experiments were conducted to include the effect of quasi-static strain rate, strain level, twist, impregnation and yarn geometry on their piezoresistivity. Strain rates affect the failure mechanisms and electromechanical properties of CNT yarns; high strain rates exhibit increased tensile strength and a positive piezoresistivity while low strain rates favor a higher strain-to-failure and a negative piezoresistivity. The sensitivity of the CNT yarn remains relatively unchanged with varying strain rates, but strongly dependent on the strain level and yarn geometry. The linearity needed for a robust sensor is favored at higher strain rates. Due to the interplay of inter-tube slippage and structural reformation with twist, there is an optimal twist level to achieve desired properties. Low-twist CNT yarns (twist angle of 10-25°) exhibit a lower breaking strength, lower strain-to-failure but higher rigidity compared to medium- (25-35°) and high-twist CNT yarns (> 35°). The piezoresistive response of the studied CNT yarns was found to be highest at medium twist. CNT yarn strain sensors made by embedding the yarn completely in a flexible substrate achieved ultimate strains of up to 50 % and gage factors greater than 1000; i.e. one and three order/s of magnitude greater than the bare yarn, respectively. This is attributed to the role of the substrate as a backbone in load-bearing and the conformance property of CNT yarns. This study demonstrates the capability of CNT yarn sensors in strain sensing and damage detection, with potential applications as torques sensors, flexible strain gauges, wearable sensors and in health monitoring of composites.