Abstract:

Objective This study aims to overcome the limitations of existing fabric-based capacitive sensors, namely low sensitivity, poor stability, and narrow application scope, by optimizing the dielectric layer modification process. The specific objectives are to investigate the influence of single-walled carbon nanotube (SWCNT) doping on the dielectric properties of 3D polyester fabrics, determine the optimal SWCNT concentrations that maximize sensitivity while ensuring sensor repeatability and stability, elucidate the underlying sensing mechanism and clarify the synergistic effect between SWCNTs and fabric thickness, and demonstrate the usefulness of the optimized sensors for monitoring diverse human motions, thereby broadening their practical applications.

Methods Two types of 3D polyester fabrics, i.e., 3D high-spacing polyester fabric (3D-HSPF) and 3D low-spacing polyester fabric (3D-LSPF), were used as matrix. SWCNTs at gradient mass concentrations (0%, 0.2%, 0.4%, 0.6%, 0.8% and 1.0%) were doped into the composite matrix by an impregnation method to form dielectric layers in the composite. Comprehensive characterization and performance tests were carried out, including analysis of weight gain rate for SWCNT loading, surface morphology observation, and evaluation of key sensor parameters which are capacitance-pressure response, sensitivity, and relative permittivity. The optimal sensors were further assessed for repeatability, stability, and hysteresis, and their application in human motion monitoring was demonstrated by recording capacitance change rates during various body movements.

Results The results demonstrated that both SWCNT concentration and fabric structure significantly influenced the sensor performance. For the 3D-HSPF sensor, a SWCNT concentration of 1.0% yielded a maximum sensitivity of 927.36%/kPa, 4.26 times that of the undoped one, and a maximum relative permittivity of 34. At 0.6%, SWCNTs were relatively uniformly dispersed on the fiber surfaces, whereas slight local agglomeration was observed at 1.0%. For the 3D-LSPF sensor, the optimal sensitivity of 925.85%/kPa was achieved at a lower SWCNT concentration of 0.4%, representing a 1.95 times improvement over the undoped one, with uniform SWCNT dispersion and no significant agglomeration. In terms of reliability, the optimal sensors exhibited excellent performance. The repeatability tests for 3D-HSPF-1.0/S and 3D-LSPF-0.4/S showed standard deviations of 3.96 and 3.72, respectively, indicating a stable response under cyclic loading. Stability tests revealed minimal capacitance drift over 2 h, with standard deviations of 0.10 and 0.17. Both sensors demonstrated good reversibility, with hysteresis rates below 13%. In practical application tests, 3D-LSPF-0.4/S effectively monitored small joint movements, with average capacitance change rates of 32.59%, 14.75%, and 76.87% for finger, wrist, and elbow movements (maximum of 82.63% for elbow movement), respectively. Conversely, 3D-HSPF-1.0/S was better suited for detecting moderate to large deformations, with average capacitance change rates of 9.59% for knee movement and 41.69% for arch movement.

Conclusion High-performance fabric-based capacitive sensors were successfully developed by modifying 3D polyester spacer fabrics with SWCNTs by an impregnation method. The optimal SWCNT concentrations were identified as 1.0% for 3D-HSPF and 0.4% for 3D-LSPF, resulting in substantial improvements in capacitance, sensitivity, and permittivity. Specifically, the sensitivity reached 927.36%/kPa for 3D-HSPF-1.0/S and 925.85%/kPa for 3D-LSPF-0.4/S, corresponding to 4.26 times and 1.95 times enhancements, respectively. These performance gains are attributed to the SWCNT-mediated regulation of interfacial polarization, which synergistically modulating the 3D fabric thickness to enhance the dielectric constant while optimizing mechanical properties. Both optimal sensors demonstrate excellent repeatability, stability, and low hysteresis. The 3D-LSPF-0.4/S sensor is particularly suitable for monitoring subtle motions of small joints like fingers, wrists, and elbows, while the 3D-HSPF-1.0/S sensor is effective for larger joint and body movements, such as knee flexion and arch deformation. These results highlight the promising potential of the proposed sensors for applications in human motion and health monitoring.

Key words: single-walled carbon nanotubes, fabric-based capacitive sensor, dielectric layer, spacer fabric, sensitivity, human motion monitoring