Design and performance of integrated capacitive sensor based on knitting

Objective The electrode of the sensor is placed in the external environment, and the surface conductive material is prone to oxidation and spalling to a certain extent during long-term use, thus affecting the performance of the sensor. Moreover, the air permeability of the packaging material is poor, which reduces the comfort of the human body. Knitting technology can knit conductive yarn into fabric structures, playing a role in improving the comprehensive performance of the sensor. This paper proposes a knitted capacitance sensor with integrated surface insulating electrode and dielectric layer. Method The sensor electrode was designed by double-yarn knitting technology, and cotton yarn and silver-plated yarn were selected for knitting. The cotton yarn always covered on the surface of the silver-coated yarn, so that the electrode was placed inside the sensor, avoiding direct contact with the external environment. Nylon monofilament was used to knit dielectric layer, which was combined with fabric electrode to form capacitive sensor. The spacer fabric capacitive sensor was further placed in the middle of the indenter of the Mark-10 tension/compression meter. A conductive yarn was drawn from each of the upper and lower fabric surfaces of the sensor to connect the positive and negative collets of the precision LCR digital bridge TH2830, making it a complete conductive path. During the test, the digital bridge monitors and recorded the capacitance change of the sensor in real time. By analyzing the relationship between capacitance and pressure, the influence of different spacer fabric thickness and spacer wire diameter on the sensor performance was investigated. Results The mechanical properties, insulation properties, sensitivity, hysteresis, response time and repeatability of the sensor were studied and analyzed. With the increase of thickness, the pressure required to achieve the same strain decreases, and the fabric was more easily compressed. The surface insulation performance of the sensor was further characterized. When the electrode pen was placed in the inner layer of the spacer fabric, the digital multimeter would detect the resistance value, indicating that the surface layer was conductive and acted as the electrode layer of the sensor. However, when the electrode pen is placed on the fabric surface layer, the multimeter would not read the resistance value of the fabric surface correctly. During the whole compression process, sensitivity showed different values in different compression strain ranges. The general trend was that as the compressive stress of the fabric increased, the capacitance change rate of the sensor was positively correlated with it. As the thickness of the spacer fabric increased, the sensitivity of the sensor increased gradually. According to the hysteresis error calculation principle, the maximum hysteresis error occurred at the stress of 6.49 kPa, and the value is 2.24%. The response time curve of the sensor showed that the response time was less than 150 ms. From the the trend of capacitance change rate with time during 2 000 cycles, it can be seen that the sensor maintained stable input and output electrical characteristics during the initial compression stage. To further verify its practical application value in the real application scenario, the capacitive sensor designed was used for the monitoring and recognition of hand movements. Based on the curve peak characteristics, the sensor was able to easily distinguish fingertip press and boxing movements. Conclusion It is found that the sensor with larger thickness and smaller diameter of spacer wire has better overall performance, in which the sensor with thickness of 8.0 mm and spacer wire diameter of 0.15 mm has the best performance with sensitivity being 0.033 kPa-1. Moreover, it has low hysteresis and fast response time, good resolution and response ability to input signals of different properties (different compression distances and different compression frequencies), and repeated stability within 2 000 cycles, showing good pressure sensing ability in hand motion recognition and liquid weighing scenarios. The sensor reduces the production cost, and the signal is stable during the application process, showing great application potential in wearable, medical monitoring and human-computer interaction interface. © 2024 China Textile Engineering Society. All rights reserved.