Investigation on the Relation between Structure Parameters and Sensing Properties of Knitted Strain Sensor under Strip Biaxial Elongation

Transcription

1 Investigation on the Relation between Structure Parameters and Sensing Properties of Knitted Strain Sensor under Strip Biaxial Elongation Juan Xie, Hairu Long Donghua University, Songjiang District, Shanghai, Shanghai CHINA Correspondence to: Hairu Long ABSTRACT Knitted strain sensors with different specifications were made from silver-coated conductive yarn and elastic insulating filament with plating-jacquard technique on a SANTONI seamless knitting machine. The arrangements between conductive loops and insulating loops in a structure unit were varied to investigate the effect of the sensor s structures and parameters on sensitivity and repeatability. The influence of fabric density was explored for designing sensing mechanisms. The sensors were also fabricated with different conductive areas by altering the size of the sensing area. The sensing properties of the sensors were assessed by testing their resistance variations under strip biaxial elongation. The results show that knitted strain sensors with various loop arrangements have shown distinct sensitivity and sensing repeatability. Sensors with different fabric densities display good resistance repeatability under both strip biaxial elongation. In addition, the increment of conductive loop wales makes larger resistance variation of sensors with equal numbers of conductive loop courses, and sensors including similar conductive wales tend to have smaller resistance variation at the highest strain level with the growth of conductive courses. INTRODUCTION Fabrics with sensing properties are known as smart fabric sensors and are sensitive to a plethora of external stimuli like changes in force, pressure, mechanical strain, resistance, temperature, etc. As part of smart fabric detectors, strain sensors can be made by different methods at various levels of the structure hierarchy. The first major method to create strain sensors is extrinsic modification, i.e., by coating and attaching discrete or self-contained sensing elements to fabric substrates such as nonwoven, woven, and knitted fabrics [1-9]. The other type of strain sensing structures are made out of conductive fiber and yarns, [9-11] where combing, coating, and melt spinning are some of the methods to produce sensing yarns. Compared with sensors with external modifications and gauges comprised of metal yarns, sensing mechanisms made out of conductive yarns, such as piezo-resistive fibers [12-15] and conductive silver-coated yarn, [16, 17] may behave better in traits like flexibility, wearable comfort, repeatability, ease of production, and washing. To enable strain sensors to be ideal for a target application, therefore, our approach to sensing properties focuses on making gauges out of conductive silver-coated yarn and two elastic insulating yarns using a plating jacquard knitting technique on a SANTONI seamless knitting machine to introduce high sensing properties to gauges. Indeed, the sensing principle of proposed strain sensors is based on the electrical change in response to strain applied during a stretching and recovering process. Hence, it is quite necessary to incorporate fabric structure parameters such as density, construction details, and flexibility with high sensing performance, i.e., sensitivity and repeatability. To date, several researchers have reported the electro-mechanical properties of flexible strain sensors. Zhang et.al [9] have evaluated the effects of two types of conductive materials on sensitivity of the knitting sensor and proposed that a fabric detector made from carbon fibers displays a higher sensitivity, repeatability, and accuracy, and that a tubular structure has a higher maximum strain range than a single warp structure but offers lower sensitivity than the latter. In addition, temperature exhibits greater influence on the sensing behavior. Furthermore, the effects of various numbers of conductive loop courses and Journal of Engineered Fibers and Fabrics 166

2 wales in the same plain knitted structure on resistance change in response to strain under relaxed and unidirectional elongation have also been explored experimentally [18, 19]. Some parameters like loop length, space between wales, conductivity of conductive yarns, and conductive regions are also responsible for sensing performance of a knitting sensor, [20] and a suitable sensing structure has been devised according to experimental results. In the literature, [21] the researchers have also investigated the influences of insulating yarns, input tension of conductive yarn, and numbers of courses of conductive loops used in the sensors on linear range and sensitivity. Apart from factors like types of conductive yarns, experimental conditions, and conductive area, however, little attention has been given to exploring various arrangements between conductive loops and elastic insulating ones of a knitting sensor, which is the main controlling factor for sensor structure and sensing mechanism during stretching. Furthermore, it is noticeable that when designed to measure body movements, electro-active devices may be loaded by in-plane tensile forces such as biaxial extension. Thus, it is necessary to analyze sensing properties of a knitting sensor under biaxial elongation and even multi-directional extension to provide a fundamental research theory required for the target application on which little data has been generated. In our previous study, the relationship between resistance variation and tensile forces and strain of a knitted sensor under strip biaxial elongation has been reported [22]. We also presented the equivalent resistance based on the length of yarn segment of a knitted strain sensor under strip biaxial elongation [23]. In this paper, the sensing properties of a knitting strain detector under strip biaxial elongation have been assessed. Initially, the arrangements between conductive loops and insulating ones in structure units have been varied so that the effect of sensing structure on the sensitivity and repeatability can be investigated. Moreover, the influence of fabric density is explored for designing sensing mechanisms. Finally, strain gauges have been fabricated with different conductive areas by altering the number of courses and wales of conductive loops, and the change of sensing properties in response to various conductive regions has been analyzed. SENSOR MANUFACTURE Materials and Fabrication Method Knitted strain sensors were manufactured by a two-color plating jacquard knitting technique on a knitting machine of SANTONI SM8 Top 2. Figure 1 illustrates the technical back of the conductive knitted fabric unit comprising one conductive loop and two non-conductive ones, where each loop comprises a face yarn visible on the face side of the fabric and a ground yarn displayed on back side. In this design, face yarn F 1 was 78 dtex/48 F polyamide elastic filaments and the other, F 2, was 110 dtex/40f silver-coated conductive yarn with a resistance of 0.5 Ω/mm. Polyamide/spandex core-spun yarn G was intended as a ground yarn to introduce elasticity into the sensing devices. In the fabricating process, three yarn-feeders were utilized to deliver face yarns F 1, F 2, and the ground yarn G to knit sensing structures. F 1 and G are knitted into non-conductive loops, while the float conductive yarn F 2 is behind these loops. According to the needle selection device, F 2 and G are selected to knit conductive loops and the float yarn F 1 is behind these loops. FIGURE 1. The technical back of conductive knitted fabric. Design of Sensing Structure Various Arrangements between Conductive Loops and Insulating Ones Sensors S1 to S9 made with course-wise density of 80wales/50mm and wale-wise density of 140courses/50mm were employed to investigate the effect on sensing performance of different settings between both conductive and non-conductive loops. The size of each fabric sample including a sensor was 16cm 16cm, in the center of which the conductive section was about 3cm 3cm. Their sensing structure units are depicted in Figure 2, where black lines represent conductive filament while white lines represent polyamide yarn. The ground yarns are not all shown for each arrangement. Journal of Engineered Fibers and Fabrics 167

3 Sensors S1 to S3 contain equal numbers of conductive and insulating loops (Figure 2a), where three arrays were employed for assessing the impact on sensing properties of two conductive loops in course and wale directions respectively. The number of conductive loops is more than that of insulating ones in each sensing structure unit of sensors S4 to S7, shown in Figure 2b. As may be seen from Table I, the total number of loops in each structure unit are displayed as well as the ratios of the number of insulating loops to that of conductive ones. In the structure units of sensors S8 and S9 displayed in Figure 2c, conductive and insulating face yarns were knitted with ground yarn in interval courses. The characteristic that there are no float conductive yarns in both sensing areas differs from that in sensors S1 to S7. Both sensors were used to explore the effect on sensing performance by altering arrays between conductive courses on the premise of no float yarns and no interlocks between conductive loops. Sensors with Different Fabric Densities The sinking depths set by loop cam were varied to fabricate sensors D1, D2, and D3. The sensing region of each sensor was developed with only conductive loops with 48 wales and 82 courses. For each sensor, insulating loops lay in a non-conductive area, possessing identical sinking depth as that of conductive loops. Table II demonstrates information on sensors D1 to D3 in terms of course-wise density (wales/5cm), wale-wise density (courses/5cm) and area loop density (loops/25 cm 2 ). FIGURE 2. Structure units of sensors S1 to S9. Journal of Engineered Fibers and Fabrics 168

4 TABLE I. Total numbers of loops in each structure unit as well as ratios of the number of insulating loops to of that conductive ones for sensors s4 to s7. Sensor type S4 S5 S6 S7 Total loop numbers in each structure unit Ratio of number of insulating loops to that of conductive ones 3-Jan 2-Jan 5-Jan 3-Jan TABLE II. Course-wise density, wale-wise density and area density of sensors D1, D2 and D3. Sensor type D1 D2 D3 Course-wise density(wales/5 cm) Wales-wise density(courses/5 cm) Area density(loops/25 cm 2 ) no change in conductive yarn length during elongation at a small strain level, the sensing mechanism depends mainly on contact resistance. The theory of contact resistance [24] is described in Eq. (1), where ρ(ω/m), H(N/m 2 ), n, and P(N) are electrical resistivity, material hardness, number of contact points and contact pressure between conductive materials. As is clearly visible in Eq. (1), contact resistance varies inversely with number of contact points and contact pressure. R c ρ πη = (1) 2 np TABLE III. The numbers of conductive courses and wales of sensors N12, N22, N32, N21 and N23. Sensor type N12 N22 N32 N21 N23 Number of courses Number of wales Sensors With Various Numbers Of Conductive Courses And Wales Each sensing area of sensors N12, N22, N32, N21, and N23 has the same conductive loop density, whose numbers of conductive courses and wales are shown in Table III. THEORETICAL Sensing Mechanism The underlying principle of a strain gauge is that resistance changes with deformation. Two main controlling factors determine resistance variation in response to strain of the knitting sensor, i.e., contact resistance which is related to number of contact points between conductive loops, and length resistance concerning initial loop length and length change of conductive yarns. The impact degree and range of both resistances on sensing mechanisms rely strongly on the structure including arrangements between conductive loops, fabric density and size of the conductive region, tensile stimuli like maximum strain level, etc. When a knitting sensor is stretched under strip biaxial elongation, yarn segments of each loop transfer mutually and the curved yarn starts to extend up to a straightened state without elongation; the elongation of the straightened yarn then begins to exhibit increasing strain. Deriving from the fact that there is Resistance Involved in Devices According to the sensing structures depicted in Figure 2, the contact regions between conductive loops involved in the above sensors can be elucidated by Figure 3. Here, contact resistance R c is in the interlocking area between conductive loops, contact resistance R h is between two adjacent head loops, and R s is between two neighboring sinker loops. When the sensor is extended under biaxial elongation, a change of contact resistance R c depends on the opposing impacts of contact force increasing and the number of contact points decreasing in the interlocking region. Moreover, both contact force and contact region reduce in response to the separation between conductive wales and legs under course-wise extension, which contributes growth in contact resistance R h and R s respectively. As displayed in Figure 4, the values of resistances such as three wale-wise resistances R w1, R w2, and R w3 and a course-wise resistance R course differ as a result of various arrangements between conductive loops, which are contained in sensors S1 to S9, and relevant contact resistances have been depicted. As for sensors with different densities and those with distinct sizes of conductive area, their sensing structures are similar to that shown in Figure 3. FIGURE 3. The contact region types between conductive loops involved in all sensors. Journal of Engineered Fibers and Fabrics 169

5 strain level before the second loading-unloading cycle started. Experiments lasted until stretching-recovering cycles were applied five times. Because the devised sensors may be intended for measuring human body postures and activities, a strain level of 30% was chosen to imitate common body movement extensions. In addition, two periods were introduced to reflect the duration of a specific body movement and the interval between two successive body positions respectively. FIGURE 4. Resistance types within sensors S1 to S9. EXPERIMENTAL Testing Setup From the experimental setup shown in Figure 5, it is seen that measurements of repeated mechanical loading and unloading under biaxial elongation were operated on a DRong X-Y Biaxial Material Tester, where two pairs of clamps were employed to fix knitted sensor samples. In Figure 5(b), the size of each specimen was 16 cm 16 cm and extension area was 10 cm 10 cm. Sensing area prepared with specific wales and courses is located centrally in each sensor sample, around which is a non-conductive region. Resistance variation of sensors were tested and recorded synchronously on a Rigol Digital Multimeter by using the four-wire sensing method. The frequency of data collection was eight times per second, or one point every 125ms. After data transmission, resistance change was displayed real-time on a panel based on the LabVIEW platform (Figure 6). Experimental Method The knitted sensors were subjected to a series of experiments of cyclic loading-unloading. Initially, strip biaxial elongation in the x direction (SBE-X) was performed, during which knitting sensors were fixed in the wale direction in order to keep strain at zero in this direction. At the first loading stage, specimens were stretched up to 30% strain level in the course direction at a tensile speed of 60 mm/min. After a five second pause at the highest strain value, samples were recovered to zero strain level at the same speed of 60 mm/min at the third stage. There was another five second lag time at the lowest FIGURE 5. The equipment for measuring sensing properties under strip biaxial elongation in x direction (SBE-X) and in y direction (SBE-Y). After that, strip biaxial elongation in the y direction (SBE-Y) was tested, in which the strain of specimens in the x direction remained zero. The experimental conditions were in keeping with those in SBE-X, namely, the maximum strain level, extension speed, and two periods. RESULTS AND DISCUSSIONS Effect of arrangement between conductive loops on sensing properties Sensing properties of sensors S1, S2 and S3 It can be seen from Figure 7 that sensor S3 offers the highest resistance variation R compared with sensors S1 and S2. Firstly, under the premise that Journal of Engineered Fibers and Fabrics 170

6 there is no contact effect between adjacent conductive wales in sensors S1 and S3, the former comprises wale-wise resistance R w2 while the latter includes R w3. In fact, contact resistances R c contained in R w3 determines more contact regions between conductive loops than that in R w2. When sensors stretched under strip biaxial elongation in x and y directions respectively, detaching occurs between neighboring sinker loops, which causes fewer contact points and a decline in contact force and growth of sensor S1 resistance. Besides, sensor S3 experiences a higher change in resistance than that of sensor S1. This may be due to the positive role contact resistance R c plays under strip biaxial elongation at a small strain level, where decreasing contact area contributes more than increasing contact pressure. FIGURE 6. Real-time resistance change of specimen during cyclic stretching-recovering. It may also be observed from Figure 7 that the resistance change of sensor S1 is not as high as that of sensor S2 since contact regions between adjacent conductive wales only exist in sensor S2. Moreover, both sensors S1 and S2 show mounting resistance variation, attributed to the fact that the number of contact points and contact pressure lessen in response to separation between neighboring sinker loops and head loops under SBE-X and SBE-Y. In addition, sensor S2 and S3 possess an identical number of contact resistances R s. As shown in Figure 8, the contact region between two adjacent parallel legs contained in sensor S3 (Figure 8b) is larger than that between two neighboring conductive head loops involved in sensor S2 (Figure 8a). Therefore, more separable contact points in sensor S3 are responsible for the larger resistance variation over sensor S2 under both strip biaxial elongation. Another observation from Figure 7 is that each sensor experiences higher resistance variation in response to strain under SBE-X than that under SBE-Y. One of the main factors is as follows: the number of courses is over that of wales in the extension area with a size of 10cm 10cm (Figure 5(b)), and each loop has more obvious stretching in the course direction than that in the wale direction under the same strain levels. Therefore, the degree of separation in contact regions under SBE-X has advantages over that under SBE-Y, leading to its higher resistance change. Resistance grows with contact area reduction when the sensor is stretched under biaxial elongation, and then decreases towards the beginning resistance value as a result of the enhanced contact of neighboring conductive loops during the unloading process. Nonetheless, deformation hysteresis occurs in knitting sensors during successive stretching-recovering which creates hysteresis of the resistance change in response to strain. The degree of hysteresis depends strongly on internal and external stimuli such as elasticity and flexibility of sensors, tensile load, maximum strain level, etc. Journal of Engineered Fibers and Fabrics 171

7 FIGURE 7. Resistance variations of sensors S1, S2 and S3 under SBE-X(a) and SBE-Y(b) respectively. FIGURE 8. Technical faces of sensors S2(a) and S3(b) at relaxed state. Journal of Engineered Fibers and Fabrics 172

8 As it is clearly visible in Figure 7, sensors S1 to S3 provide good sensing repeatability under SBE-X. The reason may be that the increased flexibility in the course direction of the knitted fabric plays a considerable role in enhancing strain and resistance repeatability, which offsets the negative effect of conductive float yarns in terms of number and length. Figure 7 also depicts that if either the conductive float yarn is longer or the number of conductive float yarns in the wale direction is larger, the sensing repeatability under SBE-Y worsens. One of the key contributors is that float conductive yarns on the technical back of the knitting sensor produce an adverse effect on elastic recovery performance. Sensors with longer float yarns like sensor S2 display slightly higher strain hysteresis during cyclic loading-unloading than those with shorter ones like sensor S1. In addition, the lengths of conductive and insulating float yarns in the sensing area of sensor S1 are equal to those of sensor S3. Under such circumstances, similar arrangements between two types of loops in the course and wale direction in sensor S1 may offer advantages in dimensional stability over anisotropic arrays shown in sensor S3. Therefore, sensing repeatability of sensor S1 is slightly better than that of sensor S3. Sensing Properties of Sensors S4 to S7 It can be found from Figure 9 that the sensors with more wale-wise resistances R w1 and less R w2 exhibit resistance change under SBE-X and SBE-Y, attributed to several factors. Firstly, resistance R w1 containing contact resistance R c has more contact regions between adjacent sinker loops than resistance R w2 does. Besides, sensors with more R w1 contain more contact points between neighboring conductive head loops. Deriving from these facts, sensor S6 with the most R w1 offers the highest resistance change in response to strain, followed by sensors S4 and S7 containing identical numbers of R w1 and R w2, and the lowest resistance variation is present in sensor S5. Furthermore, sensors S4 and S7 have approximate resistance alteration at the highest strain level under strip biaxial elongation in the x and y directions respectively, indicating that the array between R w1 and R w2 in the course direction makes little difference on sensing properties. Figure 9 also displays resistance variations of sensors under SBE-Y are not as sensitive as that under SBE-X. The reason, which has been considered and may help to elucidate this phenomenon, is that detachments between adjacent conductive loops differ in degree under two sorts of strip biaxial elongation. According to the results shown in Figure 9, the elasticity differences between conductive loops and insulating ones have little effect on sensing repeatability under strip biaxial elongation in the x direction due to the great flexibility of the knitted fabric in the course direction. As to sensing repeatability under SBE-Y, sensor S6 containing the lowest resistance R w2 has the poorest sensing stability. The main controlling factor is that wale-wise R w2 has more insulating loops than R w1 does which improves deformational stability of sensors due to its great elasticity. Sensing Properties of Sensors S8 and S9 As may be seen from Figure 10, sensor S9 alters much more strongly in terms of resistance change than sensor S8 does. Journal of Engineered Fibers and Fabrics 173

9 FIGURE 9. Resistance variations of sensors S4 to S7 under SBE-X(a) and SBE-Y(b) respectively. Journal of Engineered Fibers and Fabrics 174

10 FIGURE 10. Resistance variations of sensors S8 and S9 under SBE-X(a) and SBE-Y(b) respectively. FIGURE 11. Resistance variations of sensors D1, D2 and D3 under SBE-X(a) and SBE-Y(b) respectively. In fact, identical fabric densities were intended for fabricating sensor S8 and S9 in order to realize the same length of each loop. However, specimens are allowed to relax off the knitting machine which brings about closer contact between loops. Besides, two elastic insulating courses between adjacent conductive courses, shown in sensor S9, possess much stronger relaxation behavior and offer larger contact regions between conductive loops than that of sensor S8. Consequently, sensor S9 displays higher resistance change at the maximum strain level due to more separation between contact points under SBE-X and SBE-Y. Sensing repeatability of sensor S9 is not as good as that of sensor S8 under SBE-Y, as shown in Figure 10. The structure in sensor S9 causes obvious differences between conductive courses and elastic non-conductive courses, which plays a negative role in deformational stability. Therefore, larger resistance hysteresis between cycles is seen in sensor S9 than in sensor S8. Sensing properties of both sensors under SBE-X show little hysteresis due to great elastic recovery of specimens in course direction. Journal of Engineered Fibers and Fabrics 175

11 Effect of Fabric Density on Sensing Performance From Figure 11 the sensors with lower loop density experience much higher sensing variation in response to strain under SBE-X and SBE-Y than those with higher fabric density. The mechanism underlying this effect is that sensors like D3 with lower fabric density tend to possess larger loops and looser fabric structures, as shown in Figure 12(c), while compact structures and small loops caused by large fabric density are presented in sensor D1 (Figure 12(a)). Because the same numbers of conductive loops are contained in both sensors, contact points between conductive loops in high density sensors are more than those in lower density ones. With strain increasing under SBE-X and SBE-Y, sensor D3 exhibits the highest change, followed by sensor D2 and then sensor D1. From the graphs in Figure 11, great sensing stabilities of sensors D1 to D3 under both SBE-X and SBE-Y are depicted. There are several predominant factors responsible for this phenomenon. Firstly, only conductive loops are involved in sensing regions of these three sensors, leading to identical elasticity of loops within each conductive area. Besides, there is no conductive float yarn in the conductive area. Both factors play a positive role in enhancing dimensional stability of sensors. Therefore, little resistance hysteresis occurs in sensors D1 to D3. Effect of Number of Conductive Courses and Wales on Sensing Properties To illustrate how numbers of conductive courses and wales affect sensing properties, Figure 13 depicts resistance change under SBE-X and SBE-Y. Increasing the number of conductive wales exhibits remarkable resistance variation of sensors N12, N22, and N32, which contain equal numbers of conductive courses. This may be caused by equal wale-wise resistance involved in sensors N12, N22 and N32, where 82 conductive loops are interlocked successively in the wale direction. The number of series resistors grows with the number of conductive wales mounting. When specimens are stretched to the same strain level under either SBE-X or SBE-Y, a decrease of contact points plays a significant role in resistance change. Therefore, sensor N32 sees a larger resistance change than sensors N22 and N12 do. As may also be seen from Figure 13, equivalent numbers of conductive wales included in sensors N21, N22, and N23 exhibit the same number of series resistances of sensing area in the course direction. According to the theory of parallel resistance, the more the number of parallel resistors, the smaller the total resistance. Wale-wise resistance decreases with growth of the number of conductive loops in wale direction. Therefore, sensor N23 containing the most conductive courses among these three sensors has the smallest resistance change at the highest strain level. As shown in Figure 13, these five sensors with various numbers of conductive courses and wales show similar resistance repeatability under both SBE-X and SBE-Y. The mechanism underlying this effect is that similar elastic performance of each loop and the same fabric density in sensing areas ensure deformation stability of these five sensors in spite of various sizes of the conductive area. Little strain hysteresis enables stable recovery of resistance. FIGURE 12. Technical faces of sensors D1(a), D2(b), and D3(c) at relaxed state. Journal of Engineered Fibers and Fabrics 176

12 FIGURE 13. Resistance variations of sensors (N12, N22, N32, N21, and N23) under SBE-X (a) and SBE-Y (b) respectively. CONCLUSION This paper serves as the first exploration on the relation between structural parameters and sensing properties of knitted strain sensors under strip biaxial elongation. It can be found that knitted strain sensors with various loop arrangements have shown distinct sensitivity and sensing repeatability. Sensors with different fabric densities display good resistance repeatability under both strip biaxial elongation. In addition, the increment of conductive loop wales makes larger resistance variation of sensors with equal numbers of conductive loop courses, and sensors including similar conductive wales tend to have smaller resistance variations at the highest strain level with the growth of conductive courses. Last but not least, each sensor experiences higher resistance variation under SBE-X than that under SBE-Y due to the deformation character of knitted fabric. The data described in this paper is focused on the relation among fabric structures, configuration parameters, and sensing properties of knitted strain sensors made out of silver-coated conductive yarn and elastic insulating filaments under strip biaxial elongation. We do expect that these results in terms of fabrication parameters and sensing properties can provide foundational information for the design and optimization of high-performance strain sensors in the future. ACKNOWLEDGMENTS This research was supported by the State Scholarship Fund through a project named the China Scholarship Council (CSC) Fund for a join PhD student (No ). The authors wish to thank Engineer Ling Cen for his help in sample fabrication. REFERENCES [1] Tognetti, A, et al., Body segment position reconstruction and posture classification by smart textiles, Transactions of the Institute of Measurement and Control, 29, 2007, pp [2] Calvert, P., et al., Conducting polymer and conducting composite strain sensors on textiles, Molecular Crystals and Liquid Crystals, 484, 2008, pp [3] Kap, Jin K., et al., A novel piezoelectric PVDF film-based physiological sensing belt for a complementary respiration and heartbeat monitoring system, Integrated Ferroelectrics, 107, 2009, pp [4] Lang, E., J., Chou, T., W., The effect of strain gage size on measurement errors in textile composite materials, Composites Science and Technology,58, 1998, pp Journal of Engineered Fibers and Fabrics 177

13 [5] Wang, Xiaojun, Chung D.,D.,L., Short carbon fiber reinforced epoxy as a piezoresistive strain sensor, Smart Materials and Structures, 4, 1995, pp [6] Sau, K.,P., Chaki, T.,K., Khastgir, D., Conductive rubber composites from different blends of ethylene-propylene-diene rubber and nitrile rubber, Journal of Materials Science 32, 1997, pp [7] Lorussi, F., et al., Strain sensing fabric for hand posture and gesture monitoring, IEEE Transactions on Information Technology in Biomedicine, 9, 2005, pp [8] Wei, Y., et al., A Novel Fabrication Process to Realise Piezoelectric Cantilever Structures for Smart Fabric Sensor Applications, IEEE Sensors Proceedings, 2012, pp [9] Tao, Xiaoming, et al., Conductive knitted fabric as large-strain gauge under high temperature, Sensors and Actuators A (Physical), 126, 2006, pp [10] Gilsoo, C., et al, Performance evaluation of textile-based electrodes and motion sensors for smart clothing, IEEE Sensors Journal 11, 2011, pp [11] Wijesiriwardana, R., Inductive fiber-meshed strain and displacement transducers for respiratory measuring systems and motion capturing systems, Ieee Sensors Journal, 6, 2006, pp [12] Wijesiriwardana, R., Dias, T., Mukhopadhyay, S., Resistive fibre-meshed transducers, Seventh IEEE International Symposium on Wearable Computers, Proceedings, 2003, pp [13] Paradiso, R., Loriga, G., Taccini, N., A wearable health care system based on knitted integrated sensors, IEEE Transactions on Information Technology in Biomedicine, 9, 2005, pp [14] Huang, C., H., et al., Parametric design of yarn-based piezoresistive sensors for smart textiles, Sensors and Actuators: A Physical, 148, 2008, pp [15] Huang, C.,T., et al., A wearable yarn-based piezo-resistive sensor, Sensors and Actuators: A Physical, 141, 2008, pp [16] Li, L., Au, et al., A Resistive Network Model for Conductive Knitting Stitches, Textile Research Journal, 80, 2010, pp [17] Pacelli, M., Caldani, L., Paradiso, R., Textile piezoresistive sensors for biomechanical variables monitoring, In: Conference Proceedings Annual International Conference of the IEEE Engineering in Medicine and Biology Society (ed IEEE Piscataway) NJ, USA, 2006, pp 4. [18] Li, L., et al., Electromechanical analysis of length-related resistance and contact resistance of conductive knitted fabrics, Textile Research Journal, 82, 2012, pp [19] Li, L., Au, et al., Design of a conductive fabric network by the sheet resistance method, Textile Research Journal, 81, 2011, pp [20] Wang, J., et al., Electro-mechanical properties of knitted wearable sensors: Part 2-Parametric study and experimental verification, Textile Research Journal, 84, 2014, pp [21] Atalay, O., Kennon, W., R., Knitted strain sensors: impact of design parameters on sensing properties, Sensors (Basel, Switzerland), 14, 2014, pp [22] Xie, J., Long, H.R., The macroscopic equivalent resistance model of knitted sensor under strip biaxial elongation, Industria Textila, 65, 2014, pp [23] Xie, J., Long, H.R., Equivalent resistance calculation of knitting sensor under strip biaxial elongation, Sensors and Actuators A: Physical, 220, 2014, pp [24] Holm, R., Electric Contact Theory and Application fourth ed., Springer Verlag, AUTHORS ADDRESSES Juan Xie Hairu Long Donghua University Room 4015, No North Renmin Road Songjiang Distric Shanghai, Shanghai CHINA Journal of Engineered Fibers and Fabrics 178