Improvement of electro-mechanical properties of strain sensors made of elastic-conductive hybrid yarns

Abstract

Fabric-based strain sensors have been developed using different technologies, among which flat knitting is one of the most effective and economical methods. However, knitted strain sensors are not often used in practical applications because the sensors usually exhibit large elastic hysteresis when they are deformed and subjected to stress during application. One possible approach to overcome these shortcomings is to introduce elastic properties at the yarn level by combining the conductive materials with elastic materials. In this paper, we demostrate a hybrid yarn made of a conductive yarn that winds around an elastic core yarn in a direct twisting device. The electro-mechanical properties of strain sensors knitted from the hybrid yarns were tested in order to characterize the sensors. This study consisted of two stages: the yarn preparation and the sensor characterization. In the first stage, two kinds of elastic core components (polyamide/Lycra and polyamide) and two kinds of conductive winding yarns (Bekinox BK50/1 and Bekinox BK50/2) were selected for twisting. The twisting was done with a constant twisting speed and four different numbers of twists. Mechanical properties, that is, the tenacity, force at break and elongation at break, were tested in order to determine the optimal parameters for producing the hybrid yarns. The results indicated that among the tested yarns those with a polyamide core and Bekinox BK50/1 winding yarns at 450 twist/meter and with a polyamide/Lycra core and Bekinox BK 50/2 winding yarns at 600 twist/meter had the best properties. These were thus selected as the materials for producing knitted strain sensors. In the second stage, electro-mechanical properties of the knitted strain sensors were determined under tensile stress and multi-cyclic tensile stress. The results show that the hybrid yarns can effectively enhance the electro-mechanical properties of the knitted strain sensors without compromising processiablity and comfortability.

Keywords

smart textiles knitted strain sensor hybrid yarn electro-mechanical elasticity conductive

Introduction

Textiles can be distinguished from other materials by their flexibility and comfortability. They are the optimized materials to act as hosts for electronics or to function as electronics when electronics are applied to the human body. Recently, there has been a growing interest in using fabric-based electrodes and sensors to provide patient information in health care applications, as well as real-time performance feedback to athletes and coaches.^1–10 Successes have been achieved in the textile electrodes development; for example, the Polar Belt¹¹ and Adidas Acquires Wearable Sensors¹² can effectively indicate the temperature and the heartbeat rate of the wearers. However, the market for fabric-based sensors, especially that for knitted strain sensors, is still immature. The delay in commercialization is mainly due to the unstable properties of the knitted-based sensors. Conductive yarns used for producing knitted-based sensors generally have poor electro-mechanical properties. One of the reasons is that the existing conductive yarns made of metals or conductive polymers usually exhibit low elasticity.

When combining electronic components with textiles, it is desirable that the elastic behavior of the electronic components be enhanced. This is important not only for keeping textile products processable and comfortable, but also for stabilizing their electrical and electro-mechanical properties. For instance, a knitted fabric sensor integrated into a fitted garment for breathing monitoring purposes should be able to indicate a wearer’s breathing pattern without disturbing the wearer’s daily activities. The desired elastic property can be introduced either into the yarns or the fabric. When adding elastic properties into the conductive yarns, various methods can be used, for example by blending carbon-black particles into elastic spun yarns¹³ or coating elastic yarns with conductive polymers.¹⁴ In the former approach, both electrical and mechanical properties of the composite yarns are dependent on the amount of carbon-black fillers. A low amount of carbon-black fillers limits the conductivity of the yarns, and the flow of the conductivity can often be destroyed by applying large tensile stress. On the other hand, a high amount of carbon-black fillers complicates the spinning process, which affects the mechanical properties of the spun yarns. In the latter approach, the coated layer is not as elastic as the core yarn and cracks can be generated when the composite yarn is being stretched. The above-mentioned techniques involve chemical reactions, but there are methods to make the elastic-conductive yarns using only physical actions. One research group at Gent University successfully combined a rubber band and metallic filaments using hollow spindle spinning.^15–17 The combined yarn increased the elasticity of the metallic filaments. However,the rubber band (6000 dtex) was too thick to be used as yarns in the weaving or knitting process. In terms of the winding yarns, metallic filaments are excellent conductors, but they are too stiff and harsh as textile materials. The aim of this paper is to describe how the elastic-conductive hybrid yarns are manufactured by means of a direct twisting process and evaluate the performance of the sensors made of the hybrid yarns. We used polyamide (PA) filaments (780 dtex) and PA/Lycra yarns (780 dtex) as the core materials and stainless steel staple fibers blended into polyester yarns (Bekinox BK50/1 and Bekinox BK50/2) as the winding yarns. Both PA filaments and PA/Lycra yarns are frequently used in knitting and weaving processes, and the Bekinox yarns provide sufficient conductivity to the fabric sensors.¹⁸

The following section describes the preparation of the hybrid yarns and the knitted strain sensors, followed by the testing methods for evaluating the mechanical properties of the yarns and the electro-mechanical properties of the sensors. The third section reports the results obtained from the proposed methods and a discussion of the improvement of the sensors’ properties compared with sensors made of pure (untwisted) conductive yarns. The paper ends with a conclusion and the avenues for further research.

Materials and methods

Preparation of the hybrid yarns

The hybrid yarns were composed of a core yarn and a winding yarn. For the core yarn, a PA/Lycra blended yarn (Salzman Stretch, Switzerland) and PA filaments (Salzman Stretch, Switzerland) were chosen, because both PA/Lycra blends and PA filaments have good elasticity. As conductive winding materials, two commercially available conductive yarns, Bekinox BK50/1 (Bekintex, Wettern, Belgium) and Bekinox BK 50/2 (Bekintex, Wettern, Belgium), were selected. The Bikinox conductive yarn has relatively lower conductivity than do pure metallic yarns, but it has excellent processability. As materials for strain sensors, their conductivity is sufficiently high to produce desirable changes when it is stretched. Table 1 gives an overview of the properties of yarns that were used as input materials.

Table 1.

Materials used as core components and winding yarns of the hybrid yarns and their properties (The fineness and electrical conductivity were given by the suppliers of the yarns, other parameters were measured by the authors)

			Breaking force (cN)		Tenacity^a (cN/Tex)		Elongation at break (%)
Materials		Fineness (Tex)	Mean	std	Mean	std	Mean	std	Electrical conductivity	Yarn specification
Core yarns	PA/Lycra	78	3.59	0.21	0.46	0.03	235.3	11.68	-	#1
Core yarns	PA	78	5.46	0.14	0.35	0.01	66.70	3.34	-	#2
Winding yarns	BK50/1	200	5.97	0.74	0.30	0.04	57.60	4.90	100 Ω/cm	#3
Winding yarns	BK50/2	400	10.84	0.74	0.27	0.02	63.20	4.85	50 Ω/cm	#4

The study of the tenacity is not in the scope of this paper.

PA: polyamide.

The hybrid yarns were manufactured using a direct twisting device. In this device, the core and winding yarns were fed separately into a chamber where a certain number of twists can be applied on the combined yarn. In our experiments, the twisting speed was kept constant at 3600 twists per minute. Four different numbers of twists (twists per meter) – 150, 300, 450 and 600 t/m – were applied to all specimens. Four groups of specimens were prepared in total. The yarn specifications are shown in Table 2 and microscope pictures of specimens of group #01 and group #04 are shown in Figure 1.

Figure 1.

Microscope picture (unit: mm; magnification × 40) of manufactured hybrid yarns in group #01 and group #04.

Table 2.

The specifications of manufactured hybrid yarns

Yarn specification	Group #01				Group #02				Group #03				Group #04
Yarn specification	#01a	#01b	#01c	#01d	#02a	#02b	#02c	#02d	#03a	#03b	#03c	#03d	#04a	#04b	#04c	#04d
Core yarn	PA/Lycra				PA/Lycra				PA				PA
Winding yarn	BK50/1				BK50/2				BK50/1				BK50/2
Twists/meter	150	300	450	600	150	300	450	600	150	300	450	600	150	300	450	600

PA: polyamide.

Mechanical testing of the yarns

Testing of the elastic-conductive hybrid yarns is not the same as that applied to normal yarns. This is because the slight variation in tension of the elastic component affects its properties. In addition, the twists added to the hybrid yarn generate false twists when the yarn is in a relaxed state (Figure 1). The false twists can be generated when a high degree of twists are added to the hybrid yarn. False twists affect the actual length of the yarn, and thus affect the accuracy of the mechanical property measurement. However, false twists are temporary twists, which can be reduced by adding certain force to (stretching) the ends of the yarns. Therefore, in measuring the mechanical property of the hybrid yarns, special bench marks and weights were used to avoid the false twists, as shown in Figure 2(a). In this setup, the hybrid yarn was placed across two holders, and a calibration weight was hanging on both ends of the yarn. The total length of the yarn was 50 cm, and the space between the two holders was 26 cm. The weight was adjusted according to the fineness of the yarn (0.01 ± 0.001 cN/tex). A mark was made at the center of each holder. When placing the yarn into the grips in the tensile tester (Tinius Olsen, H10K) (Figure 2(b)), the marks were aligned to the starting point (A) and the ending point (B).

Figure 2.

The yarn marking system (a) and the placement on the tensile testing device (b).

Figure 3.

Knitted strain sensor specimens made of hybrid yarns (#03c and #02d) and the pure conductive yarns (#3 and #4).

After fixing the yarn into the grips, the upper grip moved at 50 mm/m and stopped when the yarn broke. The average forces at break, tenacity and elongation at break were measured and calculated as follows:

Average force at break (N) = \frac{\sum F}{n} (n : number of tests)

(1)

Tenacity (N / tex) = \frac{average force at break}{tex}

(2)

Average elongation at break = \frac{\sum Elongation}{n}

(3)

where F is force at break and n is the number of tests.

Ten specimens were tested in each type of the yarn, and the mean values and the standard deviations were calculated.

Preparation of the knitted strain sensors

After studying the mechanical properties of the manufactured hybrid yarns, two groups of the yarns (#02d and #03c) were selected for producing the strain sensors. The selected yarns possess the optimal combination of breaking force and the elongation at break, which guarantee good mechanical properties of the fabric. The selected yarns were made into 1 × 1 rib knitted fabrics, realized using a flat knitting machine. The wale and course density were 16 and 12 per 10 mm, respectively. The size of the sample fabric was 10 mm × 100 mm. In addition, the pure conductive yarns – Bekinox BK50/1 (#3) and Bekinox BK 50/2 (#4) – were used to produce samples with the same parameters for the purpose of comparison. Three specimens were prepared for each type. The pictures of the samples were shown in Figure 3.

Electro-mechanical properties’ tests of the knitted strain sensors

The electro-mechanical properties were measured by the tensile tester in combination with a digital-multimeter (Keithley, 6487), which was used to measure the resistance of the specimens. The specimens were subjected to a constant rate extension of 5 mm/sec. The level of total strain was 2, 5, 10, 15 and 20%. Each specimen was measured five times, and the relaxation time between measurements was 1 minute.

Resistance change ratio determination under tensile stress

In strain sensors, the resistance change ratio (RCR) is an important parameter when calculating the sensitivity of the sensors. The RCR is calculated as

RCR = \frac{R_{f} - R_{0}}{R_{0}}

(4)

where R₀ is the initial resistance (the resistance before stretching) and R_f is the final resistance (the resistance after releasing from stretching).

The RCR is revealed by means of a hysteresis curve (Figure 4).¹⁹ The positive value of the RCR means the resistance after removing the load is larger than the initial resistance, while the negative RCR means the resistance after removing the load is smaller than the initial resistance. Both positive and negative RCR are hysteresis effects of the sensors. When a sufficient load is applied to an elastic material, it will cause the material to deform. The total deformation of an object is equal to the sum of the elastic deformation and plastic deformation. An elastic deformation is reversible after the load is removed, while a plastic deformation is permanent.²⁰ The resistance change of a strain sensor indicates its total deformation, which can be divided into elastic deformation-related resistance change and plastic deformation-related resistance change. The plastic deformation-related resistance change ratio (PD-RCR) can be used as a criterion of the sensors’ elastic character. The lower the PD-RCR is, the better the sensor performs. When it is zero, it means the material can be used as a perfect strain sensor. In sensors made of composite materials, the PD-RCR can be reduced by adding high elastic components.

Figure 4.

A typical hysteresis curve of the fabric-based strain sensor. R₀: the initial resistance, R_f: the resistance after releasing.

Figure 5.

Cyclic tester along with attached sample sensor, Keithley multimeter and software.

In our experience, when a tensile stress was applied to the specimen and the specimen was thus stretched to an extended length, the specimen returned to its original length after the stress removed. This action was repeated once. The values of the RCR of the two extension cycles were calculated from the hysteresis curve, and the PD-RCR was determined consecutively.

Cyclic electro-mechanical properties determination

Electro-mechanical measurements were also performed under multi-cyclic tensile stress. The specimen was connected to a cyclic tester²¹ (see Figure 5) in order to apply repeated mechanical extension and deformation. The device has a fixed end and a moveable end, which has a variable speed between 5 and 50 mm/sec. The length of the whole testing device is 400 mm, and 64 different steps can be customized by varying the speed and position of the moveable end. In our experiments the maximum elongation varied from 2 to 20%. For each elongation, each specimen was tested for 1 minute, and the tests were repeated five times.

Results and discussion

Tensile properties of the hybrid yarn

The tensile properties of the manufactured hybrid yarns are presented in Table 3.

Table 3.

Tensile properties of the manufactured hybrid yarns

			Breaking force(N)		Tenacity(N/Tex)		Elongation at break (%)
Core yarns	Winding yarns	Fineness (dtex)	Mean	std	Mean	std	Mean	std	Yarn specification
PA/Lycra	BK50/1	278	3.45	0.19	0.28	0.04	155.10	16.72	#01a
			3.28	0.25	0.28	0.03	167.40	8.28	#01b
			3.63	1.01	0.29	0.032	172.50	8.90	#01c
			4.89	2.28	0.29	0.02	184.5	7.79	#01d
	BK50/2	478	4.02	3.44	0.27	0.03	170.60	12.14	#02a
			4.17	2.92	0.23	0.04	151.40	13.42	#02b
			7.31	2.62	0.23	0.05	144.30	18.02	#02c
			10.68	0.99	0.23	0.02	169.00	14.80	#02d
PA	BK50/1	278	4.30	1.48	0.28	0.02	69.00	6.74	#03a
			7.48	1.09	0.31	0.02	79.90	7.24	#03b
			10.38	1.99	0.31	0.02	88.70	7.83	#03c
			9.79	1.81	0.03	0.02	84.20	8.50	#03d
	BK50/2	478	4.91	1.25	0.27	0.02	74.30	6.19	#04a
			12.06	2.18	0.27	0.02	81.10	5.58	#04b
			12.12	1.85	0.22	0.03	76.60	1.84	#04c
			11.23	1.92	0.20	0.04	70.50	13.2	#04d

PA: polyamide.

By comparing the elongation at break of the hybrid yarns with the pure conductive yarns (group #3 and #4 in Table 2), we found that the hybrid yarns in general considerably improved the elongation. Groups #01 and #02 had much higher elongation compared with groups #03 and #04, regardless of the numbers of twists (Figure 6(a)), but the improvement in elongation in terms of the numbers of twists within each group was insignificant. From these observations, we concluded that the elongation at break of the hybrid yarns was highly dependent on the elastic properties of the core yarns. The numbers of twists applied on the hybrid yarns had little influence on the elongation at break of the yarns. In other words, the predominant element was the elastic core yarns rather than the numbers of twists added to the hybrid yarns.

Figure 6.

The elongation at break (a) and the breaking force (b) of manufactured hybrid yarns.

Considering the breaking force (Figure 6(b)), in the case of groups #01 and #02 where PA/Lycra was the core yarn, the breaking force of the hybrid yarns was lower than that of the pure conductive yarn in general. This is because the Lycra filament in the core yarn has a relatively low breaking force. As the Lycra filament takes up one third of the whole hybrid yarn, the hybrid yarn was considered to be broken when the Lycra filament broke. However, the breaking force increased with the numbers of the twists, and when the latter equaled 600 t/m, the breaking force was comparable with that of the pure conductive yarn. In groups #03 and #04, the hybrid yarns generally improved the strength of the conductive yarn. When the number of twists was larger than 150 t/m, the breaking force of the hybrid yarns significantly increased, and the peak was plotted at the place of 450 t/m. As a conclusion we can say that the numbers of twists applied on the hybrid yarns strongly influences the breaking force.

We also found that low twisting amounts gave relatively high standard deviations in most of the cases (Table 3). The reason for this is probably that when the numbers of twists is too low, the core yarn and winding yarn cannot totally bind together (see Figure 1, #01a and #01b) and, therefore, the performance of the hybrid yarns varies with the percentage of the bounded parts.

Electro-mechanical properties of the knitted strain sensors

The improvement of the elasticity in sensors by the resistance change ratio

Table 4 presents the RCR of the selected four groups of knitted strain sensors and we can see that the RCR in the first tension cycle is negative in most of the sensors. In the second tension cycle, most values became positive.

Table 4.

The related resistance change ratio (%) of selected fabric sensors

	#3		#4		#03c		#02d
	1st cycle	2nd cycle	1st cycle	2nd cycle	1st cycle	2nd cycle	1st cycle	2nd cycle
2%	−0.55363	6.051	2.741348	2.741348	−1.01374	−0.38304	1.247041	0.993404
5%	−2.86046	1.183	6.933496	0.872563	−2.91589	0.1278	−1.2084	2.965049
10%	−0.82803	4.01797	−13.3639	3.981541	−2.45719	1.495531	−7.63201	1.524287
15%	−2.454	7.492155	−14.7074	0.520519	−5.56367	1.082191	−9.39777	1.620909
20%	−5.65906	2.1278	−16.7656	1.676555	−6.49661	1.827135	−8.75763	2.608064
AVG	−2.47104	4.174385	−7.03242	1.958505	−3.68942	0.829923	−5.14976	1.942343
STD	2.042753	2.631095	11.00343	1.417047	2.273166	0.930774	4.839351	0.816653

The contact resistance between yarns has proved to be the key sensing elements in the knitted strain sensor.²² Strain both increases the number of the contact points^22,23 and reduces the contact resistance in each point, so that the contact resistance decreases and the overall resistance of the sensor decreases. When the stress is removed, the contact points decrease and the resistance thus increases, but it usually cannot return to the initial resistance. This is why the RCR in the first cycle was always negative, which can be explained by two physical phenomena: (1) as we mentioned earlier, there is a plastic deformation in the knitted strain sensors, which is not reversible; (2) the conductive yarns do not totally return to their initial positions rapidly, so that when the force is removed, the number of contact points are still larger than the initial ones. In the second cycle, the plastic deformation was partially or totally removed. The resistance can be totally returned to the initial one after removing the force. In most cases, the final resistance became even higher than the initial value; therefore, most of the negative RCR becomes positive.

Figure 7 shows a typical hysteresis curve of a specimen in the first and second cycles and Figure 8 illustrates the absolute value of the RCR of all tested specimens. It was evident that the absolute value of the RCR depended on the value of the total strain applied on the specimens in the first cycle (Figure 8(a)). Any increase in the strain resulted in increasing the RCR. This can also be explained by the second phenomenon mentioned above: larger elongation created more contact points among the conductive yarns in the sensors, so when the force was removed the sensors needed more relaxation time to return to the initial form. However, in Figure 8(b), we can see that the RCR value is not influenced by the total strain in the second cycle, which proves that the second phenomenon can be totally or partially eliminated in the second cycle. Therefore, elastic deformation predominated in the sensor performance after the second cycle, which means the RCR in the second cycle gives a strong indication of the PD-RCR, and thus evidenced the elasticity of the sensors.

Figure 7.

The hysteresis curve of the first (a) and the second (b) cycle of specimen #02d.

Figure 8.

The absolute value of the resistance change ratio against the elongation in the first (a) and second (b) cycle of specimens.

The RCR of the second cycle was calculated from the hysteresis curves (Figure 9). A comparison was made between groups #3 and #03c, and groups # 4 and #02d. As we can see in Table 4, the RCR of specimen #3 was 4.17 ± 2.63%, while it was 0.83 ± 0.93% for specimen #03c. This proves that the elasticity of the knitted strain sensors was indeed improved by using hybrid yarns. By comparing the standard deviation, the changes in resistance of the specimens become more stable in relation to the elongation when hybrid yarns are used, in both comparison groups. This means that the sensors made of the hybrid yarns showed a more stable performance.

Figure 9.

The hysteresis curve of the second cycle of specimen #3 (a) and #3c (b).

Cyclic electro-mechanical properties of the knitted strain sensors

The same comparison groups were chosen when studying the cyclic electro-mechanical properties of the specimens. The conductive yarns were the same in each comparison group, which guaranteed that the differences in sensor performance came from the participation of the elastic core yarns. Figure 10 shows the compared results of specimens #3 and #3c in relation to elongation from 5 to 20% (the tests failed in #3 and #4 when elongation was 2%, so they are not given in the figures). From the curves we can see that by using hybrid yarns, the sensor performance improved significantly, that is, the cyclic patterns became more regular when elongation was larger than 5%.

Figure 10.

Cyclic electro-mechanical properties of specimens #3 and #03c under elongations of 5, 10, 15 and 20%.

Comparison of specimens #4 and #02d are shown in Figure 11; when the elongation was 5%, specimen #4 produced a cyclic pattern with increased baseline shift and magnitude. This is because the pure conductive yarns have low elasticity and the PD-RCR was not fully eliminated even after several cycles. The hybrid yarns improved the elasticity when the strain was 5% by giving a more homogeneous output against the time. When the strain was equal or larger than 10%, the improvement in elasticity was not very significant; however, the signal strength was enhanced.

Figure 11.

Cyclic electro-mechanical property of the sensor made of conductive yarn (#4) and hybrid yarn (#02d) under elongations of 5, 10, 15 and 20%.

Conclusions

In this paper, we have presented the manufacturing of electro-conductive hybrid yarns and knitted fabric sensors with the hybrid yarns. The test results show that the number of twists and the elasticity of the core yarns both have considerable effects on the mechanical properties of the hybrid yarns. Knitted fabric sensors made of selected hybrid yarns reduce the plastic deformation-related RCR, thus reducing the elastic hysteresis of the sensors. In the next step, we aim to study the sensors’ performance after washing and in different climate conditions. What is more, knitted fabric sensors made of elastic-conductive hybrid yarns can be used as textile breathing sensors. Cyclic electro-mechanical properties over 24 hours, days and months should be studied in order to verify the sensor performance for long-term monitoring.

Footnotes

Funding

This work was supported by the Smart Textile Initiative at the Swedish School of Textiles in Borås.

References

Noury

Dittmar

Corroy

. A smart cloth for ambulatory telemonitoring of physiological parameters and activity: the VTAMN project. Proceedings of enterprise networking and computing in healthcare industry 2004; 155–160.

Carpi

De Rossi

. Electroactive polymer-based devices for e-Textile in biomedicine. IEEE Trans Inf Technol 2005; 3: 295–318.

Paradiso

Loriga

Taccini

. A wearable health care system based on knitted integrated sensors. IEEE Trans Inf Technol 2005; 9: 344–377.

Coosemans

Hermans

Puers

. Integrating wireless ECG monitoring in textiles. Sens Actuators 2006; A130-131: 48–53.

Wiklund

Karlsson

Östlund

. Adaptive spatio-temporal filtering of disturbed ECGs: a multi-channel approach to heart beat detection in smart clothing. Med Bio Eng Comput 2007; 45: 515–523.

Tesconi

Tognetti

Pasquale Scilingo

. Wearable sensorized system for analyzing the lower limb movement during rowing activity. Proceedings of industrial electronics 2007; 2793–2796.

Heilman

Porges

. Accuracy of the LiftShirt (vivometrics) in the detection of cardiac rhythms. Biol Psychol 2007; 76: 300–305.

Song

Lee

Kang

. Textile electrodes of jacquard woven fabrics for biosignal measurement. J Textil Inst 2008; 101: 758–770.

10.

11.

Citing Websites. Polar WearLink®+transmitter. Retrieved June 28, 2012, from http://www.polar.fi/en/products/accessories/WearLink_transmitter.

12.

Citing Websites. Textronics, Sports & Fitness. Retrieved June 28, 2012, from http://www.textronicsinc.com/sports-fitness/.

13.

Mattmann

Clemens

Tröster

. Sensor for measuring strain in textile. Sensors 2008; 8: 3719–3727.

14.

Bashir

Skrifvars

Persson

. Surface modification of conductive PEDOT coated textile yarns with silicone resin. Mater Technol Adv Perform Mater 2011; 26: 131–139.

15.

Schwarz

Kazani

Cuny

. Comparative study on the mechanical properties of elastic, electro-conductive hybrid yarns and their input materials. Textil Res J 2011; 81(16): 1713–1723.

16.

Schwarz

Kazani

Cuny

. Electrical circuit model of elastic and conductive yarns produced by hollow spindle spinning. Mater Technol 2011; 26: 121–127.

17.

Schwarz

Kazani

Cuny

. Electro-conductive and elastic hybrid yarns- the effects of stretching, cyclic straining and washing on their electro-conductive properties. Mater Des 2011; 32: 4247–4256.

18.

Guo

Peterson

Qureshi

. Knitted wearable stretch sensor for breathing monitoring application. Proc Ambience 2011; 11: 72–76.

19.

Namiranian

Etriati

Najar

. Investigation of the physical and mechanical properties of fine polyester/viscose-elastic composite roter-spun yarn. Fibre Textil East Eur 2011; 19: 28–33.

20.

21.

22.

Zhang

Tao

. Conductive knitted fabric as large-strain gauge under high temperature. Sens Actuators 2006; A126: 129–140.

23.

Guo

Berglin

Mattila

. Textile strain sensors characterization- sensitivity, linearity, stability and hysteresis. Nordic Textil J 2010; 2: 51–63.