An instrument for assessing fabric prickle propensity

Abstract

A technique has been developed for measuring the mechanical stiffness of fibers protruding from the surface of a fabric. Surface fiber stiffness is a key indicator of the propensity for a fabric to evoke a prickle sensation when a garment is worn against the skin. Using this approach, an area of a knitted fabric is assessed quickly in a single pass. A prototype device has been constructed and tested on knitted acrylic fabric samples with known fiber content and that have been ranked for prickle in forearm tests. The device produces results indicative of the coarse fiber content and that correlate well with the forearm assessment of these fabrics. Knitted wool fabrics spanning a range of average fiber diameter distributions have also been tested with this device.

When a garment is worn next to the skin there can sometimes arise an unpleasant sensation known as fabric-evoked prickle. This sensation results from the mechanical stimulation of specific nerve endings in the skin rather than an allergic reaction.^1,2 Studies have demonstrated that the particular nerve endings associated with prickle are triggered by a force applied perpendicular to the skin surface that is greater than a critical level, typically 0.75 mN, and that the presence of a relatively small number of such stimuli per unit area of the skin surface is sufficient to trigger the sensation of ‘prickle’.²

The accepted working hypothesis, consistent with a large number of published studies,¹^–¹² is that stiff fiber ends protruding from a fabric surface and contacting the skin during wear act mechanically as simple Euler rods and if they are able to sustain sufficient force before buckling they trigger the nerve endings, resulting in the ‘prickle’ sensation. These published studies include (a) experiments where fabrics manufactured from blended acrylic fiber, engineered to give a broad fiber diameter distribution, also exhibited a prickle response, thus demonstrating that prickle is not wool specific;⁵ (b) an extensive characterization of fabric-evoked prickle in wool worsted spun single jersey fabrics;⁷^–⁹ and (c) a recent study demonstrating that the model for fabric-evoked prickle is equally applicable for fabrics manufactured from Optim™ fine fiber¹² (wool that has been stretched and set to reduce the mean fiber diameter).

From an engineering perspective, these fiber ends act as simple columns under compression that will buckle at a particular threshold force applied to the top of the fiber and parallel to the fiber axis, which is proportional to ${Ed}^{4} / l^{2}$ (where E is the Young’s modulus of the material, d is the fiber diameter and l is the length of the protruding fiber end).^13,14 For example it has been estimated^1,5 that a wool fiber protruding 2 mm above the fabric surface with a fiber diameter of 30 micrometers would give a maximum force before buckling of about 0.75 mN. Early studies confirmed that the prickle response was often found to broadly correlate with the percentage of fibers greater than 30 µm for a given fabric construction.⁶ A number of detailed studies¹⁵^–¹⁸ of fiber diameter characteristics have identified the role of a critical diameter in the fiber diameter distribution of the protruding fiber ends as a key predictor of prickle.

He and Wang¹⁹ extended this idea by modeling the mechanical behavior of irregular fibers, that is, fibers with diameter variations along their length. They demonstrated that increasing the level of irregularity decreases the buckling load. Using a sinusoidal diameter variation as the input to the modeling, the size of the decrease in buckling load was found, perhaps not surprisingly, to depend on both the frequency and initial phase of the irregularity.

Prickle cannot be adequately predicted simply from a measure of the fiber diameter characteristics as many other parameters, such as fiber length, fabric construction and fabric finishing,^8,20 affect the prickle response.

Measuring the prickle propensity of a fabric or garment is a difficult task with the primary assessment being subjective (sensory) responses from wearers. From a research perspective, collection of this data is both time consuming and relatively expensive. The forearm prickle test^2,3 developed at the Commonwealth Scientific and Industrial Research Organization (CSIRO) enabled a collation of subjective responses to fabrics without the need for whole garment wearer trials. In the early work^2,3 participants placed a test fabric on their forearm and were asked to quantify the ‘prickle’ sensation on a 0–10 scale. In these studies two control samples were used to both acquaint the judges with the sensations and to define the 0–10 scale. The fabric assigned a value 10 was chosen to have a sensation well above that of any of the experimental fabrics and, similarly, a cotton fabric was used to define the zero point on the scale.² Good correlations with the results of whole garment trials confirmed the usefulness of the forearm test.⁴

The sensitivity of the forearm prickle test was refined by (a) the use of a trained and selected panel of judges and (b) the use of a multiple paired comparison protocol.⁵ In this case, two samples are presented consecutively to a judge who is forced to identify the pricklier of the two using the forearm test. This procedure has been used successfully at the CSIRO.⁵^–¹²

In the search for an instrument-based (i.e. objective) test method, Garnsworthy et al.² examined the imprints as fabrics were pressed in a controlled manner against a simulated skin surface formed by stretching Teflon plumbers’ tape over glass slides. Using 55 participants and a set of graded fabrics, this approach confirmed Steven’s psychophysical power law^21,22 between the observed prickle response and the underlying physical stimulus. A new Teflon-covered slide was required for each test, and the detailed calibration system has limited the general applicability of this approach.

Matsudaira et al.^23,24 explored the use of a modified audio pickup from a record player to measure the bending stiffness of individual protruding fibers. As an instrument-based test for prickle, they found that the signal correlated well with the subjective prickle responses of a series of woven fabrics; however, the sensitivity of the technique appears to be limited.

More recently, a ‘one-side compressing tester’ to measure fabric-evoked prickle has been briefly outlined.²⁵ In this instrument a testing plate covered with a latex film and linked to appropriate sensors is pressed against the test fabric. No experimental data from this device are reported.

The present paper reports the development of a new technique and prototype instrument for rapidly assessing the propensity for fabric-evoked prickle.

Theoretical considerations

Figure 1 illustrates two different modes of deforming a cantilever, which are well described from an engineering perspective. In the case of Euler buckling, the force is applied to the top of the fiber and parallel to the fiber axis, as illustrated, and buckling will occur when the applied force is greater than a particular threshold that is proportional to ${Ed}^{4} / l^{2}$ (where E is the Young’s modulus of the material, d is the fiber diameter and l is the length of the protruding fiber end).

Figure 1.

Measures of fiber rigidity.

Figure 1 also illustrates the bending of a fiber or cantilever. In this case the external force is applied normal to the axis of the cantilever and the bending stiffness is proportional to ${Ed}^{4} / l^{3}$ . There is a strong similarity between the equation for bending of a cantilever and the buckling force of an Euler column. Both are proportional to the Young’s modulus of the material, both increase as a function of the fourth power of the diameter of the column and both are a function of an inverse power of the column length (i.e. the protruding fiber length). Thus measuring the bending stiffness of the protruding surface fiber ends should be a good proxy for a measurement of their buckling characteristics. This is the basis of our approach for the development of an instrument.

The above has assumed a circular fiber cross-section. In the case of other cross-sectional shapes the mathematics becomes more complex; however, the general concept that bending stiffness (averaged over different orientations) is a good proxy for buckling characteristics is still valid. Further, in the case of merino wool (the predominant wool used in apparel), studies²⁶^–²⁸ have indicated that the ellipticity (ratio of major to minor axis) is approximately 1.2 or less, that is, a circular cross-section is a reasonable first-order assumption.

Methods

Two sets of fabrics, with well-characterized fiber content and known forearm prickle ranking, were chosen to assess the performance of the prototype instrument. Previous published work with these two sets of fabrics has focused on mechanistic studies of prickle and so the fiber diameter and length properties of both sets of samples have been extensively measured, including the diameter characteristics of the fiber ends.^5,7,12

Set 1 consisted of a range of acrylic fabrics⁵ that spanned a wide range of prickle ratings. These were constructed from different blends of two stretch-broken acrylic tops. By combining increasing proportions of a nominal 9 dtex top with a 3.3 dtex top (both containing fibers with a nominally circular cross-section), a range of fabrics was constructed with increasing mean fiber diameter. The yarns (R80/2 tex) were knitted to produce 14-gauge single jersey fabrics with a cover factor of 1.4 tex^½ mm^–1.

Set 2 consisted of seven pure wool single jersey fabrics manufactured from a range of different commercial wool top qualities, ranging from 15.6 to 22.8 micrometers mean fiber diameter, in increments of approximately one micrometer.¹²

Previous studies^5,12 have demonstrated that both these sets of fabrics covered a wide range of prickle sensations, with the change in prickle level between successive fabrics just detectable (at the 5% level) by the skilled panel of judges used in the sensory trials. This proved useful in evaluating the sensitivity of the instrument relative to the requirements of a commercial test instrument. The published analysis of the prickle levels determined from the forearm prickle trials uses a ‘Rank Sum’ prickle score determined from the paired comparison studies. It should be noted that these scores represent the relative prickle value between fabrics in one experiment, in this case between fabrics in each set of test fabrics. The rank sum values and least significant different values from these previously published studies^5,12 are listed in Tables 1 and 2. (In relation to the samples corresponding to Table 2, the raw data from Naylor¹² were reanalyzed to be consistent with the protocol used to obtain the data in Table 1, that is, with a low score now representing a lower prickle sensation. Also in Table 2, to include sample GRN0600, two data sets were combined using a simple linear interpolation of the three fabrics common to both.)

Table 1.

Summary of the Rank Sum relative prickle scores for the single jersey acrylic fabric samples. For this set of data the least significant difference is 22.⁵ (Fiber diameter characteristics are also reproduced in this table for convenience.)

Sample ID	Mean fiber diameter (µm)	Percentage of fibers greater than 30 µm	Relative prickle score	Difference in prickle score from previous sample
Acrylic 02	19.3	2.8	234	–
Acrylic 04	19.8	4.7	250	16
Acrylic 06	20.2	6.6	293	43
Acrylic 08	20.8	9.8	333	40
Acrylic 10	21.7	14.8	348	15
Acrylic 12	22.8	20.2	395	47
Acrylic 14	25.2	32.2	415	20

Table 2.

Summary of the Rank Sum relative prickle scores for the single jersey wool fabric samples. For this set of data the least significant difference is 25.¹² (Fiber diameter characteristics are also reproduced in this table for convenience.)

Sample ID	Mean fiber diameter (µm)	Percentage of fibers greater than 30 µm	Relative prickle score	Difference in prickle score from previous sample
GRN0600	15.6	0.2	299	–
GRN0601	16.5	0.1	327	28
GRN0602	18.6	0.6	375	48
GRN0603	19.6	1.7	386	11
GRN0604	20.4	2.5	442	56
GRN0605	21.8	4.4	498	56
GRN0606	22.8	6.6	538	40

Prior to testing on the prototype instrument, fabric samples were pretreated with a simple open steaming with the test face unrestrained.

Results

One approach to measuring the bending rigidity of a fiber is to determine the force necessary to deflect the fiber from its original position. Individual force measurements are technically possible using a very small and highly sensitive force transducer,²³ but this approach is very time consuming given the large numbers of fibers protruding from a fabric. However, the measurement of many fibers can be accomplished using a linear array of force transducers if they can be packed sufficiently close together so as not to miss any fibers. In the limit of close packing, this linear array becomes a continuum or line sensor, which responds to all fibers along its length. This approach is the basis for the development of a practical instrument outlined below.

Instrument development

A tensioned wire was drawn over the surface of a fabric so as to interact with each protruding fiber. The force exerted on the wire was determined from the deflection of the wire as it interacted with each fiber. These very small movements of the wire were detected by imposing a static magnetic field across the wire and measuring the induced current as the wire moved through the magnet field. A convenient device for implementing this approach with existing technology was to use a steel ‘string’ and a magnetic pickup commonly found in an electric guitar. When fibers with sufficient rigidity deflected the wire from its original position, a small electric signal was generated in the coil of the guitar pickup. Figure 2 shows the hand-held prototype sensor, which was constructed using the finest commercially available steel electric guitar string (D’Addario Guitar String Model PL009, diameter 0.23 mm) and an electric guitar pickup. The wire was recessed approximately 0.5 mm inside the base of the prototype device.

Figure 2.

Hand-held sensor and schematic cross-section drawing.

When the sensor was drawn over the surface of a fabric, an electrical signal was generated in the pickup in response to the movement of the wire. These very small signals were amplified, digitized and recorded as an audio file on a computer. Figure 3 shows the signal recorded for a single pass over a fabric. Fiber contact events appear as spikes in the signal, as a result of deviations of the wire from its equilibrium position. A closer inspection of an event reveals that these spikes correspond to small damped vibrations in the wire. Since the density of protruding stiff fibers is low in most fabrics of interest, the occurrence of multiple simultaneous events was found to be very rare.

Figure 3.

Signal from a single pass and a closer view of one portion of the signal.

In order to quantify the performance of this approach, a prototype instrument was constructed which provided control over the operating parameters that were not easily regulated in a manually operated hand-held sensor. The most important of these were determined to be the speed of travel, the distance moved and the compression force applied to the fabric by the sensor. The length of the sensing wire is 120 mm. The fundamental frequency (f) of the wire was set manually to 400 Hz by adjusting the applied tension according to the well-established relationship²⁹

where the wire is of length L with a mass per unit length of m and under a tension T.

Precise speed, distance and positioning of the sensor were achieved using an intelligent brushless motor with an integrated position controller. The motor was coupled to a lead screw that moved an arm attached to the measurement head. A loose coupling was used between the arm and the measuring head to ensure the head could adjust to the thickness of the fabric and to allow for slight undulations in the fabric surface. This loose coupling also meant that the compression of the fabric was minimized and determined by the mass of the measuring head (approximately 185 g, resulting in a pressure of approximately 225 Pa given that the dimensions of the measurement head are 145 mm by 55 mm). Figure 4 shows the prototype instrument with the sensor head raised to allow for easy placement of a fabric sample. The fabric sample is manually placed on the cork tile that forms the base of the instrument. The friction between the cork tile and the sample has been found to be adequate to stop any movement of the sample during the measurement, that is, further clamping of the sample is not necessary. Measurements are made with the sensor head lowered onto the fabric while the sensor traverses back and forth over the fabric at a speed of 30 mm/sec. The scan length was set at 200 mm for the experiments reported here. The amplified signal from the sensor was digitized and analyzed on an attached computer. Data were recorded for both the forward and return parts of the scan.

Figure 4.

Prototype laboratory instrument.

The signal was acquired using a National Instruments data acquisition module and processed in LabVIEW™. Band-pass digital filtering was used to limit the signal to only those features that were close to the fundamental frequency of the tensioned wire. Using available LabVIEW™ routines, the absolute value of the filtered signal was first obtained and then all peaks greater than a configurable threshold were identified and counted. This peak count was used to characterize the prickle propensity of the fabric. Unless indicated otherwise, 10 measurement cycles were recorded without moving the sample. The average and standard deviation of the instrument response from these multiple scans is reported as the instrument response to a particular sample.

One measurement covers a fabric surface area of 120 mm (the length of the wire) by 200 mm (the length of the scan). The sensitivity of the tensioned wire will vary along its length, that is, a fiber end interacting with the wire at the mid point of the span will produce a larger amplitude vibration than a similar interaction near the end of the wire. Thus, while the reported instrument response is a simple summation of the individual observed electronic peak events, it represents a more complex ‘weighted summation’ of each fiber end–wire interaction over the measurement area.

Assessment of acrylic fabrics

Figure 5 shows the mean instrument response for each of the fabrics, while the error bars indicate the standard deviation of the mean for the 10 passes. As the proportion of stiff fibers in the acrylic fabrics increased, the instrument response also increased.

Figure 5.

Instrument response to increasing coarse fiber content in the acrylic fabrics (Table 1). In this and subsequent figures, the instrument response is the number of recorded events and so has no units.

The instrument was able to rank the acrylic fabrics in exactly the same order as the ranking obtained from forearm tests. Figure 6 shows the relationship between the instrument response and the pair-wise fabric ranking obtained from forearm trials. It is noted that since the rank sum prickle value is a relative measure of the panel’s assessment of fabric pairs, there is no reason to expect a linear relationship between the rank sum values and the instrument response. A quadratic relationship, as illustrated in Figure 6, was found to more adequately fit the observed data than a linear relationship.

Figure 6.

Relationship between the instrument response and the sensory prickle data for the set of acrylic fabrics.

Assessment of wool fabrics

Five replicate fabric samples of each different wool quality were tested. The global mean instrument response for ten passes on five replicates of each fabric sample, together with its standard deviation, are presented in Figure 7. Again, the instrument response increased as the proportion of coarse fibers in the wool fabrics increased and the instrument was able to rank the fabrics in exactly the same order as the ranking obtained from forearm tests. Again, as noted above, there is no reason to expect the relationship depicted in Figure 7 to be linear.

Figure 7.

Relationship between the instrument response and sensory prickle data for the set of wool fabrics.

A comparison of the two laboratory prototype instruments (A and B) was made using the single jersey wool fabrics. Again, the prickle propensity was assessed using five replicate physical samples from each of the seven different wool fabrics (i.e. Set 2). Figure 8 shows the between-instrument comparison, giving an R² value of 0.99.

Figure 8.

Linear regression of individual replicate sample measurements for the two instruments. The different symbols represent the seven different wool qualities. The five replicates of each symbol correspond to the five replicate physical samples for each wool quality.

Each point in Figure 8 corresponds to the average of 10 passes on a single physical fabric sample. The five replicate physical samples for each wool quality are presented with the same symbol (e.g. the five solid circles represent the five different physical samples of the fabric made from wool with a mean fiber diameter of 15.6 µm). Note that each of the replicates within a wool quality was individually identified so that pairs of results from the two instruments for each physical sample were correctly formed in the creation of Figure 8. It is interesting in Figure 8, that within a particular wool quality the two instruments rank the replicate samples of the same wool quality in the same order (i.e. the apparent within-replicate sample differences are being consistently identified by both instruments).

Conclusions

A technique for assessing the stiffness of fibers protruding from the surface of a knitted fabric has been devised. A prototype instrument was constructed to provide a quick and reliable measurement of the prickle propensity of a fabric. In tests on single jersey acrylic and wool fabrics, the instrument gave results consistent with the known coarse fiber content of the fabric. In both cases, the instrument results were highly correlated with human assessment of the fabrics, ranking them in the same order as an expert panel of judges. A comparison of two laboratory prototype instruments produced nearly identical results over a total of 35 different single jersey wool fabric samples (7 wool qualities × 5 replicates), confirming the robustness of the approach and the manufacturing design. This instrument provides a new approach to the measurement of fiber stiffness and offers great potential for quickly and reliably assessing the prickle propensity for a fabric when used in knitted garments in next to skin wear.

Footnotes

Funding

This work was supported by the CSIRO and the Australian Cooperative Research Centre for Sheep Industry Innovation.

Conflict of interest statement

None declared.

Acknowledgements

The authors would like to thank Graham Higgerson for helpful scientific discussions during the course of the instrument development, Laurie Staynes for his knitting expertise and Glenda Howarth for her expert technical assistance.

References

Garnsworthy

Gully

Kandiah

Kenins

Mayfield

Westerman

. Understanding the causes of prickle and itch from skin contact of fabrics, CSIRO Division of Wool Technology Report, No 64, 1988. Also published in Australas Textil 1988; 8: 26–29.

Garnsworthy

Gully

Kenins

Mayfield

Westerman

. Identification of the physical stimulus and neural basis of fabric-evoked prickle. Neurophysiology 1988; 59: 1083–1097.

Naylor

GRS

Veitch

Mayfield

Kettlewell

. Fabric-evoked prickle. Textil Res J 1992; 62: 487–493.

Veitch CJ, Kenins P and Naylor GRS. Comparison between laboratory protocol and simulated wear conditions for the evaluation of prickle. In: Proceedings of the 8th International Wool Textile Research Conference, vol. 5, 1990, pp.286-295.

Naylor

GRS

. The role of coarse fibers on fabric prickle using blended acrylic fibers of different diameters. Wool Technol Sheep Breeding 1992; 40: 14–18.

Dolling

Marland

Naylor

GRS

Phillips

. Knitted fabric made from 23.2 µm wool can be less prickly than fabric made from finer 21.5 µm wool. Wool Technol Sheep Breeding 1992; 40: 69–71.

Naylor

GRS

Phillips

Veitch

Dolling

Marland

. fabric-evoked prickle in worsted spun single jersey fabrics. Part 1: the role of fiber end diameter characteristics. Textil Res J 1997; 67: 288–295.

Naylor

GRS

Phillips

. Fabric-evoked prickle in worsted spun single jersey fabrics. Part 2: the role of fiber length, yarn count and fabric cover factor. Textil Res J 1997; 67: 354–359.

Naylor

GRS

Phillips

. Fabric-evoked prickle in worsted spun single jersey fabrics. Part 3: wear trial studies of absolute fabric acceptability. Textil Res J 1997; 67: 413–416.

10.

Naylor

GRS

Stanton

. Time of shearing and the diameter characteristics of fiber ends in processed tops: an opportunity for improved skin comfort in garments. Wool Tech. Sheep Breeding 1997; 45: 243–255.

11.

Naylor

GRS

Oldham

Stanton

. Shearing time of Mediterranean wools and fabric skin comfort. Textil Res J 2004; 74: 322–328.

12.

Naylor

GRS

. Fabric-evoked prickle in worsted spun single jersey fabrics. Part 4: extension from wool to Optim™ fine fiber. Textil Res J 2010; 80: 537–547.

13.

Veitch

Naylor

GRS

. The mechanisms of fiber buckling in relation to fabric-evoked prickle. Wool Technol Sheep Breeding 1992; 40: 31–34.

14.

Liu

. Measurement of the mechanical stinging behaviour of single wool fibers. J Textil Res (China) 2005; 26: 61–63. in Chinese).

15.

Naylor

GRS

Phillips

Veitch

. The relative importance of mean diameter and coefficient of variation of sale lots in determining the potential skin comfort of wool fabrics. Wool Tech. Sheep Breeding 1995; 43: 69–82.

16.

Naylor

GRS

. The relationship between the fiber diameter distribution of wool top, fiber ends and yarn surface fiber. Wool Technol Sheep Breeding 1992; 40: 40–43.

17.

Naylor

GRS

Hansford

. Fiber end diameter properties in processed top relative to the staple for wool grown in a Mediterranean climate and shorn in different seasons. Wool Technol Sheep Breeding 1999; 47: 107–119.

18.

Naylor

GRS

. The coarse fiber component and fiber end diameter characteristics of Australian wool tops. J Textil Inst Part 1 1996; 87: 265–273.

19.

Wang

. Mechanical behaviour of irregular fibers, part III: flexural buckling behaviour. Textil Res J 2002; 72: 573–578.

20.

Naylor GRS. Improving the skin comfort characteristics of wool fabrics. In: Proceedings of the Textile Institute ‘Natural Fibres in Australasia’ Conference, NZ, 2009.

21.

Stevens

. On the psychophysical law. Psychol Rev 1957; 64: 153–181.

22.

Meilgaard

Civille

Carr

. Sensory evaluation techniques, 4th edn. Boco Raton, FL: CRC Press, Taylor and Francis Group, 2007.

23.

Matsudaira

Watt

Carnaby

. Measurement of the surface prickle of fabrics. Part 1: the evaluation of potential objective methods. J Textil Inst 1990; 81: 288–299.

24.

Matsudaira

Watt

Carnaby

. Measurement of the surface prickle of fabrics. Part 2: some effects of finishing on fabric prickle. J Textil Inst 1990; 81: 300–308.

25.

Li-Ming AO, Wen T, Yan-Hong Y and Chong-Wen Y. The analyzing and simulating of the interaction between skin and fabric. In: Proceedings of the Sixth International Conference on Machine Learning and Cybernetics, Hong Kong, 19–22 August 2007.

26.

Anderson

Benson

. Fibre ellipticity and its effect on diameter measurement. J Textil Inst 1953; 44: P98–P104.

27.

Blankenburg

Phillipern

Spiegelmacher

Hahnen

. Correlation of the fibre ellipticity, snippet length and embedding medium with the mean diameter of mohair and wool. IWTO Technical Committee Report No 3. Nice, 1992.

28.

Wang

Mahar

Liu

Wang

. Measurement precision and evaluation of the diameter profiles of single wool fibers. Textil Res J 2007; 77: 981–991.

29.

Sears

Zemansky

Young

. University physics, 5th ed. Addison-Wesley, Reading, Massachusetts, USA, 1976.