Human wetness perception of fabrics under dynamic skin contact

Participants

Sixteen young (22.4 ± 2.5 years) male (8) and female (8) participants, of Western European and North American origin, with no history of sensory-related disorders and active at least 4–6 hours per week, volunteered to participate in this study. The test procedure and instruments were explained to each participant verbally and through a written information form. Following on from this, participants gave written informed consent for participation. Due to the nature of the study, participants were not informed about the detailed aim of the study, experimental conditions, magnitude of the stimuli (amount of water applied) or type of fabric. The protocol and procedures involved were approved by Loughborough University Ethics Committee. The study was conducted within the confines of the World Medical Association Declaration of Helsinki for medical research using human participants.

Specimen

Eight knitted fabrics (120 × 100 mm) selected for their different structure, fiber type, ST properties, thickness, and treatments were included in this experiment (Table 1).

OPEN IN VIEWER

Table 1.

Specifications of the experimental fabrics grouped according to low (L), medium (M) and high (H) thickness and presenting different surface textures (STs) measured as surface roughness (SMD) by the Kawabata Evaluation System. Water content was defined as equal water per volume of textile. The criterion for water content manipulation is reported in the Conditions section

Fabrics	Thickness group	Thickness (mm)	Fiber type	Structure	ST (µm)	Mass (g·m−2)	Material description	Water content (µL)
	LOW
LCO4		0.60	cotton	single jersey	3.7 ± 0.18	140	‘fuzzy’ texture	2400
LPM6		0.56	polyester	single jersey	5.9 ± 1.8	160	multi-chanelled fiber cross-section	2400
LP3		0.56	polyester	single jersey	2.6 ± 0.4	160	plain surface	2400
LP6		0.60	polyester	single jersey	6.4 ± 0.6	160	profiled surface	2400
	MEDIUM
MP2		0.90	polyester	double jersey	1.9 ± 0.3	130	silicon treated ‘silk-like surface’	3600
MP3		1.00	polyester	double jersey	2.7 ± 0.2	280	untreated ‘standard surface’	4000
	HIGH
HP4		2.10	polyester	double jersey	3.6 ± 1.4	330	‘smooth’ texture	8400
HP15		2.10	polyester	double jersey French terry	15.3 ± 1.7	330	‘rough’ texture	8400

The fabrics were grouped in three main clusters according to their thickness characteristics (Table 1): low (0.56–0.60 mm; L), medium (0.90–1.00 mm; M) and high (2.10 mm; H). The results of this study will primarily be applied for the design of base-layers sportswear, usually presenting low thickness characteristics, therefore four fabric samples were included in L and only two fabric samples were included in both M and H.

Within each thickness group the fabrics presented different STs, measured as surface roughness (SMD) by the Kawabata Evaluation System ²⁰ (KES; higher ST corresponds to higher roughness) (Table 1). During the subjective assessments, the face side of the experimental samples was tested only in the wale direction; therefore the KES measurements were also performed in this direction and used for the estimation of ST.

The fabrics were coded according to thickness group (L = low thickness; M = medium thickness, H = high thickness), fiber type (CO = cotton; PM = polyester multi-channeled fiber cross-section; P = polyester) and ST (approximated ST, determined by rounding up to a whole number). For instance, MP2 stands for medium thickness group, polyester and a ST of ∼2. Table 1 reports specifications of the experimental fabric samples.

Experimental set-up

The experimental set-up consisted of: a fabric sample, an adjustable chair where each participant was positioned and a fabric motion rig (Figure 1). OPEN IN VIEWER

Fabric sample

To prevent water spreading across a fabric area larger than 120 × 100 mm, each long side of the experimental fabrics (120 × 100 mm) was fitted to a non-wicking material (200 × 120 mm). The two non-wicking materials together with the fabric, in-between, formed the fabric sample.

Fabric motion rig

Each fabric sample (Figure 1) was placed in a custom-made linear motion rig. The sample was connected to a motor drive on one side and to a counterweight on the other side. The fabric sample could run over two rollers, creating a horizontal area of stimulation. Under this area, the right forearm of each participant was placed onto a height adjustable arm rest, such that the fabric touched the ventral forearm. The latter’s setting was adjustable vertically to ensure equal pressure/contact area in different size arms.

The ventral forearm was selected as the body region of interest for practical reasons. In fact this body site allowed easy application of the fabrics in relation to the design of the motion rig, while maintaining the comfort status of the participants during the trial. Additionally, it has been indicated that the ventral forearm presents the same sensitivity to cold as the upper back, ³ therefore the results can be compared with the existing.^7,12

A dividing wall was mounted onto the fabric motion rig, approximately halfway between the forearm and the arm of each participant. With this setting the participants could not see the experimental textile samples before, during or after the application process, therefore any visual influence on the perceptual responses was prevented.

Each fabric sample was pulled bi-directionally across the skin at a velocity of 0.02 m.s⁻¹. Two fixed levels of pressure were applied: 127 Pa (LOW-P) and 236 Pa (HIGH-P). The order of this two-pressure condition was counterbalanced (the method to measure fabric-to-skin pressure is reported in the Conditions section). The range of travel of each fabric was of 5 cm per stroke, with a total of 8 strokes per fabric, 4 toward the medial forearm and 4 toward the lateral forearm.

Adjusting chair and participant position

A chair was positioned at a standard distance from the fabric motion rig. After each participant settled in the chair, the investigator adjusted the position of the participant’s right forearm on the armrest. The posterior margin of the olecranon was placed in line with the posterior edge of the armrest, the dorsal side of the forearm was positioned in contact with the armrest and the ventral side was left exposed to allow the application of the samples. Because of individual forearm shapes and size differences, to ensure a standard level of contact between the sample and the ventral forearm, each forearm was maintained at a distance of 2 cm above the rollers of the motion rig. Additionally, the height of the chair was adjusted to achieve standard position of the forearm in respect to the arm (90° angle), which varied based on the individual height of the participants.

Conditions

The fabric samples were tested in wet (WET) and dry (DRY) states. In WET, the samples were all treated with an amount of water corresponding to 50% of their total absorption capacity, according to the wetting procedure described in Raccuglia et al. ¹² This amount was shown to deliver the same quantity of water per unit of volume of the different fabrics.

Water absorption capacity was determined according to the ‘water absorption capacity test’ described by Tang et al. ¹⁹ For the test, a fabric sample (100 × 100 mm) was put into a tank of water and 5 min. was allowed for it to sink completely into water. Following on from this, the fabric was taken out by tweezers and hung on a rod vertically until there was no water dripping within a 30-sec. interval. The water gain was calculated according to

Waterabsorbed ( μ L ) = [ wetF ( g ) - dryF ( g ) ] * 1000 μ L g

The fabrics were wetted 30 min. before the experimental trial, in accordance with the order (balanced) of application during the human sensorial assessment. Each fabric was positioned on a plastic film and water was added by using a micropipette (SciQuip LTD, Newtown, UK) positioned at a fixed distance of 10 cm perpendicular to each sample and pointing at its center. When the water was in equilibrium with the fabric, (specifically, when the water spread out uniformly across the sample; this took approximately 1 min.) each fabric was placed into a plastic bag which was securely sealed to prevent water evaporation. No water dripped from the samples inside the plastic bags during the storage period. Given that within each group the experimental samples had the same thickness and the same volume, the fabrics also presented the same relative to volume water content (µL·mm⁻³) and the same, or almost the same, absolute water content (µL·mm⁻²) ¹² (Table 1).

The fabrics were also tested in the DRY condition to observe changes in texture and pleasantness sensation between WET and DRY. In the DRY condition no water was added to the fabric samples, which were in equilibrium with the environment (25℃ ambient temperature and 40% relative humidity). In DRY the pressure applied was 127 Pa (the same as the LOW-P condition, see below).

To confirm the role of resultant fabric-to-skin pressure on WP, as observed under static contact ¹² within the WET condition each fabric sample was tested at two pressure levels: a low pressure of 127 Pa (LOW-P) and high pressure of 236 Pa (HIGH-P). The two pressure conditions were achieved using two different counterweights (200 g and 300 g), attached at one end of each experimental sample, and mounted on the fabric motion rig (Figure 1). Extensive pilot testing was conducted to define the two resultant skin pressures. Results indicated that a pressure of 127 Pa represented the lowest possible pressure applicable in order to ensure enough tension in each WET fabric sample during the pulling process across the skin and to avoid sticking. The HIGH-P of 236 Pa was chosen with the aim of achieving perceivable differences from the LOW-P condition without applying excessive mechanical stimulation. Given the significantly higher weight of the two counterweights compared to the individual wet weight of each fabric sample (4.05–12.00 g), the effect of fabric weight on resultant skin pressure was negligible. To measure the pressure resulting from the application of each fabric sample plus attached clamp and counterweight, a calibrated electronic weighing scale (PSK 360-3, Kern, UK), with a maximum load of 360 g and a precision of 0.001 g, was used. A cylinder with a forearm-like shape, made of hard foam, was placed on the measuring scale and each fabric, at both LOW-P and HIGH-P, was positioned on top of it. The two weight readings (g) were recorded and from these the two corresponding applied pressures (Pa), assuming a surface contact area of 10 × 10 cm, were calculated according to

Pressureapplied ( Pa ) = Weightreading ( kg ) × 9 . 81 m . s - 2 / contactarea ( m 2 ) .

Study overview

Fabrics were assessed in one single experimental trial including three different conditions: DRY; WET LOW-P; WET HIGH-P. Replicates of the eight experimental fabric samples were tested under the three different experimental conditions, therefore a total of 24 fabric samples were tested during the experimental trial. New, fresh fabric samples were used for each participant. The fabrics were assessed using a quantitative sensory test, which consisted of placing, in counterbalanced order, the 24 samples on the right ventral forearm of each participant. Participants reported their local texture sensation, WP, stickiness sensation and pleasantness sensation on ordinal scales (see Measurements section). Prior to the experimental trial, participants were familiarized (∼15 min) with the experimental protocol, procedures and instruments used in the present study. The experimental trial was conducted immediately after the familiarization session. The experiment was performed in a climate-controlled room, maintained at an air temperature of 25.8 ± 0.2℃, relative humidity 39 ± 0.7% and air velocity <0.05 m/s to ensure thermo-neutrality of the participants throughout the trial.

Experimental protocol

In the experimental trial, participants entered the climate-controlled room and were positioned comfortably on the adjustable chair wearing standard T-shirt and shorts. Participants positioned their forearm on the armrest of the motion rig. A reference fabric sample (120 × 100 mm) was placed on the skin, with the long sides of the sample perpendicular to the longitudinal axis of the forearm, and two lines next to these two sides were drawn on the forearm to identify the fabrics’ area of application. The center of the sample was positioned 2/3 above the distal margin of the carpus; the length of the ventral forearm was ∼27 cm and it was measured from the distal margin of the carpus to the coronoid fossa. Participants were then instrumented with two thin skin-temperature sensors (see Measurements section), in the skin area in contact with the fabric, and with temperature sensors across the body to measure body skin temperature. After this, participants rested for 20 min. to allow time for skin temperature to stabilize. After the stabilization period, the investigator applied six reference fabrics on the participants’ ventral forearm, two for each ordinal scale, each corresponding to one of the two extreme points of texture, wetness and stickiness scale. The reference samples were chosen after extensive pilot studies. Specifically, a dry wool (very rough) and a dry silk (very smooth) material were selected as references for texture sensation. Two samples of the same polyester fabric were used as ‘extremely dry’ (no water added) and as ‘extremely wet’ (50% of the total saturation) references. For stickiness, a wet silk fabric (extremely sticky) and a cotton fabric (not sticky), both presenting the same thickness and water content, were chosen as references. The score of each reference fabric was reported by the investigator who also informed the participants that the intensity of the subsequent fabrics would not exceed the range of the two references for each scale. Following on from this, each experimental fabric was applied on the participants’ ventral forearm, moving for a period of 20 sec. Participants were alerted by the investigator before the application of each fabric. At the end of the 20 sec., participants were encouraged to verbally report their texture sensation, WP, stickiness sensation and pleasantness for the stimulated body area, using the four ordinal scales. After 15 sec. of application, Local T_sk was recorded. After 20 sec. the fabric sample was removed and a dry cloth was placed onto the tested body area to avoid any chilly sensation, as a consequence of the evaporation of any remaining water on the skin. The tested skin area was then gently wiped with the cloth and dried by blowing warm air; this took approximately 2 min. and allowed the temperature and hydration state of the skin to return to baseline before the application of the following experimental fabric. Additionally, since the repeated application of dynamic wet stimuli can decrease thermal and tactile sensitivity, 2 min. of rest, before the subsequent fabric application, allowed the recovery of the sensory system. The same protocol was repeated for each of the 24 fabrics. Each experiment (stabilization, familiarization and experimental trial) took approximately 2 hrs and participants were instructed to ask for a rest whenever they felt uncomfortable.

Surface texture

To characterize the ST of the experimental fabrics, SMD was measured using the KES (Kawabata Evaluation System). For the measurement a sensor is in contact with the surface of the fabric under a constant normal force. The sensor consists of a metallic rod connected, by its free end, to a thin wire with a U shape. SMD is calculated from an electrical signal generated by the vertical displacement of the sensor contacting the fabric surface.

Skin temperature

Local T_sk during contact with each fabric was measured with two fine-wire (0.025 mm diameter, time constant of 0.003 sec.) type-T thermocouples (RS Components, Northants, UK). The thermocouple temperatures were monitored and recorded every second throughout the application of the stimulus via a Grant Squirrel SQ2010 data logger (Grant Instrument Ltd., Cambridge, UK). Local T_sk was calculated from the mean of the two measured spots. Local skin temperature drop (Local T_sk Drop), resulting from the application of each wet fabric sample on the skin, was calculated according to

Local T sk Drop = PRELocal T sk - POSTLocal T sk

Before testing, the thermocouples were calibrated by placing the measuring junction of each thermocouple in a circulating water bath whose temperature was monitored with a certified mercury thermometer.

To ensure thermo-neutrality, the skin temperature of five body sites (cheek, abdomen, upper arm, lower back and back lower thigh) was measured throughout the experimental trial, with iButtons wireless temperature loggers (Maxim, San Jose, USA). From these five body sites, mean skin temperature, sampled every minute, was estimated according to the work of Houdas and Ring. ²¹

Texture

To assess perception of fabric texture, i.e. roughness and smoothness, a bipolar balanced ordinal scale was developed (Figure 2: a). To prevent forced choice, the scale had a neutral (0) point in the middle, corresponding to ‘Neither rough nor smooth’. From 0 to 9 (progressive increase in texture), the scale presented different levels of roughness, whereas from 0 to −9 (progressive reduction in texture) different magnitudes of smoothness were displayed. During the scoring process, participants were instructed to first associate the texture of the sample with one of the two attributes, i.e. rough (positive side) or smooth (negative side) and then to report the magnitude of the specific attribute chosen. OPEN IN VIEWER

Wetness

A 30-point unipolar ordinal scale (Figure 2: b) was adopted to assess fabric WP. The scale ranged from 0 to 30, presenting descriptors at points 0, 5, 10, 15, 20, 25 and 30. ¹²

Stickiness

Sensations of fabric stickiness were assessed using a unipolar ordinal scale, ranging from 0 (Not sticky) to 12 (Extremely sticky) and intermediate descriptors at points 3, 6 and 9 (Figure 2: c).

Pleasantness

A bipolar, balanced ordinal scale was developed to assess pleasantness sensation of the tested fabric samples (Figure 2: d). As with the texture sensation scale, this scale presents an unforced choice at the middle point 0 (Neither pleasant nor unpleasant). Points −2, −4 and −6 were linked to the descriptors indicating a progressive reduction in pleasantness, whereas points 2, 4 and 6 were linked to descriptors indicating a progressive increase in pleasantness.

Statistics

The independent variables were: fabric thickness, fabric ST, skin pressure (HIGH-P versus LOW-P), wet state (DRY versus WET). Dependent variables were: Local T_sk Drop, mean T_sk, texture sensation, WP, stickiness sensation and pleasantness.

Data were tested for normality of distribution and homogeneity of variance with Shapiro–Wilk and Levene’s tests, respectively.²²

One-way repeated measures ANOVA tests were conducted to assess whether mean T_sk was significantly different over time (T0–T55) and whether there were differences in Local T_sk Drop (normalized data from baseline) between fabric samples tested under WET.

Texture, wetness, stickiness and pleasantness sensation data were measured through means of ordinal scales and also violated the assumption of normality of distribution, therefore for the statistical analysis non-parametric tests were conducted.

The main effect of fabric ST on WP (WET) was tested using the Wilcoxon signed rank test for M and H (two levels of comparison: MP2 and MP3 for M, HP4 and HP15 for H) and by the Friedman analysis of variance test for LOW (four levels of comparison: LCO4, LPM6, LP3, LP6). ²² The Friedman test was also conducted to test the main effect of fabric ST on texture sensation. Where a significant effect was found, a post hoc analysis was conducted using the Wilcoxon signed rank test.

Wilcoxon signed rank tests were conducted to test the main effect of resultant fabric-to-skin pressure on WP (two levels of comparison for each fabric, i.e. LOW-P versus HIGH-P) and the main effect of wet state on texture sensation and pleasantness sensation (two levels of comparison for each fabric, i.e. DRY versus WET).

Regression analyses were performed to observe the relationships within and between objective (i.e. Local T_sk Drop, fabric total water content, fabric thickness, ST) and subjective (i.e. WP, stickiness sensation, texture sensation) variables, using data from group means. To choose the most suitable regression model, linear and second-order polynomial analyses were performed for each subject. Individual r² values for linear and second-order polynomial models were statistically compared using a paired t-test. The regression model that explained the highest variance was then chosen for the analysis of group mean data.

In all analyses p < 0.05 was used to establish significant differences. Parametric data are reported as mean ± standard deviation (SD). Data were analyzed using the software IBM SPSS Statistics (version 22) (IBM, USA).

Low pressure (LOW-P) condition

Wetness perception predictors

In order to define factors affecting fabric WP under dynamic contact, stepwise regression analysis was conducted. For this analysis, textile factors such as fabric thickness and ST, as well as non-textile factors such as Local T_sk Drop and stickiness sensation, were inputted as independent variables, and WP as the dependent variable.

Fabric ST and/or texture sensation did not appear as relevant predictors. WP was described by stickiness sensation and Local T_sk Drop as the predicting variables (Table 2, Model 1 ), giving an explained variance of 89%:

Wetness perception = 4 . 338 + Stickiness * 1 . 059 + Tsk drop * 5 . 583

OPEN IN VIEWER

Table 2.

Stepwise multiple regression analysis of wetness perception (WP) under LOW-P (low fabric-to-skin resultant pressure) condition

		Unstandardized coefficient		Standardized coefficient
Perceptual variables	Predictor variables	B	SD error	β	t	Sign.	F	r2
MODEL 1 Wetness perception	(Constant)	4.338	2.693		1.758	.006	22.051	.89
	Stickiness	1.059	.247	.642	4.290	.008
	Tsk Drop	5.583	.529	.529	3.529	.017
MODEL 2 Wetness perception	(Constant)	9.444	1.940		4.869	.005	15.322	.86
	Stickiness	1.247	.277	.756	4.493	.006
	Thickness	1.985	.717	.466	2.770	.039

Fabric stickiness sensation was the main predictor with a relatively larger Beta value at 0.64 (p = 0.008), while Local T_sk Drop was found to make a significant additional contribution to the predictive model (β = 0.53, p = 0.017).

Thickness alone did not predict WP (Figure 6), mainly because of the latter’s interaction with stickiness sensation. However, when replacing ‘Local T_sk drop’ with ‘thickness’ and including ‘stickiness sensation’ a similar prediction model of the one above is obtained (r²= 0.86, p = 0.003) (Table 2, Model 2; Figure 8). OPEN IN VIEWER

Figure 8.

Representation of the prediction model of wetness perception (WP) (fabric stickiness sensation and fabric thickness as main predictors). Fabrics grouped according to L (low), M (medium) and high (H) thickness.

High pressure (HIGH-P) condition

Texture and pleasantness sensation

Fabric pleasantness showed a significant relationship (second-order polynomial fit) with texture sensation in both DRY (r²= 0.93, p < 0.001) and WET (r²= 0.89, p < 0.001), (Figure 10). Pleasantness was also significantly related (second-order polynomial fit) with ST in DRY (r²= 0.75, p < 0.005) and WET (r²= 0.39, p < 0.05), although in WET the model presented less predictive power. OPEN IN VIEWER

Pleasantness sensation was significantly reduced in WET (p < 0.05) compared to DRY, apart from LPM6 and HP15 which did not present significant differences between the two conditions (p = 0.53, p = 0.14, respectively).

Texture sensation increased in WET compared to DRY, however the increase was significant only in samples LP3 (p = 0.05), LP6 (p = 0.001) and MP3 (p = 0.03).

Wetness perception and surface texture

Due to the critical impact of tactile sensitivity on WP, we hypothesized that fabric texture properties and/or sensation could affect WP through changes in skin tactile responses, such as stickiness sensation. In particular, we expected the wet smoother surfaces to cause higher stickiness sensation; the latter likely due to the higher number of contact points between the skin and the fabric also causing higher skin displacement compared to the rougher surfaces.

Two of the eight experimental fabrics (LPM6 and MP2) presented different WP responses compared to those fabrics presenting the same water content (Figure 3). However, these differences could not be attributed to the measured fabric texture property per se. For instance, in the low thickness group (L), LPM6 was perceived to be significantly dryer than LCO4, LP3 and LP6, and although LPM6 was rougher than LCO4 and LP3 it was not rougher than LP6 (Table 1). Additionally, LCO4, LP3 and LP6 presented the same WP scores (Figure 3), despite differences in ST. Similarly, in the high thickness group (H), despite the ST of samples HP4 and HP15 being considerably different, 3.65 versus 15.3 respectively, no significant differences in WP were observed (Figure 3). On the contrary in M (medium thickness group) WP was significantly different between samples MP2 and MP3 (Figure 3) even though the difference in ST was quite small (1.9 versus 2.7, respectively).

The above-mentioned observations are validated by the lack of correlations between stickiness sensation and ST as well as between wetness sensation and ST, indicating that changes in stickiness sensation and related wetness cannot be attributed to the ST parameter (i.e. SMD), measured with the KES. It could be argued that the ST of the samples was not different enough to show its influence on stickiness sensation (between 2.6 and 6.4, except for HP15 in which it was 15.3). However, despite these small differences in ST, the participants could sense significant differences in texture across the fabrics, suggesting that the KES is not sensitive enough or not as sensitive as humans.

The significant differences observed within the same thickness group, i.e. between fabrics presenting the same absolute (µL·mm⁻²) and relative water content (µL·mm⁻³), suggest that certain surface and texture properties might still affect the mechanical interaction between the skin and the fabric under wet conditions. However, in order to assess this, measures other than KES or more suitable means able to characterize fabric surface properties are needed.

In L (low thickness group), LPM6 performed as the best fabric in terms of WP, being perceived as drier than LCO4, LP3 and LP6. Fabric LPM6 is a polyester material in which the fiber cross-section consists of a series of closely spaced channels (either tetra- or hexa-channels) increasing the total surface area and facilitating the capillary action. As such, the theory is that moisture is wicked along the fiber surface and spread across a wider fabric surface area, enhancing evaporation. Hence, in LPM6 the faster evaporation rate should have resulted in a higher Local T_sk Drop from baseline, however this was not the case (Figure 3, panel A). Therefore, it is possible that the fuzzy structure of LPM6 reduced skin adhesiveness during the application process, causing lower stickiness sensation (Figure 5) and WP. On the other hand, differences in fiber type between LCO4 (cotton) and LP6 or LP3 (polyester), did not determine changes in ST such as to affect stickiness sensation and related WP. Similarly, the substantial difference in ST between LP3 and LP6, as well as between HP4 and HP15, did not influence stickiness sensation or WP. These two pairs of fabrics had different knit structures, but the yarn type was identical. The latter suggests that changing the knit structure might not affect the mechanical interaction between the skin and the fabric – however, the effect of yarn shape was not investigated. Finally, MP2 was perceived to be significantly wetter and stickier than MP3. Even in this case these differences could not be attributed to the texture parameters measured by the KES, given that the difference in ST was minimal (1.9 in MP2 versus 2.7 in MP3). Nevertheless, it is likely that the silicon treatment applied to MP2 caused higher adhesiveness with the skin under wet conditions, resulting in higher stickiness sensation and WP.

The role of tactile sensitivity on fabric wetness perception

Unlike fabric ST, stickiness sensation was related to WP. When the fabric and/or the skin is wet the higher adhesiveness ²³ increases the frictional force between the two surfaces.^14,23 In normal wear conditions, the higher adhesiveness occurs in response to the increase in the size of the cells of the stratum corneum when it is wet, which result in a higher number of contact points between the fabric and the skin. ¹³ This higher frictional force may cause greater skin displacement, sensed by the cutaneous mechanoreceptors as higher stickiness and perceived as greater wetness. Skin stickiness sensation was not related to fabric water content, likely because the experimental fabrics were tested under the same saturation level (50%). Additionally, given that the same pressure condition was applied, the individual weight of the fabric pressing on the skin did not influence stickiness sensation. Based on this, the skin mechanical stimulation when in contact with a wet material might be affected by various factors and it does not seem as straightforward to identify a single parameter triggering stickiness sensation.

The contribution of tactile sensitivity to WP is corroborated by the significantly different responses between HIGH-P and LOW-P conditions, also observed under static conditions. ¹² In fact, almost all of the experimental fabrics were perceived to be significantly wetter under higher compared to lower pressure conditions. Conversely, Filingeri et al. ²⁴ observed a diminished WP when increasing the contact pressure of the wet stimulus applied to the skin. In Filingeri et al.’s study (2013) a significantly higher contact pressure was applied compared to the current study, (10000 Pa versus 260 Pa) suggesting that there might be a U-shape relationship between WP and contact pressure. However, a contact pressure of 260 Pa seems more realistic for the current applications, therefore reducing the fabric-to-skin pressure is recommended for the design of clothing with reduced WP.

The significant relation between WP and stickiness sensation, as well as the role of fabric-to-skin pressure, indicate that WP can be manipulated by changing the tactile stimulation of the skin. In practice, using fabrics with reduced stickiness sensation features, together with the use of lightweight materials can help the clothing industry in designing garments with reduced moisture discomfort.

The role of thermo-sensitivity on fabric wetness perception

In line with the earlier results obtained in static applications, ¹² a significant relation was observed between WP and Local T_sk Drop. With the increase in fabric water content, the drop in Local T_sk also increases, the latter sensed as higher cooling and associated with greater WP. Fabric water content is mainly influenced by fabric thickness. ¹² Because of the important relation between these two parameters, fabric thickness has been indicated as a critical factor affecting WP under static fabric-to-skin contact. ¹²

In the current dynamic condition no significant relation was observed between WP and fabric thickness. However, when examining the model, it is evident that the lack of relation was mainly caused by two fabrics: LPM6 and MP2. These two fabrics did not fit in the regression line because of their significantly higher and lower stickiness sensation, respectively. The latter suggests that under dynamic skin contact fabric thickness can predict fabric WP only when considered in combination with stickiness sensation. This was shown by the multiple regression analysis, which indicated stickiness sensation and fabric thickness as valid predictors of WP (r²= 0.86).

Because of the correlation between Local T_sk Drop and thickness/water content (Figure 4) a similar prediction model is obtained when replacing the variable ‘thickness’ with ‘Local T_sk Drop’. Indeed, when using Local T_sk Drop, instead of thickness, together with stickiness sensation as variables, a stronger prediction model is obtained (r²= 0.89). This means that Local T_sk Drop is a better predictor than fabric thickness in dynamic conditions, pointing toward the temperature drop being the mechanism of action, and fabric thickness showing an effect due to its correlation with this, based on water content for evaporative cooling.

Pleasantness and texture sensation of dry and wet fabrics

Pleasantness and comfort are criteria commonly used by the consumers when selecting fabrics and clothing. Pleasantness was significantly reduced when fabric texture sensation increased (Figure 10). The significant relation between texture sensation and pleasantness indicates that fabric texture is an important parameter to consider in terms of clothing acceptability, in addition to WP and thermal comfort. Interestingly, under wet state, fabric texture sensation significantly increased compared to dry and resulted in a concomitant reduction in fabric pleasantness sensation. In line with this, Gwosdow et al. (1986) indicated that fabrics felt more textured as skin wetness rose above 20%. Therefore, judgments of fabric texture and associated pleasantness can change in relation to the hydration state of the skin and/or fabric moisture content. As such, evaluations of fabric/clothing texture and related acceptability should be conducted under both dry and wet conditions.

Due to practical reasons and to prevent the effect of personal, environmental and clothing factors on the outcomes, in this study we studied comfort-related properties of fabrics only at the ventral forearm. We speculate that the current results could show a similar trend at different body regions; however it is unknown to what extent the outcomes would be different across the body. For instance, according to the mechanisms underlying skin WP, body regional differences would likely depend on human sensorial factors, such as thermal and tactile sensitivity, as well as anatomical factors, i.e. hair distribution and differences between glabrous/non-glabrous skin. In terms of an overall garment, clothing factors such as the air gap between the skin and the fabric as well as clothing fit (both influencing the level of fabric-to-skin contact) represent additional variables that would influence WP responses across the body. Hence, future researches are necessary to understand how these initial results relate to an overall garment.