Artificial skin for sweating guarded hotplates and manikins based on weft knitted fabrics

Abstract

Measurement devices such as sweating manikins, cylinders or hotplates are used for testing thermal and moisture transfer properties of clothing or textiles. A critical feature of these measurement devices is the design of the outer covering fabric that tightly enfolds the device like a skin. The artificial skin principally has to match individual requirements because the different sweating devices have different sweating systems and surface compositions. In this study knitted fabrics with different fiber and yarn types are proposed to be used as an artificial skin. Thermal and moisture properties of the fabrics were measured to obtain skin-like characteristics and a mathematical model for the quantification of thermal and moisture-management properties based on geometrical characteristics was developed. The results show that the thermal and moisture-management properties of the fabrics do not only depend on the fiber properties but also relevantly on their geometrical properties such as thickness, diameter and number of stitch pores. For example, thermal resistance is significantly affected by the stitch pore diameter, and evaporative resistance by the fabric thickness. Furthermore, water content and drying speed are determined by the capillary structure, and therefore, are more influenced by yarn and fabric structure parameters, whereas contact angle and wettability are more influenced by the fiber type. In conclusion, the tested fabrics satisfy all the requirements to match the anatomical properties of the human skin; however, two fabric types, PES_19f30_SET and PES_28f48_GL, exhibited superior characteristics suitable for application as artificial skin on measurement devices.

Keywords

knitting fiber yarn fabric formation manufacture properties wetting wicking artificial skin

According to international standards and scientific literature, there is a variety of measurement devices used for the investigation of thermodynamic processes in textiles and clothing, such as radiation, conduction, convection, moisture evaporation.¹ As a common basic principle, these thermodynamic measurement devices simulate the metabolic heat loss. A critical feature of such measurement devices is the design of the outer covering fabric that is tightly enfolding the device like a skin. These fabrics are not standardized and a great variety of materials and designs are currently used.

Most common thermodynamic measurement devices for textile and clothing are alambeta and permetest,^2,3 guarded hotplates,^4–6 thermal cylinders^7,8 and thermal manikins.^9–11 They comprise a great variety of materials, such as metals (copper, aluminum, bronze)^4,11,12 or polymers,^13,14 flexible material such as membranes, woven or knitted fabrics,^4,15,16 paintings or coatings^12,17 as their surface finish. The analysis of the literature shows that there are no consistent standards addressing the surface finish.^4,11,18–21 Beside their surface finish many thermal manikins are additionally covered by fabric skin.^14,22 Sweating guarded hotplates are commonly covered by a vapor permeable PTFE-membrane combined with a hydrophobic woven PES-fabric according to the standard DINCEN/TR16422-2012.²³

The thermodynamic properties of the fabric skins are assessed by measurement devices such as the Moisture Management Tester (MMT) or the heat and mass loss method according to ASTM F 2370.²² Different knitted fabrics (cotton, polyester and hydrophilic polyester) used as artificial skin for manikins have been investigated with regards to moisture transfer property measured on fully wetted skin.²⁴ In some cases such as thermo-physiological human simulators, transient measurements with regards to temperature and sweat rate are also applied. To be able to evaluate the behavior of the fabrics in different phases of sweating (onset of wetting, fully developed sweating and drying) four fabrics (cotton with elastane, polyester, polyamide with elastane and fabric skin provided by manikin manufacturer Thermetrics) have been investigated by Koelblen et al.²⁵ Findings of the study showed similarly good performance for the three tested fabrics, however, the recommendation was to choose the type of fabric that best meets the requirements of the sweating system. For the wetting phase, the spreading of a homogeneous liquid on the fabric’s surface is proposed to guarantee the cooling effect as at the human skin. For evaporative resistance measurements, according to standards, a high moisture content with a low drying rate is suggested to ensure a sufficiently steady state to take measurements.

The ability of knitted fabrics to transport heat and mass depends on the fiber properties and the porosity of the yarn and fabric structure. Wetting is determined by the fiber surface and the wetting liquid, whereby the wicking process is mostly affected by the capillaries between the fibers and yarns, thus, by the arrangement of the fibers in yarns and in knitted fabrics. Therefore, the pore size and number of pores in a fabric’s structure are crucial parameters in determining its thermodynamic properties.^26,27 The fabric pore structure also influences further physical properties such as thermal conductivity, capillary flow by the use of intermolecular forces between the liquid and surrounding fiber surfaces, water adsorption within yarn pores and the ability of hygroscopic fibers to swell.^28–31

Research groups investigated how thermal resistance strongly correlates with thickness, mass per unit area and porosity in fabrics related to air entrapped in the fabric structure,³² and that the amount of water wicked from inside to the outer surface strongly correlates with thickness and pore size.^29,33

The geometrical properties of the fabrics used as artificial skin have to match the anatomical properties of the human skin to simulate a human’s heat and mass release realistically by using thermodynamic devices.

The human skin interacts between the body and the environment as a multilayer interface to regulate temperature and therefore heat and mass transport.³⁴ The human skin’s outermost hydrophobic layer, the epidermis, is made up of horny cells, and suet and sweat glands.³⁵ Its thickness varies between 0.05 ∼ 1.2 mm³⁶ and it is estimated to have the ability to contain a water amount of 72% of its skin weight.³⁷ The thermal conductivity λ of the skin varies between 0.5 and 2.8 W m⁻¹K⁻¹ depending on blood circulation and moisture in the hydrophilic body tissue.³⁸ The human skin’s moisture permeability of 0.003 W m⁻²Pa⁻² is the inverse of evaporative resistance R_e.³⁹ Moisture evaporates through sweat glands to the skin’s surface. Sweat glands have an average diameter of 0.4 µm and an average pore number of about 120 per square centimeter depending on the body segment.³⁶ The insensible water loss defines the measured quantity of water passing from inside the body through the epidermis to the ambient atmosphere via diffusion and evaporation processes and depends on several anatomical factors such as anatomical site and thickness or others such as lipid content, blood flow or skin temperature.³⁴ The daily insensible water loss for humans (surface area 1.8 m²) varies between 0.6 and 2.3 l, hands and feet losing the most (50–160 g h^-1), the head and neck losing 40–75 g h⁻¹ and all remaining other parts between 15–60 g h⁻¹.⁴⁰ The capacity for the active secretion of sweat is very large, whole-body sweat losses can exceed more than 2 l h⁻¹ during physical activity,⁴¹ with rates of about 3–4 l h⁻¹ during intense exercising in the heat for short durations.⁴²

In this study, different knitted structures were developed to get skin-like thermal properties to be used as an artificial skin in a relaxed state for thermal measurement devices. For this purpose the designed knitted fabrics were developed to match anatomical properties of the human skin, such as a thickness between 0.05 ∼ 1.2 mm, a pore size of 400 µm and number of pores about 120 cm⁻². Different fiber and yarn types were selected and characterized with regards to pores in knitted fabrics and their influence on measured heat and mass transfer in different sweating phases (wetting, fully developed sweating and drying). Additionally, a functional model was developed using multiple regression analysis and correlations to investigate the relationship between geometrical and heat and mass transfer properties particularly in fabric skins.

The fabric skin as a potential artificial skin for sweating thermal devices was recommended by performing systematic studies to get the best match with human skin.

Specific parameters of fibers, yarns and knitted fabrics such as density, diameter, thickness, yarn length in knitted fabrics, fiber or yarn fineness in knitted fabrics, number of yarns or stitch pores form the basis for the calculation of geometrical structure in the investigated textile fabrics. The symbols and abbreviations for important technical terms are compiled in a combined list (Table 1).
Table 1.
Symbols and abbreviations for the most important technical terms

Acronym Description Unit

ρ_f fiber density g cm⁻³

ρ_y yarn density g cm⁻³

m_f fiber fineness dtex

V_y yarn volume cm³ 10 km⁻¹

V_fy fiber volume in yarn cm³ 10 km⁻¹

V_py pore volume in yarn cm³ 10 km⁻¹

V_pyr relative pore volume in yarn %

l_yt yarn length km m⁻²

ρ_t fabric density g cm⁻³

V_t knitted fabric volume cm³ m⁻²

V_yt yarn volume in knitted fabric cm³ m⁻²

V_ft fiber volume in knitted fabric cm³ m⁻²

V_pyt yarn pore volume in knitted fabric cm³ m⁻²

V_pt total pore volume in knitted fabric cm³ m⁻²

V_ptt stitch pore volume in knitted fabric cm³ m⁻²

V_tt total fabric volume cm³ m⁻²

d_ptt stitch pore diameter µm

m_y yarn fineness tex

n_y number of yarns –

m_yn yarn end-fineness tex

m_t Weight of knitted fabric g m⁻²

h_t fabric thickness mm

n_ptt number of stitch pores per square centimeter cm⁻²

Λ thermal conductivity W m⁻¹ K⁻¹

R_ct thermal resistance m² K W⁻¹

R_et evaporative resistance m² Pa W⁻¹

Θ contact angle °

T wetting time s

R wetting radius mm

SWC surfacial water content g m⁻²

WC water content g g⁻¹

DS drying speed g h⁻¹

SDS surfacial drying speed g m⁻² h⁻¹

Methods

Pore size calculation

To determine the geometrical properties of the knitted fabrics, such as thickness h_t [mm], stitch pore diameter (pore size) d_ptt [µm] and number of stitch pores n_ptt, the entire pore structure in the fabric [cm³ m⁻²] was computed based on the pore and fiber volume [cm³ m⁻²] in individual yarns and in knitted fabrics.⁴³ The entire pore structure in the fabric consists of meso-pores occurring between the fibers in the yarn and makro-pores occurring between interconnected yarns in the fabric that are known as stitch pores. Physical parameters of fibers, such as density ρ_t [g cm⁻³] and the fineness m_f [dtex] of yarns such as diameter d_y [µm] and fineness m_y [tex], and of fabrics such as mass m_t [g m⁻²] and thickness h_t [mm], were measured. The yarn end-fineness m_yn [dtex] is the product of the fineness m_y [tex] and number of yarns n_y.

The specific volume of the fabric V_t is the reciprocal mathematical function of the density ρ_t which was calculated according to Equation 1 and is an important value for the characterization of the fiber and pore volume in the fabric:
$\begin{matrix} ρ_{t} = \frac{m_{t}}{h_{t}} * 0.001 \end{matrix}$
(1)

To characterize the pore structure in the fabrics, the pore and fiber volume [cm³] were calculated for a fabric area of 1 m² and to a yarn length of 10 km. For determining the volume ratio from fibers to pores, the fabric volume V_t [cm³ m⁻²] was calculated as follows:
$V_{t} = \frac{h_{t}}{10} * 10.000$
(2)

The yarn volume V_yt in the knitted fabric [cm³ m⁻²] results from the product of the yarn volume V_y [cm³ 10 km⁻¹] as a result of the yarn fineness m_y [dtex] divided by the yarn density ρ_y [g cm⁻³]. The yarn length in the knitted fabric l_yt [km m⁻²] results from the mass of fabrics m_t [g m⁻²] and yarn end-fineness m_yn [dtex]:
$V_{yt} = V_{y} * \frac{l_{yt}}{10.000}$
(3)

The product of the fiber volume in yarn V_fy [cm³ 10 km⁻¹] and yarn length in the knitted fabric l_yt [km m⁻²] is the fiber volume in the textile V_ft [cm³ m⁻²]:
$V_{ft} = V_{fy} * \frac{l_{yt}}{10.000}$
(4)

The pore volume V_pyt in the fabric [cm³ m⁻²] results from the product of the yarn pore volume V_py [cm³ 10 km⁻¹] and yarn length in the fabric l_yt [km m⁻²]. The pore volume in yarn V_py [cm³ 10 km⁻¹] is calculated by multiplying the yarn volume V_y [cm³ 10 km⁻¹] by the relative pore volume in the yarn V_pyr [%], which is the result of the yarn density ρ_y [g cm⁻³] divided by the fiber density ρ_f [g cm⁻³] minus 1:
$V_{pyt} = V_{py} * \frac{l_{yt}}{10.000}$
(5)

By subtracting the fiber volume V_ft [cm³ m⁻²] from the total volume V_t [cm³ m⁻²] the total pore volume V_pt [cm³ m⁻²] of the knitted fabric is calculated as follows:
$V_{pt} = V_{t} - V_{ft}$
(6)

The stitch pore volume V_ptt [cm³ m⁻²] is calculated analogous to the total pore volume V_pt [cm³ m⁻²] included in the knitted fabrics:
$V_{ptt} = V_{pt} - V_{pyt}$
(7)

To summarize, the sum of the pore and fiber volume [cm³ m⁻²] results in the entire fabric volume balance V_tt [cm³ m⁻²] as,
$V_{tt} = V_{ft} + V_{pyt} + V_{ptt}$
(8)

The number of stitch pores $n_{ptt}$ is calculated by multiplying the stitch pores in wales by the stitch pores in course-direction referred for a fabric area of 1 cm².

Given the assumption that all stitch pores are cylindrical with a circular cross-section, the equivalent pore diameter d_ptt [µm] can be calculated as follows:
$d_{ptt} = 2 * (\frac{\frac{V_{ptt} * 1.000}{n_{ptt} * 10.000}}{\frac{h_{t}}{π}})^{0.5} * 1.000$
(9)

Selection of yarns

As shown in Table 2, yarns made of six different fiber polymers, such as polyamide 6 (PA6), polyamide 6.6 (PA6.6), polyester (PES), polypropylene (PP), rayon (CV) and cotton (CO) were selected and characterized. The chosen fiber types differed in polymeric composition, fiber shape (staple fiber or filament) and texturing processes. A wide selection of different yarns was purposefully chosen to represent knitted fabrics with different geometrical and fiber properties in order to have broad selection of properties to compare to human skin. Fiber types such as PA, PES and CO are already used for fabric skins; CV is known to be highly water absorbing and was, therefore, selected as an additional reference fiber type. PP-fiber is known to be a polymer that does not absorb any moisture, which is most similar to the horny layer of the epidermis.
Table 2.
Characterization of yarn types

Code Polymers Fiber type
Yarn fineness m_y (tex)

staple filament

PA6_17 polyamide 6 x 17

PA66_23f26/2 polyamide 6.6 x 23

PA66_11f34_HE* polyamide 6.6, HE x 11

PA66_11f34_SET* polyamide 6.6, SET x 11

PA66_18f104_GL* polyamide 6.6, GL x 18

PA66_24f48_GL* polyamide 6.6, GL x 24

PES_20 Polyester x 20

PES_19f30 Polyester x 19

PES_19f30_SET polyester, SET* x 19

PES_24 Polyester x 24

PES_28f48_GL polyester, GL* x 28

PP_20 polypropylene x 20

CV_24 rayon x 24

CO_20 cotton x 20

*
GL (flat), SET (fixed), HE (highly extensible)

The fiber types included natural or synthetic fibers with varying length as staple fiber and continuous fibers as filaments. The number of filaments varied between 30 and 104 (e.g. f30 in Table 2). The labeling also marks the applied texturing processes. The addition GL indicates total oriented yarns with a characteristically low elastic and volumetric quality (non-textured filament yarns). The additions SET and HE indicate drawn-textured yarns or fully drawn yarns with a medium elastic/volumetric quality (SET) or a high elastic/volumetric quality (HE), respectively (textured filament yarns). The yarn fineness m_y is specified in the unit [tex].

Manufacturing weft knitted fabrics

Preliminary manufacturing tests were conducted on the knitting machine for a great variety of knitting structures to meet the geometrical properties of human skin (thickness h_t, stitch pore diameter d_ptt and number of pores n_ptt) and ensure mechanical stability. Therefore, different machine settings such as needle gauge, number of stitches in wales and course-direction, knitting structure and sinking depth were altered and provided crucial parameters for the knitting process to achieve the fabric geometrical properties desired for mimicking the human skin. The selected 18 weft knitted fabrics matched the requirements of the human skin structure with defined thickness h_t, stitch pore diameter d_ptt and number of pores n_ptt.

The 14 different yarn types were used to manufacture the 18 weft knitted fabrics on a flatbed knitting 12-gauge machine (type: STOLL CMS330.6 TC). The machine was set to knit 200 stitches in wales (x) and 400 stitches in course-direction (y), so all fabrics were comprised of precisely the same number of stitches or stitch pores. The fabrics were manufactured in double knit structure (both needle beds activated) to obtain a 1 × 1 rib structure. The smallest sinking depth as the specific parameter number given by the knitting machine manufacturer (dimensionless number 8.3) was used on this type of flatbed knitting machine to manipulate the loop length and stitch density and achieve the desired geometrical values such as thickness h_t and stitch pore diameter d_ptt. Eleven fabrics were knitted using the single yarns and seven fabrics were knitted using the double yarn described as the number of yarns n_y and marked with an ending ‘2’ in Table 4. For statistical reasons, all knitted fabrics were reproduced three times and a sample was cut out of each fabric.

After knitting, all fabrics received standard care by washing (easy-care washing program, washing temperature of 40℃, 1000 U min⁻¹) according to the standard EN ISO 6330-2001.⁴⁴ Human skin is hydrophobic at the surface, however, in order to be able to measure mass transfer, the surface has to be fully wetted, therefore the fabric skin has to be hydrophilic. To assess the wettability of fabrics, a simple test was performed by dripping a water droplet on the fabric’s surface from a syringe. The eight fabrics made of the fiber types PES and CV turned out to be highly hydrophobic and were additionally treated by using a padding machine (Mathis, Switzerland), an aqueous polymer solution ARRISTAN AIR (CHT, Germany) and a lab dryer (Mathis, Switzerland) to turn them into hydrophilic specimens as required for the artificial skin.

Experimental design

The size in width and length of all washed and hydrophilic-treated fabrics were measured to define the stitch pore number n_ptt, whereby the total number of counted stitch pores (wales multiplied by course-direction) is divided by the measured fabric area m². The mass m_t according to EN 12127⁴⁵ and thickness h_t according to ISO 5084⁴⁶ of the fabric were measured and the end-fineness of yarns m_yn was calculated by multiplying the yarn fineness m_y according to EN ISO 2060⁴⁷ by the number of yarns n_y. In addition, the pore volumes were calculated as outlined before.

Thermal and wicking properties related to the behavior of the fabrics in different sweating phases, such as onset of wetting, fully developed sweating and drying were measured in the knitted fabrics. The thermal properties, such as conductivity λ and thermal resistance R_ct were determined to evaluate the thermal parameters. These properties are relevant to calculate the thermal resistance R_c of the air layer on the device for reference measurements or of the tested fabric or clothing sample. Since the sensors of the (sweating) guarded hotplate or manikins are actually under the skin, the temperature on the outer side of the skin is lower than that on the sensors. By using thermal parameters, further measurements can be corrected. The evaluation of onset of wetting was indicated by the contact angle θ between liquid and surface and by the water spreading ability on fabrics investigated by the MMT. Therefore, wetting and wicking are influenced by the yarn structure and contact angle. Information about the wetting and wicking properties in fabrics is provided by measuring the time t that is needed to moisten the fabric successively from top to the bottom. The shape of water spot occurring on the bottom surface, determined by measuring the diameter of the spot r, provides information about water spreading ability. The role of evaporative resistance R_et was to allow sufficient evaporation from within the fabric and from its surface, since evaporation takes place in the entire thickness of the fabric in its pores. The maximum water content was determined to evaluate fabric behavior in the fully developed sweating phase. The drying speed DS is a basic indication of fabric behavior in the drying phase. For thermal devices, the surface area is one of the crucial parameters and therefore both properties, maximum water content WC and drying speed DS, are provided in relation to surface named as surfacial maximum water content SWC or surfacial drying speed SDS. All fabric samples were conditioned for 12 hours at the required environmental conditions (temperature of 20 ± 0.5℃ and relative humidity of 65 ± 5%). Tap water was used to pre-wet fabric samples.²⁵

Evaluation of thermal and moisture properties on the artificial skins

Thermal conductivity

The instrument KALCOS was used for measuring conductivity according to standard DIN 52612⁴⁸ at an air temperature of 20 ± 0.5℃ and relative humidity of 65 ± 5%. Each fabric sample with the dimensions of 20 x 20 cm was placed in-between the measuring plates under a defined temperature gradient of 5℃ (upper plate 25℃, lower plate 20℃, normal pressure 2 kPa). The heating power at steady state (maintained for at least 30 min) was measured and the sample’s thermal conductivity λ (W m⁻¹K⁻¹) was calculated. The fabric thickness h_t needed for the thermal conductivity calculation was determined using a thickness meter (Universal Micrometer, accurate to 1 µm, Frank PTI-GmbH, Germany) at the same normal pressure of 2 kPa that was applied during measurement.⁴⁶

Thermal and evaporative resistance

The thermal and evaporative resistances (R_ct and R_et) of three samples of each of the fabrics (25 x 25 cm) were determined using a (sweating) guarded hotplate simulating a human’s heat production by metabolism and heat release by conduction, radiation, convection and evaporation. The tests were performed according to ISO 11092⁴ in a climatic chamber at an air temperature of 20 ± 0.5℃, relative humidity of 65 ± 5%, air speed parallel to the sample of 1 ± 0.05 m s⁻¹ for the thermal resistance, and an air temperature of 35 ± 0.5℃, relative humidity of 40 ± 5%, and air speed of 1 ± 0.05 m s⁻¹ for the evaporative resistance.

Contact angle

The contact angle between the fabric surface and water droplet on the fabric was determined by using the drop shape analyzer DSA-25 (Krüss GmbH, Hamburg, Germany) under climate controlled conditions (air temperature of 20 ± 0.5℃, relative humidity of 65 ± 5%).⁴⁹ For the tests, a water drop with a volume of 4 µl was applied on three samples measuring 1 x 1 cm with a syringe. A camera recorded the wetting process. The angle between fabric surface and drop contour was measured by digital picture analysis. The maximum measurement time was set to 25 s. Surfaces with contact angles between 0° and 90° are defined to be wettable or hydrophilic and surfaces with contact angles between 90° and 180° are defined to be non-wettable or hydrophobic.

Wetting and wicking

The MMT (M290MMT, SDL Atlas, USA) was used to characterize the wetting and wicking processes in three samples of each fabric (ø 8 cm) at an air temperature of 20 ± 0.5℃, relative humidity of 65 ± 5%, and air speed below 0.1 m s⁻¹. Principally, the time to moisten the fabric sample successively from top to the bottom and the radius of wetted spots were measured. The spreading time for the upper fabric surface was set to 20 s. The results were classified according to the terms of the device manufacturer.⁵⁰

Water content and drying speed

The maximal moisture content was determined by weighing the dry fabric samples (10 x 10 cm) and putting them in a water bath at an air temperature of 20 ± 2℃ and relative humidity of 65 ± 5% for four hours. Afterwards, the saturated fabrics were fastened with a clamp on a stand, which was placed on a scale (Mettler AE240-S, accurate to ±0.02 g, Mettler-Toledo, Switzerland) and covered by a shield to prevent weight fluctuation due to air movement in the air-conditioned laboratory. The wet fabrics were observed until water stopped forming drops due to gravitational moisture displacement within the sample and then the actual measurement started. The scale was connected to a computer to record the weight change over time and the total drying time.

Statistical analysis

By using a multiple linear regression model, the relationship between the thermal and wicking properties to the geometrical properties of the fabric were described in the statistical software SPSS (IBM Microsoft, Version 22). This method offers the possibility to assess geometrical properties, their influence on measured heat and mass transfer in different sweating phases and to identify the most influential properties (predictors).

Independent variables determined experimentally and used in the model include thickness h_t, stitch pore diameter d_ptt and number of stitch pores n_ptt (Table 3). Dependent variables for which the regression equations were obtained include thermal resistance R_ct and evaporative resistance R_et and drying speed DS (Table 3). In preliminary correlation analysis and regression equations it was observed that thermal conductivity λ, water content WC, wetting time t and radius r did not correlate with the geometrical properties. And thus, no prediction can be made for these properties using the geometrical properties because the mean variation is very large and the correlation coefficient R² less than 0.2.
Table 3.
Description of the variables and constants in the model

Variables Unit Variable types Symbol in model Regression coefficients

Thickness h_t mm Independent x₁ a₁

Stitch pore diameter d_ptt µm Independent x₂ a₂

Number of stitch pore n_ptt – Independent x₃ a₃

Thermal resistance R_ct m² K W⁻¹ Dependent y₁ –

Evaporative resistance R_et m² Pa W⁻¹ Dependent y₂ –

Drying speed DS g h⁻¹ Dependent y₃ –

Constant – – – b_y1, b_y2, b_y3

The relationship type between independent and dependent variables (linear or others) was determined based on scatter plots of the experimental data for individual predictors and was tested statistically:
$y = a_{1} * x_{1} + a_{2} * x_{2} + a_{3} * x_{3} + b$
(10)
where a and b are sought coefficients for the defined dependent variables and each predictor (independent variables). The regression coefficient indicates the slope of the function. Unstandardized coefficients and the standard error indicated the degree each predictor affected the outcome if the effects of all other predictors were held constant. The p-value showed the significance of contribution to predicting the outcome for each influential predictor (p-values less than 0.05 are significant). To assess the goodness-of-fit the root-mean-square deviation (RMSD), which represents the average difference between the measured results of the thermal and wicking properties and the corresponding calculated properties based on the geometrical properties of the fabrics. It is defined as:⁵¹
$rmsd = \sqrt{\frac{\sum (y_{measured} - y_{model})^{2}}{n}}$
(11)
where y-measured is the measured value of the thermal and wicking properties (thermal resistance R_ct, evaporative R_et resistance and drying speed DS), y-model is the calculated value using the regression equations, and n is the number of experimental data points. The RMSD is an indicator of model goodness-of-fit that can be assessed by comparing RMSD values and the standard deviation of the measured data. The average difference between trend line and points is considered acceptable when the RMSD-value is smaller than the average standard deviation of the data set.

Results

Physical and thermal properties of fabrics

The results of the measurements of the end-fineness of yarns m_yn, mass m_t, thickness h_t and density ρ_t of the fabrics are presented in Table 4. The calculated numbers n_ptt and diameter d_ptt of stitch pores are also presented in Table 4.

Table 4 describes the structure of the manufactured fabrics. The end-fineness m_yn of the yarns varied between 11 and 49 tex and correlates with the fabric weight m_t, which varied between 140 and 416 g m⁻². The fabric thickness h_t varied between 0.91 and 1.50 mm and the fabric density ρ_t varied between 0.14 and 0.35 g cm⁻³. The fabrics made of textured filament yarns (marked with ‘x’ in Table 4) represented densities ρ_t smaller than 0.25 g cm⁻³, which was on average smaller than many fabrics made of non-textured filament yarns. The double-yarn processed fabrics had higher mass m_t and density ρ_t than the single-yarn processed fabrics. The stitch pore number n_ptt ranged between 190 cm⁻² (for PES_24) and 362 cm⁻² (for PA66_11f34_HE) by a factor of about two. Fabrics made of filament yarns tended to have more stitch pores than fabrics made of staple fiber yarns due to decreasing yarn fineness m_yn, and consequently, a smaller stitch pore diameter d_ptt as shown in Table 4. The stitch pore diameter d_ptt ranged from 225 µm (for PA66_11f34_SET_2) to 710 µm (for PES_24).
Table 4.
Physical properties of the fabrics

Code Textured m_yn (tex) m_t (g m⁻²) h_t (mm) ρ_t (g cm³) n_ptt (cm⁻²) d_ptt (µm)

PA6_17_2* 34 304 1.25 0.24 226 584

PA66_28f26/2 28 254 1.30 0.20 283 472

PA66_11f34_HE x 11 162 1.05 0.16 362 393

PA66_11f34_SET x 11 140 1.02 0.14 351 390

PA66_11f34_SET_2* x 22 237 1.13 0.21 294 225

PA66_18f104_GL 18 171 0.91 0.19 254 554

PA66_18f104_GL_2* 35 394 1.14 0.35 318 334

PA66_24f48_GL 24 275 1.17 0.23 326 402

PES_20_2* 39 352 1.14 0.31 213 560

PES_19f30 19 172 1.14 0.15 244 554

PES_19f30_SET x 18 154 1.07 0.14 202 632

PES_19f30_SET_2* x 37 306 1.21 0.25 207 470

PES_24 24 200 1.32 0.15 190 710

PES_24_2* 49 416 1.50 0.28 215 567

PES_28f48_GL 28 244 1.07 0.23 236 513

PP_20 20 164 1.16 0.14 193 672

CV_24_2* 48 401 1.36 0.29 199 555

CO_20 19 178 1.04 0.17 213 644

*
_2 implies the number of yarns n_y

In order to validate the calculated equivalent diameter d_ptt and number of stitch pores n_ptt for a fabric area of 1 cm², a comparison between the microscopic and graphic display of these geometrical parameters are shown in Figure 1.
Figure 1.
Comparison between the microscopic and graphic display of stitch pore diameter d_ptt and number of stitch pores n_ptt in manufactured knitted fabrics.

The results from the conductivity and resistance measurements to investigate thermal properties and the measurements of contact angle, moisture-management testing, water content and drying speed to investigate moisture properties are shown in Table 5.
Table 5.
Thermal and moisture properties determined in fabrics

Fabric sample λ (W m⁻¹K⁻¹) R_ct (m² K W⁻¹) R_et (m² Pa W⁻¹) θ (°) t (s) r (mm) SWC (g m⁻²) WC (g g⁻¹) DS (g h⁻¹) SDS (g m⁻² h⁻¹)

PA6_17_2 0.059 ± 0.001 0.043 ± 0,000 5.92 ± 0.14 150 ± 26 7.1 ± 1.3 14.0 ± 2.2 971 ± 29 3.26 ± 0.10 14.4 ± 2.7 1593 ± 266

PA66_23f26/2 0.052 ± 0.000 0.027 ± 0.002 4.30 ± 0.27 129 ± 6 15.3 ± 3.3 8.0 ± 2.7 941 ± 28 3.72 ± 0.11 14.1 ± 0.9 1413 ± 85

PA66_11f34_HE 0.047 ± 0.000 0.026 ± 0.000 2.69 ± 0.08 0 ± 0 3.7 ± 0.3 12.0 ± 2.7 614 ± 18 4.60 ± 0.14 9.1 ± 0.6 917 ± 60

PA66_11f34_SET 0.046 ± 0.000 0.027 ± 0.001 2.91 ± 0.03 144 ± 6 4.7 ± 2.0 9.0 ± 4.2 602 ± 18 3.95 ± 0.12 9.1 ± 0.7 913 ± 72

PA66_11f34_SET_2 0.050 ± 0.001 0.019 ± 0.000 3.20 ± 0.20 180 ± 0 5.8 ± 0.9 15.0 ± 5.0 631 ± 19 3.13 ± 0.09 9.2 ± 2.9 977 ± 289

PA66_18f104_GL 0.041 ± 0.003 0.019 ± 0.001 2.59 ± 0.03 138 ± 22 4.7 ± 1.0 9.0 ± 4.2 354 ± 11 2.11 ± 0.06 10.1 ± 1.8 923 ± 175

PA66_18f104_GL_2 0.062 ± 0.001 0.019 ± 0.001 4.35 ± 0.11 180 ± 0 6.7 ± 1.0 12.0 ± 2.7 740 ± 22 1.92 ± 0.06 11.9 ± 2.4 1170 ± 244

PA66_24f48_GL 0.052 ± 0.000 0.016 ± 0.001 3.75 ± 0.02 0 ± 0 5.4 ± 0.4 16.2 ± 2.2 424 ± 13 1.59 ± 0.05 9.5 ± 0.4 943 ± 38

PES_20_2 0.057 ± 0.000 0.035 ± 0.000 4.45 ± 0.10 0 ± 0 3.3 ± 0.1 21.0 ± 2.2 832 ± 25 2.30 ± 0.07 15.5 ± 0.9 1547 ± 95

PES_19f30 0.042 ± 0.002 0.024 ± 0.002 3.03 ± 0.16 0 ± 0 2.9 ± 0.1 24.0 ± 2.2 575 ± 17 3.15 ± 0.09 9.6 ± 0.4 960 ± 40

PES_19f30_SET 0.045 ± 0.002 0.027 ± 0.002 2.70 ± 0.06 0 ± 0 3.4 ± 0.2 19.0 ± 2.2 686 ± 21 4.23 ± 0.13 11.6 ± 0.3 1150 ± 30

PES_19f30_SET_2 0.051 ± 0.000 0.019 ± 0.001 3.46 ± 0.02 0 ± 0 3.9 ± 0.2 20.0 ± 0.0 775 ± 23 2.38 ± 0.07 13.9 ± 0.2 1383 ± 23

PES_24 0.046 ± 0.001 0.046 ± 0.000 4.26 ± 0.05 0 ± 0 2.9 ± 0.4 27.0 ± 2.7 777 ± 23 3.38 ± 0.12 12.3 ± 0.6 1233 ± 57

PES_24_2 0.054 ± 0.000 0.035 ± 0.001 5.07 ± 0.07 0 ± 0 3.7 ± 0.1 20.0 ± 0.0 874 ± 26 2.10 ± 0.06 17.2 ± 0.3 1713 ± 32

PES_28f48_GL 0.046 ± 0.003 0.018 ± 0.000 2.97 ± 0.19 0 ± 0 3.0 ± 0.2 27.0 ± 2.7 364 ± 11 1.46 ± 0.04 8.9 ± 1.2 957 ± 124

PP_20 0.041 ± 0.001 0.036 ± 0.000 4.03 ± 0.10 138 ± 31 4.7 ± 1.7 13.0 ± 2.7 589 ± 18 2.37 ± 0.10 14.2 ± 2.0 1193 ± 197

CV_24_2 0.057 ± 0.000 0.027 ± 0.001 4.20 ± 0.02 0 ± 0 4.5 ± 0.2 15.0 ± 0.0 893 ± 27 2.50 ± 0.07 16.7 ± 0.4 1670 ± 40

CO_20 0.053 ± 0.001 0.038 ± 0.002 3.97 ± 0.17 180 ± 2 7.4 ± 2.5 11.0 ± 4.2 672 ± 20 4.77 ± 0.14 10.2 ± 1.3 1073 ± 127

λ
– thermal conductivity; R_ct – thermal resistance; R_et – evaporative resistance; θ – contact angle; t – wetting time; r – wetting radius; SWC – surfacial water content; WC – water content; DS – drying speed; SDS – surfacial drying speed.

Thermal conductivity

Fabrics consisting of staple fiber yarns resulted in higher values for thermal conductivity λ compared to fabrics consisting of filament yarns. The fabrics consisting of filament yarn PA66_11f34 textured as HE and SET showed only a difference of about 0.001 W m⁻¹K⁻¹.

In the fabrics consisting of the yarn PA66_11f34_SET the thermal conductivity λ differed by about 0.004 W m⁻¹K⁻¹, which was related to the number of yarns n_y. In comparison, for the fabrics consisting of non-textured filament yarn PA66_18f104_GL, thermal conductivity λ differed by a five times higher value of about 0.021 W m⁻¹K⁻¹.

The thermal conductivity λ of human skin is higher by more than one power of ten than that of the fabrics.

Thermal and evaporative resistance

Fabrics made of staple fiber yarns systematically showed values higher than 0.027 m² K W⁻¹, fabrics made of filament yarns showed values lower than 0.027 m² K W⁻¹. The double-yarn processed fabrics had lower values for thermal resistance R_ct than the single-yarn processed fabrics.

The evaporative resistance R_et ranged between about 2.5 and 6.0 m² Pa W⁻¹. Particularly, fabrics made of staple fiber yarns showed values higher than about 4.0 m² Pa W⁻¹, and fabrics made of filament yarns showed values lower than 4.0 m² Pa W⁻¹, except for PA66_100f26/2, which had a value of 4.30 m² Pa W⁻¹. The double-yarn processed fabrics had higher values for evaporative resistance R_et than the single-yarn processed fabrics.

Thermal resistance R_ct is basically the inverse of thermal conductivity λ. Human skin has different values depending on its blood circulation and moisture compared to the fabrics, which varied between 0.016 and 0.046 m² K W⁻¹. In comparison to the human skin moisture permeability of 0.003 W m⁻²Pa⁻², the inverse of the evaporative resistance R_et of the fabric varied in a range between 0.16 and 0.39 W m⁻²Pa⁻².

Contact angle

In Table 5 it is shown that the treated fabrics made of PES- and CV-fiber and two PA-fiber fabrics (PA66_11f34_HE and PA66_24f48_GL) were highly hydrophilic with contact angles θ of 0°, while the other fabrics were hydrophobic with contact angles θ greater than 90°. Contact angles between 0° and 90° were not observed.

Wetting and wicking

In the top-to-bottom surface wicking process in the tested knitted fabrics, the wetting time of the bottom surface was slightly delayed compared to the top surface (since the water droplets were applied from the top in the MMT device). The hydrophilic fabrics (contact angle θ = 0°) mostly showed a short top-to-bottom wetting time of less than about 5 s. Most of the hydrophobic fabrics (contact angle θ > 100°) showed a wetting time between about 5 and 15 s (medium). The wetting radius on the top and bottom surface of the fabric, respectively, was approximately the same size in both hydrophilic and hydrophobic fabrics. The hydrophilic fabrics (θ = 0°) mostly showed a wetting radius on the bottom between about 15 and 28 mm (large). The wetting radius on hydrophobic fabric surfaces (θ > 100°) was slightly smaller and ranged between about 10 and 15 mm (medium). The hydrophobic fabrics PA66_23f26/2, PA66_11f34_SET and PA66_18f104_GL showed radii smaller than 10 mm. The effect of both transversal and lateral wetting is related to the water affiliation properties of the fabrics (e.g. contact angle) for initial wetting and the capillary forces within the fabric structure for further spreading of the moisture within the fabric. The particular results obtained for the fabrics considered in this study follow this logic consistently.

Water content and drying speed

The maximum surfacial water content SWC was determined and varied between 364 and 971 g m⁻² with high standard deviations. The relationship between wetted and dry weight referred to m⁻² was derived from determining the water content. The single-yarn processed fabrics showed higher values than double-yarn processed fabrics. Fabrics from textured filament yarns and usually staple fiber yarns showed water content WC values higher than 2.30 g g⁻¹ compared to fabrics from non-textured filament yarns with, unexceptionally, only values smaller than 2.30 g g⁻¹ and thus, maximum water content WC was about three times higher in comparison to human skin with an average of 0.75 g g⁻¹.

The surfacial drying speed SDS of the knitted fabrics varied from 917 to 1713 g m⁻² h⁻¹. The fabrics of staple fiber yarns usually had surfacial drying speeds higher and fabrics of filament yarns lower than 1000 g m⁻² h⁻¹. The surfacial drying speed SDS decreased by the number of yarns n_y.

The daily insensible water loss for humans varying between 22 and 89 g m⁻² h⁻¹, exceeding up to 1000 g m⁻² h⁻¹ during physical performance, with rates up to about 1600 g m⁻² h⁻¹ lies within the same range as the surfacial drying speed SDS of fabrics.

Statistical analysis

The conducted correlations between predictors related to dependent variables are presented in Figure 2, whereby the predictor thickness h_t was related to evaporative resistance R_et and drying speed DS showed coefficients of determinations R² of about and higher than 0.5, also the predictor diameter of stitch pores d_ptt was related to thermal resistance R_ct with an R² of about 0.5 and RMSD values smaller than the average standard deviation of the given data set.
Figure 2.
Correlations between predictors and dependent variables and the corresponding coefficient of determination.

By replacing the sought coefficients a and b, the specific model for thermal resistance R_ct, evaporative resistance R_et and drying speed DS was defined and the most influential predictors identified. In all regression models the standard error was very low.

The specific model for thermal resistance R_ct was defined as:
$\begin{matrix} R_{ct} = 0.023 * h_{t} + 6.50 * 10^{- 5} * d_{ptt} \\ + 5.49 * 10^{- 5} * n_{ptt} - 0.046 \end{matrix}$
(12)

The value a₂ indicated that as the diameter of stitch pores d_ptt increased by one unit, thermal resistance R_ct increased by $6.50 * 10^{- 5}$ units. The diameter of stitch pores d_ptt was identified as the most influential predictor $(p - value = 0.008)$ . The fit of this multiple regression model was assessed with the coefficient of determination R² = 0.60 and root-mean-square deviation RMSD = 0.002 m² K W⁻¹ (the mean standard deviation of measured R_ct approximated 0.001 m² K W⁻¹ for Rct values in the range of 0.016–0.046 m² K W⁻¹).

The specific model for evaporative resistance R_et was defined as:
$R_{et} = 4.48 * h_{t} + 0.001 * d_{ptt} + 0.001 * n_{ptt} - 2.32$
(13)

The value a₁ indicated that as thickness h_t increased by one unit, evaporative resistance R_et increased by 4.48 units. The thickness h_t was identified as the most influential predictor (p-value = 0.004). The fit of this multiple regression model was assessed with R²= 0.62 and root-mean-square deviation RMSD = 0.15 m² Pa W⁻¹ (the mean standard deviation of measured R_et approximated 0.10 m² Pa W⁻¹ for R_et values in the range of 2.59–5.92 m² Pa W⁻¹).

The specific model for drying speed DS was defined as:
$R_{DS} = 12.58 * h_{t} + 0.00 * d_{ptt} - 0.018 * n_{ptt} + 2.09$
(14)

The value a₁ indicated that as thickness h_t increased by one unit, drying speed DS increased by 12.58 units. The thickness h_t was identified as the most influential predictor (p-value = 0.002). The predictor number of stitch pores n_ptt showed a negative value, indicating a negative relationship to the outcome. The fit of this multiple regression model was assessed with R² = 0.62 and root-mean-square deviation RMSD = 1.3 g h⁻¹ (the mean standard deviation of measured DS approximated 1.1 g h⁻¹ for DS values in the range of 8.9–17.2 g h⁻¹).

Discussion

The geometrical properties of the weft knitted fabrics developed in this study are similar to the anatomical properties of the human skin. In particular, the thickness h_t (from 0.90 to 1.20 mm), stitch pore diameter d_ptt (from 200 to 700 µm) and stitch pore number n_ptt (from 190 to 360 cm⁻²) of the knitted fabrics are very similar to the characteristic distribution of thickness between about 0.05 ∼ 1.2 mm, sweat gland number between about 80 and 500 cm⁻² or sweat gland diameter of about 400 µm of the human skin. The water content WC of the fabrics varied between 354 and 971 g m⁻² (1.46 and 4.77 g g⁻¹), and thus, it was about three times higher in comparison to human skin with an average of 0.75 g g⁻¹. The wetting contact angle θ of the fabrics ranged from 0° (highly hydrophilic) to 180° (highly hydrophobic), whereby the contact angle θ of human skin is averagely 104° (hydrophobic). Interestingly, the tested hydrophobic fabrics PA66_23f26/2 with 129 ± 6°, PA66_11f34_SET with 144 ± 6°, PA66_18f104_GL with 138 ± 22° and PP_20 with 138 ± 31° best met the hydrophobic top layer of human skin (epidermis). Nonetheless, the main interest of this study focuses on the functionality of fabric skin as the imitation of human-like sweating, which implies good moisture spreading properties for efficient skin wetting. In this case the hydrophilic fabrics have a clear advantage to fulfill the expected wetting and wicking performance of the fabric skin. Another unexpected phenomenon was that the fabrics made of PA-fibers expected to be highly hydrophilic turned out to be highly hydrophobic. The different yarn structure and number of yarns in the fabrics might explain the hydrophobicity.²⁸

The conductivity λ of the fabrics investigated, which ranged from 0.041 to 0.062 W m⁻¹K⁻¹, increased by the increase of fabric density ρ_t and had a quantity between air (0.025 W m⁻¹K⁻¹) and water (0.598 W m⁻¹K⁻¹)⁵² saying that knitted fabrics as porous material mainly consist of pores formed by fibers filled with air or water. The thermal resistance R_ct of the fabrics investigated ranged from 0.016 and 0.046 m² K W⁻¹ and is basically the inverse of the thermal conductivity, however the thermal resistance values are about five times higher because different measurement conditions such as compression and air movement on fabric influence its surface structure. The thermal resistance R_ct of air with 0.022 m² K W⁻¹ ⁵³ had a quantity in the middle of the fabric’s range. The thermal conductivity λ and resistance R as insulation properties are higher by more than one power of ten for the human skin consisting of hydrophilic body tissue and hydrophobic epidermis depending on blood circulation and moisture compared to fabric skins. However, the knitting structure provides geometrical structure to transfer mass for simulating the wicking process.

The thermal properties, such as conductivity λ and thermal resistance R_ct, are relevant to calculate the thermal resistance R_c of the air layer on the device for reference measurements or of the tested fabric or clothing sample. Since the sensors of the (sweating) guarded hotplate or manikins are actually under the skin, the temperature on the outer side of the skin is lower than that on the sensors. By using thermal parameters further measurements can be corrected.

The role of evaporative resistance R_et was to allow sufficient evaporation from within a fabric and from its surface since evaporation takes place in the entire thickness of the fabric in its pores. In comparison to a human skin moisture permeability of 0.003 W m⁻²Pa⁻², the inverse of the evaporative resistance R_et of the fabric varied in a range between 0.16 and 0.39 W m⁻²Pa⁻².

The wettability of fabrics affects moisture behavior, whereby hydrophilic fabrics (contact angle θ = 0°) show short wetting times and large wetting radii on the bottom of measured fabrics. The cooling effect that occurs on top of human skin is imitated on the bottom of fabric skins where the actual cooling and wetting take place. According to Koelblen et al.²⁵ the cooling effect shows a desirable characteristic for sweating simulation on fabric skins. They also discussed how measured moisture properties might be different in horizontal and vertical orientations of the measurement surface or skin, and the cooling effect might be smaller when testing samples in a vertical orientation due to water flowing downwards under the influence of gravity, and also forces and couples acting in-between the fibers might manipulate moisture behavior. Therefore, fabric samples with lower downward wetting ranges are recommended due to moisture migration and the dripping of water.

Fabrics from textured filament yarns and usually staple fiber yarns reach a water content based on dry weight higher than 2.30 g g⁻¹, while non-textured filament yarns have a water content based on dry weight lower than 2.30 g g⁻¹. The water content WC is about three times higher in comparison to human skin with an average of 0.75 g g⁻¹. That means that the critical thickness of the water layer on the skin's surface and water droplets start to drip off laying between 364 and 971 g m⁻² and are hardly comparable with human skin with about 37.6 g m⁻².⁵⁴ Due to water on fabrics surface traps inside the fabric skin structure while water on humans’ skin forms droplets and runs off the surface. Also, literature data were given from only one test person, which is not conclusive enough for statistical reliability.

Fabrics made of staple fiber yarns usually show drying speeds higher than fabrics made of filament yarns, and the double-yarn processed fabrics have higher drying speeds in comparison to single-yarn processed fabrics due to the pore structure in yarns and fabrics. The drying speed DS is related to the amount of captured water within the fabric, as human skin is related to the blood circulation and water loss.

The daily insensible water loss for humans varying between 22 and 89 g m⁻² h⁻¹, exceeding up to 1000 g m⁻² h⁻¹ during physical performance, with rates up to about 1600 g m⁻² h⁻¹, lies within the same range as the surfacial drying speed SDS of fabrics varying between 917 and 1593 g m⁻² h⁻¹, which indicates that the fabrics ensure continuous wicking and moisture transfer for sweating simulation.

The statistical analysis of measured properties in Figure 2 shows correlations between the geometrical properties such as thickness h_t, which are related to evaporative resistance R_et and drying speed DS with coefficients of determinations R² of about and higher than 0.5, also the diameter of stitch pores d_ptt is related to thermal resistance R_ct with an R² of about 0.5. Thermal conductivity λ, water content WC, wetting time t and radius r do not correlate with geometrical properties with R² being below 0.2. And thus, it is not possible to accurately predict these parameters using the geometrical properties because the mean variation is very large. In the resultant regression equations (equations 12–14) the thermal and evaporative resistance R_ct and R_et, and drying speed DS of the knitted fabrics are predominantly affected by the most influential geometrical properties, such as thickness h_t, diameter of stitch pores d_ptt and stitch pore number n_ptt, whereby the p-value smaller than 0.05 proved their significance. The stitch pore diameter d_ptt significantly contributes to thermal resistance R_ct with a coefficient of $6.50 * 10^{- 5}$ , whereby the thickness h_t significantly contributed to the evaporative resistance R_et with a coefficient of 4.48 and also the drying speed DS with a value of 12.58. Finally, the R² values were higher or equal to 0.60 indicating a medium-strength correlation. On the other hand, the RMSD values were very close to the mean standard deviation of the measured values (RMSD = 0.002 m² K W⁻¹, mSD = 0.001 m² K W⁻¹ for R_ct; RMSD = 0.15 m² Pa W⁻¹, mSD = 0.10 m² Pa W⁻¹ for R_et; RMSD = 1.3 g h⁻¹, mSD = 1.1 g h⁻¹ for DS, respectively) and smaller by several orders than the measured values (below 6%, 9% and 15% for R_ct, R_et and DS, respectively), which indicates relatively strong predictive power of the developed regression equations.

The different sweating devices have different sweating systems, e.g. water supply through horizontally oriented porous metal plates, sweat glands distributed on vertically oriented cylindrical shapes or water supply on artificial skin draped on measurement surface. Therefore, the artificial skin properties need to be chosen individually to match the requirements of the specific devices. Furthermore, pre-wetted skins used on non-sweating devices require high water content WC and a low drying speed DS to provide a longer steady state for measuring, whereas a fast drying speed DS and low water content WC are beneficial to mimic artificial skin used on devices including continuous water supply. The requirements are related to thermal properties (conductivity, thermal and evaporative resistance) and moisture behavior (water content, drying speed) investigated on the fabrics, and are influenced by geometrical characteristics and the wettability of the fabrics.

The fabric sample PES_19f30_SET made of fiber polymer PES and textured yarn (FDY) provides highly desirable thermal and moisture-management properties to mimic human skin to be used as pre-wetted artificial skin on measurement devices. The hydrophilic fabric presents low wetting times and a large wetting radius in comparison to a high water content of about 21 g g⁻¹ and a relatively low surfacial drying speed of 1150 g m⁻² h⁻¹. However, to mimic human skin for use as artificial skin on measurement devices with a continuous water supply, the fabric type made of non-textured yarn (TOY) is preferred. Therefore, the hydrophilic fabric sample PES_28f48_GL made of fiber polymer PES meets the requirements with low wetting times, a large wetting radius and even a low water content of about 11 g g⁻¹ and related fast surfacial drying speed of 957 g m⁻² h⁻¹.

A comparison of the results on proposed fabric skins with existing fabric skins (25) will be subjected to further studies. However, the focus of this study was to create fabrics with specific properties based on the developed theoretical model and in this way, the fabrics’ performance can be improved by changing production parameters.

The authors Postle and Munden^55,56 investigated how the fiber and pore structure in relaxed plain-knitted fabrics are influenced by flexural and torsional strains in fibers, which implies a function of force distribution that varies the effective yarn diameter and mesh pore diameter. That approach has not been considered in this study yet because the equivalent pore diameter d_ptt was calculated from the relevant geometric modeling of knitting structures. Additionally, further investigations are necessary on artificial skins based on weft knitted fabrics fashioned as tight-fitted garments dressed on cylindrically shaped and upright positioned devices, e.g torso or sweating manikin in stretched conditions. Therefore, a redistribution of geometrical characteristics such as stitch pore diameter d_ptt and thickness h_t might occur and affect the thermal and moisture behavior.

Conclusion

The tested fabrics satisfy all the requirements to match the anatomical properties of the human skin, such as thickness, pore size and number of pores, used as artificial skin on measurement devices measuring thermal and moisture properties.

The regression analysis offers the possibility to assess the impact of geometrical properties thickness h_t, stitch pore diameter d_ptt and number of stitch pores n_ptt on measured heat and mass transfer in different sweating phases and to identify the most influential predictors. The stitch pore diameter d_ptt significantly increased thermal resistance R_ct although with an extremely low value, whereby the thickness h_t significantly increased the evaporative resistance R_et and also the drying speed DS. The knitting structure provides geometrical structure to transfer mass for simulating the wicking process.

Two fabric samples are proposed for different types of sweating systems. The fabric sample PES_19f30_SET is considered for the use as artificial skin on measurement devices without any water supply due to its high wettability and potential to absorb and store high amounts of water rapidly in combination with its low drying speed. The fabric sample PES_28f48_GL is recommended for artificial skin on measurement devices with a water supply due to its high wettability and the ability to absorb water rapidly in combination with low water content and related fast drying speed.

Those two ideal fabric samples provide possible alternatives to previously used artificial skin for measurement devices such as guarded hotplates and thermal manikins to determine thermodynamic properties of clothing textiles.

Acronym	Description	Unit
ρ_f	fiber density	g cm⁻³
ρ_y	yarn density	g cm⁻³
m_f	fiber fineness	dtex
V_y	yarn volume	cm³ 10 km⁻¹
V_fy	fiber volume in yarn	cm³ 10 km⁻¹
V_py	pore volume in yarn	cm³ 10 km⁻¹
V_pyr	relative pore volume in yarn	%
l_yt	yarn length	km m⁻²
ρ_t	fabric density	g cm⁻³
V_t	knitted fabric volume	cm³ m⁻²
V_yt	yarn volume in knitted fabric	cm³ m⁻²
V_ft	fiber volume in knitted fabric	cm³ m⁻²
V_pyt	yarn pore volume in knitted fabric	cm³ m⁻²
V_pt	total pore volume in knitted fabric	cm³ m⁻²
V_ptt	stitch pore volume in knitted fabric	cm³ m⁻²
V_tt	total fabric volume	cm³ m⁻²
d_ptt	stitch pore diameter	µm
m_y	yarn fineness	tex
n_y	number of yarns	–
m_yn	yarn end-fineness	tex
m_t	Weight of knitted fabric	g m⁻²
h_t	fabric thickness	mm
n_ptt	number of stitch pores per square centimeter	cm⁻²
Λ	thermal conductivity	W m⁻¹ K⁻¹
R_ct	thermal resistance	m² K W⁻¹
R_et	evaporative resistance	m² Pa W⁻¹
Θ	contact angle	°
T	wetting time	s
R	wetting radius	mm
SWC	surfacial water content	g m⁻²
WC	water content	g g⁻¹
DS	drying speed	g h⁻¹
SDS	surfacial drying speed	g m⁻² h⁻¹

Code	Polymers	Fiber type	Yarn fineness m_y (tex)
PA6_17	polyamide 6	x		17
PA66_23f26/2	polyamide 6.6		x	23
PA66_11f34_HE*	polyamide 6.6, HE		x	11
PA66_11f34_SET*	polyamide 6.6, SET		x	11
PA66_18f104_GL*	polyamide 6.6, GL		x	18
PA66_24f48_GL*	polyamide 6.6, GL		x	24
PES_20	Polyester	x		20
PES_19f30	Polyester		x	19
PES_19f30_SET	polyester, SET*		x	19
PES_24	Polyester	x		24
PES_28f48_GL	polyester, GL*		x	28
PP_20	polypropylene	x		20
CV_24	rayon	x		24
CO_20	cotton	x		20

Variables	Unit	Variable types	Symbol in model	Regression coefficients
Thickness h_t	mm	Independent	x₁	a₁
Stitch pore diameter d_ptt	µm	Independent	x₂	a₂
Number of stitch pore n_ptt	–	Independent	x₃	a₃
Thermal resistance R_ct	m² K W⁻¹	Dependent	y₁	–
Evaporative resistance R_et	m² Pa W⁻¹	Dependent	y₂	–
Drying speed DS	g h⁻¹	Dependent	y₃	–
Constant	–	–	–	b_y1, b_y2, b_y3

Code	Textured	m_yn (tex)	m_t (g m⁻²)	h_t (mm)	ρ_t (g cm³)	n_ptt (cm⁻²)	d_ptt (µm)
PA6_17_2*		34	304	1.25	0.24	226	584
PA66_28f26/2		28	254	1.30	0.20	283	472
PA66_11f34_HE	x	11	162	1.05	0.16	362	393
PA66_11f34_SET	x	11	140	1.02	0.14	351	390
PA66_11f34_SET_2*	x	22	237	1.13	0.21	294	225
PA66_18f104_GL		18	171	0.91	0.19	254	554
PA66_18f104_GL_2*		35	394	1.14	0.35	318	334
PA66_24f48_GL		24	275	1.17	0.23	326	402
PES_20_2*		39	352	1.14	0.31	213	560
PES_19f30		19	172	1.14	0.15	244	554
PES_19f30_SET	x	18	154	1.07	0.14	202	632
PES_19f30_SET_2*	x	37	306	1.21	0.25	207	470
PES_24		24	200	1.32	0.15	190	710
PES_24_2*		49	416	1.50	0.28	215	567
PES_28f48_GL		28	244	1.07	0.23	236	513
PP_20		20	164	1.16	0.14	193	672
CV_24_2*		48	401	1.36	0.29	199	555
CO_20		19	178	1.04	0.17	213	644

Fabric sample	λ (W m⁻¹K⁻¹)	R_ct (m² K W⁻¹)	R_et (m² Pa W⁻¹)	θ (°)	t (s)	r (mm)	SWC (g m⁻²)	WC (g g⁻¹)	DS (g h⁻¹)	SDS (g m⁻² h⁻¹)
PA6_17_2	0.059 ± 0.001	0.043 ± 0,000	5.92 ± 0.14	150 ± 26	7.1 ± 1.3	14.0 ± 2.2	971 ± 29	3.26 ± 0.10	14.4 ± 2.7	1593 ± 266
PA66_23f26/2	0.052 ± 0.000	0.027 ± 0.002	4.30 ± 0.27	129 ± 6	15.3 ± 3.3	8.0 ± 2.7	941 ± 28	3.72 ± 0.11	14.1 ± 0.9	1413 ± 85
PA66_11f34_HE	0.047 ± 0.000	0.026 ± 0.000	2.69 ± 0.08	0 ± 0	3.7 ± 0.3	12.0 ± 2.7	614 ± 18	4.60 ± 0.14	9.1 ± 0.6	917 ± 60
PA66_11f34_SET	0.046 ± 0.000	0.027 ± 0.001	2.91 ± 0.03	144 ± 6	4.7 ± 2.0	9.0 ± 4.2	602 ± 18	3.95 ± 0.12	9.1 ± 0.7	913 ± 72
PA66_11f34_SET_2	0.050 ± 0.001	0.019 ± 0.000	3.20 ± 0.20	180 ± 0	5.8 ± 0.9	15.0 ± 5.0	631 ± 19	3.13 ± 0.09	9.2 ± 2.9	977 ± 289
PA66_18f104_GL	0.041 ± 0.003	0.019 ± 0.001	2.59 ± 0.03	138 ± 22	4.7 ± 1.0	9.0 ± 4.2	354 ± 11	2.11 ± 0.06	10.1 ± 1.8	923 ± 175
PA66_18f104_GL_2	0.062 ± 0.001	0.019 ± 0.001	4.35 ± 0.11	180 ± 0	6.7 ± 1.0	12.0 ± 2.7	740 ± 22	1.92 ± 0.06	11.9 ± 2.4	1170 ± 244
PA66_24f48_GL	0.052 ± 0.000	0.016 ± 0.001	3.75 ± 0.02	0 ± 0	5.4 ± 0.4	16.2 ± 2.2	424 ± 13	1.59 ± 0.05	9.5 ± 0.4	943 ± 38
PES_20_2	0.057 ± 0.000	0.035 ± 0.000	4.45 ± 0.10	0 ± 0	3.3 ± 0.1	21.0 ± 2.2	832 ± 25	2.30 ± 0.07	15.5 ± 0.9	1547 ± 95
PES_19f30	0.042 ± 0.002	0.024 ± 0.002	3.03 ± 0.16	0 ± 0	2.9 ± 0.1	24.0 ± 2.2	575 ± 17	3.15 ± 0.09	9.6 ± 0.4	960 ± 40
PES_19f30_SET	0.045 ± 0.002	0.027 ± 0.002	2.70 ± 0.06	0 ± 0	3.4 ± 0.2	19.0 ± 2.2	686 ± 21	4.23 ± 0.13	11.6 ± 0.3	1150 ± 30
PES_19f30_SET_2	0.051 ± 0.000	0.019 ± 0.001	3.46 ± 0.02	0 ± 0	3.9 ± 0.2	20.0 ± 0.0	775 ± 23	2.38 ± 0.07	13.9 ± 0.2	1383 ± 23
PES_24	0.046 ± 0.001	0.046 ± 0.000	4.26 ± 0.05	0 ± 0	2.9 ± 0.4	27.0 ± 2.7	777 ± 23	3.38 ± 0.12	12.3 ± 0.6	1233 ± 57
PES_24_2	0.054 ± 0.000	0.035 ± 0.001	5.07 ± 0.07	0 ± 0	3.7 ± 0.1	20.0 ± 0.0	874 ± 26	2.10 ± 0.06	17.2 ± 0.3	1713 ± 32
PES_28f48_GL	0.046 ± 0.003	0.018 ± 0.000	2.97 ± 0.19	0 ± 0	3.0 ± 0.2	27.0 ± 2.7	364 ± 11	1.46 ± 0.04	8.9 ± 1.2	957 ± 124
PP_20	0.041 ± 0.001	0.036 ± 0.000	4.03 ± 0.10	138 ± 31	4.7 ± 1.7	13.0 ± 2.7	589 ± 18	2.37 ± 0.10	14.2 ± 2.0	1193 ± 197
CV_24_2	0.057 ± 0.000	0.027 ± 0.001	4.20 ± 0.02	0 ± 0	4.5 ± 0.2	15.0 ± 0.0	893 ± 27	2.50 ± 0.07	16.7 ± 0.4	1670 ± 40
CO_20	0.053 ± 0.001	0.038 ± 0.002	3.97 ± 0.17	180 ± 2	7.4 ± 2.5	11.0 ± 4.2	672 ± 20	4.77 ± 0.14	10.2 ± 1.3	1073 ± 127

Footnotes

Acknowledgement

The authors wish to thank scientists from BioMemTex at Empa, Dr Emel Mert and Barbara Koelblen for the introduction to and quality assurance of the fabric tests, Dr Agnieszka Dabrowska for consultation on the properties of hydrated human skin. We further acknowledge the support rendered by the scientist from Laboratory Knitting Technology at DITF, Uwe Röder, for the technical development of the knitting fabrics.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by Fundamental Research Funds of the DITF German Institutes of Textile and Fiber Research and Empa Swiss Federal Laboratories for Materials Science and Technology.

References

Havenith

Richards

Wang

et al.

Apparent latent heat of evaporation from clothing: attenuation and “heat pipe” effects. J Appl Physiol 2008; 18: 142–149.

Hes

. Instruction manuals of the Alambeta, Permetest Instruments Czech Republic: SENSORA, 1990.

ISO 9920 . Ergonomics of the thermal environment – Estimation of thermal insulation and water vapour resistance of a clothing ensemble, Geneva: International Organization for Standardization, 2007.

ISO 11092 . Textiles – Physiological effects – Measurement of thermal and water-vapour resistance under steady-state conditions (sweating guarded-hotplate test), Geneva: International Organization for Standardization, 2014.

Mc Cullough

Huang

Kim

. An explanation and comparison of sweating hot plate standards. J ASTM Int 2004; 1: 121–133.

Scoarnec

Hameury

Hay

. A new guarded hot plate designed for thermal-conductivity measurements at high temperature. Int J Thermophys 2015; 36: 540–556.

Psikuta

Wang

Rossi

. Prediction of the physiological response of humans wearing protective clothing using a thermophysiological human simulator. J Occup Environ Hyg 2013; 10: 222–232.

Zimmerli

Weder

Protection and comfort – a sweating torso for the simultaneous measurement of protective and comfort properties of PPE. In: Stull

Schwope

(eds). Performance of protective clothing, 6th ed. Philadelphia, PA: American Society for Testing and Materials, 1997; pp. 271–278.

ASTM F1291 . Standard test method for measuring the thermal insulation of clothing using a heated manikin, West Conshohocken, PA: ASTM International, 2015.

10.

ASTM F2370 . Standard test method for measuring the evaporative resistance of clothing using a sweating manikin, West Conshohocken, PA: ASTM International, 2015.

11.

Holmér

. Thermal manikin history and applications. Eur J Appl Physiol 2004; 92: 614–618.

12.

ASTM D1518 . Standard test method for thermal resistance of batting systems using a hot plate, West Conshohocken, PA: ASTM International, 2014.

13.

Havenith G. Thermal comfort for building occupants, https://www.cibse.org/getmedia/f501c997-46b3-4110-b02e-1d0e8b113d6b/Thermal-Comfort-Experiments-George-Havenith.pdf.aspx (2008, accessed March 2017).

14.

Wang LW and Dai XQ. Effect of skin designed for sweating thermal manikin NEWTON on measuring the evaporative resistance of clothing. 2014.

15.

Huang

Zhang

Qian

. A simple test method for measuring water vapor resistance of porous polymeric materials. Polym Test 2013; 32: 1037–1040.

16.

Zhang

Wang

et al.

A new method for evaluating heat and water vapor transfer properties of porous polymeric materials. Polym Test 2010; 29: 553–557.

17.

Bräuer

English

MJM

Sander

et al.

Construction and evaluation of a manikin for perioperative heat exchange. Acta Anaesthesiol Scand 2002; 46: 43–50.

18.

Psikuta

Kuklane

Bogdan

et al.

Opportunities and constraints of presently used thermal manikins for thermo-physiological simulation of the human body. Int J Biometeorol 2015; 60: 1–12.

19.

Varheenmaa

Meinander

. A new generation sweating thermal cylinder for the evaluation of the thermoregulation properties of protective clothing materials. In: Konola S and Kähkönen P (eds) Active Wears – Perspectives on Arctic Clothing , 21–24 August 2015; pp. 154–161.

20.

Tamura

. Development of a two-layer movable sweating thermal manikin. Ind Health 2006; 44: 441–444.

21.

Psikuta

Richards

Fiala

. Single-sector thermophysiological human simulator. Physiological Measurement 2008; 29: 181–192.

22.

Wang

Gao

Kuklane

et al.

A study on evaporative resistance of two new skins designed for thermal manikin ‘Tore’ under different environmental conditions. J Fiber Bioeng Informatics 2009; 1: 301–305.

23.

DIN16422 . Classification of thermoregulatory properties Berlin: German Institute of Standardization, 2012.

24.

Wang

Dai

. A study on material skin for measuring clothing evaporative resistance. J Fiber Bioeng Informatics 2015; 8, pp. 221–228.

25.

Koelblen

Psikuta

Bogdan

et al.

Comparison of fabric skins for the simulation of sweating on thermal manikins. Int J Biometeorol. 2017; 61: 1519–1529.

26.

Morent

DeGeyter

Vansteenkiste

CLE

et al.

Measuring the wicking behavior of textiles by the combination of a horizontal wicking experiment and image processing. Rev Sci Instrum 2007; 77.

27.

Ramachandran

Kesavaraja

. A study on influencing factors for wetting and wicking behaviour. J Institut Engineers (India) 2004; 84: 37–41.

28.

Yanilmaz

Kalaoğlu

. Investigation of wicking, wetting and drying properties of acrylic knitted fabrics. Text Res J 2012; 82: 820–831.

29.

Yoon

Buckley

. Improved comfort polyester Part I Transport properties and thermal comfort of polyester cotton blend fabrics. Text Res J 1984; 54: 289–298.

30.

Oğlakcioğlu

Marmarali

. Thermal comfort properties of some knitted structures. Fibers Text East Eur 2007; 15: 64–65.

31.

Wong

Tao

Yuen

CWM

et al.

Wicking properties of linen treated with low temperature plasma. Text Res J 2001; 71: 49–56.

32.

SalopekČubrić

Skenderi

Mihelic-Bogdanic

et al.

Experimental study of thermal resistance of knitted fabrics. Exp Therm Fluid Sci 2012; 38: 223–228.

33.

Crow

Osczevski

. The interaction of water with fabrics. Text Res J 1998; 68: 280–288.

34.

Dabrowska

Spano

Adlhart

et al.

The relationship between skin function, barrier properties and body-dependent factors. Skin Res Technol. 2018; 24: 165–174.

35.

Lippert

. Lehrbuch Anatomie, 8th ed. München: Urban & Fischer GmbH, 2011.

36.

Schaal

Kunsch

. Der Mensch in Zahlen: Eine Datensammlung in Tabellen mit über 20000 Einzelwerten, 4th ed. Berlin: Springer Verlag, 2016.

37.

Stüttgen

Schaefer

. Funktionelle Dermatologie, Berlin: Springer-Verlag, 1974.

38.

Fercher

. Medizinische Physik, 1st ed. Wien: Springer-Verlag, 1992.

39.

Fanger PO. Thermal comfort – Analysis and applications in environmental engineering. Copenhagen: Danish Technical Press, 1970.

40.

Taylor

NAS

Machado-Moreira

. Regional variations in transepidermal water loss, eccrine sweat gland density, sweat secretion rates and electrolyte composition in resting and excercising humans. Extrem Physiol Med 2013; 2: 4.

41.

Rehrer

Burke

. Sweat losses during various sports. Aust J Nutr Diet 1996; 53: 13–16.

42.

Bean

Eichna

Ashe

et al.

Performance in relation to environmental temperature. Bull Johns Hopkins Hosp 1945; 76: 25–58.

43.

Mark

Bauer

Gresser

. Quantification of hierarchic multimodal pore structures by the example of knitted fabric structures. In: International Textile Conference, Aachen, 27–28 November 2015.

44.

EN 6330. Domestic washing and drying producers for textile testing. Brussels: European Comittee for Standardization, 2001.

45.

EN 12127. Textile fabrics – Determination of mass per unit are using small samples. Brussels: European Comittee for Standardization, 1997.

46.

ISO 5084 . Textile – Determination of thickness of textiles and textile products, Geneva: International Organization for Standardization, 1996.

47.

EN2060-04. Textiles – Yarn from packages – Determination of linear density (mass per unit length) by the skein method. Brussels: European Comittee for Standardization, 1995.

48.

DIN 52612. Determination of thermal conductivity by means of the guarded hot plate apparatus. Berlin: German Institute for Standardization, 1984.

49.

Gocławski

Urbaniak-Domagała

. The measurement of wetting angle by applying an ADSA model of sessile drop on selected textile surfaces. Fiber Text East Eur 2008; 16: 84–88.

50.

AATCC. Liquid Moisture Management Properties of Textile Fabrics. North Carolina: American Association of Textile Chemists and Colorists, 2011.

51.

Field A. Discovering Statistics Using SPSS, 4th Edition. London: SAGE Publishing, 2013.

52.

Cengel

Ghajar

. Heat and mass transfer, 5th ed. New York: Mc graw Hill Education, 2015.

53.

ISO 6946 . Building components and building elements – Thermal resistance and thermal transmittance – Calculation method, Geneva: International Organization for Standardization, 2007.

54.

Berglund LG. A theoretical and experimental study of the thermoregulatory effects of salt accumulation on human skin during seating. PhD Thesis, Kansas State University, Manhattan, Kansas, 1971.

55.

Postle R and Munden DL. 24-Analysis of the dry-relaxed knitted-loop comfiguration: Part I: Two-dimensional analysis. The Journal of The Textile Institute 1967; 58: 329–351.

56.

Postle R and Munden DL. 24-Analysis of the dry-relaxed knitted-loop configurations: Part II: Three-dimensional analysis. The Journal of The Textile Institute 1967; 58: 352–365.

Code	Polymers	Fiber type		Yarn fineness m_y (tex)
Code	Polymers	staple	filament	Yarn fineness m_y (tex)
PA6_17	polyamide 6	x		17
PA66_23f26/2	polyamide 6.6		x	23
PA66_11f34_HE*	polyamide 6.6, HE		x	11
PA66_11f34_SET*	polyamide 6.6, SET		x	11
PA66_18f104_GL*	polyamide 6.6, GL		x	18
PA66_24f48_GL*	polyamide 6.6, GL		x	24
PES_20	Polyester	x		20
PES_19f30	Polyester		x	19
PES_19f30_SET	polyester, SET*		x	19
PES_24	Polyester	x		24
PES_28f48_GL	polyester, GL*		x	28
PP_20	polypropylene	x		20
CV_24	rayon	x		24
CO_20	cotton	x		20