Three-dimensional stretchable knitted design with transformative properties

Abstract

This paper proposes the use of digital knitting to fabricate three-dimensional stretchable fabric with transformative properties. The research focus is on the application of the curling effect and the resultant stitch structures. Rib, purl, and links structures, which have alternating face and back loops in the structural knit cell and produced the curling effect, are investigated. Thirty-six samples based on the three structures along with 12 different stitch combinations are produced through the digital knitting process. The properties of the samples, including the dimensional changes, surface texture, weight and thickness, and tensile properties, are subsequently evaluated and compared by using both quantitative and qualitative assessment methods. The results show that the developed knitted fabrics that use the curling effect through alternating face and back loops are significantly three-dimensional in surface texture and have considerable stretchability. These effects in general increase with the growth of stitch number in a structural knit cell. However, only the links structures show considerable extensibility along both course and wale directions, while the rib and purl structures have enhanced stretchability in only the course and wale directions, respectively. Therefore, this design-led textile study provides a simple but effective means for producing new materials that offer both function and aesthetics for fashion with transformable designs. The work here also provides a technical-based integrated approach for innovative textile and fashion developments.

Keywords

stretchable textile three-dimensional digital knitting curling effect transformative properties transformable designs

Transformable design in fashion features the sustainability and adaptability of garments for accommodating different fashion demands and lifestyles. The term “transformative properties,” which derives from the design concept, refers to the ability of clothing to alter, adapt, or evolve its shape or state. Rahman and Gong¹ stated the transformable garment is a practical solution that allows a piece of garment to be transformed into different looks to serve diverse individual needs and proposes. Koo and Ma² summarized four basic categories of the transformable garment, including reversible and folded designs, modular designs, and smart clothing with transformable color or patterns, as well as do-it-yourself and multi-life designs. On account of the diversity of design approaches for transformable garments, the specific categories of transformative properties of resultant products are also different. Currently, most design research and practices in transformative properties are based on using woven textiles and, as a result, the diversity of the produced knitted garments has been limited.^1–4 However, the properties of knitted fabric offer many advantages, such as extensibility and dimensional instability, so that knitting might be conducive to creating a new transformable design for fashion, particularly when digital knitting is used as the fabrication method. In this study, the transformative properties that are generated from the structural construction of knitting are focused upon. With this transformability, a fashion design collection has been developed using the proposed knitted fabrics, which are expected to offer considerate adaptability for different individuals.

One of the primary properties of weft knitting is the curling effect, which can influence the structure and performance of weft-knitted fabrics in different ways and, as such, has attracted scholarly interest. However, previous research works have mainly used two methods to examine curling. One is theoretical in nature, with the intention to examine or predict the effect of curling through mathematical or geometric analyses.^5–9 The second is application-oriented, with the aim to investigate and utilize the unique characteristics of the resultant knitted fabrics with curling, particularly in terms of the auxetic effects and foldable properties.^10–15 Almost all of these research studies, from an engineering perspective, have focused on functional performance. In fact, the curling effect not only imparts specific properties onto the fabric, but also produces distinctive surface characteristics and aesthetic attributes for garment design. However, design-led research on the curling effect and its resulting structures, especially the surface texture and variations, is still lacking in the literature.

This study, from a design viewpoint, aims to investigate the variations of stretchable three-dimensional (3D) knitted fabrics that are created by the curling effect from three aspects: dimensional properties, surface texture and other relative properties (the characteristics of the surface texture, fabric weight, and thickness), and extension abilities. The study also examines the relationships between the changes and the number of stitches of a structural knitted cell (SKC; repeated units in a knitted structure). The results of this study demonstrate that a series of unique knitted fabrics with a distinctive 3D texture as well as greater stretchability can be produced by employing the curling effect. With different stitch pattern designs (different stitch combinations), the resultant fabrics exhibit different properties, which in turn provide a new creative mean for fabricating transformable fashion knitwear.

Principle of the curling effect

The edge of weft-knitted plain fabric (single jersey) can show significant curling, which is caused by the loss of elastic deformation of the bending yarn that is found in the 3D configuration of the knit. Figure 1(a) shows the cross-section of the fabric along the course direction, in which Loops 1-1 and 2-2 are the needle loops, while Loops 3-3 and 4-4 are the sinker loops. Attempts were made to straighten the bent Loop 3-3, resulting in the deviation of the angle of Loop 1-1, which is in the tension-free state. Similarly, attempts were made to straighten Loops 1-1, 2-2, and 4-4, which resulting in curling of the edge loop toward the back side of the knitted fabric, as illustrated by the dotted line in Figure 1(a). Figure 1(b) shows the cross-section along the wale direction, in which there is curling of the edge loop toward the face of the knitted fabric. Since the direction of curling in the course-wise and wale-wise directions is different, the curling forces are balanced at the four corners of the knitted fabric. As shown in Figure 1(c), the fabric generally curls toward the face at the top and bottom of the fabric and toward the back at the sides.¹⁶ The curling deformation of weft-knitted fabric generally increases with an increase in the bending rigidity of the yarn and a reduction in loop length.^5,6 The main factors that determine the bending rigidity of the yarn include the structure, material, and count of the yarn itself.¹⁷

Figure 1.

Principle of the curling effect.

The effect of curling is also important in the formation of different knitted stitch structures. The rib structure, which is a basic stitch structure in weft knitting, is composed of loops formed in opposite directions. The simplest rib structure is the one-by-one rib, which is formed by alternating the face and back loops along the course direction. When rib fabric is in its natural state without extra tension from the course-wise direction, the alternate wales come in contact with one another, which is caused by the natural curling effect taking place in the intermediate section of the loop that connects the face and back loops within the structure.¹⁸ As such, this structure provides a higher degree of extension allowance along the course direction as well as imparts a 3D fold on the surface. Similar characteristics can also be found with a purl structure, in which there is an exchange of face and back loops of the SKC along the wale direction.

Consequently, it is anticipated that alternating the face and back loops during knitting construction could impart a greater degree of extension as well as a 3D surface texture. Many studies in the literature and practices of designers have reported that, based on this principle, various surface effects can be achieved in association with different stitch pattern designs.^11,13,19,20 In these cases, the utilization of the curling effect can produce an alternative 3D stretchable textile to realize transformative properties in design.

Experimental details

Design parameters and sample design

Thirty-six weft-knitted samples were produced based on three basic knitted structures (rib, purl, links) and 12 stitch combinations in SKCs (1 × 1, 2 × 2, 3 × 3, 6 × 6, 9 × 9, 12 × 12, 15 × 15, 18 × 18, 21 × 21, 24 × 24, 27 × 27, and 30 × 30) were used. All of the samples were then placed into one of three groups based on their structure (Group R, P, or L based on rib, purl, or links, respectively). Table 1 provides an overview of the sample designs in this study.

Table 1.

Overview of sample designs

Sample		Group R	Group P	Group L
Stitch structure	Knitted structure	Rib	Purl	Links
	Structure example and notation	1 × 1 Rib	1 × 1 Purl	1 × 1 Links


	Alternating face and back loops	Only in course direction	Only in wale direction	In both course and wale directions
Stitch number combinations	1 × 1, 2 × 2, 3 × 3, 6 × 6, 9 × 9, 12 × 12, 15 × 15, 18 × 18, 21 × 21, 24 × 24, 27 × 27, and 30 × 30

F: front needle bed; B: back needle bed; 1: the first row of knitting; 2: the second row of knitting.

The three selected basic stitch structures are all composed of alternating face and back stitches, but in different specific arrangements. Hence, it was possible to observe and compare the variations in the knits caused by the curling effect. Furthermore, different stitch combinations were selected to trace the variations in the surface with increases in the number of stitches of the SKC. Since there are many possibilities of stitch number combinations, 12 stitch combinations are selected for use in this study as examples, as shown in Figure 2. The 36 SKCs created in this study are shown in Figure 3.

Figure 2.

Possible stitch number combinations and the selections. Note: shaded boxes denote the combinations used in this study.

Figure 3.

Structural knit cell of each knit sample categorized into three groups.

Materials and knitting specifications

All of the samples were knitted with four ends of 100% polyester filament yarn on a SVR123SP (14 gauge) computerized flat knitting machine (Shima Seiki, Japan). Each end was 100D/4 F (100 denier, with four filaments). Sample size was fixed as 160 stitches in the wale-wise direction and 200 stitches in the course-wise direction. Each SKC was filled with a certain number of repeats. A knitting template was developed on the SDS-ONE APEX 3 design system (Shima Seiki, Japan), which can combine waste knitting and binding-off for each sample, as shown in Figure 4. The aim is to minimize external influences that deform the fabric, such as the take-down force, as well as enhance the accuracy of the data obtained from each sample in the subsequent experiments. The loop tension and mechanical parameters were set based on common industry practices and remained constant. Three of the same samples were prepared for each of the 36 designed structures, which provided 108 samples in total.

Figure 4.

Knitting template of the samples.

Characterization and evaluation

Both quantitative and qualitative tests were used to evaluate the properties of the samples. Prior to measurement, all of the knitted articles were separately placed onto a flat and smooth surface for relaxation and drying in air in laboratory standard conditions (temperature of 20℃ and relative humidity (RH) of 60%) for more than 72 hours to release knitted stresses, and were gently moved once a day to prevent possible friction between the fabric and surface of the table.

Dimensional properties

The dimensional characteristics were evaluated based on two factors: linear (fabric length and width) and areal changes. The length of the fabric is the wale direction, and the width of the fabric is the course direction. Measurements were taken in three different areas – in the middle as well as between the middle and the both borders of each fabric sample by using a ruler.

The irregular shape of the fabric after relaxation required a special method to calculate the areal dimensions of the knitted samples so that their dimensional characteristics can be assessed. The developed method is as follows. Firstly, a self-standing scanner was used to scan each sample in the same direction. The same ruler used for measurement was scanned along with each sample to act as a dimensional reference for calculation purposes. Secondly, a computational program developed via MATLAB was used to capture the area of the fabric sample as a polygon area by outlining the perimeters of each sample on the scanned images. Each sample of the same design was measured thrice on three separate specimens and the obtained values were averaged.

Surface texture and relative properties

The evaluation of a physical property, the surface texture of the samples, consisted of two tasks that used a mixed approach of both quantitative and qualitative methods. Firstly, the different textures were visually recorded and qualitatively compared by examining the scanned images. Then, the measured fabric thickness and weight were assessed to identify their correlation with the surface texture. The thickness (T) values were measured by using a self-standing caliper to measure the fabric in the free state. For each sample design, each measurement was made on three separate specimens and the resulting values were averaged. Two indexes, which are the weight of the fabric or gravity (G) and density of the fabric (D), were used as the weight values. Here, G is the total weight of the fabric sample and D is the surface weight of the fabric. For each stitch design, three separate samples were weighed on an electronic balance together thrice and then the different weights were averaged to obtain the G value. The D value was then calculated by dividing the G value of each fabric sample by its total area.

Extension ability

The extensibility of the knitted samples was measured by using standard test method ASTM D2594-04 with a universal testing machine (Instron). The test was conducted in the constant-rate-of-load (CRL)²¹ mode, in which the maximum load was 15 N with a constant speed of 30 mm/min. Generally, loading at a specific extension is more appropriate to determine the wear properties.²² Each specimen was extended by force and released for recovery for three rounds. Two principal directions, parallel to the course and wale directions of each sample, were respectively measured. The obtained load–extension curves and the values of the tensile strain at the maximum load (TM) were investigated. A statistical analysis was then used to gain additional insight, which moreover provides a reference that could be used to support future design applications.

Results and discussion

Dimensional properties

In general, the shrinkage found in all three groups of fabric samples continuously increases with number of stitches in the SKC. The scanned images of each sample in Groups R, P, and L are provided in Figures 5–7. Dimensional shrinkage is mostly observed to be along the course-wise direction for the Group R samples. These samples evolve into a long ribbon shape. The areal shrinkage of the Group P samples is mainly found along the wale direction. These samples generally evolve into a reverted U-shaped rectangle. Finally, the shrinkage of the Group L samples is found in both the course-wise and wale-wise directions. These samples generally evolve into a rectangle with irregular borders.

Figure 5.

Images of shrinkage of Group R samples.

Figure 6.

Images of shrinkage of Group P samples.

Figure 7.

Images of shrinkage of Group L samples.

All three groups of samples first showed a steep reduction in fabric area, and then this reduction became constant, as shown in Figure 8. Among the samples with fewer stitches (less than nine), the fabric area of the Group L samples is always the largest, followed by the samples in Group R and then Group P, which has the smallest fabric area. However, with increases in the number of stitches (more than nine), the fabric area of all three groups of samples tends to be relatively the same and constant.

Figure 8.

Fabric area of samples for Groups R, P, and L.

Generally speaking, with increases in the number of stitches in the SKC, the fabric length (L) and fabric width (W) of each group of samples were continuously reduced, except for R1 × 1, L1 × 1, and L2 × 2; see Figures 9 and 10, which show the linear dimensional changes of the three groups of samples. A careful examination shows that the differences between the L and W values of Groups R and P are significant, particularly the samples in Group P. However, the difference is minimal for the samples in Group L, expect for L1 × 1 and L2 × 2. It can also be observed that the L of the Group R samples and W of the Group P samples show the same trend of change, except for the 1 × 1 sample, and have relatively high values consistently compared to the other samples. Similarly, the W of the Group R samples and the L of the Group P samples also show a similar trend of change, in which they both have relatively low values after a reduction in size from 1 × 1 to 6 × 6. In contrast, the changes in both the L and W of the samples in Group L remain moderate most of the time.

Figure 10.

Fabric width of samples for Groups R, P, and L.

Figure 9.

Fabric length of samples for Groups R, P, and L.

The results reveal that the dimensional characteristics of the samples are mainly affected by the curling effect and the consequent shrinkage. This shows that when there are only alternating face and back loops in one direction (Groups R and P), the curling effect is only produced along this direction, thereby causing the shrinkage of the fabric in this direction. Furthermore, when there are alternating courses and wales (Group L samples), the curling effect is consequently produced from both directions, which in turn results in the bidirectional shrinkage of the fabric. However, the degree of shrinkage from both directions is always not as high as the shrinkage from one direction.

The results also indicate that as the number of stitches in the SKC increases, the effect of curling on fabric shrinkage becomes continuously enhanced, which is because the increase in use of one type of stitching (front or reverse) in a certain SKC increases the impact of the curling effect for single jersey. However, when the alternating front and reverse stitches exceed a certain number of stitches, such as 20 stitches, the influence of curling reaches a certain limitation, and thereby the dimensional differences between the samples become less significant.

In addition, it is known that generally R1 × 1, or one-by-one rib, is considered as double jersey, which needs to be knitted with a tighter loop tension compared to the tension used for common single jersey in this study. The drastic increase in loop spacings and the changes in the characteristics of the filament yarn used both cause loosening of the fabric, especially in the course-wise direction. It is also evident that the L values are higher than the W values for L1 × 1 and L2 × 2, which is obviously different from the other samples in Group L. It could be that when the number of stitches in the SKC is fairly low (1 or 2) in the Group L samples, the influence of the curling effect on the dimensional changes is fairly limited. In addition, a very balanced fabric structure is produced on account of the alternating front and reverse stitches in both directions. Therefore, the L1 × 1 and L2 × 2 samples ultimately show similar characteristics to ordinary single jersey.

Besides, it is worth noting that the specific shape of the fabric samples is also unavoidably affected by the tensile force during machinery knitting. For instance, the two ends of the purl-based samples show an inverted U-shape. Since the initial experimental design tried to minimize the impacts (by using a knitting template), the overall impact is not very significant.

Surface texture and relative properties

It was found for all three groups of samples that the 3D surface texture is increasingly correlated to an increase in the number of stitches in the SKC; see Figures 11–13 for the surface texture of the samples in Groups R, P, and L, respectively. Overall, the surface texture of Groups R and P shows uniform horizontal and vertical stripes, respectively. In contrast, the Group L samples are much less uniform, as they show a more 3D, irregular, and distorted textured effect. It can also be observed that when there are a fairly large number of stitches in the SKC, the surface texture of each sample increases in similarity and it becomes difficult to differentiate between the different samples. All three groups of samples show this result. The reason is similar to the practically indiscernible difference in dimensional shrinkage, as discussed above.

Figure 11.

Surface texture of Group R samples.

Figure 12.

Surface texture of Group P samples.

Figure 13.

Surface texture of Group L samples.

The G values of the samples in the three groups are plotted in Figure 14. It can be observed that the slopes are close together and the plotted G values of Groups R and P initially show a steep decline and then become constant followed with a slight decrease. In contrast, the G values of the Group P samples show consistency and are relatively low.

Figure 14.

Fabric gravity of samples for Groups R, P, and L.

The main factor that affects the G values is the frequency of alternating of the front and reverse stitching along the course direction found in certain samples. This is the case because this kind of alternating requires extra yarn length to bridge the needle bed gap between the front and back needle beds. The increased frequency in exchange of stitches in correspondence to the increased number of connections between the needle beds can result in more yarn used in a sample and vice versa. In this case, the increase in the number of stitches in a SKC reduces the number of connected stitches along the course direction and, therefore, the G value is lower. Therefore, when there are an equal number of stitches in a SKC, the frequency of the alternating of the face and back loops of the rib structure samples (Group R) and links structure samples (Group L) is also the same in the course-wise direction. There is no such exchange of stitches in any of the purl samples (Group P), and thus the G values of this group remain almost unchanged.

The D values of all three groups of samples show a similar trend of increase as the number of stitches in the SKC is increased, as shown in Figures 15. Meanwhile, the D value of each sample with the same number of stitches in all three groups is also relatively close, particularly those in Groups R and P. On the other hand as shown in Figure 16, it can be observed that when there is a fairly small number of stitches in the SKC (less than six stitches), the difference of the T values of the three groups is insignificant, and the T values of both Groups P and R subsequently show a gradual increase (more than six stitches). In comparison, a rapid increase of the T values can be observed in the Group L samples from 6 × 6 and the values are significantly higher than those of Groups R and P.

Figure 15.

Fabric density of samples for Groups R, P, and L.

Figure 16.

Fabric thickness of samples for Groups R, P, and L.

The results show that, compared to the samples in Groups R and P, the 3D surface effects of the Group L samples become especially significant as T is increased, particularly when the stitch number is up to six. It also indicates that, with an increase in the number of stitches in a SKC, the higher shrinkage of fabrics always contributes to the quantity of yarns in a certain area of the fabric and, hence, the weight of the unit per area (D value) and thickness of the fabric (T value) subsequently increase.

Extension ability

The extension–load curves of all of the samples in both directions for Groups R, P, and L are plotted in Figure 17. It can be observed that each curve shows a similar narrow slope, which is typical of stretchable fabric. It can also be observed that, relatively speaking, the plotted curves of the Group L samples are evenly distributed in the two directions. In contrast, the curves of Groups R and P in the course-wise and wale-wise direction are significantly dispersed and uneven.

Figure 17.

Extension–load curves of samples in Groups R, P, and L.

The values of the tensile strain at the maximum load (TM) of each sample are plotted in Figure 18. It can be observed that the TM of the Group R samples in the course direction (TMc_R) and that of the Group P samples in the wale direction (TMw_P) are significantly higher in comparison to those of the other samples. The value of the TMc_R rapidly increases from 1 × 1 to 3 × 3 to a particularly high value, and then slowly increases, whereas the value of TMw_P shows a gradual and continuous increase from a relatively low to high value. In comparison, the TM values of Groups R and P from the opposite directions (TMw_R and TMc_P) always remain low, particularly TMw_R.

Figure 18.

TMc and TMw of Groups R, P, and L.

In contrast, the TM values of the Group L samples (TMc_L and TMw_L) are fairly similar and remain at a moderate level compared to the other samples. It can be observed that in the first half of the samples (1 × 1 to 9 × 9), the TMc_L values are always higher than TMw_L. Then, the two TM values of L12 × 12 show an approximation. After that, TMw_L starts to increase more than TMc_L in the second half of the samples (15 × 15 to 30 × 30).

Such differences in the extension of the samples, on the one hand, are caused by the basic configuration of the knitted loops. It is well known that the stretchability of weft-knitted fabric along the course direction is generally higher than it is along the wale direction on account of the degree of constraint imposed onto each knitted loop due to the intermeshing. Thus, when there is no alternating of the face and back loops, the TMc_P is consistently higher than the TMw_R and remains at about twice the TMw_R, except for the special case of R1 × 1, as explained earlier.

On the other hand, the higher extensibility resulting from the specifically designed knitted stitch structure is another and the most important influential factor. As the number of stitches of each SKC increases, the shrinkage of the fabric also increases, which corresponds to the increase of the extension ability of the fabrics. The for this reason is the same as that for the dimensional changes. As a result, higher shrinkage and folding result in more potential of the fabric to unfold and stretch.

For the Group R samples, there are alternating face and back loops along the course direction for all the rib structures, together with the higher stretchability of the basic construction units of knitted loops assembled in the course-wise direction. Thus, the highest TM values are recorded in the course direction for this group of samples. For the Group P samples, their stretchability in the wale-wise direction is not significant in the initial stages due to the basic construction unit of the knitted loop. However, the stretchability increases rapidly in correspondence with an extended period of alternating stitches. For the Group L samples, there are alternating stitches in both directions, and thus the extension ability of the samples in both the course-wise and wale-wise directions is considerable. However, when the shrinkage produced by the stitch structure is distributed in both directions, the resultant impact is not as high as the shrinkage is mostly in one direction. The situation is similar to L and W, the TM values of the Group L samples remain moderate compared to the other groups of samples.

Independent sample t-testing was carried out to determine whether there is a statistical difference between each TM value. The results of the two TM values of the Group R samples are shown in Table 2 as an example. It can be seen that, using the unequal variances test, the t-value is 11.681 and has a corresponding p-value of 0.000, which is less than the significant level (0.05). Thus, it can be concluded that TMc and TMw values of Group R samples are significantly different.

Table 2.

Independent t-tests and results of Group R samples

	F	Sig.	t	df	Sig. (2-tailed)	Mean difference	Std. error difference	95% confidence interval of the difference
	F	Sig.	t	df	Sig. (2-tailed)	Mean difference	Std. error difference	Lower	Upper
TM of Group R
Equal variances assumed	11.415	0.003	11.681	22	0.000	522.941	44.768	430.097	615.784
Equal variances not assumed			11.681	11.412	0.000	522.941	44.768	424.839	621.042

Other relevant results of the t-tests of the TM values are summarized in Tables 3 (by direction) and 4 and 5 (by group). The tables show that the TMc and TMw values of the samples in both Groups R and P are significantly different. In contrast, the TMc and TMw values of the samples in Group L are not significantly different (p = 0.863). Meanwhile, the TMc values in all three groups of samples are significantly different from each other. The same is true for TMw.

Table 3.

Summary of t-test results for TMc and TMw of Groups R, P, and L samples

	TMc_R	TMc_P	TMc_L
TMw_R	S (p = 0.000)
TMw_P		S (p = 0.000)
TMw_L			NS (p = 0.863)

Notes: S denotes significant (p ≤ 0.05); NS denotes not significant (p > 0.05).

Table 4.

Summary of t-test results for TM of Groups R, P, and L: course direction

	TMc_R	TMc_P	TMc_L
TMc_R		S (p = 0.000)
TMc_P			S (p = 0.000)
TMc_L	S (p = 0.000)

Notes: S denotes significant (p ≤ 0.05); NS denotes not significant (p > 0.05).

Table 5.

Summary of t-test results for TM of Groups R, P, and L: wale direction

	TMw_R	TMw_P	TMw_L
TMw_R		S (p = 0.000)
TMw_P
TMw_L	S (p = 0.001)		S (p = 0.003)

Notes: S denotes significant (p ≤ 0.05); NS denotes not significant (p > 0.05).

Linear or curvilinear regression equations were used for the analysis of the TM values of each group of samples and the number of stitches used, respectively, and the results are graphically presented in Figures 19–21. There is a high correlation of the curvilinear equation between TMc_R and the number of stitches. Meanwhile, there is only a slight correlation of the curvilinear equation between TMw_R and number of stitches with to the special case of R1 × 1. At the same time, the TM values of the samples in both Groups P and L in both the wale and course directions are observed to have a significant correlation. The results of the regression show that the TM values can be predicted with the used number of stitches to a certain extent.

Figure 19.

Regression results of TM values in course and wale directions and number of stitches used: Group R samples.

Figure 20.

Regression results of TM values in course and wale directions and number of stitches used: Group P samples.

Figure 21.

Regression results of TM values in course and wale directions and number of stitches used: Group L samples.

Design application

A series of 3D stretchable knitted textile designs were developed based on the results of the experimental study. Different combinations of stitch patterns and stitch structures were designed, produced, and investigated. It was found that a variety of distinctive 3D surface effects could be achieved based on the study results. In addition, the borders of the fabrics were also re-shaped due to the different amount of shrinkage with the different stitch structures. The different directions and amount of tensile force can create fabrics with different properties. As shown in Figure 22, (a) and (c) exhibit fabrics in the natural state, while (b) and (c) show their deformations under extra forces, both unidirectional and bidirectional. When subjected to external tension, the fabric tends to unfold while the 3D surface texture partially becomes part of the elongated section of the fabric.

Figure 22.

Textile design of stretchable three-dimensional fabric.

To further explore the design capability of the 3D stretchable fabric, a traditional Chinese wedding-inspired dress was designed to demonstrate the transformative properties of the fabric (Figure 23). Manipulating the stretchability and textured surface of the fabric for different parts of the garment was key in this design creation. That is, a piece of 3D stretchable knitted material that can accommodate the different curvatures of the body was used to accommodate the curvatures through elongation. At the same time, the material was used to embellish other parts of the dress by manipulating the distinctive surface texture (folds). As such, no cutting of the fabric was needed for the entire construction process of the dress. Therefore, different fashion statements can be made to accommodate the different body shapes of wearers, which reflect the transformative properties and adaptability, of the material in fashion applications.

Figure 23.

Wedding dress with three-dimensional stretchable textile: (a) front view; (b) back view.

Conclusion

A design-focused study is presented here, which examines the surface effects of knitted structures produced by the curling effect to develop transformative properties for an innovative new textile. This study shows that stitch structure designs with the curling effect can be used to develop specific types of knitted fabrics that have a 3D surface texture with greater stretchability. The trends in the differences of the fabric samples with the use of different stitch combinations and surface effects are examined through both quantitative and qualitative methods. The solid technical findings in this study contribute as a reference for future designs of 3D stretchable textiles with the use of digital knitting. A mixed approach that is innovatively premised on a technical-based methodology can be used to provide both aesthetics and function and is also a contribution of this study.

Footnotes

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 the receipt of the following financial support for the research, authorship and/or publication of this article: The authors would like to express their appreciation to The Hong Kong Polytechnic University for the financial support of this study from the Research Grants Council in the form of a postgraduate award.

References

Rahman

Gong

. Sustainable practices and transformable fashion design–Chinese professional and consumer perspectives. Int J Fashion Des Technol Educ 2016; 9: 233–247.

Koo H and Ma YJ. Exploration of transformable garment design strategies on dresses for sustainability. International Textile and Apparel Association Annual Conference Proceedings. Santa Fe, New Mexico, UAS: Iowa State University Digital Press, 2015.

Hur E and Thomas B. Transformative Modular Textile Design. In: Sarhangi R and Spin C, (eds.). Proceedings of Bridges 2011: Mathematics, Music, Art, Architecture, Culture University of Coimbra, Portugal: Tessellations Publishing, 2011, p. 217–224.

Sato

Meier

. Issey Miyake: Making things, Zurich: Scalo, 1999.

Minapoor

Ajeli

Hasani

, et al. Investigation into the curling behavior of single jersey weft-knitted fabrics and its prediction using neural network model. J Text Inst 2013; 104: 550–561.

Mozafary

Payvandy

Rezaeian

. A novel approach for simulation of curling behavior of knitted fabric based on mass spring model. J Text Inst 2018; 109: 1620–1641.

Basiri

Najar

Yazdan-Shenas

, et al. A new approach to de-curling force of single jersey weft-knitted fabric. J Text Inst 2010; 101: 941–949.

Munden

. The geometry and dimensional properties of plain-knit fabrics. J Text Inst Trans 1959; 50: T448–T471.

Walker

Doyle

. Some fundamental properties of hosiery yarns and their relation to the mechanical characteristics of knitted fabrics. J Text Inst Proc 1952; 43: P19–P35.

10.

Wang

Liu

. Development of auxetic fabrics using flat knitting technology. Text Res J 2011; 81: 1493–1502.

11.

Liu

Lam

, et al. Negative Poisson’s ratio weft-knitted fabrics. Text Res J 2010; 80: 856–863.

12.

Knittel

Nicholas

Street

, et al. Self-folding textiles through manipulation of knit stitch architecture. Fibers 2015; 3: 575–587.

13.

Glazzard

Breedon

. Weft knitted auxetic textile design. Physica Status Solidi (b) 2014; 251: 267–272.

14.

Blaga M, Ciobanu AR, Pavko-Cuden A, et al. Production of foldable weft knitted structures with auxetic potential on electronic flat knitting machines. Melliand International 2013; 19: 220.

15.

Darja

Ciobanu

Blaga

, et al. Compression of foldable links-links knitted structures. Tekstil ve Konfeksiyon 2014; 24: 349–355.

16.

Spencer

. Knitting technology: A comprehensive handbook and practical guide, Cambridge: Woodhead Publishing Limited, 2001.

17.

Park

J-W

A-G

. Bending rigidity of yarns. Text Res J 2006; 76: 478–485.

18.

Brackenbury T. Knitted clothing technology. Oxford: Blackwell Scientific Publications, 1992.

19.

Black

. Knitwear in fashion, London: Thames and Hudson, 2002.

20.

Pavko-Čuden A and Rant D. Multifunctional Foldable Knitted Structures: Fundamentals, Advances and Applications. In: Kumar B and Thakur S (eds) Textiles for Advanced Applications. London: IntechOpen, 2017.

21.

Wang X, Liu X and Hurren C. Physical and mechanical testing of textiles. In: Hu J (ed.) Fabric testing. Cambridge: Woodhead Publishing Limited, 2008, pp.90–124.

22.

Hurd

JCH

Doyle

. Fundamental aspects of the design of knitted fabrics. J Text Inst Proc 1953; 44: P561–P578.