Study on Optical Properties of -like Fabrics By Ikuko Maekawa*,Toshihiro Gunji* and TsuneyoTsuboi**, Members, TMSJ *OtsumaWomen's University , Sanban-cho12, Chiyoda-ku,Tokyo, 102 **AichiInstitute of Technology, Yatsugasa-machi,Toyota-shi, Aichi Pref., 470-03 Basedon Journal of theTextile Machinery Society of Japan, Transactions, Vol. 35, No. 8, T117-T125(1982-8) Abstract Testswere carried out with a gonio-photometer toinvestigate optical properties ofsilk-like synthetic fiberfabrics. To compare silk-like fabrics with natural silk fabrics, habutai and crepe de chine of silk- likesynthetic as wellas thoseof naturalsilk fibers were used as samples.The following results wereobtained : 1. Thedifference in optical properties of silk-likeand natural silk fabrics can be distinguished by comparingtheir reflected light distribution curves and quantity of reflectionobtained by revolving specimens. 2. Littledifferences have essentialy been found between silk-like and natural silk fabrics. However a detailedobservation has revealed that a silk-likefabric has little amount of diffusedreflected light and thespecular reflected light of a silk-likefabric tends to gather. This causes the visual sensory differences in luster. 3. Sensory tests were carried out of a differencein luster of a silk-likeand a naturalsilk fabric. The crepede chine of a silk-likesynthetic fibers can be distinguished from that of natural silk fibers by the differenceof degree of creping. By contrast, a habutai of silk-like synthetic fibers can not completely be distinguishfrom that of naturalsilk fibers. In otherwords, a habutai of polyesterfibers is much similar invisual luster to that of natural silk fibers.

1. Introduction density, and processes. There have been many reports~l-57 published on the sub- 2) Increase of intra- spaces by sodium hydroxide re- duction of polyster fabrics. ject of silk-like synthetic fibers. Efforts to improve the tex- ture and the dyeability of marketed silk-like fibers and fab- 3) Production of soft texture by the use of oil. rics have focused on the following: Through the use of the above techniques, it has now be- A. Fibers come possible to produce all syntheric fabrics with silk-like 1) Alteration of polymer structure. gloss and texture. 2) High Young's modulus by altering spinning, drawing, The authors have hitherto investigated~s-13) the light and heat-setting processes. reflection and the transparency characteristics of both single 3) Alteration of fiber surface properties. and multifilament fibers, and the relationship between luster 4) Good luster by making the fiber cross-section irregular. and goniophotometric curves of fabrics. In the present re- 5) Soft touch by the use of fine-denier multifilament fibers. search, the optical properties of marketed silk-like synthetic B. fiber fabrics produced by the techniques mentioned above 1) Application of silk-like twists to yarns by texturing. were investigated using both a usual goniophotometer and a 2) Moderate intra-fiber spaces by blending fibers with special goniophotometer to which an integrating sphere high and low shrinkage factors. was attached for measuring reflectivity and transparency. Results obtained were compared with those obtained on 3) Twisting of fibers having different denier, dyeability, similar real silk fabrics. and physical characteristics. Research is also being carried on into both the number and the direction of twists, 2. Samples and Experimental Method with the aim of altering the characteristic texture, and of producing silk-like finish. 2.1 Samples C. Fabrics The following fabrics of almost identical specification were I) Production of silk-like finish by modifying weave, yarn used : marketed (hereafter PET) plain-weave habu-

18 Journal of The Maclhine,~' Society of Japan Table 1 Fabric Samples

Fig. 1 Surface of fabrics (habutai) (No. 1)

Vol. 30 (No. 1) (1984) 19 Fig. 1 Surface of fabrics (crepe de chine) (No. 2)

Fig, I Cross-section of fabrics (habutai) (No. 3) tai and crepe de chine, and real silk plain-weave habutai habutai sample, Fig. 1(a), clearly shows that the aligned flat and crepe de chine (see Table 1). Surface and cross-section surfaces of uniformly-thick tri-lobal cross-section fibers micrographs of samples were made with a scanning electron give rise to smooth fabric surface. Moreover, it can be seen microscope, and shown in Fig. 1, in which (a) is PET plain- from Fig. 1(a") that the uniform tri-lobal fibers of the PET weave habutai; (b) silk plain-weave habutai; (c) PET crepe plain-weave habutai interlock together neatly, producing de chine; (d) silk crepe de chine. Micrographs of samples (a) compactness. In contrast, the silk plain-weave habutai sam- to (d) taken at an inclination of 60° are shown in Figs. (a')- ple shows thickness variation and natural twisting of individ- (d'). ual fibers, and fibers separate each other resulting from In plain-weave habutai fabrics there is not twisting in sericin reduction. either warp or filling yarns. Especially, the PET plain-weave In the crepe de chine fabrics, the combination of virtually

20 Journal of ' The Textile Machinery Society of Japan Fig. 2 Cross-section of yarns untwisted warp yarns and strongly twisted filling yarns 2.2 Experimental Method gives rise to surface undulation. In the PET crepe de chine, A. Goniophotometric Curve Measurement it can be seen that the flat surfaces of tri-lobal fibers are Figure 3 shows a schematic diagram of the goniophotome- aligned on the yarn surface in the same way as in the PET ter (Murakami Color Research Laboratory Model CP-1R) plain-weave habutai. Because of the strong twist applied to used in the experiment. filling yarns, there would appear little essential difference be- As shown in Fig. 4, the apparatus was set to give incident tween PET and silk yarns. In the PET crepe de chine, how- angles of 45° and 60°, and could be continuously adjusted to ever, strongly-twisted and heat-set filling yarns gives rise to give reflected angles of from -35° to 90°, and from -50° to a comparatively smooth surface with no crimping. In con- 90°. Variations in reflected light values were recorded, and trast, the silk crepe de chine shows both more irregularity of goniophotometric curves were derived. Furthermore, taking single-fiber arrangement arising from variation in silk denier into consideration the surface properties of the samples, the and natural twist, and wider intra-fiber spaces resulting from diameter of the incident beam was set at 10 mm. de-gumming. Fig. 2(a) shows a cross-section of polyester B. Angle Measurement of Sample Rotation and Varia- fibers used in the present experiment. It can be seen that the tions in Reflected Light fiber cross-sections exhibit a regular tri-lobal form. Using the method developed by R. Jeffries,115Jboth inci-

Fig. 3 Schematic diagram of goniophotometer.

21 Vol. 30 (No. 1) (1984) Fig. 6 Method to measure reflectivity by integrating sphere. Fig, 4 Optical diagram to measure the goniophotometric reflec- tion curve.

Fig. 5 Measurement of sample rotation. Fig. 7 Method to measure reflectivity by integrating sphere. dent and reflected light angles were set at 60° (45°), and the ured. From this, diffuse reflectivity values at vertical inci- samples were rotated from 0° to 360°, as shown in Fig. 5. dence were derived. Continuous goniophotometric measurements were made of b. Measurement of Total and Diffuse Reflectivity at 45~ the relationship between the angle of sample rotation and Incidence. reflected light values. As shown in Fig. 7, sample holders were inclined at 45°, C. Measurement of Reflectivity and Transparency Using and standard white plates attached to Sample Holders an Integrating Sphere. No. 2 and No. 3. After adjustment to give a value of 100, the Hitherto, two methods have been used in the measurement sample was attached to Sample Holder No. 2 in such a way of luster: the two-dimensional method's and the three- as to bring filling yarns of the fabric parallel to the incident dimensional method.t17~ To investigate three-dimensional light flux. Total reflectivity at 45° incidence was measured. reflectivity, the authors constructed an integrating sphere Next, the standard white plate was removed from Sample attached to a goniophotometer. Measurements were made Holder No. 3, and in its place was affixed a trap of fully of reflectivity at vertical incidence, of both diffuse and total black color (to absorb specular reflected light). Total diffuse reflectivity at 45° incidence, and of overall transparency val- reflected light values at 45° incidence were measured. ues. The internal diameter of the integrating sphere was 120 c. Measurement of Overall Transparency mmz5, and a silicon photocell (SP-R20) manufactured by After removing Sample Holder No. 1 in Fig. 6, standard Kodenshi Industrial Research Laboratories was used as the white plates were attached to Sample Holders No. 2 and light receptor. Measurement accuracy was + 1 %, and the area No. 3, and their values set at 100. The sample was affixed to illuminated has a maximum diameter of 20 mm. Further- Sample Holder No. 1, and the value obtained was taken as more, the ratio of the internal surface area of the integrating the overall transparency. sphere to that of the aperture was 144:1. The flux aperture 3. Experimental Results and Observations was adjusted to the diameter of 10 mm. a. Measurement of Diffuse Reflectivity at Vertical Inci- 3.1 Measurement of Reflected Light Distribution dence. If fibers, including both PET and silk, are considered to be In Fig. 6, after removing Sample Holder No. 1 of the in- dielectric substances, then reflectivity can be determined by tegrating sphere, standard white plates were attached to using Fresnel's Coefficient : Sample Holders No. 2 and No. 3. Sample Holder No. 3 was R _ 21 tan2(Bit -O) + sln2(Oi -dt) adjusted to receive a vertical light flux, and the reflected an2(Bi+Bt) sln2(ei+et) light value set at 100. A sample was then attached to Sample di : angle of incidence Holder No. 3, and the amount of light reflected was meas- do: angle of refraction

22 Journal of The Textile Mac hinery Society of Japan Using this formula, reflectivity R was calculated taking n" l59, n = 1.54 and n" - 1.73,,i = I .53 as the refrac- tion values~18' for silk and PET fibers, respectively. Figure 8 shows these results, indicating that the effects of variations in reflected light caused by refraction differences upon the goniophotometric curves can be considered to be extremely small when measured in unpolarized light. Consequently, it was decided to study the effects of the fabric surface tex- ture and of the cross-sectional shape of fibers upon the goni- ophotometric curves. Figures 9 and 10 show the goniophotometric curves of PET and silk plain-weave habutai fabrics at 45° and 60° an- gles of incidence. The goniophotometric curves of both PET

Fig. 10 Goniophotometric reflection curves of polyester & silk fabric habutai. (angle of incidence: 60°)

and silk plain-weave habutai fabrics bear a marked similari- ty, and it can be seen from the ratio of specular to diffuse reflected light, in other words, from the ratio of luster~7,16' at 45° and 60°, that both PET and silk plain-weave habutai fabrics have luster. When the incident light was parallel to the PET filling yarns, reflected light values exhibited a peak in the direction of specular reflection. This can be attributed to the fabric texture of the PET plain-weave habutai men- tioned before; namely, to the relative by smooth fabric sur- face caused by the alignment of the flat surfaces of tri-lobal Fig. 8 Relation between angle of incidence and reflectance. filaments along the yarn surface. Comparison of Figs. 1(b) and (b') shows that the silk plain-weave habutai exhibits varieties in twist and thickness of individual fibers, irregulari- ties in yarn alignment, and fiber slackness, compared with the PET plain-weave habutai. All these factors contribute to a high proportion of light entering the interior of the yarn and emerging again after diffusion, thus increasing total light diffusion. It is for this reason that the goniophotometric curves of silk exhibit a characteristic roundness. The effect of silk fibrillar structutre~19,20~ upon goniophotometric curves was not investigated in the present research. Figures 11 and 12 show the goniophotometric curves de- rived for PET and silk crepe de chine at 45° and 60° angles of incidence. Because the filling yarns of crepe de chine are highly twisted, the amount of reflectivity when the luminous flux is parallel to the filling yarns is much smaller than when the luminous flux is parallel to the the untwisted warp yarns, and diffusion is greater. In addition, although the goniopho- tometric curves of both PET and silk are similar, the silk Fig. 9 Goniophotometric reflection curves of polyester & silk fabric habutai. (angle of incidence: 45°) curves show slightly larger received light values at all angles, and the characteristic roundness of the goniophotometric

Vol. 30 (No . I) (/984) 23 curve. When luminous flux was parallel to the filling yarns, can be attributed to the fact that the warp yarns are un- the maximum reflectivity values on the goniophotometric twisted, and the wave form is composed of many surfaces curve were in the region near 0° angle of reflection, an reflec- inclined at 25°-35° to the overall surface. tivity values decreased in inverse proportion to the angle of 3.2 Investigation of Reflectivity Variation by Sample Rota- light reflection. In other words, they resemble Lambert's tion Cosine Curve. This is because crepe de chine is a fabric of low luster with much irregular fabric surface (Figs. 1(c), (d)) Figure 13 shows the curves obtained for the reflectivity of resulting from weave, density, and yarn processing. How- ever, when goniophotometric curves were derived from measurements made with luminous flux parallel to warp yarns (angle of incidence 60°), the maximum reflectivity val- use arised in the region near 0° for PET and 10° for silk. This

Fig, 13 Reflectivity of polyester fabric and silk fabric by sample rotation.

PET and silk plain-weave habutai and crepe de chine at 60° angles of incidence and reflection, using the sample rotation method. Sample rotation angles of 0° and 180° are for lumi- nous flux parallel to filling yarns, and these of 90° and 270° for luminous flux parallel to warp yarns. In both PET and silk plain-weave habutai fabrics, the maximum values were obtained when the incident light flux was parallel to either warp or filling yarns, and the minimum values were obtained when the incident light flux was at an intermediate angle of 45° to them. In the PET plain-weave Fig. 11 Goniphotometric reflection curves of polyester & silt habutai, however, the maximum value when light flux was fabric crepe de chine. parallel to filling yarns was greater than when parallel to (Angle of incidence: 45°) warp yarns, and there is a sharp bend in the curve near the maximum value. This can be attributed to the fact that the trilobal fibers of uniform thickness form a flat surface (see Figs. 1(a), (a"), and furthermore the relatively low filling density, as compared with the warp density, gives rise to flattening of yarns. In contrast, the silk yarns (see Fig. 1) ex- hibit a continuous, smoothly rounded, symmetrical curve. This demonstrates that the high degree of light diffusion by silk yarns results in the minimal difference between warp and filling yarns. To determine the degree of specular reflected light varia- tions caused by the sample rotation, the luster ratios~15a for different angles of reflected light were derived from Fig. 13. The PET plain-weave habutai had the ratio of 1.46, higher than that of the silk plain-weave habutai, 1.37. This can be considered one of the prime factors determining luster. The reflected light curves produced by sample rotation of PET and silk crepe de chine are similar, but the silk crepe de chine shows higher reflectivity. This is further made clear by the goniophotometric curves for 60° angles of incidence (Fig. Fig. 12 Goniophotometric reflection curves of polyester & silk 11), which show that at around the 60° angle of reflection, fabric crepe de chine. the silk crepe de chine gives a higher value than the PET (Angle of incidence: 60°) crepe de chine.

24 Journal of The Textile Muchinerv Society of J apa11 With crepe de chine, the fillings of which are highly twisted lower than that of the silk plain-weave habutai. On the other the fabric diffuses much light in parallel to the filling yarns, hand, although the difference was small, the PET crepe de and the specular reflection is extremely small. There is also chine showed a lower value than the silk crepe de chine. little specular reflection parallel to the warp yarns, and there 3.4 Visual Differentiation of Material and Comparison of is no maximum value. In fact, the reflected light value is Luster smaller than that obtained when light flux is at 45°. This is In investigating the optical difference between PET and because crepe de chine (Figs. 1(c), (d)) is a fabric of high sur- silk, the following visual test was made: face irregularity, low specular reflectivity, and low luster. (1) Panel. As the samples used in the present experiments When luster ratios for different angles of light measure- were fabrics widely used is blouse manufacture, the panel ment were derived in the same way as for plain-weave ha- was formed of 10 members of the merchandizing division of butai fabrics, the values of 3.0 for PET crepe de chine and 1.9 a blouse manufacturer. for silk crepe de chine show that the large value for PET (2) Samples. The samples consisted of 20 x 20 cm pieces of crepe de chine has an important bearing on luster. PET and silk plain-weave habutai, PET and silk crepe de 3.3 Investigation of Reflectivity and Transparency Using chine, and they were spread on a 40 x 25 cm cardboard ( Y: the Integrating Sphere 16.2, x : 0.314, y : 0.313). The results by the method described in 2.2 C are given in (3) Questioning. To facilitate visual judgement, panel Table 2. The silk plain-weave habutai shows higher diffuse members were interviewed individually in a room giving a reflectivity with vertically incident light than the PET plain- sample illuminated approximately at 500 lx. Each member weave habutai. In contrast, there is no noticeable difference was asked to sit by the same table, was shown one pair of between PET crepe de chine and silk crepe de chine. For total PET and silk plain-weave habutai, and then another pair of reflectivity at 45° of light incidence, the PET plain-weave PET and silk crepe de chine fabrics, and asked the following habutai value was smaller than that for the silk plain-weave questions: habutai, and the PETcrepe de chine showed largervalue than i. Which is silk and which is PET? the silk crepe de chine. For diffuse reflectivity at 45° of light ii. How do they differ in appearance? incidence, the PET plain-weave habutai value was smaller iii. Which has more luster? than that for the silk plain-weave habutai, but there was little iv. How does the luster differ? difference between the PET crepe de chine and the silk (4) Results. crepe de chine. From the above results it can be seen that the i) Out of 10 panel members, 4 were able to discriminate PET plain-weave habutai has rather lower total reflectivity correctly between two plain-weave habutai fabrics. All than the silk plain-weave habutai. In addition, it is evident panel members were able to discriminate between two crepe that there is extremely little difference in reflectivity be- de chine fabrics correctly. tween the PET crepe de chine and the silk crepe de chine. The ii) Even those panel members, who were able to discriminate overall transparency of the PET plain-weave habutai was correctly between two plain-weave habutai fabrics, were

Table 2 Reflectivity and Transparency

Vol. 30 (No. 1) (1984) 25 unable to explain clearly the external differences between The results obtained were as follows: the two. Of the two crepe de chines, the silk crepe de chine 1) The optical differences between silk-like synthetic fiber was graded more crimping. These crimp differences are illus- fabrics and real silk fabrics can be clearly demonstrated by trated in Figs. 1(c) and (d). the goniophotometric curves using the sample rotation iii) Those panel members who differentiated correctly agreed method. that the PET plain-weave habutai had greater luster than the 2) The goniophotometric curves and the sample rotation silk plain-weave habutai, and that the PET crepe de chine reflectivity curves obtained for both synthetic fiber fabrics was slightly more lustrous than the silk crepe de chine al- and silk fabrics are extremely similar. though both crepe de chines had luster. 3) However, a close investigation of the results of 2) indi- iv) Panel members, who were able to differentiate correctly, cates that there still remain some differences in the optical said that the silk fabrics had natural, soft, deep luster in con- characteristics of PET and silk plain-weave habutai fabrics. trast to the glittering metallic luster of the PET. The results further indicate that this marked similarity in 3.5 Optical Characteristics of Silk-Like Synthetic Fiber the surface texture of both PET and silk crepe de chine due Fabrics. to highly twisted filling yarns is responsible for the similar- Figures 1(a) and (a") show that the individual fibers of ity in the optical properties of two fabrics. the plain-weave habutai are of uniform thickness, and that 4) Although the panel was rather small, the test results of the flat surfaces of the tri-lobal cross-section are arranged subjective visual differentiation of PET and silk plain-weave relatively smoothly on the yarn surfaces. Consequently, there habutai fabrics and their luster showed that 4 out of 10 tends to increase the light flux reflected specularly by these members could correctly discriminate between PET plain- surfaces. In other words, at 60° of light incidence, peaks weave habutai and silk plain-weave habutai. These results are produced at -30° and 60° points along the goniophoto- show that silk-like synthetic fiber fabrics are very similar to metric curve. The diffuse reflected light component is, how- silk fabrics, visually, though tactile differentiation is not ever, low in comparison with silk. (cf. 3.3) included. This finding agrees closely with the opinions of panel The present report is a synopsis of the report made at the members who correctly differentiated between PET and silk 34th Annual Meeting of this Society. plainweave habutai fabrics, and said that the PET habutai had more glittering luster than the silk habutai. Literatures Cited To remove this glitter, it is necessary to investigate further [ 1] Harada, Nakai ; J. Text. Mach. Soc. Japan, 32, p. 589 the cross-sectional shape of the fibers, to find ways of pre- (1978). venting the formation of smooth yarn surfaces, and to ran- [2] Y. Matsukura; J. Soc. Fib. Sci. Tech. Japan, 36, p. 320 domize the arrangement of fibers within the yarn. (1980). In the crepe de chine, the diffuse reflected light is large, [3] Kosaka, Iwakura; J. Soc. Fib. Sci. Tech. Japan, 37, p. 77 and luster is rather small. As the goniophotometric curves (1981). of Fig. 12 make clear, luminous flux parallel to warp yarns [4] S. Ishizaki; J. Soc. Fib. Sci. Tech. Japan, 37, p. 92 (1981). gives a specular reflected light peak in the region near - l0°, [5] Fib. Sci. Tech. Soc. Japan; Text book (1981, 1, 23). but there is no maximum value along the direction of specu- [6] Gunji, Nihira, Tsuboi; J. Text. Mach. Soc. Japan, 24, lar reflection when the angle of incidence is 60°. This shows T149 (1971). that the irregular fabric surface is not optically flat, and has [7] Gunji, Nihira, Tsuboi; J. Text. Mach. Soc. Japan, 24, quantitively little luster. The PET crepe de chine has, how- T33 (1971). ever, rather larger specular reflectivity than the silk crepe de [8] Tsuboi, Nihira, Gunji; J. Text. Mach. Soc. Japan, 25, chine. It may be considered that this is why the PET crepe de T19 (1972). chine seems more lusterous than the silk crepe de chine. Al- [9] Tsuboi, Nihira, Gunji; J. Text. Mach. Soc. Japan, 26, though the optical comparison of PET and silk fabrics seems T103 (1973). to show small differences, the PET fabrics exhibit less light [10] Tsuboi, Nihira, Gunji; J. Text. Mach. Soc. Japan, 27, diffusion, and tend to increase specular reflection flux. This T110 (1974). may account for the visually-perceived difference in luster. [11] Tsuboi, Nihira, Gunji; J. Text. Mach. Soc. Japan, 27, 4. Postscript T118 (1974). [12] Nihira, Tsuboi, Gunji; J. Text. Mach. Soc. Japan, 30, In investigating the optical characteristics of polyester T23 (1977). silk-like synthetic fiber plain-weave habutai and crepe de [13] Nihira, Tsuboi, Gunji; J. Text. Mach. Soc. Japan, 30, chine fabrics widely used in women's dresses and blouses, T77 (1977). a comparative study of almost identical silk fabrics was made [14] S. Hagiwara; Jap. Soc. Seric. Sc., 12,1(1941). using a goniophotometer and an integrating sphere.

26 Journal of The Textile Mac{iinerv Society of Japan [15] Jeffries, R.; J. Text. Inst., 46, T391 (1955), 46, T759 [19] M. Minagawa; Jap. Soc. Seric. Sci., 28, 48 (1959). (1955), 47, T319 (1956). [20] H. Matsumura; 2nd ser. Silk Yarn Const.,180, (1980), [16] JIS Z 8741 (1962). Shinshyu Univ. [17] M. Sawaji; J. App!. Phy., 29, 804 (1960). [18] Japan Cot. Spin. Soc.; Text. Tech. News, No. 266, p. 14, (1962-3-15).

Vol. 30 (No. I) (1984) 27