Subjective fabric evaluation - MIRALab

35 downloads 13843 Views 3MB Size Report
Moreover, an increasing automation of. Christiane Luible ... Route de Drize 7, CH-. 1227 Carouge (phone: 0041-22-379 0115; fax: 0041-22-379 0079; e-mail: ... challenging task, if for example a panel of people is assessing the fabrics.
471911

1

Subjective fabric evaluation Christiane Luible, Minna Varheenmaa, Nadia Magnenat-Thalmann, Harriet Meinander

Abstract— The number of various existing fabric materials for different usages is unlimited. Therefore it is important to judge each textile material regarding quality and suitability before any manufacturing process. Related fabric characteristics can be subjectively assessed or objectively measured. A new field of research tries to imitate the subjective fabric evaluation method by virtually simulating the touch of fabrics with new haptic and tactile technologies. However, the today existing technology does not allow the rendering of complex interactions between hand and fabric, as it occurs during the real assessment method. Thus, the subjective fabric evaluation needs to be simplified to allow a direct comparison of the real and the virtual process. The main difference of the traditional subjective assessment, where the fabric is touched with both hands, to the simplified one lays basically in the fixation of the fabrics, so that the specimen can be judged with two fingers. The main mechanical properties such tensile, shear, bending, compression, friction, surface and weight have been assessed with the new test arrangement. The results of the assessment have been reported for all tested properties. Index Terms—Fabric assessment, fabric hand, evaluation, rating, haptic/tactile interface, virtual touch

W

I. INTRODUCTION

ITH the development of new fibres and the continuous enhancement of new material structures, the variety of different fabric materials has become immense over the years and is still increasing (smart textiles) [1]. Fabrics for different activities, climates or tendencies and trends have been developed to guarantee an optimal comfort of garments. However, the development of such a variety of different materials makes it difficult to evaluate the fabrics according their quality and suitability. Each textile possesses specific characteristics, which are advantageous for some types of garments, but can be unfavourable for others, regarding comfort. The use of unsuitable or inferior fabrics is often the fundamental aspect that determines the success or failure of a textile product [1]. Moreover, an increasing automation of

Christiane Luible, Miralab - University of Geneva, Route de Drize 7, CH1227 Carouge (phone: 0041-22-379 0115; fax: 0041-22-379 0079; e-mail: [email protected]) Minna Varheenmaa, Smart Wear Lab, Sinitaival 6, FIN-33720 Tampere, (phone: 00358-3-3115-2958; fax: 00358-3-3115-2955; e-mail: [email protected]) Nadia Magnenat-Thalmann, Miralab - University of Geneva, Route de Drize 7, CH-1227 Carouge (phone: 0041-22-379 7769; fax: 0041-223790079; e-mail: [email protected]) Harriet Meinander, Smart Wear Lab, Sinitaival 6, FIN-33720 Tampere, (phone: 00358-3-3115-4456; fax: 00358-3-3115-4515; e-mail: [email protected])

apparel manufacturing processes demands a more precise control of fabric characteristics [3]. Thus, it became more and more important to judge textiles before the production of a garment. The concept of ‘fabric hand’, ‘handle’ or simply ‘hand’ is an important method of fabric assessment which was introduced by the apparel and textile industry. Fabric hand refers to the total sensation, experienced when a fabric is touched or manipulated in the fingers. The attractiveness of a fabric’s handle depends on its end use [4], as well as on possible cultural and individual preferences of the wearer [3]. Fabric hand attributes can be obtained through subjective assessment or objective measurement [3], [5], [6], [7], [8] [9]. Recently, researchers worked on the relation and prediction of fabric handle parameters through the help of so-called “neural fuzzy networks” [10] [11]. A new field of research tries to imitate the subjective fabric evaluation method by virtually simulating the touch of fabrics using new haptic and tactile technologies [12] [13]. But, today’s existing haptic/tactile devices do not allow the rendering of complex interactions between hand and fabric. The applied real time simulation is not able to handle the occurring complex collisions between the hands and the fabric, as it occurs during real assessment processes. Thus, the virtual touch is simplified to a manipulation of the cloth with only two fingers [14]. Consequently, in order to conduct a valuable comparison of the real and the virtual touch, the real process needs to be adapted to the virtual one.

II. TRADITIONAL SUBJECTIVE FABRIC EVALUATION The subjective fabric evaluation constitutes the traditional textile assessment method, referring to the total sensation, experienced when a textile is touched, squeezed, rubbed or otherwise handled and rated by experts. The felt sensations about a textile are described as components of fabric hand. The subjective evaluation of fabrics becomes notably a challenging task, if for example a panel of people is assessing the fabrics. Individual evaluators maybe perform the test procedure differently and a consistency in the test results can not be guaranteed any more. The same problem appears for subjective assessments, which are conducted at different places or at different times. In order to counter this problem, the organisation AATCC (American Association of Textile Chemist and Colourists) has published guidelines for the standardization of the subjective fabric assessment, which propose possible evaluation conditions [14].

471911

2 correlate with the objective measurements, we can assume that the newly sought method is precise enough in order to be compared with the virtual touch.

Fig. 1: Subjective fabric hand assessment

The proposed guidelines prescribe the detailed manner how the specimens have to be prepared, how the evaluator is presented the specimen and how he is asked to handle the specimen in a prescribed sequence. Different methods for the fabric rating or ranking are proposed. The AATCC guidelines also suggest terminologies for each fabric property, in order to accurately express the felt sensations about different textiles. Nevertheless, this AATCC protocol constitutes the only standard about the subjective fabric evaluation and there is no evidence that this standard is commonly accepted and widespread. As a consequence, it can be assumed that there is not only one single and clear technique for the evaluation of a certain textile property.

A. Constraints resulting from the virtual process During the virtual feel of fabrics, a certain material with its distinctive properties is simulated in a virtual environment. The user has then the possibility to interact in real time with the virtual simulated textile using a haptic/tactile device. However, the needed real time simulation of the fabric requires a simplified hand movement. Therefore a set of reasonable and feasible handling actions, which allow an adequate evaluation of the selected relevant textile mechanical properties, needs to be found. Simulated mechanical fabric properties are: • • • • • •

Tensile properties Shear properties Bending properties Compression properties Surface properties Weight property

For the array of the haptic/tactile interface it was found that the optimal way of assessing different fabric properties without manipulating the material too much is by handling it with two fingers. Hereby, different kinds of stimuli to the user fingertips are responsible for various sensations [15].

III. MODIFIED SUBJECTIVE FABRIC EVALUATION Because of limitations resulting from the real time simulation system, the haptic/tactile interface imitates the real subjective fabric assessment in simplified way by using only two fingers. However, in order to be able to validate this new high-tech fabric assessment method, it is inevitable to compare it with a corresponding real process. The main objective of this work is thus a redesign of the traditional subjective fabric evaluation, in order to match it to the constraints of its imitating virtual process and hence, to allow their direct comparison. To accomplish this task, a new evaluation protocol have been established according to new demands and tested with 10 different kinds of textiles. To better understand the new demands, possibilities and limitations existing haptic/tactile devices for the simulation of the touch of fabrics were studied. Where possible, the AATCC guidelines for the subjective fabric evaluation were followed. However, before the actual association of the real and the virtual processes, it also needs to be assured that the results of the new subjective evaluation method are accurate from the rating point of view. For this, the correlation of the subjective rating and the objective measurements will be tested. Only if the results of the new subjective evaluation

Fig.2: Scheme of the haptic/tactile Interface [16]

But, in order to be able to properly touch a fabric with only two fingertips, the textile needs to hang in space with the upper border fixed where the rest of the fabric is deformable. The user is then able to draw up his fingers to the fabric, pinch and rub it, stretch and pull it or handle it otherwise. The virtual fabric characteristics are then simulated with a complete simultaneous force-feedback and tactile-feedback.

471911

3 1) Test sample specification Ten test fabrics have been chosen among a library of 52 textiles, which already have been objectively measured with the Kawabata evaluation system for fabrics (KES-F) [5]. According to the objective measurement, five of the ten chosen fabrics possessed very different characteristics, whereas the other five fabrics were more similar. The chosen test samples were the following:

Fig.3: Virtual touch with two fingers [16]

B. New assessment arrangement The real subjective fabric evaluation process has now to be adapted to the arrangement of the virtual system. According to the AATCC standard, the tested fabric should be placed on a surface, from where the evaluator can grasp it and manipulate it with both hands. However, regarding our constraints from the virtual system, this guideline could not be followed. In our arrangement, the samples were attached to a stand from where the evaluator can assess the fabric with two fingers. It was agreed that the thumb is the decisive finger when making the evaluation and that the evaluation is made on both sides and in both directions of the specimen, if needed. During the assessment process only one fabric should be placed at a time to avoid interfering of other samples during the evaluation. The stand is placed on a table, whereas the evaluator is sitting on the other side. The facilitator changes the samples on the opposite side (fig. 4/5).

1. 33_Men’s plain woven suit fabric, 60% WO, 38% PES, 3% EL 2. 34_Men’s herringbone suit fabric, 100% WO 3. 36_Men’s twill woven overcoat fabric, 59% CO, 25% PAN, 11% WO, 5% PES 4. 37_Plain weave outdoor leisurewear fabric, 100% PES 5. 38_Weft knitted jersey fabric, 48% CO, 48% CMD, 4% EL 6. 39_Weft knitted terry fabric, 55% CV, 45% PES, Weft 7. 43_Warp knit for car seats, 89% PES, 11% EL 8. 44_Warp knit for car seats, 100% PES 9. 45_Warp knit for car seats, 100 % PES 10. 46_Warp knit for car seats, 100% PA The AATCC guidelines recommend using test specimen of a certain size so that the evaluator can hold them in both hands. Regarding the size of the fabric sample for our experiment, the same size of specimen as for the virtual experiment was used. This rather small dimension is necessary for the real time computation of the virtual fabric deformation. But in agreement with the guidelines, all used specimen had the same size and shape. The length and the width direction of each specimen were clearly identified and each specimen was only used once. In addition, all test fabrics have been conditioned 24 hours prior to the evaluation. The same conditioning of specimen has been performed prior to the objective fabric measurements, from what the mechanical fabric parameters for the virtual simulation of touch have been derived.

Fig. 4/5: Test arrangement

2) Evaluator Half an hour before the assessment, the evaluators washed the hands and dried them with a paper towel, following the recommendation of the AATCC standard. The evaluator also avoided activities before the testing, was relaxed, did no extreme exercise and did not expose his hands to extreme temperature changes. As the evaluator should fully concentrate and be in contact with no other materials than the fabrics, he was assisted by a facilitator, who was giving the instructions about the assessment and who noted all the results.

The AATCC guidelines recommend to block the view to the specimen, to make sure that the evaluator is not influenced by the visual fabric information. This guideline however could be followed and thus, it was agreed for our new specification to do first a non-visual 1st and 2nd day evaluation (fig.3) followed by one visual evaluation.

3) Rating Regarding the rating or ranking for the specimen, the official AATCC guidelines, which propose several methods, were followed as well. Generally the comparison is made with pairs or sets and judged for the direction and magnitude of the differences. The following techniques are proposed:

471911

4

- The fabric standard can be established and other specimen be rated against the reference. Hereby, the used terminology is important (Smoother, not as smooth, etc.). - The specimen can also be ranked in a comparative way such as for example “most” (rough), “least” (rough) or “moderately” (rough). - Perception scales for the description of change in a constituent element of hand can be developed, when comparing an original fabric sample against a processed, treated or other finished specimen. - Another rating method and the method which has been applied in our experiment consist in the establishment of two extremes for each property of interest. At this, an arbitrary numerical value is assigned, for instance 1 and 5 for both extremes. Values between 1 and 5 will then be assigned to the assessed fabrics. The rating scale was ran through for each property separately and the corresponding values of the standard references. 4) Standard references for extremes of each property Before the actual experiment of new subjective evaluation, the standard references were selected from the database of test fabrics. Based on these objective measurements, the standard references were selected from samples whose test results were among the lowest or highest values (not necessarily the lowest or the highest value). Also the availability of the fabric and the easiness/unambiguity of interpreting the hand of the fabric affected to the choice. To avoid the interfering of the test specimen with the reference samples, the two references samples were lying on the table in front of the evaluator. The evaluator was not allowed to see the reference samples before the evaluation.

Rate 1 Rate 5 PROPERTY Sample No. Sample No. Bending 7 10 Shear 7 32 Tensile 39 10 Roughness 24* 10 Friction 24* 16 Compression 27 10 Weight 7 32 Drapeability 10 23 *in warp direction on face side

the hands are to be pulled apart noting the ease of extending the specimen. However, as in our experiment we fix the fabric on the stand and manipulate it with only two fingers, this part of the assessment is again different from the guidelines. In the following, the new way of assessing each property is described. The first evaluation round was more a learning process for both evaluator and facilitator. The facilitator gave instructions about which side or direction of the sample was under evaluation and also the position of hand and the hold on the sample was discussed and ran through from instructions.

a)

Bending

For the assessment of the bending property, the fabric was hold with both fingers and lifted up so that the fabric created a fold. The bending was evaluated separately in forward (face) and backward (back) directions without detaching the sample and without changing the hand position (fig. 6/7). The bending in warp direction was evaluated so that the warp was in vertical position (warp yarns were bending) and correspondingly in weft direction. Standard references for bending can be seen in fig. 8/9.

Fig. 6/7: Forward and backward bending

Parameter B G EMT/E SMD MIU EMC g/m2 visually

5) Evaluation The AATCC standard proposes precise guidelines about how the fabric should be manipulated for the assessment of each property. For example to assess the ease of stretch of a textile, the specimen is to be held so that there is at least 9 cm of fabric between the hands. With elbows close to the body,

Fig. 8/9: Manipulating the standard references of bending. On the left, reference corresponding to value 1 (very weak resistance) and on the right, reference corresponding to value 5 (very strong resistance).

471911 b)

5 Shear

To assess the shear property, the fabric was hold with two fingers and moved to the left and to the right side. Shear was evaluated in the warp and the weft direction, where the sample faced the evaluator. The assessed direction was in vertical position. Shearing movements and standard references are seen in figures 10/11.

Fig. 14: Evaluation of roughness on face side (image on the left), Fig. 15: Standard references of roughness. On the left, reference corresponding to a very smooth surface and on the right, reference of a very rough surface.

e)

Fig. 10: Shearing to right and left, Fig. 11: Manipulating the standard references of shearing. On the left, reference corresponding to value 1 (very weak resistance).

c)

Tensile

During the assessment of the tensile property, the fabric is elongated with both fingers. The tensile property was evaluated in warp and weft direction, where the sample faced the evaluator. The assessed direction was in vertical position. Stretching movement and standard references is shown in figure 13.

Fig. 12: Stretching movement for tensile evaluation, Fig. 13: Standard references of tensile (image on the right). On the left, reference corresponding to value 1 (very weak resistance).

d)

Roughness

To assess the roughness property, both fingers stroke over the fabric. Roughness was evaluated on both sides in warp and weft direction separately. In the case that the evaluator noticed a difference in roughness when moving fingers downwards (D) and upwards (U), separate values for each direction were given. To ease the handling, the evaluator held the sample from the lower edge with the other hand. Roughness was evaluated with shorter finger movements than friction. Face side and back side of the sample were evaluated without detaching the sample so that the evaluator changed her hand position (first thumb on face side then thumb on back side) (Fig. 14). Standard references are shown in figure 15.

Friction

The assessment of friction is similar to roughness, whereas the movement of the fingers is longer. Friction was evaluated on both sides in warp and weft direction separately. In case the evaluator noticed difference in friction when moving fingers downwards and upwards she gave separate values for each direction. To ease the handling evaluator held the sample from the lower edge with the other hand. Friction was evaluated with longer finger movements than roughness. Face side and back side of the sample were evaluated without detaching the sample so that evaluator changed her hand position (first thumb on face side then thumb on back side) (Fig. 16). Standard references are shown in figure 17.

Fig. 16: Evaluation of friction on back side, Fig. 17: Standard references of friction. On the left, reference corresponding to a very slippery surface and on the right, reference of a very rough surface.

f)

Compression

The compression property was assessed by pressing both fingers together. It was evaluated in one sample position. Evaluation of compression and standard references are shown in figure 18/19.

Fig. 18: Evaluation of compression,

Fig. 19: Standard references.

471911 g)

6 Weight

d)

For the assessment of the weight property, the fabric was lifted up with both fingers. The weight property was evaluated also non-visually. Deviating from the instructions of the preliminary plan the sample was fixed in the same way as with other parameters and the evaluator lifted up the lower edge of the sample (Fig 17). To the weight property an own scale was applied: 1=very light, 5= very heavy. Standard references are shown in fig. 20.

Generally, the facilitator found that it is important to follow the position of the hand and the hold on the sample all along the evaluations (now this was done only in the beginning). e)

During all the three evaluation cycles, the standard references were not handled exactly in the same way as the hanging samples. Therefore, it would better to also fix the standard references on the stand or maybe a second stand. f)

Some difference between the results of the first and second non-visual evaluation could be stated. The evaluator felt that during the first round the samples were more compared to each other than to the references. V. CONCLUSION Fig. 20: Evaluation of weight,

Fig. 21: Weighing of standard references

IV. IMPROVEMENTS After each test cycle and again at the end of all the three test cycles, the new assessment protocol was evaluated and improvements made. The following remarks are a summary of the learning process of this experiment. a)

In the beginning, the evaluator started automatically to compare the samples to each other during the rating. Thus, the facilitator has to stress more clearly from the very beginning that the rating is based on the standard references and that the samples should be compared to them. In addition, it should be mentioned more clearly that the same rates can be given to more than one sample if necessary b)

During the first test cycle, the evaluator was only informed after the first five samples that there are another five samples to evaluate and that these two sets are of different types. This caused some kind of confusion for the evaluator regarding how to rate the second set. Thus, it was found out that the general result of the assessment would be better, if the facilitator states in the very beginning that the total number of samples is ten and that the same rates can be given to more than one sample if necessary. c)

After the two blind assessments cycles, the evaluator stated that the rating scale was too rough, especially for the similar types of fabrics (Scale 1 to 5 was used for both sets.) Thus, it is important to use a more detailed rating scale.

The present experiment was conducted to propose a new form of subjective fabric evaluation for the final goal to compare it with the virtual touch of fabrics. Possibilities and limitations of the virtual system have been studied and copied to the subjective fabric assessment. Guidelines of the AATCC standard have been applied where possible. New evaluation methods have been sought pour each tested fabric property, based on the new two finger assessment. A rating method using two standard references and five scales in-between was applied. The ratings showed some medium correlation to the objective measurements. For the future, new developments on the side of the virtual touch have to be respected and the corresponding subjective processes adapted. At the end of this development, the virtual evaluation of the touch of fabrics should reach the capacities of today’s fabric hand assessment to be able to partly replace those processes. ACKNOWLEDGMENT This research was funded by the European Project Haptex. This research was funded by the European Project HAPTEX – Haptic sensing of virtual Textiles, Project Nr.: IST-6549, Future and Emerging Technologies (FET - IST - FP6). We also want to thank our research partner Percro for the sketches of the haptic/tactile device. REFERENCES [1] [2]

[3]

[4]

Lam Po Tang S., Stylios G., “An overview of smart technologies for clothing design and engineering”, International Journal of Clothing Science and Technology, Vol. 18, No. 2, 2006, pp 108-128 Mäkinnen M., Meinander H., Luible C., Magnenat-Thalmann N., “Influence of Physical Parameters on Fabric Hand”, Proceedings of Workshop on Haptic and Tactile Perception of Deformable Objects, Hanover, 2005 P. G. Minazio, “FAST – Fabric Assurance by Simple Testing”, International Journal of Clothing Science and Technology, Vol. 7, No. 2/3, 1995 Kawabata S., Niwa M., “Influence of fibre process parameters on product performance”, Proceedings Fibre to Finished Fabrics, Fibre

471911

[5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]

[16]

Science/Dyeing & Finishing Groups Joint Conference, The Textile Institute, Dec., 1998, 1. Kawabata, S. “The standardization and analysis of hand evaluation (2nd edition). The hand evaluation and standardization committee”, The Textile Machinery Society of Japan, Osaka, 1980, Peirce F. T., “The Handle of Cloth as a Measurable Quantity”, Journal of the Textile Institute, Vol. 21, 1930, pp 377-416 J. Lindberg and B. Dahlberg, "Mechanical Properties of Textile Fabrics, Part III: Shearing and Buckling of Various Commercial Fabrics," Textile Research Journal, Vol. 31, No. 2, 1961, pp 99-122 Stylios G., “New measurement technologies for textiles and clothing”, International Journal of Clothing Science and Technology, Vol. 17, No. 3/4, 2005, pp 135-149 Grineviciute, D., Daukantiene, V., Gutauskas, M.: Textile Hand: Comparison of Two Evaluation Methods. ISSN 1392-1320 MATERIALS SCIENCE (MEDZIAGOTYRA), Vol. 11, No. 1, 2005. Hui C. L., “Neural Network Prediction of Human Psychological Perceptions of Fabric Hand”, Textile Research Journal, Vol. 74, 2004 Stylios, G., Cheng, L.: A Neural-fuzzy Modelling for the Prediction of Fabric Handle Values. May 1996. Bishop, D. P.: Fabrics: Sensory and Mechanical Properties. The Textile Institute, Textile Progress, Volume 26, Number 3, 1996. Cho, G., Kim, C., Casali, J. G.: Sensory Evaluation of Fabric Touch by Free Modulus Magnitude Estimation. Fibres and Polymers 2002, Vol. 3, No. 4, 169-173. Gladstone, K., Graupp, H, Avizzano, C. A.: Assessing the Utility of Dual Finger Haptic Interaction with 3D Virtual Environments for Blind People. IDCVRAT, September 2002. www.aatcc.org, “Fabric Hand: Guidelines for the Subjective Evaluation”, Developed in 1990 by AATCC Committee RR89, editorially revised and reaffirmed 2001, AATCC Technical Manual 2002, http://haptex.miralab.unige.ch/, Deliverable D4.1. “Specification of the Whole Haptic Interface”.

7