Indi an Journal of Fibre & Research Vol. 28, March 2003, pp. 23-28

Direct determination of snarliness

A Primentas" School of Textile Industries, The University of Leeds, Leeds LS2 9JT, UK

Received 5 October 2001; revised received and accepted J J March 2002

The formation of snarls that takes pl ace in the in post-spinning processes, such as winding, warpin g, weaving and knitting, has been studied. This defect is due to the excessive yarn twist liveli ness (torsional buckling) and causes great trouble to thc manufacturers of yarns and fabrics. Biased testin g techniques are used for the measurement of this defect. A snarling testing apparatus PRIANIC and an unbiased testing technique have been developed and are reported. The investigation of factors such as yarn package form, yarn twist factor, time elapsed between production and processing stages of yarns and yarn conditi oning show the significant role th ey play in the reduction of yarn snarliness.

Keywords: Knitting, Snarliness, Twist li ve liness, Warping, Weaving, Winding, Yarn package, Yarn twist factor

6 1 Introduction twist direction . Newly produced yams exhibit a great The twisted shape of the fibres forming a yarn tendency to untwist when freed from restraint. When blings to the fore two of their main physico-mechanical such a yarn is prevented from untwisting, by being held properties: torsional rigidity and torsional buckling. at both ends approaching each other, it forms an arc. The higher the torsional rigidity, the stronger is the tor­ The nearer to U-shape this arc gradually comes, the sional buckling effect, also known as yarn twist liveli­ easier is snarl formation (Fig. 1), whereas in highly ness. It is the tendency of a newly spun yarn to untwist twisted yarns, snarls are formed very fast before an arc spontaneousli, a detrimental yarn characteristic hav­ appearance (Fig. 2). Recent research7 has shown that ing two representatives - the snarliness that occurs in the torque required for a yarn to snarl is independent to the yarns and the spirality that appears in knitted fab­ the yarn twi st level. Snarls appear in thinner yarn parts rics. On the contrary, the crepe phenomenon, appears where twist accumulation occurs and/or in a likel y part in special woven fabrics, is considered as a beneficial of the yam when the numerical difference (yarn weight result of the yarn twist liveliness. minus untwisted stress) has a negative value. Yarn twist liveliness is affected by the twist factor, Many problems originate from the yarn snarliness yarn fineness2 and retractive forces, which, in turn, are and appear in post-spinning yarn processes. In yarn determined by the torsional and bending stresses in the winding, a snarly yarn might be accumulated behind 3 fibres , and the torque generated during yam twisting. the finger guide (knife) of an electronic slub catcher, The latter is constituted by three components, viz. tor­ resulting in yarn breakage and worsening its quality. 4 sion, bending and tension of the fibres , whereas its level depends on the degree of fibre migration that takes places. The twist liveliness of fal se-twisted and set yarns depends on their structure and, in particular, s on the properties of the helical parts of their filaments .

1.1 Snarl The snarl is due to yarn torque and takes place dur­ ing the torsional buckling effect, where the yarn retires into itself and simultaneously is twisted in the opposite

a Present address: Department of Textile Engineering, Faculty of Applied Technologies, TEl of Piraeus, 250 Thivon & P. Ralli, Athens 12244, Greece Phone: + 3-2 10-4515137; Fax: + 3-2 10-4122977 ; E-mail: aprim@ teipir.gr Fig. I- Snarl formation of a normally twi sted yarn 24 INDIAN 1. FIBRE TEXT. RES. , MARCH 2003

midd le of a known length of yarn. As soon as the two yarn ends are brought together, a loop (s narl ) is formed as rotated unti l reaching the immobility state. The number of turns of the loop and/or th e measurement of the di stance between the two yarn ,s end s .' 6 give the yarn snarliness level. The drawbac k ! of th ese methods is th e constraint of the snarl j ! ~------formation in a specifi c yarn pl ace, a fact that does not permit the draw ing of true conclusions about the objecti ve snarling tendency of the yarn s. Furthermore, there is confusion about the mass of the hanged object and the length of the yarn sampl es. Also, th ere is no mention about yarn pretension. Al l th e above high lighted th e necessity of usin g a simple new method and an apparatus for the measurement of snarliness, overcoming most of the

I sources of errors and shortcomings of th e earl ier

----~~- ...... - -.--- ...... ~. -- -.- -- - . meth ods. Therefore, in the present work, a snarl in g testing apparatus PRIANIC and an unbi ased testin g technique have been developed and are reported.

Fig.2-Snarl fo rmatio n of a hi ghl y twisted yarn 2 Materials and Methods 2.1 Snarling Testing Apparatus PRIANIC Due to lack of tensions, the snarl s are fo rmed in sin gle The apparatus shown in Figs 3 and 4 consists of a yarns as they are with drawn over the top of bobbins base where two stands are firmly fixed. A th readed during the fo lding-twisting processes and especially rod with two ball bearings is placed on th e top of in two-for-one twisting where thi s problem becomes these stands. A motor on the base drives the threaded more serious. In weaving, snarl s in weft yarns appear rod where beneath the latter there is a scale with when the shuttle returns to the shed from the box. li near metre (l00 cm). On th e right side of the Some of these snarl s are entrapped into the cloth and apparatus, a movable part (J) with clamp jaw and a do not open out even wh en th e weft is sub seq uentl y pointer is in contact with th e threaded rod , whereas a tensioned as picking proceeds, makin g necessary th e stationary jaw (2) with a special yarn pretension cloth mending for their removal. As the hanks made system ex ists on the left side. The pretension system from hi ghl y twisted yarns are removed from the reel consists of an adjustable co unterbalance and a swift , th ey shrin k and fo rm numerous snarl s, causing graduated rod (0-200 g) that is based on the TEX great trouble during their positioning on the rods system , with tension uni t of tex/2 ± 10% g. (buttons) of the hank-dyeing machi ne. Moreover, during the winding of these dyed yarn hanks to cones, 2.1.1 Principle the formed snarls can cause snatching, tension peaks The operation principle of the dev ice is simple. A and yarn breakage. yarn length of one metre is clamped on the two jaws It is a virtue for a yarn, with its optimum twist of the dev ice. The mo vab le part approaches th e factor, to have the minimum snarliness level that ca n stati onary one as it is driven by the rotat ing threaded be achieved by several ways such as storin g9 at rod. The yarn figures a U-shape loop and according to adequately high temperature and rel ativ e humidity its twi st li ve liness, a snarl is fo rmed. (70-75 %) or by steaming lO th at resu lts in a fast yarn relaxati on. 2.1.2 Testing Procedure The device is set to its ini tial condit ions. The 1.2 Existing Methods for Testing Yarn Snarliness movable part (1) is returned to the ri ght side of the Several wo rkers3.11. 14 , in an endeavour to measure device, where the pointer in dicates th e va lue 100 on yarn snarliness, have described various methods that the scale. As the lin ear density of the yarn is known, are more or less based on the same principle. In th ese the appropriate adjustmenr is made on th e pretension methods, a li ght mass object is suspended fro m th e system. After discarding a substanti al quantity of yarn PRIMENTAS: DIRECT DETERMINATION OF YARN SNARLINESS 25

Fi g.3-Snarling testing apparatus "PRIANIC"

Guide Movable Jaw (1) Pulley·

Threaded Rod \ /

Fig.5-Effect of yarn unwinding method on yarn twi st (a-over­ Yarn Specimen end, b-sideways, and c-under-end) Scale standard operation speed. The movable jaw moves towards the stationary jaw at a constant speed of 0.5 cmls, giving the total maximum time of 200 s for testing a dead yarn. In the existing various apparatus and methods, there is no reference about the speed of the movable jaw. Moreover, th e testing time using the PRIANIC apparatus has been reduced remarkably in

Pointer rel ation to the time of the other methods that ranged approximately between 10 min and 15 min. Fig.4-Schematic presentation of the snarling testing apparatus "PR1ANIC" Furthermore, the apparatus is suppli ed with a proper yarn pretension system (tex/2 ±1O% g) that is strongly from a bobbin, one metre yarn specimen is fastened in recommended for testing yarn characteristics under the grip of the stationary j aw (2) while holding the dynamic conditions. The most important feature of other end of the yarn to prevent the loss of twist. The the PRIANIC device and the accompanied method is yarn is stretched until no contact of the graduated rod that the yarn forms freely a snarl in the most likely pointer of the pretension system can be detected and part of it, confirming its unbiased testing principle. fastened in the grip of the movable part (1). By The snarliness value given by the measurement of the switching on the motor, the threaded rod rotates, distance between the two yarn ends at the starting forcing the movable part to move towards the point of the snarl formation makes this method stationary jaw. Thus, the yarn becomes loose and unrivalled. figures a U-shape loop. By the time the first snarl is formed, the operation of the device is suspended. The 2.2 Yarn Unwinding pointer of the stopped movable part indicates a valu e The simple experiment shown in Fig. 5 was carried (e.g. 73.6 cm) that represents the snarliness value of out to investigate the effect of the yarn unwinding the examined yarn specimen. The higher the value, manner on yarn twist and twist liveliness. Two the more snarly is the yarn. parallel roving stripes were manually wound on a cop. Ten yarn specimens from each sample are The over-end unwi nding added some twist in thi s considered as the most appropriate for testing. The assembly. The added number of turns per unit length atmospheric testin g conditions must be 65 ± 2 % RH of the unwound assembly with Z direction was equal at 20 ± 2 DC under which the yarn samples must be to the figured spirals of that length on the cop (Fig. kept for conditioning for 24 h. Sa). On the contrary, the under-end unwinding (Fig. 5c) showed exactly the same behaviour apart from 2.1.3 Advantages of the Apparatus and the Method that the twist direction was opposite, i.e. S direction. The PRIANIC apparatus has the ad vantage of the When the assembly was unwound sideways, no 26 IND IAN J. FIBRE TEXT. RES ., MARCH 2003

alteration in its stru cture was observed (Fig. 5b). into two lots. One of these lots was kept in a standard The over-end unwinding of the yarns from their testing laboratory environment of 20±2 °C and 65% cop package adds most of the times some twist and RH and was referred to as conditioned lot. The other probably increases sli ghtly the twist li veliness of th e lot was kept in environment of non-standard yarns. It was shown that this twist increase was conditions and was referred to as unconditioned. negligible when referred to yarn from cone packages After the yarn samples production, their linear 7 and therefore did not affect the twist li veliness . Thus, densities and twist levels were tested and the means it was decided fo r the various experiments to adopt are presented in Table 1. The effect of the yarn the conveni ent, generally accepted and widely used package form of the unconditioned lot, the first day over-end yarn unwinding. after the yarn production, on the yarn snarliness is presented in Fig. 7. Table 2 shows the readings of 2.3 Preparation of Yarn Samples and Testing of Snarliness yarn snarliness, taken from all the yarn samples of 2.3.1 Long-Staple Yarns both lots after one and one hundred days elapsed from Yarn samples of 32 tex were produced from 100% the yarn production day. acrylic long-staple fibres with four different twist factors. Measurements of th e yarn snarliness of these 3 Results and Discussion samples were made on th e PRIANIC apparatu s, with From Fig. 6 it is obvious that the increased twist set pretension of 15 g, every two days for a period of level of the long-staple yarn samples of both linear 20 days. The obtained results are shown in Fig. 6, densities resulted in an in crease of th eir snarliness. On whi ch clearly shows the signi ficant role of the storage the other hand, all the four yarn samples showed the time on the relaxation of the twist liveliness. dramatic decrease of the snarliness tendency on th e very first two days after their production. This 2.3.2 Short-Staple Yarns Experiments were carried out to examine any 98 · 1 112 Twi st factor, turns.cm .tex alterations of the snarliness level of short-staple yarns 96 24 .0 --- 27.5 . 1 due to the possible effect of the following four /-.--+- 29.3 --+- 30.8 factors: E (i) the shape of the yarn package (cop, cone); u,94 (/) (/) (i i) the twist factor of the yarn samples; Q) c (iii ) the time elapsed from the yarn production up 'E 92 co c to the snarliness testing; and (f) (iv) the atmospheric conditions in which the yarn ~ 90 samples were stored. >-

Cotton singles ring-spun Z-twisted yarns of 88 nominal linear densities 29 tex and 39 tex were produced and wound onto cops and cones. For both 86+------,------,------,------, these yarn linear densities, three nominal twist factors 5 10 15 20 V o (32, 34 and 39 turns.cm,l.tex ,) were chosen. Time , days Quantities of these six yarn samples were made up Fig.6- Effect of yarn storage on snarl iness value

Table 1- Effect of yarn package on short-staple yarn tw ist

Yarn Linear density Twist, turns.n'-' Actu al twist factor sample tex A B C turns.cm·'.tex in

S, 29.3 603.9 592.7 598.3 32.4 S2 29.3 656.5 63 1.2 643.9 34.9 S3 29.4 7 11.2 7 11.5 711.3 38.6 S4 39.4 522.3 505.5 5 13.9 32.3 S5 38.9 545.6 526.8 536.2 33.4 S6 39.5 632.0 625.6 628.8 39.5 A-Yarn on a cop package; B-Yarn on a cone package; and C-Average value of cop and cone yarn twi st readings PR IMENTAS: DIRECT DETERMINATION OF YARN SNARLINESS 27

Table 2- Effect or yarn package, conditi oning and conditioning time on short-staple yarn snarlincss

Yarn Linear density Twist factor Yarn Yarn snarliness ,cm sample tex turns. cm · 1.t ex Yl package C D I" 100" I " 100 " A 78.2 72.3 67 .0 52.6 S, 29.3 32.4 B 43.8 41.9 43 .7 4 1.6 A 74.0 70.7 78.3 73.0 S2 29.3 34.9 B 55.7 49.3 59.9 52. 1 A 88. 1 86.8 89.9 80.4 S) 29.4 38.6 B 67.0 64.4 61.9 59.5 A 76.1 75.0 78.6 69. 1 S4 39.4 32.3 B 39.4 37 .7 56.4 37.3 A 79.9 74.9 73 .5 67 .6 Ss 38.9 33.4 B 58.8 49.3 45.1 4 1. 2 A 89.6 88. 1 90.5 82.4 S6 39.5 39.5 B 77.2 76. 1 65.1 52.5

A-Y3rn on a cop package; 13-Yarn on a cone package; C-Unconditioncd yarn; and D--Conditioncd yarn. a Number of days intcrvcned between yarn production and testing.

100 Yam 29 In Yam 39 lex With an intervened time of 100 days from the 90 ~oCone 80 production day, the yarns from conditioned cones 5 70 showed, in most of the cases, smaller tendency to 60 form snarls. This may be due to the combination of "' 50 "c 40 various factors such as the more efficient conditioning ~c III 30 of the yarn during the wi nding stage where the most ..E >- 20 important factor was likely to be the wind , and th e 10 application of longitudinal tensile stress on the yarn. 32.4 34.0 38.6 32.3 33.4 39.5 Furthermore, a cross-wound cone had a much more Twist Factor . tums.cm·'.tex Y. open structure. The better conditioning of the yarn in the cone form might also be due to lower tension Fig.7- Erfect of the yarn packagc form on yarn snarliness existing in the yarn forming coils around the yarn package, resulting in a softer package when compared reduction in snarliness became smoother as the time with the cop form of the yarn package. Furthermore, from the production day was steadily elapsed. It could on comparing the readings obtained by measuring the be also stated that yarn snarliness followed an almost yarn snarliness after a period of 100 days with those linear trend. obtained one day after the yarn production, it can be The data in Tables 1 and 2 aBd Fig. 7 show that al­ seen that the time factor contributed slightl y to the though the differences in twist amount were very snarliness reduction of unconditioned yarns. On the small between cops and cones of the same yarn sam­ other hand, following the appropriate statistical ples, the yarns from the cones in all the cases exhib­ analysis that was based on the fac torial design ited a relatively high reduction (20-40%) in snarli­ analysis of variance of these data, it could be stated ness. This was probably due to the tension applied to that, in some cases, apart th e great significance of the the yarn resulting from the winding speed and ten­ factor yarn package form, the factors conditioning and sioning devices on the winding machine. This tension time were of noticeable importance for yarn seemed to extend slightly the yarn, resulting in a rear­ relaxation with a significance level above 95 %. rangement of the structure of the yarn cross-section. Moreover, the same tension may be responsible for 4 Conclusions the significant reduction in torque produced by the Twist li veliness remall1s a great problem, twist insertion during the spinning process. manifesting itself as snarls in the yarns and spirality in 28 INDIAN J. FIBRE TEXT. RES., MARCH 2003

knitted fabrics. The relative effectiveness of the 2 Banerjee P K & Alaiban T S, Text Res J, 58 (5) (1988) 287. various apparatus used for determining yarn 3 Zhu Y, An examination oj the twist liveliness oj trapped-twist textured yarns, M.Sc. dissertation, The University of Leeds, snarliness and their disadvantages, concerning the Leeds, 1989. adoption of the biased testing, led to the development 4 Bennett J M & Postle R, J Textlnst, 70 (4) (1979) 121. of the snarling testing apparatus PRIANIC in an 5 Denton M J, J Text Inst, 57 (7) (1966) T372. attempt to overcome most of the limitations of the 6 Primentas A, Contribution to the determination oj yarn earlier instruments. The validity of this instrument as snarliness, M.Sc. dissertation, The Univcrsity of Leeds, well the adopted simple testing method was shown Leeds, 1991. after testing a numerous series of yarns at different 7 Primentas A, Mechanical methods Jor the reduction oj spiral­ ity in weft kniued Jabrics, Ph.D. thesis, The University of twist factors. Because this apparatus is still in the Leeds, Leeds, 1995. embryonic stage of development, many improvements 8 Beevers H, Practical spinning on the BradJord system (Na­ need to be made. These could be: (i) an electronically tional Trade Press, London), 1954, 145. adjusted pretension system that will be helpful mainly 9 von Bergen-Mauersberger, American Handbook (Tex­ during the yarn stretching test. This test is considered tile Book Publishers, New York), 1948,618. as a sequential part of the snarl test. Except the 10 Twist setting: The usc of vacuum' steaming and cquipment, presence of the advanced pretension system, the Wool Record, 6 September (1951) 26. reverse movement of the movable bracket at various I I Ingham J, J Text Inst, 43 (I) (1952) P44. speeds is considered as essential; and (ii) the use of a 12 Methods oj test Jor , B.S. Handbook No II (British Standards In stitution), 1974,3/31. counter with a rotating grip fixed on the movable 13 Cheung H, Some properties oj draw-textured yarns, bracket offers the opportunity for yarn twist testing. M.Sc. dissertation, The University of Leeds, Leeds, 1974. The twist alteration of the examined yarn will give the 14 Nigam J K, Th e influence oj processing variables on the capability for examining the character of the new yarn physical properties oj continuous filament textured yarns, status during the snarling test. Ph.D. thesis, The University of Leeds, Leeds, 1972. 15 Chow K C, The structure, mechanics and twist liveliness oj References Jalse-twist textured yarn, M.Sc. dissertation, The University I Textile Terms and Definitions, 10th edn (Textile Institute, of Leeds, Leeds, 1976. Manchester, UK), 1995. 16 Araujo M D & Smith G W, Text Res J, 5S' (5) (1989) 247.