Ellis Developments Limited
Nottingham, United Kingdom
|
Ellis Developments Limited Nottingham, United Kingdom
|
|
Fabric Quality Systems These old course notes are incomplete, have not been fully checked since retyping, and do not include the diagrams, but may be of some assistance in home study. FABRIC AND GARMENT DIMENSIONS - YARN, FABRIC GEOMETRY AND FINISHING CONSIDERATIONS INTRODUCTION
There are three dimensions of major importance:-
Weight is made up from the weight of the total length of yarn in the unit area of fabric or unit lot of garments. Clearly this weight is related to the count of the yarn. The length of yarn involved is related to the mean loop length knitted and total number of loops knitted in this unit area of fabric or unit lot of garments. The total number of loops will depend on the number of loops per unit area of either fabric or garments.
SIMPLE FABRIC GEOMETRY AND THE IMPORTANCE OF LOOP LENGTH
Extensive studies of knitted fabric geometry with yarns not subject to shrinkage on finishing - such as cotton and worsted spun yarns including fully relaxed high bulk acrylics - and with simple fabric structures such as plain knitting or 1x1 rib, have shown that the number of loops per unit area - the stitch density, - is inversely related to the loop length squared - viz. S = k/l2 where S is the stitch density, 1 is the loop length and k is a constant for the particular structure, yarn and state of relaxation. Whilst there is a somewhat different loop shape in these fabrics depending on state of relaxation (e.g. finishing), on the type of yarn, and on the structure (whether rub or plain), there is for a given yarn, structure or finish, a distinct loop shape, so that if loop length and finishing processes are adequately controlled, S will be consistent, and also the courses and wales per unit length (inch or cm). In other words, all three dimensions will be consistent. These dimensions can be calculated from loop length and yarn count. S, and courses or wales per unit length are independent of yarn count over the ranges normally used in apparel, and unless count itself varies, the weight and widths and lengths will be consistent from delivery to delivery, if there is adequate control of loop length and of finishing.
This is not to say that variations in yarn properties do not influence the knitted loop length. However, if machines are correctly adjusted to avoid machine variation, and knitting adequately monitored as it proceeds, loop length variation can be minimised. From fabric geometry considerations, loop length control and finishing control are vital.
Courses and Wales per cm or inch
Courses and wales per cm or inch are measured by placing an inch or centimetre glass on the fabric, and counting the number of courses and wales, which are contained within the area. The values vary if the fabric is distorted as the example shows.
The product of the two, that is the stitch density (represented by S) is less variable than either c or w alone.
Measuring Loop Length with a Course Length Tester
The length of one loop is very small and it is easier to measure the length of yarn in a number of loops. The length of yarn unroved from fabric can be measured on a Course Length Tester.
A weight is suspended on one end of the yarn to remove yarn and knitting crimp of crimped yarns or the knitting crimp of spun yarns.
The appropriate weights or tensions to be used are: -
Notes
The above tensions are quoted in B.S. 5441.
The resultant tex or count of the yarn (i.e. depending on the No. of plies) must be used.
IMPORTANCE OF FINISHING AND RELAXATION
Finishing affects the state of relaxation and therefore the loop shape. The ultimate relaxed state is the "fully relaxed" state, in which the loops are in an equilibrium state. When this is achieved - such as by washing and tumbling - no further shrinkage occurs, except felting. However, it is difficult to reach this state
However, this state is likely to be reached if the article is to be washed in service. A feature of this state is that contraction of the loop is always involved owing to the increased three dimensional curvature, (S is at a maximum) so that some area shrinkage always occurs on developing full relaxation.
Therefore the finishing processes should be designed as far as practical costs permit, to provide a relaxation which is as nearly complete as possible, if shrinkage in service (i.e. domestic washing) is to be avoided.
Subsequent processing must be designed to avoid excessive tension on the material especially with single jersey and rib structures relatively sensitive to distortion.
One of the great assets of the knitted structure is its relative extensibility, i.e. movement and distortion of the knitted loop. Whilst a high degree of recovery can be achieved after this distortion, it is not generally complete. If the tension is maintained for a period, a degree of set develops. So in this respect finishing of knitted goods needs perhaps greater control than with woven materials.
FABRIC GEOMETRY OF COMPLEX YARNS AND STRUCTURES
Continuous practical studies of various fibres and yarn types, including continuous filament textured yarns and of knitted structures more complex than plain and 1x1 rib knitting, have led to the following general conclusions: - The dimensions of a fabric knitted from a particular yarn depend mainly upon three factors -
Only in complex structures and with large changes in yarn count is there any significant influence on dimensions from count variation.
It is therefore apparent that in all knitted fabrics it is of paramount importance that consistency be maintained as far as it is economically possible in these three factors and that the spinner, knitter and finisher have a major contribution each to make to the achievement of consistent dimensions.
IN CONCLUSION It is difficult to understand how knitted goods behave on finishing and during domestic washing, because of the inter-action of so many factors.
All knitted articles comprise a number of knitted loops. The knitted loops can be regarded as the structural unit and a rectangle can be drawn to contain this unit. Whatever affects the shape and size of this rectangle, and however it is affected, will affect the size of the knitted article correspondingly. If the effect on this rectangle is known or has been determined, and since the number of these in the width of the article and in the length can be determined, the dimension of the article can be predicted.
This unit rectangle containing the knitted loop has relevant two parameters:
The shape is mainly determined by the treatment history of the article. Knitted goods are produced in the knitting machine under tension. In service they are also subjected to tensions both widthways and lengthways. The great virtue of knitted goods is their extensibility, - their ability to accommodate both shape and movement. On removal of these tensions, the structure in theory, should spring back, to its "natural" shape and dimensions. In practice friction at the various points of contact between loops - the shaded areas in the diagram - and between fibres in the yarn, precludes the complete return to the original state. During movement, scouring, and various relaxation procedures, the fibres and loops are jostled and friction progressively overcomes, enabling the "natural" relaxed shape to be gradually achieved. Even on just standing, some degree of relaxation occurs, the unit rectangle changing shape.
Diagram showing points of contact (shaded) between the dotted loop and its neighbours.
The size of the unit rectangle is mainly determined by the length of yarn in the loop - e.g. the dotted portion in the diagram, i.e. the loop length. If this increases, say, by 10% then the height and width of the rectangle both increase by 10%.
This is why it is important to control the knitted loop length to within reasonable tolerance limits.
Relaxation also affects size. Consider a fully relaxed fabric: if this is stretched width-way by 10%, then the wales per cm will fall by 10%. The compensatory increase in courses per cm is, however, much less, - between 5-8%. Thus the unit rectangle has grown in size. A similar effect occurs if the fabric is stretched length-way. The rectangle containing the loop increases in size on any stretching of the fabric. During relaxation this effect is reversed. The knitted loop is three dimensional and as the fully relaxed state is approached the yarn in the loop is accommodated in a progressively smaller rectangle.
PRACTICAL EXAMPLE
A long roll of knitted fabric was taken from a knitting machine, knitting satisfactorily from one batch of yarn and on positive feed. It was immediately cut into 5 separate lengths which were then treated as described below. A short length was cut from each of these lengths for the determination, by unroving and measuring, of the average loop length of the feeders for that particular separate length of fabric. The courses and wales per cm were determined for each separate length by counting under a piece glass in several places to obtain a representative average. The treatments and results are shown in the following table. The fabrics were handled carefully to minimise errors through distortion or premature relaxation.
Although the loop length varies by less than 1% the other measurements vary over a range of 25% or more and have no real meaning until the fabric is in a stable, relaxed condition such as fabric C). These figures will obviously also influence the fabric weight per square metre.
As a guide, we can make the following approximating statements: -
DIAGRAMS go in here
SUMMARY
The loop shape (i.e. shape of the rectangle):
The above considerations are modified by the yarn relaxation occurring after knitting. The higher the crimp effect (e.g. Shirley Tube Test figure), the smaller the unit rectangle. Also, the denser the structure the less is the effect of crimp variation. The relationship is a very complex one, usually resolved by fabric development trials.
Again, however, control of finishing, knitted loop length, and yarn consistency give the best control of fabric consistency. RELATIONSHIP OF WEIGHT PER SQUARE METRE IN THE GREY AND STITCH DENSITY IN THE GREY Frequently weight (mass) per square metre is used as a routine quality control check on grey fabric. On finished fabric this is perfectly meaningful as the fabric should have been relaxed and stabilised in finishing. In the grey, however, the fabric is still relaxing and the figure obtained will vary with the history of the sample of the grey fabric. Wt. of a roll of a stipulated no. of revs. is much more meaningful. If it is not feasible to doff rolls at a stipulated predetermined no. of revs, then weight per square metre can be used if it is in conjunction with the stitch density.
The weight per square metre is determined by cutting circles of a known area from the fabric, at different places across the width. Weighing these and then dividing by the total area of the circles gives the weight per square metre.
The stitch density of these self-same circles, must be determined in the same state as when cut, i.e. at the same time and with careful handling to avoid distortion.
The weight/sq. metre divided by stitch density will equal a constant for a particular fabric so long as it is of correct quality, and this can constitute a Q.C. check.
Example
Data: - Test fabric: -
Specification: -
Calculation of the fabric constant for the specification: -
Tolerance = + 5%, i.e. + 0.6067 x 5/100 = + 0.0303
Constant limits are 0.6370 - 0.5764
Test fabric figures: -
Mass/sq. metre 7.765 x 100 = 155.3g 5
If + 5% tolerance is applied here with no regard to the fact that fabric may have relaxed, 5% tolerance on 145 is 152g i.e. fabric is regarded as too heavy.
Stitch density from specification is 43.5 x 49.5 = 239.2
Stitch density from test fabric 42.2 x 54.4 = 256.2 3 3 which is high
Therefore: - Calculate fabric constant: -
155.3 = 0.6062 which is within tolerance limits. 256.2
Therefore, fabric is satisfactory for weight and not over weight as would otherwise have been deduced. By the same token, fabric can be registered as correct when in fact it is incorrect for weight. Again calculation of the fabric constant will protect you from this.
THE NEED FOR YARN TENSION AND ITS EFFECT ON FABRIC QUALITY AND ON KNITTING PERFORMANCE.
THE NEED FOR YARN TENSION
Because yarns for apparel are highly flexible materials, the only way of keeping them under control is by keeping them taught. In the absence of tension the yarn is uncontrolled and would fail to locate in the needle hook and to knit.
LATCH NEEDLE KNITTING - CIRCULAR MACHINES
Here a low yarn tension before the feeder, - say 3-5g - is used. However, from the yarn package to the point of maximum knock-over depth, the total angle of wrap (i.e. bends in the yarn path) may be of the order of five complete turns even on a single cylinder machine and higher on a dial and cylinder machine.
DIAGRAM LATCH NEEDLE KNITTING SEQUENCE
At each one of these contact points friction causes a build-up of tension. Also variations in the extensibility (modulus) of the yarn cause variations in tension as the yarn loop is drawn by the needle down to maximum knock-over depth. The tension itself is not critical unless it is in the region of the breaking load of the yarn. It is the degree of variation that matters, and we are depending on the quality of the yarn package.
The effect of tension variation can be understood by reference to needles, P, Q, R and S in the diagram. The former two needles are moving down to maximum knock-over depth and therefore taking yarn into the knitted loop. The latter two needles are moving back to the rest position and therefore spare yarn is available. This yarn is taken up by needles P and Q which therefore, in forming the knitted loop, take yarn partly from the yarn package (feeding) and partly from the loops in needles R and S (robbing-back). An increase in take-down tension reduces the degree of robbing-back and an increase in input tension increases the degree of robbing-back markedly.
The take-down tension should be set to retain the knitting on the verge. The yarn input tension is as light as possible consistent with adequate yarn control. This keeps needle and cam wear to a minimum and reduces the occurrence of yarn breaks. The knitting loop length control is effected by adjusting the knock-over depth. However any significant change in yarn input tension will inevitably change the loop length. Therefore in developing the knitting specification for the fabric, it is important to select yarn for sample knitting that is believed to be as nearly average as possible. The usual 2-3% tolerance will then absorb much if not most of the yarn input tension variation. Positive feed avoids the effect of yarn tension variation on those feeders on which it can be employed. Many types of yarn storage feed, remove off-wind tension variations and reduce variations due to friction.
LATCH NEEDLE KNITTING - FLAT MACHINES
Yarn tension and take-down tension have a similar effect on the knitting, but there is an additional factor. This is the yarn take-up spring, which has to be set up to maintain yarn tension at each carriage direction reversal. The disk tension is set up to prevent the take-up spring drawing yarn from the package in between knitted courses. The yarn input tension is therefore much higher than for circular knitting, e.g. 10-20 grams. To off-set this, it is desirable that the friction of the yarn is low. Also, it is desirable for it to be uniform, since positive feed is not available.
Yarn storage feeds or other devices can be fitted to reduce yarn tension variations.
STRAIGHT-BAR (FULLY FASHIONED) KNITTING
No robbing-back occurs, and therefore the knitting is less variable, and take-down has very much less effect on the course length. Increasing yarn input tension will, however, reduce the course length knitted. This is shown in the diagram. The action of coarse gauge machines (15g and below) is relatively simple and will serve to illustrate the effects of yarn tension.
DIAGRAM
As the yarn tension increases, the pressure of the yarn against the needles is much tighter and the course length knitted shorter. A further increase in tension causes the needle stems to bend towards the plane of advance of the sinkers, again reducing the course length knitted.
EFFECT OF KNITTING MACHINE SETTINGS ON YARN TENSION
Generally with all weft knitting, the longer the loop length knitted, the more tension is generated. Because of the increased number of yarn/knitting element contact points encountered, higher friction is created.
If the yarn breaks, it is the loops in the new course where rupture occurs, resulting in a fabric hole and a local press-off.
Where positive feed is employed, the cams are adjusted to achieve the correct input tension and not stitch length, this being controlled by the positive feed. The correct tension immediately after the positive feed device is one slightly greater than that occurring just before the device. Failure to achieve the proper tension can cause a number of problems through high tensions or inadequate yarn control at the needle.
With double-bed latch needle knitting, delayed timing, - in which the back-bed needles travel to maximum knock-over depth while the front-bed needles are returning to their rest position, - can help to reduce the knitting tension, because yarn is not being drawn into both sets of needles simultaneously. The knock-over depth of those needles first supplied with yarn must be sufficient to furnish yarn for both needles, otherwise tensions will increase rather than diminish with delayed timing. The correct settings of knock-over depth and timing for a given course length and positive feed setting are ascertained by experience with a particular type of machine.
When the course length knitted is very short (tight knitting) the action of drawing yarn though the previous loops also demands more energy and gives higher tensions due to the higher friction generated.
With positive feed, the cams are adjusted to achieve the correct input tension.
When knitting tightly, if the yarn breaks it is generally the loop in the previous course, (the course being knocked-over) which ruptures, resulting in a hole but not a local press-off.
With finer gauge machines, the effect is substantially the same, if the appropriate count of yarn for the gauge of machine is used. However, the yarns generally are weaker at finer counts, and therefore breaks will occur at lower tensions.
SOME IMPORTANT FACTORS AFFECTING LOOP LENGTH
Even is positive feed is used, problems can sometimes arise in maintaining loop-length control.
High yarn tension can arise from: -
If the tape surface is too smooth, yarn will slip and positive control ceases. A checking and replacement schedule is needed. The frequency of the schedule will largely depend on the yarn being knitted.
Regular checking of the course length with a yarn length meter, and noting the positive feed setting, will indicate when the feed is incorrect and needs changing.
In a well run installation these checks need not be frequent (say monthly) but must not be overlooked.
The instrument used to check the positive feed can be wrong through any of the following causes:
Looking at each in turn: -
The positive feed checks the knitting; the instrument checks the positive feed; the unroved course length checks the instrument. Thus while no individual check is totally reliable, one can trust the total system of checks. Also, on this basis, the need for checking the instrument by unroved course length checks becomes very infrequent, say, every 3, 6 or possibly 12 months, yet must not be overlooked.
The procedure is as follows: -
Each instrument in use and yarn type and source should be checked at regular intervals. Say there are 2 instruments, 10 different yarn types, on average from 2 spinners then, checking each combination say once a year, would involve 40 checks a year. Experience would teach which type of yarn is likely to give inaccuracies, and so a priority order can be made.
When starting the check, choose an accessible feeder, on a machine with as few a number of feeders as possible, and mark the yarn with a little contrast coloured thread.
Insert the instrument into this yarn path. Run the machine and note the yarn length over a set number of revs. Repeat a few times to ensure the reading is consistent. If not, the instrument must be serviced.
Calculate the course length reading from the average yarn length reading and the number of revs.
Stop the machine and remark the chosen yarn.
Knit the fabric through, and remove fabric including the two marked courses in the fabric cut off.
Unrove, discard the unmarked feeders, and determine the course length as usual. Take a number of readings to ensure the reading is reliable.
Compare the unroved course length with the reading from the instrument.
If virtually the same, no action.
If significantly different (e.g. more than 0.5%) this may be due to factors not affecting the instrument, for example the knitting tension and the tension used in the unroving test being different. So long as the % difference between the instrument reading and the unroved course length remains constant, this is allowed for and there is no problem. Check if the observed % difference is different from that observed in the previous check with this yarn. If not, then no problem. If different get the instrument serviced.
Use of a count testing wrap-reel.
If a wrap-reel is at hand, a much simpler procedure is available as follows: -
Place a cone of yarn being knitted on the count tester's creel, threading-up as for count testing, and adjust the reeling tension to that used in course length tests by adjusting the reeling tensioners and checking the tension with a tensiometer. The reading on the tensiometer will fluctuate quite wildly, but observation of the reading over a period of a few seconds will enable the average figure to be estimated, and the tension adjustment is then made to bring the figure to that required with this yarn. The wrap reel is now zeroed and the yarn length counter inserted into the yarn path at the same place as where the tensiometer has been inserted. Make a known number of wraps (e.g. 100) and note the instrument reading. If the reel has a metre girth, the instrument should read 100 meters if the girth is 1 yard it should read 100 yards. If not, then the instrument is giving incorrect readings.
Repeat the reeling to ensure that there has been no error in reeling.
Because the needles are drawing yarn from the packages into the knitted loop while moving down to maximum knock-over depth, it is possible for yarn to be also drawn into the loop from the previous adjacent needle while this needle is moving back to the rest position. This latter affect is called robbing back.
If the yarn tension increases, the needle moving to maximum knock-over depth will take more of its yarn from the previously formed loop so increasing the degree of robbing back and giving a shorter course length.
If the take-down tension increases, this tension will apply to all the needles, but with the needle moving back to rest, the tension will be mainly on its newly formed loop because it has just knocked-over. With the needles moving down to maximum knock-over, the tension will be mainly on the old loop, until the new loop knocks-over. Therefore the increased take-down tension tends to increase the yarn tension in the loops after reaching maximum knock-over and this reduces the degree of robbing back, giving a longer course length.
Take-down tension should be set to be minimal, to reduce wear, to suit the particular knitting machine, i.e. control the knock-over of the old loop and to hold the knitting plane onto the verge.
Yarn tension should also be minimal to reduce wear and the risk of yarn breakage and hence machine down time, but be sufficient to control the yarn adequately, i.e. to effect correct loop formation.
WITH STRAIGHT-BAR (FULLY FASHIONING) MACHINES
Here an increase in yarn tension can: -
Because the loop length is, in effect, inserted before knitting, take-down, operated mainly on the knocked over loops and has less significant effects than with latch needle knitting. However, an increase in this tension can still have some effect, as it will also tend to flex the needle stems towards the plane of advance of the sinkers, i.e. reduce the distance between the stem and the advanced sinker. Thus with straight bar knitting, increased takedown tension is likely to reduce the course length. Again take-down should be adjusted to suit the machine.
FACTORS AFFECTING THE YARN TENSION
The above effects leave us to consider factors influencing yarn tension. These can be itemised as below: -
It can be tested for, along with count testing, quite simply.
An increase tends to increase the friction (but not always) and reduces the modulus. That is why if the yarn is too dry it is difficult to knit. This is best controlled by buying from a spinners and storing yarn in undamaged, unopened containers in a dry area not subject to vast changes in temperature.
CORRELATION AND CALIBRATION OF CERTAIN INSTRUMENTS FOR KNITTING Q.C.
Pulley, i.e. mechanical tensiometers, such as Zivi, Mercer, Schmidt are quite suitable for checking yarn tension in machines where there is a reasonably steady run-in of yarn, e.g. in circular knitting.
Whilst experienced knitters and mechanics can level the tension between adjacent feeders by manual assessment ("fingering"), it is found that such assessments are unreliable to level over a suite of machines. One can say the "finger" levels the feeders and the tensiometer keeps the finger level.
However, the tensiometer can become unreliable, through wear, misuse, accident, or lack of use, and this should be checked periodically, - say once a month. This is carried out by: -
Generally the reading of a yarn length counter will not drift. However, once in a while it is as well to confirm it. The positive feed checks the knitting: the instrument checks the positive feed; the unroved course length checks the instrument. While no individual check is totally reliable, one can trust the total system of checks. The need for checking the instrument by unroved course length is very infrequent, say, every 6 or 12 months, MUST NOT BE OVERLOOKED.
The procedure is as follows: -
Each instrument in use and yarn type and source should be checked at the agreed frequency. Say there are 2 instruments, 10 different yarn types, on average from 2 spinners then, checking each combination say once a year, would involve 40 checks a year. Experience would teach which type of yarn etc. is likely to give inaccuracies, and so a priority order can be made.
Choose an accessible feeder, on a machine with as few feeders as possible, and mark the yarn with a little contrast coloured thread.
Insert the instrument into the yarn path. Run the machine and note the yarn length over a number of revs. Repeat a few times to ensure the reading is consistent. If not, the instrument may not be properly in the thread path.
Calculate the course length reading from the average yarn length reading and the number of revs.
Stop the machine and remark the chosen yarn.
Knit the fabric through, and remove fabric including the two marked courses. Unrove, discard the unmarked feeders, and determine the course length as usual. Take a number of readings to ensure the reading is reliable.
Compare the unroved course length with the reading from the instrument.
If virtually the same, no action.
If significantly different (e.g. more than 0.5%) this may be due to factors not affecting the instrument, for example the knitting tension and the tension used in the unroving test being different. So long as the % difference between the instrument reading and the unroved course length remains constant, this is allowed for and there is no problem. Check if the observed % difference is different from that observed in the previous check with this yarn. If not, then no problem. If different get the instrument serviced.
Use of a count testing wrap-reel.
If a wrap-reel is at hand, a much simpler procedure is available as follows: -
Place a cone of yarn being knitted on the count tester's creel, threading-up as for count testing, and adjust the reeling tension to that used in course length tests by adjusting the reeling tensioners and checking the tension with a tensiometer. The reading on the tensiometer will fluctuate quite wildly, but observation of the reading over a period of a few seconds will enable the average figure to be estimated, and the tension adjustment is then made to bring the figure to that required with this yarn. Zero the wrap reel and insert the yarn length counter into the yarn path in place of the tensiometer. Make a known number of wraps (e.g. 100) and note the instrument reading. Thus if the reel has a metre girth, the instrument should read 100 meters. If not, then the instrument is giving incorrect readings.
Repeat to ensure that there has been no error in reeling.
2.1 CALCULATION OF THE CORRELATION FACTOR BETWEEN THE YARN LENGTH COUNTER AND THE COURSE LENGTH TESTER
If in the correlation checks, the instrument is found to be quite consistent, i.e. the readings are consistent within themselves and with the results obtained on the previous checks the instrument can be regarded as satisfactory even if those readings are not identical with the figures obtained on the knitting machine during the course length check.
Some result must be regarded as standard, since the results on the Course Length Tester or Wrap Reel are carried out at a defined tension. On the knitting machine, the tension when forming the loop is not defined, so the latter cannot be an absolute standard. We therefore relate all results to that obtained in the course length test.
To calculate the % correlation factor: -
If L is the average course length figure from the course length or wrap reel test, and L1that obtained on the knitting machine then: -
corrected figure is actual reading x 103.86 - 100 3. Yarn Speed Meter Correlation Some Yarn Speed Meters can be switched to yarn length counting, and since it is the same pulley drive, these can be checked as yarn length counters as in 2 above.
Otherwise it is difficult to check a yarn speed meter, since yarn speed is a difficult parameter to measure. Yarn speed meters can be checked, but only
a) against yarn length counters (either a separate instrument or by switching to the yarn length counting mode if this facility is available).
or
b) as course length meters, if the actual speed (no. of revs.) of the machine can be determined.
Example A - Use against yarn length counter.
Insert the yarn length counter in an active feeder and note the yarn length reading over, say, 10 courses. This reading - 10 equals the course length. Immediately insert the yarn speed meter in the same feeder at the same point and with the same configuration of the yarn path. Note the reading. The machine speed will not have changed in this small interval of time. Use this figure of the yarn speed to level all the corresponding feeders around the machine, to within the agreed tolerance (e.g. + 2%).
Data: - Yarn length per 10 revs. 43.8 metres
Upper limit of yarn speed = 1314 + 1314 x 2 = 1340 cm/sec. 100 Lower limit of yarn speed = 1314 - 1314 x 2 = 1288 cm/sec. 100
Note the time for 100 revs. or no. of revs. in 1/2 minute. preferably 1 min. Repeat till a consistent figure is obtained. From the specified course length and tolerances, calculate the upper and lower limits for the yarn speed.
Data:- Specified course length 4.38 metres = 438 cm
Upper tol. 129.1 + 129.1 x 2 = 131.7 cm/sec
Lower tol. 129.1 - 129.1 x 2 = 126.5 cm/sec
There are two main reasons why these two instruments could give slightly different readings even in the hands of skilled operatives.
The resultant discrepancy could vary from yarn type to yarn type, and from yarn unroved from grey, steamed, scoured or set fabrics, but within any one of these categories should be constant. Therefore even in situations where there is a discrepancy between readings from these two instruments, this can be resolved by correlation.
Example: -
Data: 22g interlock knitted from 150 dtex or 1/67's metric cotton spun polyester.
Av. course length - Hatra Course Length Tester 9.14 m Av. Yarn length/100 m - Shirley Crimp Tester 38.05 cm Machine diam. 36", no. of feeders in cyl. 2484
Course length from Shirley Crimp Tester = 38.05 x 2484 = 945.2 cm
% correlation of Hatra figures/Shirley figures is 9.14 x 100/9.452 = 96.7%
Shirley figures = Hatra figures x 100/96.70.
The method often used to zero the instrument is to stand the instrument on a flat surface on the machine and rotate the scale until the pointer reads zero. Although the foot be pressed firmly on the surface slight rocking of the instrument causes significantly different readings to be displayed.
To avoid this, a calibration jig is made. It comprises a horizontal and a vertical plane arranged like a capital "L". The back of the instrument is pressed against the vertical face of the jig, and the foot of the instrument is pressed against the horizontal face. The scale is rotated to give a readout of zero.
To avoid any possible skewness of the dial of the knitting machine causing inconsistent readings to be obtained, the instrument is always inserted at the same dial segment on any given machine.
One of the main reasons for using the Knock-over Depth Gauge is to level the needle movement over every feed, between dial and cylinder (or back and front beds) of every machine knitting to a particular specification.
Normally the tolerance on knock-over depth is + 0.1 mm. However, to set realistic tolerances, the stitch length is determined at a particular knock-over depth with the yarn and machine gauge concerned. The cam settings on both beds are then readjusted by a uniform amount until a stitch length just out of tolerance results, and the new knock-over depth is noted. The difference between this and the original knock-over depth gives the maximum limit of tolerance.
In order to reduce the cumulative effect of variables, it is a wise precaution to endeavour to apply closer tolerances than that determined above, say, half.
Example: -
Data: Original course length 244 cm, + 2.0%
ESTABLISHING CHECKS ON YARN DELIVERY WEIGHTS
Before materials control is secured, yarn delivery weights need to be continuously monitored. In the absence of checks serious losses may be incurred without their magnitude being realised. Accurate costings are impossible in these circumstances.
All yarn entering the factory should do so over a weigh bridge or scales. Thus the gross weight of every case of yarn is known and is entered on a tally sheet by the Stormed and attached to the Delivery Note.
Example Tally Sheet
Spinner Yarn Description Entry No. Advice No. Date
--------------------------------------------------------------------------------------------------------------- Case No. Gross Wt. Case No. Gross Wt. Case No. Gross Wt.
The tare weights given by the spinner on the Advice Note are rarely incorrect and can be used to obtain the next weight as received of each case of yarn. If any tare weight is abnormal a Tare Check is undertaken. The contents of the case concerned are knitted, all wrappings, bags, core formers, case liners etc. being reserved and weighed. This overall weight is the correct tare weight.
The tare wt. of each case is deducted from each gross weight to give the nett weight of each case, as received.
This is checked against the Spinner's declared nett weight.
If, over the delivery, there is a discrepancy above an agreed amount, a claim is made. The discrepancies are recorded as a trend graph, to keep a cumulative check as the contract continues.
If the yarn is hygroscopic (e.g. has a regain of 5% or over) such as wool or cotton and their blends with other fibres, before a claim is made, the tally, and advice note with yarn discrepancy note are sent to the lab.
Those cases in the delivery showing the greatest discrepancies are sampled - say, 2 cones per case - and the % correct weight is determined of each cone. If the results are in poor agreement 2 further cones are tested and the overall average taken. The four cones coming from the top, middle and bottom of the cone. If there is poor agreement between the cases, further cases are sampled. From this the % correct weight figure is calculated by which the received nett weight of each case is adjusted to give the nett weight at correct condition.
A conciliation for the whole delivery is made between the case nett weights so found and as declared by the Spinner. From this the final discrepancy is calculated, and claims are based on this.
If the yarn is oil spun, the samples taken for condition testing are secured free of oil and then oven dried to get the clean over dry weight. This is adjusted by an agreed figure based on this weight, representing the adjustment for both oil and water. In securing, the liquor is poured off through a fine gauze filter and any lint recovered.
Example Condition Test Form
-----------------------------------------------------------------------------------------------------------------
CONDITION TEST
Spinner Yarn Description Entry No. Advice No. Date ------------------------------------------------------------------------------------------------------------------ Case Sample Weight Agreed % correct Nett Weight
Total Nett Weight at Correct Condition (kg)** Invoice Weight (kg) Discrepancy (kg) -------------------------------------------------------------------------------------------------------------------
% Correct Weight or "wt. adjustment factor" is calculated as follows: -
Sample wt. oven dry x (100 + % regain) Sample wt. as received
** Nett weight at correct condition is calculated as follows: -
Nett wt. as received x % weight adjustment factor - 100
The regain to be used in the above calculations is the standard regain or the "% addition to the dry clean mass or wt." as given in and BS 4784: 1973 depending on the fibre and method cleaning and if cleaned. With fibre blends the resultant % regain allowance, or % addition to the dry mass, is determined from the figures for each component fibre and the declared or determined composition of the yarn.
These checks, together with count tests (on staple fibre yarns) check that the correct amount of pure fibre at the right count is being received.
Both weighing in and count testing are necessary before this is secure.
The count testing should be carried out under the same conditions as that employed in weighing in - viz.
The results should be recorded on a trend graph or on a frequency distribution chart (i.e. a histogram), as count and yarn discrecpancy respectively.
The state of affairs can then be assessed quickly at any time as the contract proceeds. The overall or cumulative average is the most important single check in this respect, because the deviation of this from the costed average gives the picture of any overall loss or gain developing during the contract, so avoiding contracts and knitting programmes terminating with deficiencies in yarn. This cumulative average is readily determined from a histogram because the peak occurs sufficiently near to the average for routing purposes.
RECORD SHEET FOR COUNT TESTS
DIAGRAM
RECORD SHEET FOR YARN DISCREPANCY
DIAGRAM
The above scheme operates successfully in the firms where it has been installed. It should be appreciated, however, that agreement with the spinner would be required before installing such a scheme.
TOLERANCES, THE AVERAGE, STANDARD DEVIATION AND THE SPECIFICATION INTRODUCTION
All products are subject to variation militating against consistency, due to variations in raw materials processes, operatives, conditions, measuring etc. The likelihood, therefore, of a product always being spot-on, is very remote. However, the variations are just as likely to occur giving results too high as too low, - the results will vary in a random manner. This variation is not chaotic; it obeys a definite law of nature which we can exploit. This law is so fundamental to an understanding of what Q.C. is all about, that there is wisdom in a brief study, even when one is less heavily involved with measurements and their tolerances than with faults and their assessment or comparison against standards.
LAW OF RANDOM DISTRIBUTION
This is the law to which we are eluding. If repeated measurements of any particular factor are taken as production proceeds, the results will be found to cluster around a figure we call the average, with a spread of results on either side with diminishing frequency as we move away from the average. In fact if we draw a graph for a particular set of measurements, - i.e. plot the results - with the range of spread (from well below the average to well above) shown horizontally, and the frequency or number of occasions of a particular figure occurring as a result, we get a curve like that shown below, to which the vertical height or frequency increases rapidly to a maximum as we approach the average and then falls off again as we pass beyond this. It is for this reason, as we shall see later, that the specified tolerances ought to bear some relation to this peak in the curve.
DIAGRAM
TOLERANCES
Now if we fix limits within which we will accept a product as satisfactory, then clearly, the further apart and on each side of the average we set these limits or "tolerances", the greater will be the proportion of production that is accepted.
From the above curve, a position can be ascertained at which we can set the tolerances, so that the seriously wrong articles can be rejected while retaining the minimum reject rate. If tolerances are planned so as to occur at the positions equivalent to c and c on the graph, it can be seen that the curve flattens out quickly at this point as we move further away from the average and steepens quickly if we move towards the average from c. Here generally is the optimum position for the tolerance limits, and here 95-96% of the production becomes acceptable, 2% rejected with measurements regarded as too low and 2% rejected with measurements too high.
Conversely, if tolerances are fixed, e.g. by agreeing on a level of acceptance, then the reject rate ensuing can be calculated and accurately costed into a process.
STANDARD DEVIATION
The spread of results will affect the curve, - from being a tall narrow hump, the ideal situation - to being a low flat lump, an undesirable situation. The Law of Random Distribution enables us to define what governs particular curve or distribution of results. There are, in fact, only two factors, - viz. the average and a factor called Standard Deviation. This factor can be calculated from the curve or from a set of results, and is important because many decisions about measurement, such as location of tolerance limits, stem from a knowledge of standard deviation.
The higher the figure the greater the spread of results, and it is calculated in the same units as the results or measurements to which it refers, or as a percentage of the average of the measurements. It therefore represents a certain zone of spread or deviation area, in which a certain proportion of the results will always occur. In the curve on the previous page zones have been drawn at a, b, c and d, each side of the average equal to 1 x SD, 2 x SD 3 x SD and 4 x SD and the zones marked 1 to 4. The proportion of the results always occurring between the limits bounded by these zones is also shown. It can be seen that where we discussed locating the tolerances was, in fact, at a position equal to 2 x SD.
Example to show the principle behind calculations of SD - I the SD of a set of results from a sample of measurements.
Consider two deliveries of yarn of nominal count of 2/32's: -
A
Test results Deviation Square of deviation of count (x) (x-x) (x-x) 2
32.3 max 0.3 0.09 32.0 0.0 0.00 31.9 -0.1 0.01 32.2 0.2 0.04 Standard deviation 32.0 0.0 0.00 of test results. 31.9 -0.1 0.01 31.9 -0.1 0.01 = 0.028 32.2 0.2 0.04 31.8 min -0.2 0.04 = 0.17
31.8 -0.2 0.04 Mean 32.0 = x Total 0.28
Test results Deviation Square of deviation of count (x-x) (x-x) 2 32.4 0.4 0.16 32.1 0.1 0.01 31.5 min -0.5 0.25 32.2 0.2 0.04 Standard deviation 31.6 -0.4 0.16 of test results. 31.9 -0.1 0.01 32.6 max 0.6 0.36 = 0.134 31.5 -0.5 0.25 32.3 0.3 0.09 = 0.37
31.9 -0.1 0.01 Mean 32.0 = x Total 1.34 Mean 0.134
II the SD over a total delivery or contract
The best estimate of the standard deviation of total delivery (total population) is obtained by dividing by one less than the number of test results. This number is 9 in the above two examples, since the total number of tests were 10 for each delivery. Therefore the standard deviation of the deliveries are: - A 0.28 = 0.033 = 0.18
B 1.34 = 0.149 = 0.39 9
Standard Deviation The calculation of standard deviation can be expressed for any measurements and any number of test results.
Notation x an individual measurement x average or mean of a set of measurements total n number of measurements in the set sd standard deviation Calculation
ACTION ON MEASUREMENT IN OR OUT OF TOLERANCE
Tolerances can be used to decide courses of action.
If the machine average drifts, the machine will be producing work with an abnormally high reject rate. Tolerances will indicate when adjustment is needed and by how much.
DIAGRAM
The right hand curve represents work from a machine whose average has drifted. The increased work out of tolerance can be seen.
In order to contain and check any such drift without the dangers of over correction, the required action can be summarised as follows:
1. If work is within tolerance: -
normally no action (see under 3 below).
2. If work on a check is out of tolerance: -
*Before making any actual adjustment, however, it is wise to check that no obvious fault has occurred - e.g.: yarn path dislodged, dirt or lint caught -, and to rectify this and recheck. If still out, then make the appropriate adjustment.
3. If work on regular checking is consistently within tolerance but near one of the limits: -
Although a certain percentage of work is always going to be outside the tolerances, - depending on their location compared with the spread of results, - the whole point of a Q.C. scheme is to prevent, as far as is practical, any drift in the average what so ever. In other words, there is no tolerance on the overall average.
EXPLOITATION OF STATISTICS IN THE DEVELOPMENT OF THE SPECIFICATION
In the above section we have seen how keeping the average of the knitting machines under control can effectively reduce the volume of work out of tolerance. We are in effect reducing inter-machine and inter-feeder variation.
This can be taken a step further and we can come closer to the adage "being right first time" with its implied message of giving a further reduction in waste through work being out of tolerance. If we could select yarn for a machine in use for sampling that is "average" in its properties, and again do this when first setting up knitting machines on a production run, we would have all the machines knitting correct work, so long as the yarn was average. This would be marvellous!
We know however that yarn has a degree of variability, even when bought from a reliable source, and that this variability can be frequent (e.g. dye lot changes). Therefore machines known to be correctly set up can suddenly knit incorrect fabric and that the changes can swing rapidly. In this situation, the yarn can change by the time you have correctly adjusted the machine so you are "chasing your tail", and yet if you don't attempt machine adjustment you can be in trouble. So what should you do?
Yarn bought from a reliable source, though variable, will have a reliable average. Long term variation is likely to be under control.
General experience, coupled with tests and pilot knitting trials will indicate what yarn (blend or colour) knits "average" i.e. neither tight nor slack. It is a good plan to liase with the spinner here and to tell him of your plans, and he will respond because it is in his interests also to maintain satisfaction.
This "average" yarn should be selected for sampling and for setting up production machines.
The yarn properties (e.g. count, coefficient of friction, twist) should be ascertained and incorporated into the yarn specification, i.e. the properties of this "average" yarn.
The sample machine settings should be ascertained and noted and these incorporated into the knitting specification.
If production machines are of a different type from the sample machine, from experience and pilot knitting any variation in the settings from those of the sample machine are ascertained and entered into the knitting production specification.
Now from our study of random distribution and fixing of tolerances for yarn and for knitting, 95-6% of the work WILL BE correct from all the machines and all the yarn covered by these specifications provided and only provided we have taken the trouble to set up machines to be "spot-on" with average yarn.
Only 4-5% of the work will be found to be out of tolerance and even here there is a choice, because knitting machines, kept in good order, with the yarn path clear, do not change. If out of tolerance we can: -
All such adjustments should be noted, so that the machine can be adjusted back to the average settings as soon as this yarn is knitted, and the trouble taken in getting rid of inter-machine variation will not have been lost.
The principle of this concept is that huge amounts of time, "admin", "aggro", yarn waste and lost production are involved through incorrect knitting due to inter-machine variation and yarn variation, and yet this could be reduced to the statistical 5% simply by sampling and setting-up on "average" yarn.
The trouble and cost of specifying and selecting "average" yarn is likely to be more than off set, by the reduction in waste during production. The attraction is that the extra time in sampling and setting-up is compensated by time saving in the reduction of production problems, so that the knitting order is completed on time but with less waste. Personnel take more time and trouble in sampling and setting-up to save more time and trouble in production, and the firm saves money on raw materials.
This is the attraction, and therefore there is a strong argument to put in such schemes on a trial basis.
RESULTANT STANDARD DEVIATION OF SEVERAL VARIABLES
The resultant standard deviation, although increased, is not the sum of the standard deviation of each variable. It is less, and is given by the equation:
Example: 3 variables of S.D. 1, 1½ and 2 respectively.
Resultant S.D. = 1½2 + 1+ 22 = 7.25 = 2.77
and not the straight sum, viz.:
1 + 1½ + 2 which = 4.5
This principle is important in calculating the resultant S.D. and hence tolerance limits on such parameters as weight per doz. and fabric roll weight per set no. of machine revs.
PRINCIPLES OF LOOP LENGTH CONTROL The action/no action sequence of routine checking in a control scheme will follow the flow diagram below. Routine Measurement And Record Inside limit outside limit Just outside well outside No action Take second Inspect yarns Measurement and machines And record No abnormality abnormality Inside outside Limits limits Rectify, recheck and record Next adjust machine routine and recheck check until within outside and limits, and limits recording record Inside limits
Experience has shown that limits set at +- 2% of the average are generally satisfactory. However, in the absence of pre-fixed limits (e.g. commercial directories or customer stipulation), when experience of a control scheme has been gained, a good guide is to set limits so that about 4-8% of routine checks come out of tolerance on the initial measurement of each routine check.
When an Electronic Course Length Meter or other accurate instrument is available for checking knitting quality during knitting - i.e. on the knitting machine, a modification to the above scheme can give much better control with the consequent likelihood of a reduction in the number of times a machine needs adjustment. Routine Measurement And Record Inside limit outside limit Just outside well outside Take second Inspect yarns, No Measurement machines and Action And record fabric No abnormality abnormality Inside outside rectify, Limits limits recheck And record
Next Adjust machine and record:- outside Routine 1. If average of all feeders limits Is outside limits, adjust All feeders by same amount To bring average inside limits inside Limits 2. If only certain feeders are outside limits, change packages, and recheck. Retain such packages for rejection or knitting speciality. By this technique, machine adjustments are kept to a minimum consistent with maintaining knitting quality. Rogue yarns can be identified by exchange of positions with yarns on other machines knitting satisfactorily and noting effect on fabric and course length at the feeder with the rogue yarn. They can be collected for knitting especially at the end of the order or for rejection.
Experience will give guidance on the frequency of routine checks and when to increase vigilance, e.g. with yarns of known difficulty. Close attention immediately after setting up machines, and especially at yarn changes will enable machines gradually to be corrected until they knit correctly with average yarn. The number of times a machine needs adjustment will fail.
As a consequence of the machines knitting correctly with average yarn, those packages of yarn which now fall on the upper or lower tolerance limit will produce fabric within tolerance, whereas previously there was a much greater probability of the fabric being out of tolerance with those packages.
LOOP LENGTH AND QUALITY CONTROL ON HALF-HOSE MACHINES
Loop length control together with machine links checking, are used to control the production of half-hose machines. The step-by-step procedure for this on these machines is described for: -
1. YARN LENGTH COUNTER ONLY To set up machine
l. Check goods coming off machine for wt/doz. pr. off machine.
Routine control Work through items e. - h. as frequently as necessary to keep the production within required limits. Item l., used regularly, provides a valuable routine check on overall stitch length, correct links and correct yarn count, in combination.
To set up machine
Routine control
Work through items e. - i. as frequently as necessary to keep the production within the required tolerances. Again item j. provides a valuable routine check if used regularly.
To set up machine As for a. - k. in 1. Yarn Length Counter only.
Routine control Work through items e. - g. for 2. Lateral Stretch Device, as frequently as necessary to keep the production within the required tolerances. Again, item 2.j. provides a valuable routine check if used regularly. If measurement is outside the tolerances, check on machine using the Yarn Length Counter, and proceed as already described for setting up. Specification must include figures with tolerance ranges determined by both instruments. Make routine checks as frequently as necessary to keep the production within the required tolerances.
4A. ELECTRONIC COURSE LENGTH METRE ONLY To set up machine
Routine Control Work through items b. - e. as frequently as needed to keep production within the required limits. Item k., used regularly, again provides a valuable routine overall check on stitch length, correct links and yarn count.
4B. ELECTRONIC COURSE LENGTH METER AND LATERAL STRETCH DEVICE
To set up machine
As for a. - j. in 4A. Electronic Course Length Meter only.
Routine control
Work through items e. - g. for 2. Lateral Stretch Device, as frequently as necessary to keep the production within the required tolerances. Again item 4A.k, provides a valuable routine check if used regularly.
If measurement is outside the tolerances, check on machine using the Electronic C.L.M. and proceed as already described for setting up, items a. - e.
Specification must include figures with tolerance ranges determined by both instruments.
Make routine checks as frequently as necessary to keep the production within the required tolerances.
Notes
If the machine being set up is cold, re-check when warm. Routine checks must be made on warm machines.
The stud and plain links on each machine must also be checked as a routine. Weekly or perhaps even monthly checks may suffice.
For jacquard machines the machine may be set up with plain knitting using the Yarn Length Counter, or, by taking care to measure in the same pattern area, a satisfactory scheme may be developed using the Lateral Stretch Device.
There is an optimum frequency at which routine checks should be made. This will be determined by the conditions existing on a given plant. However, wt./doz. pr. provides a simple valuable routine check.
This optimum frequency of checks can be assessed by the process of Analysis of Results.
LOOP LENGTH CONTROL ON JACQUARD MACHINES
Loop length cannot be used as the normal parameter for knitting quality because the jacquard patterning gives varying lengths of the cross ties, i.e. that part of the yarn interlooping between the dial and cylinder, due to the patterning on the face of the fabric.
Loop length can be used in setting up a machine, by bluffing out the jacquard motion during the check. The actual yarns required in the ultimate fabric must be used although the fabric produced during the test will be a waste. Thus the method is not suitable for normal routine checks. It is used, however, in setting up a machine.
For routine tests, recourse is made to weight per square yard (or square metre).
The procedure recommended is described with illustrations in Knit Knacks April 1971, Vol. 16 No. 1.
If machine is rotating cam box type, f. and g. will be effected (i.e. levelling cylinder cams to give correct loop length at each feeder) by the technique detailed in QK18 Rot. C/B M/c part 2, d. to j.
h. Restore the jacquard pattern and commence knitting.
2. Routine Control Scheme
b. Examine fabric for bariness.
a. and b. are better checks.
The frequency of these checks is ascertained from analysis of results over a period.
If any machine goes slightly out of tolerance, check yarn path and tensions. Rectify as necessary, followed by a weight check of a set number of revs. approximating to a yard or metre length of fabric. If still out, make a small adjustment to dial height, followed by a weight check as above. Repeat until weight checks come within tolerance.
If any machine is grossly out, the above procedure, under "Commencement" c. to h., must be undergone.
LOOP LENGTH CONTROL ON STRAIGHT BAR MACHINES
Loop length control is the technique used to control the production of knitting machines.
The step by step procedure for this on the above machines is as follows: -
a. Select specification for the garment to be produced.
b. Check that corrects cards and trims, correctly doubled on transfer bars, are set up.
e. Adjust needle bar until the correct stitch length is obtained, and then set machine onto knitting programme.
The following techniques are used in running a Q.C. scheme: -
a. Inspect fabric for faults and unlevelness.
Check a. and b. above are the most important routine checks. c. should also be carried out periodically.
If measurements come out of tolerance, check any or all of the following as appropriate: -
Stitch length as under c. above.
Yarn tension, yarn path for obstructions.
Yarn count etc. No. of courses and wales in each part.
Adjust as needed and recheck, until correct figures are obtained.
If the machine being set up is cold, re-check when warm. Routine checks must be made on warm machines.
There is an optimum frequency at which routine checks should be made. This will be determined by the conditions existing on a given plant.
This can be assessed by the process of analysis of results. 3. Hatra Thorn Automation Stitch Length Meter
With this instrument stitch length measurements are simplified, being done while knitting, instead of unroving and measuring course length after knitting.
The procedures under headings 1 and 2 can now be simplified as follows: -
1A. Commencement
d. Put machine into knitting programme and start up.
e. Insert Stitch Length Meter detector head into yarn path and arrange readings to occur: -
f. Note stitch length reading and adjust needle bar until correct.
2A. Control
Item C. in 2 above is simplified to -
Periodically check for stitch length while knitting by carrying out 1A. e, and f above as appropriate.
LOOP LENGTH CONTROL ON POWER FLAT MACHINES
Loop length control is the technique used to control the production of knitting machines. The step by step procedure for this on the above machine is as follows: -
*For running tensions of 10-20g, a 10g weight is suitable. This is achieved using a 6" (15.2cm) length of 1/8" (0.32cm) diameter brass rod bent into a ring or other compact shape. The yarn is correctly threaded-up and the weight hung on after a guide as near to the feeder as possible.
** In the absence of a K.O.D.G., the only accurate method of balancing loop lengths on both beds is to follow c. by the following: -
The following techniques are used in running quality control scheme.
Check a. and b. above are the most important routine checks. c. should also be carried out periodically.
If measurements come out of tolerance, check any or all of the following as appropriate: -
Stitch length as under c. above.
Total needles and courses in each structure.
Yarn counts etc. No. of courses and wales in each part.
Check yarn paths for blockages.
Set up machine again as under 1d onwards.
If the machine being set up is cold, re-check when warm. Routine checks must be made on warm machines.
There is an optimum frequency at which routine checks should be made. This will be determined by the conditions existing on a given plant.
This can be assessed by the process of analysis of results. LOOP LENGTH CONTROL ON REVOLVING CYLINDER MACHINES
Loop length control is the technique used to control the production of knitting machines.
The step by step procedure for this on the above machine is as follows: -
1. Commencement
Procedure in the absence of a yarn Length Counter: -
Alternative procedure to items f. - i. when a Yarn Length Counter is available: -
Check the consistency of the positive feed: -
Although this device reduces the course length variation enormously, when knitting to fine tolerances, many a firms hove found it to be worthwhile periodically to check the positive feed and note, as a correlation factor, the difference between the course or loop length as shown by the positive feed setting and that actually being knitted.
This can be achieved by unroving and measuring, but if there are a large number of feeds, then the employment of the procedure 1A. f. to i. or j. to m., and then noting the discrepancy between the mean measured course length for each positive feed in action (there may be more than one intended course length in a given fabric) and the setting shown on the positive feed. Finally, adjust the setting as appropriate to correct the discrepancy. Re-check as needed.
The above techniques are used in running a quality control scheme.
They apply both to setting up a machine and for routine Q.C. of the machine's production.
If the machine being set up is cold, re-check when warm. Routine checks must be made on warm machines.
Roll or fabric weight of a set number of machine revolutions (either a complete roll, or a number of revolutions equivalent to approximately a running yard) gives a valuable check on stitch length and yarn count consistency.
When checking a machine newly set up and commencing production, carry out the weight check as follows: -
When making routine checks, roll weight of known number of machine revs. checked against specified figure should suffice.
When out of tolerance, return to Section 1 and work through items c. - m. of part A. or c. - f. of Part B. as appropriate.
There is an optimum frequency at which routine checks should be made. This will be determined by the conditions existing on a given plant.
This can be assessed by the process of analysis of results
KNITTING QUALITY CONTROL ON SEAM-FREE HOSE MACHINES Hose machines may be controlled by the use of the Hosextender or the Lateral Stretch Device or the Electrostatic Course Length Meter. The step by step procedure is described for
Setting up a machine
Routine Control Work through d. - g. frequently as necessary to keep the production within the required limits. Item h. used regularly provides a valuable routine check.
Setting up a machine
Routine Control Work through d. - g. as frequently as necessary to keep the production within the required limits. Item h. used regularly provides a valuable routine check.
Setting up a machine
Routine control Wt./doz. pair, off-machine, provides a valuable routine check, and should be backed-up carrying out any of the following procedures: -
Because such light input tensions are used on hose machines, the increase in input tension caused by inserting the meter for checking could significantly alter the course length being knitted while the instrument is inserted. This can be checked by inserting the instrument for a short period and measuring the course length on the hose blank so knitted on the feeder on which the instrument was used. The particular yarn can be identified by marking with a little swarf. The operation on the same feeder is repeated without using the instrument and the average of, say, 20 courses in the same part of the garment blank determined and the two averages compared. If the % difference is negligible compared with the % knitting tolerances, then the course length can be calculated from the speed meter reading and the machine speed (determined by counting no. of revs. in 30 secs. or 60 secs.) by the equation. Course length = Yarn speed reading (cm/sec) x 60 (cm) revs. per min Thereafter the procedure is the same as for 3. If the Yarn Speed Meter significantly alters the course length being knitted while it is in use, then proceed as follows. The instrument will affect the course lengths knitted, although consistently over the feeders. It can therefore conveniently be used to level the machine.
Setting up a machine
The work required in 4d. above will be much less tedious than that required in 1d., e. and f. or 2d. e. and f. because of the pre-levelling of the feeders in 4b. and c.
Routine Control
Notes
If the machine is cold when it is set up, recheck when warm. Routine checks should be made on warm machines.
It is advisable to check the chain on a routine basis to ensure that no errors have been made in adding or subtract links for size changes. Weekly or perhaps even monthly checks may suffice.
An alternative method to detect incorrect links in the chain is to use a combination the hosextender and the lateral stretch device, since a correct reading of LSD and an incorrect reading of hosextender will indicate that the number of links in the chain is wrong. For run resist hose the lateral stretch will depend on the amount of yarn in the short course.
If the measurements fall outside the limits with either the Lateral Stretch Device or Hosextender, the machine is adjusted and tests repeated. If necessary, the course length at each feeder and any part of the blank can be checked by unroving and measuring on the Course Length Tester.
With the Electrostatic Course Length Meter any machine giving work out of tolerance can be checked directly by the simple procedure 3.d. above.
There is an optimum frequency at which routine checks should be made. This will be determined by the conditions existing within a plant.
This can be assessed by the process of Analysis of Recorded Results, see QK13.
SETTING UP HOSE MACHINES TO MEET STIPULATED STRETCH MEASUREMENTS There is a relationship between course length, cam settings and stretch measurements on lateral stretch or hosextender measuring instruments. However the relationship is not simple because of yarn characteristics and machine characteristics also affect stretch. The proposed method of setting up to over-come these variables is as follows. The stretch achieved with a particular yarn is determined largely by the course length achieved. The longer the course length the higher the stretch figure, and vice-versa. The degree of bulking (crimping) whilst affecting size and cling, does not have a major effect on the stretch figure, because the crimp is largely extended in the test.
NOTE There is usually a difference in course length figures obtained for a given quality of knitting depending on whether measured by a yarn length meter or course length tester. However the difference should be consistent for a given yarn and knitting conditions.
CONTROL OF FABRIC QUALITY DURING AND AFTER KNITTING Once a knitting machine has been set up to the specification and an initial sample been checked, the machine should continue to produce satisfactory goods. However, drift, for one reason or another, can set in. Quality Control measurements related to loop length will monitor this and can be incorporated into the overall quality control scheme in the department. Even so; this is not complete, and other control measures for quality are required, both during and after knitting. A. DURING KNITTING There are three areas of control here. 1. The Yarn. Although systematic testing of incoming yarn will reduce the chances of knitting unsatisfactory yarn, an inspection of the yarn on the machine is still desirable in order to: -
2. The Machine The following checks are suggested: - a) Inspection of the knitting head.
Excessive machine speed for certain yarns can result in high yarn tension and yarn breakage.
Correct functioning of the revolution counter ensures the accurate control of piece weights and the correct recording of machine running times for wages payment purposes.
3. Fabric as knitting.
If noticed, check course length and/or change yarn packages. If a lab. has been established, cones and fabric cutting (full fabric tube of several feeder repeats) should be sent to lab. b. Rate of inspection A suggested frequency of this inspection is determined by the following formula: Inspection = Length of fabric knitted per hour per hour Length of fabric hidden by machine
This formula permits all faults created at any time up to the start of the previous inspection to be visible at the current inspection. B. AFTER KNITTING a. Types of faults A suggested list of faults to identify at this stage is - 1. Ends out. This covers the breakage of the yarn at the feeder. If the feeder is knitting on either dial or cylinder needles a press off will not result. If the feeder is knitting on dial and cylinder needles only, a press off would normally result. This fault cannot be mended. 2. Drop stitch A single dropped stitch that is not accompanied by a yarn break. This fault should be mendable. 3. Cut A drop stitch accompanied by a yarn break within the fabric. The damage will appear on one face only. This fault can often be mended. 4. Hole A hole appearing through both faces of the fabric. Not normally mendable. 5. Needle fault A fault appearing down the length of a wale due to the malfunctioning of one needle. This can be caused by a broken latch, sticking or misalignment. This fault may be mendable. Needle faults are defined only by the number occurring and not by their length. It is the number of faults that is relevant in checking knitting efficiency. The length of this fault is merely a reflection of the standard and frequency of routine inspection by the knitter. b. The expected faults per 100lbs. of fabric knitted have been found to be as tabled below:-
b. Fabric appearance.
This is checked against the stipulation in the fabric specification and the standard fabric. Quality standards are also desirable to fix the tolerable limits of departure from this standard fabric.
c. Fabric yield and dimensions.
These measurements are also needed as a measure of quality.
Yield is obtained by: -
i. Weighing small squares or circles of the fabric. ii. Weighing a roll or length of fabric and measuring the width. iii. Weighing a roll of known number of machine revolutions.
The first two methods suffer from inaccuracies in measuring length and width truly representative of the whole piece. The recent trend towards the fitting of rev. counters enables an accurate number of courses to be knitted per piece, and the piece weight then becomes a good quality guide. If the weight is determined after finishing the finished yield and dimensions can then be calculated.
d. Other tests
These may be required in order to conform with licenses or customer requirements, e.g., Woolmark, Crimplene, Wear dating and chain store specs.
These generally are carried out as required, or as stipulated by the above agencies.
QUALITY TESTS AND CHECKS REQUIRED TO ENSURE PRODUCTS AND PROCESSES MEET SPECIFICATIONS
The specification requirements and the checks and tests needed in the three defined areas of yarn, knitting and making up and examining are set out below. 1. CHECKS AND TESTS - YARN
Specifications These should cover -
Spinner should be required (if possible) to supply (certificate) evidence of compliance with each blend.
Tests - description, recommended frequency, records and remedial action
A. Essential tests
B. Secondary tests - usually only required for investigating problems arising in production.
2. CHECKS AND TESTS - KNITTING AND FABRIC
Specifications These should cover -
Tests
A. Essential tests
1) Stitch length - by unroving
Record: Checks made but confirmed out of line checks only by machine against date Remedial action: After confirmation by second check; report immediately need to adjust specific m/c.
Knock over depths using gauge for sampling and accurate setting up.
2) Weights and dimensions - by weighing and measurement Frequency:
Record: as for stitch length Remedial action: Report immediately so that appropriate action can be taken in relation to yarn or knitting.
Fabric appearance - visual assessment paying particular attention to frequency of faults e.g. holes as a basis for investigation of cause.
B. Secondary tests - normally used for investigations and setting up
3. CHECKS AND TESTS - CUTTING, SEAMING AND FINISHED GARMENTS
Specifications These should cover -
Checks - virtually all visual to check operative performance and finished garment quality (finished garment checks by Q.C. should not be confused with production examination - essentially a non productive quality operation).
A. Cutting - operative performance
Remedial action: none immediately as historic cause by time of receipt, but invaluable as a basis for remedial action in conjunction with other fault records an identifying thresholds of customer rejection, and hence tolerances and Quality Standards updating.
TESTS AND ACTION TO BE TAKEN WHEN KNITTED FABRIC QUALITY IS INCORRECT
INTRODUCTION
As a result of testing incoming yarn, one may discover that the yarn is out of tolerance but can be knitted into acceptable fabric is appropriate adjustments are made so to the knitting machines. Also if fabric is discovered to be incorrect, there are certain procedures to be undertaken, to ascertain the cause and hence the appropriate action to be taken.
The present article deals with these two aspects.
The first premise made here is that the knitting machines have been adjusted to knit correct fabric with normal yarn. The problem is to determine whether incorrect fabric is due to yarn or machine variation and what to do if it is known that the yarn is the cause of abnormality.
This is due to the yarn at one feeder giving a different cover than yarn at the other feeders, either due to a change in yarn properties or due to a change in the cam setting. To ascertain which, change the cones round and check if the ring or bar sequences changes. If still not certain, replace the yarn at the feeders giving bars and recheck for barriness. Then carry out sequence in the diagram.
This manifests itself by giving wrong dimensions (e.g. width, length, course or wale count) wrong appearance, wrong weight per dozen blanks or roll or set number of machine revs. or wrong wt. per unit area. See diagram.
If the yarn cannot be rejected the following actions may be taken, to reduce the amount by which the fabric would be out of tolerance.
Since the amount out of tolerance is known, an adjustment to stitch length can be made to compensate. This makes an assumption that maintenance of fabric density (i.e. wt. per unit area and wt. per doz. is the first priority and that over a range of at least +- 10% this accommodation will give fabric of acceptable quality. The stitch or course length is adjusted by the same percentage as that by which the average count of the delivery is out from the specified figure. For example if the count is 6% coarser, the course length is increased by 6% while knitting this delivery. The number of courses or machine revs. in the garment or roll is reduced by 6% and in flat knitting the number of needles per course by 6%.
The wt. per doz. blanks is used as the criterion of fabric quality with flat machines (F.F., V-bed and Flat-bed). With circular knitting, because the no. of needles cannot be altered, the wt. per doz. or roll wt. per set no. of m/c revs. will increase or decrease by the amount by which the course length has been increased or decreased. This new wt. per doz. or roll of set no. of revs. is used as the criterion of fabric quality for the yarn in question. If the production fabric is out of tolerance from this criterion, recheck the weight. If still out, check that the no. of courses or revs. has been corrected to the new figure, (and also the needles with flat knitting). If the weight is still out check that the course length has been correctly altered. If still out, then the count in this lot is not typical of the average of the delivery. Recheck count and readjust course length etc. accordingly or reject yarn as too variable.
In this respect it is as well to recognise that the tolerances on weights, being a function of stitch length and count, will be wider than on count or knitting alone. For instance, if these latter have been running at, say, +- 2.5% and +- 2.0%, then the tolerance on weight will be +- 3.2%.
This action will not involve any change in weight of yarn consumed with flat knitting, where a proportional adjustment in the number of needles can be made. In circular knitting, where this cannot be done, there will be proportional change in the weight of yarn consumed in maintaining constant fabric quality. Here recourse to the Q.C. and Laboratory records will give evidence where an approach to the spinner for redress is envisaged.
Here there is not the direct relation that exists with count, but if the bulking figure is higher, the course length should be lengthened and vice versa. The amount of course length adjustment can only be found by trial. However as a guide let us suppose the average bulking figure for the delivery is 6% too high, i.e. running at 36% for 30%. This by proportion is (36 - 30) x 100/30 = 17%. Since opening the stitches (because of lengthening the stitch length) will further increase the bulking power of the yarn, an adjustment of 17-20% is used as a start, and the dimensions (i.e. stitch density or wt./unit area) after a relaxation treatment is used as the criterion for correct quality achievement. If the bulking figure had fallen by 17%, we would shorten the stitch length. This will further impede the bulking power of the yarn and so again an adjustment in excess of 17% should be tried as a start, and again the same criterion used.
The relaxed stitch density (courses per cm x wales per cm) is the main criterion and the aim is to keep this to the specified figure. This will maintain the relaxed garment and pattern repeat dimensions. The readjustment we then make to the course length to achieve this, causes a proportional change in weight per doz. or roll wt. per set no. of m/c revs., and in the relaxed wt./sq. metre, and this has to be accepted. Recourse to the Q.C. and Lab. records may give evidence for an approach for redress to the yarn supplier.
If no adjustment of course length is made to accommodate yarn variation except when the course length itself varies, one is faced with the effects of count variation and bulking variation.
Count will have a direct proportional effect on the wt. per doz. lots, roll wt. of set no. of revs. and on the relaxed wt. per unit area. When fine this will give fabric too lean and weights too light for the customer. When coarse this will give fabric too dense, and since yarn is bought by weight and in effect is sold by length, a given contract weight will give rise to a short-fall in goods knitted. There will be practically no effect on dimensions.
Bulking will have no effect on wt. per doz. lots or roll of set no. of machine revs. The dimensions will be profoundly affected and hence the wt. per unit area.
Adjustment of course length to maintain fabric density (i.e. wt. per unit area and leanness and appearance) is likely to be the most important criterion if the yarn cannot be replaced.
Count variation with flat knitting by proportional adjustment of course length and then needles and courses knitted will maintain fabric density and the weight and dimensions. There will be no change in the weight of yarn consumed and no short-fall in the goods knitted. With circular knitting, where the no. of needles is fixed the relaxed wt. per unit area, i.e. fabric density and garment length will be maintained, but the weight per doz. or roll of set no. of revs. and relaxed garment and roll widths will vary proportionally. In garment blanks the width variation may not be crucial, but with fabric if the width setting is maintained by an adjustment in stentering etc. the fabric will no be of a fully relaxed state and shrinkage problems may ensue, and the fabric wt. per unit area, which the adjustment in course length was designed to maintain, will also change. Thus a calculated width change will then have to be accepted, dependent directly on the change in course length.
To recapitulate: - adjustments to course length to compensate for bulking variation and maintain fabric density, will maintain dimensions and wt. per unit area, but the wt. per doz., roll wt. per set no. of revs. will change. Therefore if the bulking increases, the course length is increased to compensate and the wt. of yarn consumed will increase, except where the number of needles can also be altered, as in flat knitting.
RECORDING SYSTEMS
Reasons for recording
Recording provides
It is essential for any Quality Control System that adequate records should be devised and kept. Such records are required to supply essential information when goods being manufactured are failing to conform to the specification and the standards required. In the making up room records are needed for
Seam Control checks
These checks should be made on a regular basis usually by the quality control staff; to ensure that the seams being produced are meeting the technical specification laid down.
An example for a Seam control check card is given overleaf.
Operative quality checks
These checks should be made not less than once a day by each supervisor. These are checks to ensure that each operative is producing seams that appear satisfactory. In most factories one of a supervisor's duties will be the responsibility for the quality of the work produced in that section.
An example for an Operative quality check card is given. This would be kept by the machine.
Entries on the card would only be made if one or more checks under any one day, or item, are substandard.
SEE RECORD CARD FOR WEEKLY CHECK OF OPERATORS' PERFORMANCE
As each fault is noted it can be entered against the appropriate item no. in the column for the particular day as five-barred-gates, from which the totals, daily and weekly, for each item can be quickly assessed.
See similar card for Supervisor's record
To obtain the all-important overall picture of operatives over a period, it may well be a time saver in the examination of these weekly reports, to enter the weekly totals on a Trend Chart.
See diagram.
EXAMPLE SINGLE OPERATION CHECK
See diagram
EXAMPLE REPAIRS RECORD
See diagram
It is often helpful for a Supervisor to have a checklist for each operation under her control, as a reminder.
SUPERVISOR'S CHECK LIST
See diagram
ANALYSIS OF RESULTS OF QUALITY CONTROL CHECKS
Compare: -
Courses of action
CONSISTENCY OF MATERIALS -INCOMING RAW MATERIALS
Yarn is the major raw material, and raw materials represent 40-70% of a knitter's manufacturing costs. So that checks to ensure that these items are satisfactory should not be overlooked.
The basic premises on which any such testing is to be founded are: -
Economic savings can be made if tests are carried out to check that a company is receiving what it is paying for. In addition, a company known to be testing is likely to receive better quality yarn.
First an estimate must be made of the losses in the absence of all yarn testing. The method of undertaking this is discussed under "Economic Aspects".
Armed with this estimate not only is the scope for savings known, but also the areas, which will give the most immediate benefits. It is in those areas where testing should be undertaken first.
The next step is to decide the appropriate testing activities overall, guided by the potential savings indicated by the estimate. The tests required, frequency of tests is then considered. A cost balance is generally successful where the total testing costs amount to 50% of the estimated losses.
Some of the tests involve the use of expensive equipment and require well-trained technicians to perform the work. Testing Houses are willing to perform such tests. A small unit would normally confine itself to simple tests such as, count, twist, oil content, sensitive dyeing tests, colour fastness to water, rubbing etc.
Where cost permits, speed is an obvious advantage in carrying out tests on your own premises. If the cost of equipment and wages for 2 years internal testing is likely to be less than the fees charged for the equivalent number of outside tests, then internal testing should be considered. However even with internal testing, correlation should be checked for drifting by the occasional, parallel test performed by an outside authority.
As with all Quality Control testing, yarn test results should be recorded, and the results analysed.
KNITTING QUALITY
THE ACTION PLAN
The purpose of this Action Plan is to provide guidelines as an approach to improving quality management. This can be considered under: -
It is highly desirable that during the development of a style a fully adequate specification is evolved at this stage. Ironing out the shortcomings will be far more expensive generating, in all probability, losses at the production stage in both time and materials and in administration. Often greater detail and care at development can enable production knitting to be achieved so much more efficiently as to permit earlier completion of orders.
Examine your specifications for the last 3 months in respect of the following aspects: -
Analyse the inadequacies and make recommendations.
For all machines recently set up (i.e. within the last week) assess: -
Problems arising from yarn variation will be significantly reduced in terms of fabric problems if between-machine variation is minimised.
Areas with scope for immediate benefit
If there are areas in which straight forward action is likely to have immediate beneficial effect these should be implemented immediately. Irrespective of any financial benefit, there will be a psychological benefit of rapid action being seen to be taken.
Four such areas should be considered; suggestions are made for each.
Review the present systems. Consider whether any modifications would: -
In both blanks and fabric a set number of loops have been knitted. If records are kept on a running basis - i.e. a continuity chart then this weight can become a meaningful parameter for routine control. A chart should be kept for each machine with shift, yarn and style changes recorded.
With roll weight there are two other benefits: -
Preset counters and stop motions can be fitted to machines for preset doffing. However the effectiveness of the system can be influenced by your present payment system, so careful consideration is required if it is to be introduced.
This unlike roll weight per unit revs. Is not a fully reliable routine check in the unfinished state. This is because of variable degrees of relaxation. However, where doffing at a set of number of machine revs. is not possible, a check using weight per square metre can be used. This involves determining weight and stitch density in the same cutting of grey fabric, and at the same time. This combination can be much more reliable than weight per square metre on its own.
Areas where benefits would take time
There are also other actions, where any benefit is likely to take time to materialise, which should be considered.
The cost of introducing and operating such potential procedure or record must be assessed beforehand for cost effectiveness.
Quantify in each case how much is due to yarn and how much is due to machines.
In the case of machines how much is due to poor adjustment, bad servicing, or incorrect gauge i.e. unsuitable machine.
Where necessary, set up a trial, analyse results, in order to make a recommendation for permanent introduction.
Great care should be exercised in the yarn used to produce samples. The yarn should be truly representative in every technical aspect i.e. count, twist etc. as the style specification will be developed from the details produced at sampling. If the yarn is not truly representative then future problems in knitting are inevitable.
Similarly, the machine used to knit samples must also be truly representative of production machines - otherwise again problems are inevitable.
Much yarn variation is short term with rapid fluctuations, so that machine adjustment in production for this variable is no use. However, the usual tolerances in knitting would accommodate most of this yarn variation if the knitting specification was evolved as recommended above.
It is therefore very essential to review current procedures to ensure that at sampling potential production problems are not being generated. There is, however, the possible problem of changing machines for sampling hence this may be longer term.
Operative awareness of agreed quality standards is essential. It is important to know whether or not there is any financial benefit to the knitter in achieving these quality standards. The long term objective must be accountability for quality by the knitter.
The quality performance of each knitter should be analyses. This can be approached through records of fabric faults due to knitting, dimensional changes, and weight inconsistencies. While the knitter may not be responsible for controlling the factors governing these parameters he needs to be aware of them in order to report problems and motivated to do so.
In order to assess the economics of machine replacement, a comparison must be made between the production costs of garments at present knitted on the worst machines in the plant, and those knitted on hypothetical new machines, or preferably on actual new machines.
The following points must be taken into account: -
When the Quality Control and Fault Analysis systems are in full operation, some of the evidence will become available and it should be possible to estimate the costs of the other points of comparison; if necessary by carrying out a short investigation.
The facts will then be available on which to base an economic assessment.
PLANNING A TOTAL Q.C. SCHEME
The aim of any such scheme is to maximise quality (i.e. minimise losses at minimum cost.
Twelve headings are suggested under which to plan a scheme.
Development
QUALITY CONTROL REQUIREMENTS - PRIORITY ORDER
FIRST PRIORITY
SPECIFICATIONS - tolerances ad Quality Standards, shade swatches
RAW MATERIALS - Shade
MANUFACTURING PROCEDURES
SECONDARY PRIORITY
PRODUCT TESTS -Performance -colour fastness -stability -abrasion etc.
THIRD PRIORITY
INCOMING MATERIALS CHECKS -Main Yarn - fibre content and quality -yarn twist, friction and oil content
YARN/FABRIC UTILISATION
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
Julian Ellis will be delighted to hear from you. Telephone on +44 (0) 1636 812100 email: info@ellisdev.co.uk We are based at The Stables, Hall Farm, Kirklington, Newark, Nottinghamshire, NG22 8NN,, United Kingdom |
