SE-664 Use of AFIS Histograms 1
SE-664 Use of AFIS Histograms 1
SE-664 Use of AFIS Histograms 1
Application Report
All and any information contained in this document is non-binding. The supplier
reserves the right to modify the products at any time. Any liability of the supplier for
damages resulting from possible discrepancies between this document and t he
characteristics of the products is explicitly excluded.
Contents
1 Introduction ............................................................................... 4
2 AFIS Modules............................................................................. 6
2.1 AFIS – Nep .................................................................................. 6
2.2 AFIS – Length ............................................................................. 7
2.3 AFIS – Maturity............................................................................ 7
2.4 AFIS – Trash ............................................................................... 8
10 Conclusion ............................................................................... 26
1 Introduction
Since the introduction of the AFIS instrument in the late 80’s it has been
found its place within the process control of the cotton spinning mills. Over
850 units are installed worldwide.
Fig. 1
Worldwide distribution of
AFIS instrumentation
All over the world spinning mills use the features of nep testing, individual
fiber length measurement as well as trash testing. The advanced fiber in-
formation systems has proven its usefulness.
All AFIS installations have the Nep (N) module installed. New on t he
USTER® AFIS PRO 2 is the categorization of fiber neps in addition to the
seedcoat neps category that already existed. The occurrence of seedcoat
neps is increasing due to changes in seed varieties. A high number of seed
coat neps may cause problems in the spinning preparation processes, and
if the fiber handling in opening and cleaning is not optimal, even increase in
number. If they are not removed, they can cause a higher number of yarn
breaks in spinning.
The length parameters including the maturity can be checked with the
AFIS-L&M (Length and Maturity) module. There are some processes in the
mill, where f iber length and i t’s distribution is a priority. Not only is it the
main parameter for specifiying the cotton quality – it also affects yarn quali-
ty, roller settings within the whole mill, and determines the waste level to a
great extent. Three processing stages have an i mpact as well as are af-
fected by fiber length and l ength distribution – the blowroom, the carding
and the combing process.
The AFIS-T (Trash) module measures the number and well as size of dust
and trash particles. With this information, the cleaning efficiencies through-
out the cleaning processes of a mill can easily be determined. Monitoring
cleaning efficiency is critical to controlling cost and quality.
As the world adopts modern cotton seed, harvesting, and ginning technolo-
gy, the risk for higher amounts of trash and seedcoat fragments increases
compared to hand-picked cottons that do not require as much cleaning and
are often roller ginned at lower production rates.
When the spinners try to clean these cottons, they run the risk of increasing
short fiber and neps by aggressively removing the trash and seedcoat
fragments. Therefore, gentle cleaning of cotton has become a priority.
Rotor spinning was always sensitive to dust and trash. Ring and Compact
spinning are more sensitive to short fiber content levels. The quality proper-
ties of cotton and the handing of fiber during processing are interlaced.
When cleaning fiber to remove trash, dust, and seedcoat fragments, there
will be an impact on the nep content and fiber length. How much they are
impacted will be determined by the technology utilized and the settings and
production rates employed.
Fig. 2
Interdependency of
parameters
One also needs to bear in mind, that the productivity of a mill plays an im-
portant role. The delivery speeds of spinning machinery keeps increasing
over the years – nevertheless the quality of the yarn in most cases has im-
proved as well. This is only possible if the machine settings are finetuned
and the condition of the machinery and technological elements are kept
aligned.
Fig. 3
Machinery influences and
impacts
2 AFIS Modules
The USTER® AFIS PRO 2 with different modules allow the simultaneous
testing of fiber samples. When utilizing the Autojet – the samples are fed in
automatically into the instrument, and the lab technician can perform other
tasks whilst the USTER® AFIS PRO 2 is testing the samples automatically.
Fig. 4
Fiber Nep, Seed Coat
Fragment
In this module not only are neps and their sizes evaluated – according to
their signal form in the sensor, seed coat neps are also detected counted
and sized.
Fig. 5
Staple diagram and fiber
length definitions
Fibers in USTER® AFIS PRO 2 are measured individually for their fiber
length – this measurement can be classed as an end oriented testing
method. However, the applied level of short fibers differs. In the western
hemisphere the short fiber limit of ½” – or 12.7 mm is commonly used – in
China and Russia the limit is 0.65” – or 16.5 mm is practiced.
Fig. 6
Immature Cotton and
Cotton fabric with white
speck
The fibers are tested also for their maturity. Dead or immature fibers have a
very thin cell wall, and appear as a thin fiber ribbon. Mature fibers develop
thicker cellulosic walls, and tend to convolute.
Depending on the voltage (=height) of the signal, the fiber fineness can be
determined. Fiber fineness of cotton fibers can range usually from 120 –
180 mtex.
Fig. 7
Dust, Leaf and Trash
particles – magnified
Dust and trash particles are heavier and l arger than fibers, and therefore
are seperated at the opening roller from the fiber flow. The number and size
are determined in the optional trash sensor.
3 Types of results
3.1 Numerical results in tables
The user can select the summary histogram (average of all samples tested)
– or if desired see only the histogram of an individual repetition. They can
be quickly checked on the screen of the instrument – or as a reference
printed out in a electronic pdf-format or as a paper-hardcopy.
A B C D E F G
®
Table 1 Print-out of USTER AFIS PRO 2 with nep result columns
Other abbreviations:
Mean : Mean result of the test repetitions
Std. Dev. : Standard deviation of the test repetitions
CV% : Coefficient of variation of the test repetitions
Q99% : Range of results within 99% of confidence level
Min : lowest value of the tested repetitions for that parameter
Max : highest value of the tested repetitions for that parameter
USPTM : Uster Statistic Percentile [2007] or [2013] of the parameter
(only listed if available)
Fig. 8
®
USTER AFIS PRO2
Nep Histogram (by count)
The neps and seed coat neps are not only counted but classed according
to their size. Due to their different behavior in the sensor, seed coat neps
are identified. These are labeled in the histogram in red color.
In the nep histogram we can see clearly the percentage of seed coat neps
in the sample. As more red / or shaded particles are within the samples
there is a higher probability that they will cause ends down in the spinning.
The graph has two scaling options – one is showing the actual count – the
other is relating the counted number to percent (“Percent of total”).
The first option is shown in the above displayed graph. It states “Nep Count
Histograms” – and the scale range is up to 350.
In this real-world example, samples from bale to comber sliver were evalu-
ated. The bale test on HVI showed a M icronaire value of 3.8, a staple
length of 37.5 mm and a bundle strength of 43.8 cN/tex.
In the example histograms we can see the increase in neps from the bale
to card mat. It is also apparent that large neps were created in the blow-
room. Already in carding all sizes of neps and all but very few seed coat
neps were removed. Here, a glance at the histogram helps to determine
how well the seed coat neps are eliminated in carding or if many remain in
the card sliver.
The effect of the combing process relates also to the noil percentage – as
more fibers are combed out, the nep level will diminish accordingly.
In this example, the nep content was increasing in the blowroom, whilethe
card and c omber were reducing the nep c ontent. This effect can be illus-
trated by calculating removal efficiency.
Fig. 9
Nep efficiencies in spinning
process
For the carding process a cleaning efficiency of 89% is good, and quite
respectable – this however is not only due t o the high level of neps, but
also to machine state, condition, speeds as well as settings.
But before fully acknowledging the nep l evel reduction in our case – the
other quality parameters need to be checked and considered as well, which
we do in the following chapters.
5 AFIS Length
5.1 Numerical results
H J K L M N O P Q R S
®
Table 3 Print-out of USTER AFIS PRO 2 with length result columns H to P
K+O The selected level of short fiber content (1/2 “ or 0.65”) is given
on the printout
Differences between those two types of evaluation are the frequency in the
individual length classes are by either gravimetrical weight or mass – or by
the counted number of fibers. So the same fibers are evaluated by the
number of events in the class – or by the weight those fibers would have on
a balance. As short fibers weigh less than longer ones, those frequencies
of short lengths are lower in the by weight distributions.
Fig. 10
Left side:
Histogram length by weight
Right side:
Histogram length by number
The level of short fibers most commonly referenced worldwide is 12.7 mm.
In China a level of 16.5 mm is often selected. The customer can chose his
own level, and will get numerical results as well as the histograms with his
selected level. In the histogram the fibers shorter than this selected limit are
highlighted in red bars.
The same samples that were evaluated for neps were also tested for their
length values. The extra long staple cotton was intended to spin fine yarn
counts for mens shirting, and therefore was used in combed ring spinning.
Length
38.8 33.9 32.6 34.8
UQL (w) [mm]
Length
32.0 27.0 25.6 28.3
L (w) [mm]
Short Fiber
4.1 8.0 10.1 3.6
SFC (w) [%]
Reduction of
short fibers
The length values in the table show that the extra long staple cotton, is be-
ing damaged in the blowroom, and then again at the card. The increase in
short fiber content by weight was almost 4% from bale to card mat, and
another 2% from card mat to card sliver.
At the combing process the short fibers were removed, bringing short fiber
level back to the level of the original bale. The changes are visible in the
histograms of the by weight distributions, but the shape of the distribution
maintain a similar shape.
However the breakage in the blowroom is considerable – loosing 5 mm in
average length, leads to the conclusion that settings were extremely close,
to operating speeds too high or any other parameters that are out of tune.
Due to this reduction in average length, the short fiber of course increased
considerably.
Normally we expect in a standard and well adjusted process a short fiber to
decrease at each process stage.
Neps
210 550 61 17
[Cnt /g]
Length
25.8 20.6 19.3 24.0
L (n) [mm]
Short Fiber
18.4 26.4 30.2 10.9
SFC (n) [%]
Reduction
of short
fibers
Even subtle change in fiber length will be seen first in the by number distri-
butions before they will be visible in the by weight distributions. For a quick
check to determine if fibers are damaged in processing a glance at the
length by number distribution is recommended. This is even more critical
when processing a new material or applying new settings to a process.
In this example the mill has used a rather aggressive opening and cleaning
process for such a delicate fiber. The neps and the short fiber content in-
creased considerably. In carding, trying to achieve a high nep removal effi-
ciency, the settings had led into breaking fibers. The mill should consider
testing different settings at the card flats.
6 AFIS Maturity
6.1 Numerical results
H J K L M N O P Q R S
®
Table 6 Print-out of USTER AFIS PRO 2 with length result columns Q to S
Fig. 11
Left side:
Histogram Fiber Fineness
Right side:
Histogram Cell Wall Thick-
ness
Cell Wall: The fibers which have a cell wall less than 0.25 are considered
as immature - and are highlighted by the red bars in the histogram. These
fibers reflect also the immature fiber content (IFC).
Fibers that have developed a cell wall of 0.25 to 0.5 are considered as thin
walled fibers.
The higher the immature fiber content the more problems will occur in dye-
ing. Immature fibers cause “white specks” in dyed fabrics, as they will not
absorb the dye stuff. These immature cotton fibers have no cellulose in the
wall, and t herefore the dye stuff can not be taken up. Therefore, a mill
should monitor fiber maturity level and its variability through processing to
prevent problems later in the fabric.
IFC
4.8 6.0 7.2 3.7
[%]
With the histograms above, one can easily see what happened in the spin-
ning process in our example study. Within the openening and carding pro-
cess fibers were damaged, and therefore the IFC content increased slight-
ly.
Immature fibers are also weak and t herefore break easily during pro-
cessing. These broken fibers then get distributed throughout a bl ended
sample causing more problems for dyeing.
The effect on the maturity ratio is also clearly visible. It drops in the two
stages card mat and card sliver.
However in the combing process - where the short fibers are combed out –
also the immature fibers were reduced as they were likely part of the short
fibers removed. This allows us to draw the conclusion, that often the short
fibers are also the immature ones.
If the comber noil would be tested on the instrument – one would see –
besides the high amount of neps and the high short fiber content that the
maturity within the noil would also be low, indicating a high amount of im-
mature fibers..
Additionally the dust and trash content can be considerably high in comber
noils.
T U V W X Y Z
®
Table 8 Print-out of USTER AFIS PRO 2 with trash result columns
Fig. 12
Left side:
Histogram Trash Particle by
Count
Right side:
Histogram Trash Particle by
percent of total
All particles smaller than 500 µm are classed as dust particles – the ones
above this limit are classed as trash particles.
The visible foreign matter, the VFM is given as a pe rcentile value. It is
based on an algorithm, which considers the number of events, their sizes
and the density.
For judging the removal of neps or trash particles and their sizes, it is rec-
ommended to use the histograms by count. So the effect of the processes
e.g. of the card or of the comber can be seen easily.
Please note, that when comparing graphs within a s pinning process one
should compare always the same type of graph, as they are two options
selectable, and the histograms look similar, but are not the same.
Fig. 13
Trash efficiencies in
spinning process
In general, the performance assessment of blowroom, carding and combing is based on the results of
neps, fiber length and trash. Not one by one, but all together.
Fig. 14
Application report:
Critical nep size
For 100% cotton yarns in ring and r otor spinning the report of the critical
nep size shows if the average nep size will cause imperfections in the yarn.
If the dot is above the blue line, it will be plotted in red, meaning the critical
nep size is exceeded for the selected yarn count. If the dot is plotted in red,
it means the average nep size is not critical for that particular yarn count.
Fig. 15
Application report:
Roller space setting
The roller space setting report gives a recommended roller space setting as
a starting point for draw frame roller distances (3 over 3 rollers) in a spin-
ning mill. These values are based on the 5% length by number. A sugges-
tion is made for distance of the pre-draft zone, and the main-draft zone.
Fig. 16
Application report:
Efficiency report of different
cards
These reports are available in the report section of the instrument and the
parameter can be selected individually by the user.
The pairs (input vs.output) need t o be s pecified, e.g. card mat can be
paired up with the corresponding card slivers.
Efficency reports can be calculated for short fiber content, neps and trash
values.
The level of removal efficiencies however also depends highly on the start-
ing level trash, neps or short fiber in the raw cotton. With a hi gher trash,
neps or short fiber content normally higher removal efficiencies can be
achieved.
Besides the histograms, the numeric data should not be forgotten. For rotor
yarn, a l evel of V.F.M. below 0.1 % needs to be achieved, as high dust
counts or trash particles will affect the running performance on OE-
machines.
Also trash and seed coat fragments are responsible for yarn breaks, espe-
cially in warping, since they are weak places, and thus replace fibers in the
yarn cross-section.
Fig. 17
Control Chart Neps of dif-
ferent cards
Here we do not have a calculation like in the efficiency graph, just the
graphical display of the originally tested values. The customer can select
manually the positions he wants to have displayed in the same graph.
With this type of graph also a view of a specific machine over time (long-
term-report) can be generated. To make this report effective, it is recom-
mended that useful and reproducible naming and filing of the tested sam-
ples be utilized.
When we always use at each test the same name, such as CARD No. 1,
we can obtain a c ontrol chart with the date on the scale of the x-axis.
Needs for card grinding or rewiring can then be determined.
Both types, the actual status of all cards – or an individual card over time-
allows to identify individual cards out of the complete line, which do not
perform up to the required level. Settings or the grinding or rewiring sched-
ule can be adjusted accordingly.
Exceptions can be recognized fast and easy with these reports. The long-
term quality therefore can be maintained in the desired tolerances.
Even though we might be satisfied with the nep level, and nep removal effi-
cienies – the short fiber content or the trash might be not behave like ex-
pected. A look and comparison of the different histograms of the individual
process step helps us in identifying those areas.
The spinner in our example used previously was damaging his fiber in the
blowroom and later had to remove the created short fiber at combing just to
regain his orginal fiber length. His loss in fiber length is dramatic, and can
be expressed in monetary loss, assuming he could use a shorter staple to
begin with, and handle this with care.
With better settings / speeds in the area of blowroom and card the mill
would be able to maintain their fiber quality. Therefore, yarn quality would
get better as well, or he could spin finer and higher quality yarns out of the
extra long staple quality.
Fig. 18
Impact of noil level on
financial gains
10 Conclusion
With the help of the single fiber testing USTER® AFIS PRO 2 in the spin-
ning mill, the processes can be optimized – not only the blowroom, carding,
and combing, but also in the drawing and roving level.
There first indications of fiber damage can be seen in the length by number
distributions. Also, well-intended approaches, such as increase of noil per-
centage can be evaluated before even getting to the yarn stage.
The subtle balance between fiber quality parameters and machine settings
can be adjusted and fine tuned to the customer needs and requirements.
Neps
Length Trash
Fig. 19
Subtle balance of fiber qual-
ity and machinery impacts.
The USTER® AFIS PRO2 offers a means to check the fiber quality for all
relevant parameters in the spinning process. The impact on fiber quality
due to the machinery state can be easily assessed and measured to opti-
mize quality and productivity.
Fiber testing in the spinning process helps the customer to optimize quality,
retain it and therefore maintain margins.
11 Literature / References
Uster Technologies AG
Sonnenbergstrasse 10
CH-8610 Uster / Switzerland
www.uster.com
textile.technology@uster.com