DYNAMIC ANALYSIS AND DEVELOPMENT OF CARD-SPINNING SYSTEM

Investigators: Youjiang Wang (leader), Mary Lynn Realff, Radhakrishnaiah Parachuru, J.

Lewis Dorrity, Fred L. Cook, Jim Brazell, Xi Chen (Georgia Tech)
Bhuvenesh C. Goswami, David A. Zumbrunnen (Clemson)

Students: Shumin Zheng, Fan Zhao, Hua Huang (Georgia Tech)

Kai Liu, John Mike Wolf (Clemson)

PROJECT GOAL

The goal of this project is to create the fundamental knowledge needed to develop more efficient
machines and novel processes that can potentially decrease the number of process steps by a
factor of 2 to 3 in short staple yarn manufacturing.

ABSTRACT

To develop more efficient machines for short staple yarn manufacturing that can significantly
shorten the processing line, we are currently conducting research in the following two
interrelated areas: (1) to analyze the motion dynamics of fiber clumps and particles in an airflow
field, and (2) to develop a single-step process (Card-Spinning) to incorporate multiple spinning
heads into a carding unit. This involves a fundamental analysis and understanding of web
uniformity monitoring, measurement, and regulation, web division, and feeding of web ribbons
into yarn spinning heads positioned at the front end of the card.

We have built a prototype of the yarn spinning head system and have modified opening and
carding machines to demonstrate the concept. A series of yarns has been made with the spinning
head system by varying the process settings. Evaluation of these yarns indicates that the
technology is promising and further development is underway.

We have frequently interacted with our industrial partners including yarn producers and
machinery manufacturers to share information on the project and to learn the state of the art
practice in the industry. Studies are underway on theoretical and experimental analysis of
airflow, and on the development of the Card-Spinning system.

INTRODUCTION

In the past decades, yarn manufacturing has transformed from a labor intensive operation to a
capital intensive one. This change has allowed the US to remain competitive in the global
marketplace. Although the productivity of current manufacturing systems continues to increase,

National Textile Center Annual Report: November 2001

F98-G15
1
a significant improvement requires major breakthroughs and new innovations, which can only be
achieved with a better understanding of the fundamental principles of the processes.

Staple yarn manufacturing involves four essential operations: fiber separation, parallelization,
attenuation, and consolidation. Currently, several machines (about 10) in sequence are needed
for these tasks. We are developing the fundamental knowledge base that can lead to more
efficient machines, shorter production lines, and novel processes. Our first major task in this
project is the study of the motion dynamics of fiber clumps and particles in an airflow field. The
second major task aims at developing a novel process (Card-Spinning) by incorporating multiple
spinning heads on the carding machine itself.

AIRFLOW IN OPENING AND CLEANING

The opening and cleaning process is the first part in a staple yarn manufacturing line. It opens
fiber bales into small tufts, removes most of the impurities, and blends fibers together. Airflow
plays a very important role in determining the efficiency and quality of the opening/cleaning
operations. Fiber opening, cleaning, and transport in the opening and cleaning processes are
achieved through interactions among airflow, inertia forces, gravity, fiber/particle configurations,
barrier arrangements (mote knives, grid bars, plates, etc.), beating actions, carding actions,
among others. Although some studies have been conducted in this area (e.g., [1] on opening and
cleaning and [2,3,4] on carding), the depth has been limited due to the complexity of the
problem.

The airflow with particles and fibers in these processes is typically a two-phase, unsteady
viscous flow. Our studies include experimental analysis using flow visualization techniques and
computational fluid dynamics (CFD) analysis. The experiments were designed to determine the
effect of fibers on airflow pattern, and to determine the condition when fibers can be neglected in
a CFD analysis. Some of the test results have been used to build a two-phase flow model in
CFD. One of the uses for the model is for predicting the flow characteristics and improving the
design of textile machinery.

Test setup and pressure distribution

The test setup, shown in Figure 1, is designed so that the test section can be changed to obtain
typical flow patterns such as parallel, diverging, converging and flow over a bent section. It is an
enclosed system consisting of a fan, test section and connection pipes. The test section is a
rectangular channel (1.5 m X 0.15m X 0.31m), the top of which can be adjusted or replaced to
change the shape of the channel. It is built with plexiglass panels to enable flow visualization
using a high-speed video camera. Forty pressure sensors arranged in four rows are installed to
monitor the flow pattern.

National Textile Center Annual Report: November 2001

F98-G15
2
 Pressure Distribution along X Dir.(V=28m/s)

-14.5
0 10 20 30 40

-15

Pressure (mmH2O)
-15.5

Bottom
-16
Z=76.2mm

-16.5 Z=152.4mm

Z=118.6mm
-17
X axis (inch)

Figure 1: Test setup for airflow studies Figure 2: Pressure distribution along the path
of the test setup predicted by CFD model.

The pressure in the test section was measured with the flow channel in the expansion type flow
configuration. With this configuration, a pressure increase should be seen along the flow
direction. The experimental results were compared to those of the CFD model (Figure 2). The
trends seen in the experimental results agreed well with those of the model, however, a
difference in the magnitude of the pressures was seen. The difference in magnitude was not of
concern, since a more meaningful comparison is done by looking at the slopes of the
experimental and model results. This comparison shows good agreement between the CFD
model and the experimental observations. Upon further investigation, a temperature increase
was observed in the test system. This causes a reduction in the density of air, which changes
flow rate. When the influence of this temperature increase is included in the modeling, the
results are only 12% different than the experimentally observed data.

Influence of amount of fiber in the flow and velocity of fibers in air flow

A series of tests were conducted to determine the sensitivity of the flow field to increases in the
amount of fiber present. The amount of fiber in the test setup was changed (approximately 35 g,
69 g, 98 g,119g and 141 g) for each series of experiments. To determine the velocity of fibers in
the airflow in the test section, a high-speed video camera is used to capture images of airflow
with a foam particle which is similar in dimensions to the fiber clumps. Results showed that the
velocity of the foam particle ranged from 7.52 to 16.8 m/sec with an average of 12.6 m/sec. The
average velocity of the air flow in the test setup without the particle was 11 m/sec. These results
have been analyzed and found to be reasonable.

In summary, we have built a test setup that captures many of the characteristics of flow fields
present in textile machinery. We have demonstrated that CFD models can be used to predict the
trends seen in the pressure distribution for the test setup for both air and air-fiber flow. The
trends in the experimental data for varying amounts of fiber in the flow were captured by the
model. However, the models do not predict the exact values of the pressures. Further

National Textile Center Annual Report: November 2001

F98-G15
3
development of the models will lead to better predictions of the experimental data. In the future,
the results of the models will be used to design new textile processing equipment.

CARDING

In a traditional process, carding individualizes fibers, removes fine impurities and neps, and
forms a sliver. By incorporating multiple spinning heads on the carding machine itself, the novel
Card-Spinning machine will replace the traditional card, two draw frames and the yarn-spinning
machine.

Web monitoring

One of the major problems in spinning multiple yarns directly from a card web is that the cross-
machine web density must be monitored and the divided sections must be controlled. Although
some studies on web uniformity have been conducted for nonwoven processes [5,6,7,8,9], there
is no available data on cross-machine uniformity for the card in yarn production line since the
web is typically condensed into a single sliver and only the overall linear density is important.
When one decides to split the web into multiple sections to feed directly to spinning heads, the
variability of the web takes on great importance.

A system has been integrated with the card to monitor the web density on-line. It employs a
computer, line-scan camera and high-speed frame grabber board (Figure 3). This system will
enable real-time monitoring across the full width of the web. This is important for monitoring
the input quality of the web to the yarn-spinning head. The system is capable of monitoring
100% of the web or just sampling a subset of the web. An image of the web is created line-by-
line with the scanning rate of the camera synchronized with the web motion when 100%
monitoring is required. In this way a picture of the web is obtained. For process monitoring and
control, a statistical sampling can be taken to analyze variation in any given section (strip) of
web. An extensive literature search did not find published data on web variability and this work
should add to the body of knowledge.

Web splitting

A proof-of-concept device has been designed and the first prototype was built. The principles
involve positive control of the web sections during division via belts that squeeze the web, and
through differential speeds, shear the web longitudinally in a short distance. Servomotors with
digital control were added to the device (Figure 4). Initially, the belts were divided into two
groups with slower and faster speeds, each set driven by one servomotor. It was found that the
slower motor’s speed was influenced due to frictional drags caused by the adjacent faster moving
pulleys. The device was then modified to use timing pulleys and belts to ensure an accurate
speed difference. Effort to synchronize the speeds of the web divider, card, and yarn spinning
heads is underway. When the concept is finalized, the apparatus can be easily expanded to 10 or
more strips. The limitation, of course, is in the uniformity of the web across the

National Textile Center Annual Report: November 2001

F98-G15
4
 Line-Scan Camera

Carding

Splitting Device

Figure 3. Photograph of experimental setup of a card equipped with web slitting, web
monitoring and control devices. The carding machine with a fine opener and chute feed
is donated by American Truetzschler, our industry partner.

Figure 4. Web divider with servomotors and digital control

National Textile Center Annual Report: November 2001
F98-G15
5
machine. It is expected that uniformity of polyester staple fibers will produce a reasonably
uniform web. A cotton web would be much more of a challenge. The web will be made as
heavy as possible in order to create the linear density needed for a coarse yarn.

YARN FORMATION BY CARD-SPINNING

In a traditional staple yarn spinning system, fibers are opened, cleaned, carded, drawn and
reduced to a roving before spinning (twisting). One of the objectives of this project is to design a
system where the short staple fibers in the carded web can be spun directly into a yarn before the
web is consolidated into a sliver. A literature research (e.g., [10,11,12,13]) shows that efforts
have been made to shorten the systems to convert short staple (cotton type) fibers into yarns, and
that in the shortened systems, carded or drawn sliver is processed into either a ring spun, a rotor
spun or an air-jet spun yarn, thus eliminating the extra step of drawing and/or roving.

We have assembled a flexible "hybrid nozzle" spinning device with the objective of straightening
the fibers, twisting the bundle with high efficiency and low air consumption, and producing a
smooth and strong yarn. Figure 5 shows a schematic drawing of the spinning device, and Figure
6 shows a photograph of the device. The device is designed to have the maximum flexibility in
processing conditions. It is placed on a height-adjustable table with wheels. The speeds of the
feeding rollers, delivery rollers and the winder can be independently set.

In conventional airjet spinning machines, a draft system is used to draft and to feed the fiber
strand into the spinning nozzle. For example, MJS801 air-vortex spinning machine by Murata
has 3-roll drafting system while MJS802 has a 4-roll system. In our Card-Spinning system,
because the card web is split to provide the necessary linear density, there is no need to draft the
web ribbon. Therefore, a pair of normal feeding rollers are used to control the rate of feeding the
web ribbon into the spinning nozzle. The vortex air spinning unit is composed of a Murata 802
MJS, twin air jet spinning nozzle, and utilizes a service air system. A negative air pressure is
formed in the first nozzle to aspirate the fiber sliver. The second nozzle twists the outer fibers
free from fiber core to wrap around the core and form a yarn. A pair of delivery rollers and a
winder is used to wind the yarn into a cone shape that is suitable for further experimental tests
and production.

Before this spinning device is directly attached to the web splitting device under development
(see previous section), we tested its operation by feeding it with separately prepared thin slivers.
The feed strand is aspirated into the spinning unit via a transfer channel. The fiber bundle is
twisted to form a yarn, which is then wound onto a paper cone.

In a conventional air-jet spinning process, drawn slivers are used as the feeding stock. Whereas
for the Card-Spinning device, card web will be fed to the spinning device directly. To study the
effect of feed stock, we have conducted a study to compare the yarn characteristics using both
drawn sliver and card sliver as the feed material. The feed slivers, of the same linear density,
were fed into our spinning device to produce yarns under the same spinning parameters: air-

National Textile Center Annual Report: November 2001

F98-G15
6
 Deliver
Roller Winder
Feeding
Roller
8.48 49cm
Control
Box
90cm

Feed roller Winder

Motor Motor

56cm Condensing Roller

Deliver Roller
Doffer Motor

Figure 5. schematic drawing of the spinning device

Figure 6. Photograph of the spinning

device

pressure of spinning nozzles, feed speeds, delivery speeds and windup speeds. The analysis on
the final yarns shows that yarns spun from card sliver are not as uniform as that of yarns from
drawn sliver. However, the card sliver, due to its loose structure, is easier to twist.

The main factors influencing the final yarn qualities include: super feeding ratio (affecting
winding tension); condenser size; distance from front roll to spinning nozzle and air pressures.
Three air-pressure conditions and two super feeding ratios were used to produce yarns with
drawn sliver and card sliver (Table 1). The effect of feed ratio, pressure ratio, and feeding stock
on yarn breaking strength is shown in Figure 7.

For the carded sliver yarns, a nozzle pressure ratio of 0.57 yielded a yarn with better final yarn
quality than the yarn produced with other ratios. For example, the tensile strength of yarns with

National Textile Center Annual Report: November 2001

F98-G15
7
 Table 1. Yarn spinning parameters: Nozzle pressures (N1, N2) and draft ratios (K1,K2,K3)
for yarns produced from card sliver and drawn sliver.
Yarns N1 N2 P= K1 = .97 K2 = .98 K3 = 1.04
(psi) (psi) N1/N2
Yarn From 60 20 0.57 Yarn_001 Yarn_002 Yarn_003
Card Sliver 40 70 3.0 Yarn_005 Yarn_006 Yarn_007
50 28 1.78 Yarn_009 Yarn_010 Yarn_011
Yarn From 60 20 57 Yarn_101 Yarn_102 Yarn_103
Drawing 40 70 3.0 Yarn_105 Yarn_106 Yarn_107
50 28 1.78 Yarn_109 Yarn_110 Yarn_111

P= 0.57 ranged from 1.6 to 2.0 kgf, while those produced with P = 1.78 and P = 3.0 were
significantly weaker (0 to 0.5 kgf). Yarn elongation and tenacity exhibited a similar trend.

For the drawn sliver yarns, a value of P=1.78 yielded the highest tenacity and elongation. A feed
ratio of 0.98 was shown to produce the highest quality yarns.

In analyzing the effects of drawing on yarn characteristics, it was shown that for P ratios of 1.78
and 3.0, the drawn sliver produces a better quality of yarn. However, for P = 0.57, the carded
sliver yields better yarn quality. This was determined by evaluating the yarn strength, yarn
elongation and yarn tenacity. This result is significant. It means that if the fiber alignment and
desired regularity of a sliver are achieved by carding machine modification, the spinning of yarn
directly from carded web may be practical. The expensive drawing process could then be
eliminated from the staple yarn production line. Further details are presented in the thesis
derived from this systematic study of processing parameters [14].

SUMMARY

This project focused on the motion dynamics of fiber clumps in an airflow field and the
development of a prototype Card-Spinning unit, which combines carding and yarn spinning in a
single machine. We have demonstrated its feasibility in producing coarse yarns. Continuing
effort is underway on experimental and theoretical analysis of air flow, web uniformity
regulation, system integration/enhancement, and yarn characteristics studies.

We are pleased that the project has received significant interest and support from the textile and
the machinery industries. The interaction with industrial partners plays a significant role in the
success of this project.

National Textile Center Annual Report: November 2001

F98-G15
8
 P=3 (N1=60, N2=30 psi)

breaking Load(kgf)
0.8
0.6
Card
0.4
Drawing
0.2
0
0.97 0.98 1.04
Feed Ratio

P=1.785 (N1=50, N2=28 psi)

breaking Load(kgf)

0.8
0.6
Card
0.4
Drawing
0.2
0
0.97 0.98 1.04
Feed Ratio

P=0.57 (N1=40, N2=70 psi)

breaking Load(kgf)

2.5
2
1.5 Card
1 Drawing
0.5
0
0.97 0.98 1.04
Feed Ratio

Figure 7. Effect of feed ratio, pressure ratio, and feeding stock on yarn breaking strength

REFERENCES

1 W. Bostock, S.M. Freeman, S.A. Shorter and T.C. Williams, “An aerodynamical study of
the opening and cleaning of cotton by existing machinery”, J. Tex. Inst., 1955, T171-190.
2 P.M. Strang, “Air shearing force separates fibers”, Am. Wool & Cotton Reporter, August
1949, p12.

National Textile Center Annual Report: November 2001

F98-G15
9
3 G. Merenyi, “Methods of control of strips and effect of air currents in carding”, Tex.
Manufacturer, 1957, p27.
4 A.L. Miller, R.S. Brown, and R.A. Ruca, “An investigation of air pressures in the cotton
carding machine”, Tex. Res. J., July 1958, p593.
5 P. Boeckerman, “Meeting the Special Requirements for On-line Basis Weight
Measurement of Lightweight Nonwoven Fabrics” TAPPI Nonwovens Conference, 1992
169.
6 L. Crook, and K. Staats, “Carded Webs—An Overview of Methods Used to Produce
Parallel and Random Webs for Print Bonding and Thermal Bonding” TAPPI Nonwovens
Conference, 1984, p21.
7 T.R. Wan, G.A.V. Leaf, C. Iype, “A New Objective Method for Assessing Fibre
Arrangement in Fibrous Webs.” J. Text. Inst. 86 No. 4, 1995, p649.
8 A. Cherkassky, “Analysis and Simulation of Nonwoven Irregularity and
Nonhomogeneity.” Textile Res. J. 68(4), 1998, p242.
9 R.K. Aggarwal, W.R. Kennon, and I. Porat, “A Scanned-laser Technique for monitoring
Fibrous Webs and Nonwoven Fabrics.” J. Text. Inst. 83 No.3, 1992, p386.
10 E. Pallay, "Method of and apparatus for spinning yarn in an air vortex in a Spinning tube",
USA patent: 4174605, Nov, 20,1979
11 S. Didek, L. Fajt, M. Reymanova, M. Markova, "Method of and apparatus for spinning
yarn from staple fibers in an air vortex", USA patent:4553383,Nov 19 ,1985.
12 D. A. Bahov, H. A. Bahov, "Vortex air nozzle for yarn spinning from staple fibers". USA
patent: 4689948,Sep, 1,1987.
13 H. Stadler, R. Ammann, "Method of producing a rotating air layer and false-twist air jet
nozzle for practicing such method", USA patent: 4934133, Jun 19,1990.
14 Kai Liu, “Yarn Spun Directly from the Card,” MS Thesis, Clemson University, December
2000.

Project Web Site

http://www.tfe.gatech.edu/faculty/wang/CardSpin/

National Textile Center Annual Report: November 2001

F98-G15
10