It is important to both detect and remove foreign matter during the cotton spinning process, say B J Hamilton, K A Thoney, and W Oxenham, who provide an overview of the available automated foreign matter detection devices and their underlying principles and also detail some of the recent and current research in this area.
The cotton spinning industry has always striven to increase productivity and efficiency in order to meet demands and maximise profits. These improvements have been achieved throughout the years with advances in spinning technology. As the job of creating yarn from cotton fibres has progressed to the modern age, much human labour has been replaced with automation.
In hand spinning and then manual wheel spinning, a skilled worker would spin just a single bobbin of yarn at a time (Shaikh, 2005). Later, the spinning jenny would allow one spinner to operate a machine that produced eight bobbins simultaneously (Morton, 1962). The self-acting spinning mule represented the first fully automatic system, capable of creating yarn without a full-time human operator (Linton, 1963). Presently, ring-spinning, open-end (rotor) spinning, and air-jet spinning represent the major methods of spinning for short staple fibres such as cotton. These methods utilise highly automated processes capable of high efficiency and productivity, producing a consistent product while minimising costly manual labour (McCreight et al, 1997).
Foreign materials in cotton spinning
One drawback of high automation is the relative dearth of human inspection across the different stages of development (Van Nimmen & Van Langenhove, 1998). Where previously a foreign material, which is anything other than cotton fibres, could be easily spotted and removed by any number of spinning operators, it is now incumbent upon the machinery itself to identify and eliminate foreign materials.
Both the cotton fibres and the trash that can be found within are naturally diverse. Trash can vary by type (Table 1), amount, behaviour, and adhesion to the cotton fibres (Foulk & McAlister, 2005). This makes it difficult to have a single strategy that separates all trash from cotton fibres.
Some foreign materials stand out more than others, with sizable or contrastingly coloured contaminants obviously being easiest to detect. Some foreign matter is apparent only after it is too late. For example, a polypropylene (PP) fibre might be well camouflaged in the cotton fibres until after the dyeing process when it now sticks out due to lack of colour (Schlichter & Loesbrock, 1997). By this point, the financial loss is very high because of all of the value-added work stages the product has already gone through (Van Nimmen & Van Langenhove, 1998). Difficulty in dyeing is one of the main problems caused by foreign matter along with the reduction in the strength of the yarn. These issues can become evident even with a low content of foreign material (Yang et al, 2009).
The presence of foreign matter in cotton is detrimental throughout the supply chain. For the fibre producer, foreign matter will reduce the grade and price of the cotton as well as reduce efficiency of ginning operations. For the yarn and fabric producers, foreign materials will harm efficiency as well as the quality of the yarn and of the resultant knitted or woven fabrics (Himmelsbach et al, 2006). This will carry over to the quality and performance of the end-use product as well.
While foreign matter in cotton spinning has always been an issue, the demand for improved measurement and control has grown in recent history because of demand for higher quality yarns and the prevalence of raw cotton production in areas of Asia and Eastern Europe, where there is a tendency for heavy foreign particle contamination (Schlichter & Loesbrock, 1997). Chinese textile companies incur a large financial hit due to returns on products with foreign fibre contamination (He et al, 2008). In contrast, cotton produced in the United States and Australia is invariably machine-picked, which results in a less significant foreign matter problem (Textile World, 2009). In fact, in August of 2008, mills from the US and abroad were surveyed regarding contaminants in cotton. The results indicated that 53% of mills believe that US cotton is significantly less contaminated than cotton from other countries (Thompson et al, 2009).
This paper will examine the current industry practice regarding foreign matter detection by reviewing presently available products and technologies. Additionally, potential future improvements will be observed in a discussion of current research.
Methods of detection
Ideally, detection of foreign materials in cotton yarn spinning would transpire as early in as possible. Figure 1 shows the stages of an open-end spinning process. Ring spinning has additional stages including roving and winding. The longer such materials proceed through the spinning process, the greater their impact on the quality of the process and product. For example, foreign matter not discovered until the yarn stage will have been broken up in the various machinery and spread throughout the product, multiplying the original number of particles (Jiao, 2009). This fibrillated foreign material will also be finer and more difficult to detect (Jia & Ding, 2005).
The method of removing foreign particles from yarns is to cut out a section of the yarn and then splice the ends back together (Van Nimmen & Van Langenhove, 1998). This requires time and reduces the productivity of the process as well as the quality of the yarn. The particles in the yarn can also cause end-breaks due to weakening of the yarn.
The repair of end-breaks is another area of short staple ring-spinning that has been improved through technology. For instance, spinning machinery manufacturer Oerlikon Schlafhorst offers a combination of technologies in its Zinser Modular Concept 351 ring-spinning system called FilaGuard and RovingGuard. FilaGuard monitors each individual spindle and activates an optical signal the instant a yarn-break occurs. At the same time, RovingGuard interrupts the roving feed within milliseconds, minimising material loss and preventing dangerous lapping (Oerlikon Schlafhorst, 2011).
Detecting foreign matter is not as simple as one might think, especially not when cotton is the fibre in question. The cotton itself makes things difficult because, being a natural fibre, its colour and luminescence can vary. The colour can be influenced by rain, frost, insects, and fungi during growing; and by temperature fluctuations and humidity during storage (Loepfe Brothers Ltd, 2011d). The profile of a cotton fibre can also vary depending on its degree of opening as it proceeds through the different processing stages (Schlichter & Loesbrock, 1997). This is important because any detection method must distinguish foreign matter without falsely identifying cotton as a contaminant.
Currently, manual sorting of foreign materials is still widely popular, especially in Asia. However, this antiquated method is time consuming and labour intensive (Jia & Ding, 2005). In fact, a mid-size cotton textile enterprise might employ 300 to 400 workers to pick out foreign fibres every day (He et al, 2008). Given the degree of automation used throughout the remainder of the spinning process, it is desirable to have a foreign matter separating system than can keep up with the efficiency found throughout the rest of the plant.
Selecting an automated foreign matter separator
A variety of automated foreign matter separators are available to cotton spinners. A common detection principle, one which would identify all foreign matter in cotton, does not exist (Himmelsbach et al, 2006). However, a suitable system can be chosen based on the needs of a particular plant. The most important factors to consider are presented in.
The differences between various foreign matter detection systems can be the sensor itself, the type of illumination, the presentation of the material to the sensor or any combination thereof. Depending on the needs and resources of a particular plant, an automated foreign matter separator system can be chosen with a combination of these characteristics, which are summarised in Table 3.
A photo sensor is the simplest and cheapest sensor that can be used to detect foreign materials. It simply detects differences in brightness as the fibres pass by. Ultrasonic sensors work on the principle of sound; solid objects are detected because they have solid surfaces, which will reflect more sound. However, these sensors do not differentiate between cotton and other fibres. Charged couple devices (CCD), or colour sensors, are able to recognise colours. 1-CCD cameras have one microchip that analyses all colour components (red, green, and blue), while 3-CCD cameras (Figure 2) possess a different chip for each of these colours (Textile World, 2009). By separating the three colour components using a beam splitter prism, each one can be evaluated individually and simultaneously (Truetzschler Spinning, 2011b).
Type of illumination
Regular fluorescent tubes are used in differentiating between cotton, which does not reflect light, and materials such as transparent foil, which reflect the light back at the colour sensor (Schlichter & Loesbrock, 1997). Ultraviolet (UV) light is utilised in detecting those objects with stronger UV reflection than regular cotton, such as polyester, polypropylene, or bleached cotton fibres (Textile World, 2009). The UV lights can come from tubes or, as is the case for the VTect by Jossi Systems, an LED circuit (Melliand International, 2011). Polarised transmitted light can be used toward detecting objects with varying degrees of transparency, such as polyethylene. This system makes use of the physical properties of plastics that make them appear coloured in polarised light (Truetzschler Spinning, 2011a).
Presentation of material
The classic way of presenting the material to the sensor is to monitor the fibres as they pass through a regular card chute. This method does not put additional stress on the fibres. However, because the speed the fibres are moving at any time is undefined, excess material has to be eliminated by the air nozzles to make sure foreign materials are completely removed. Presenting the material on a rotating needle roll solves this issue. It also opens up the material, thereby exposing more foreign materials. Simultaneously, it puts all of the material at a uniform depth, making the illumination intensity the same for all fibres. The major disadvantage of this system however, is that mechanical stress is placed on the fibres with a needle roll (Textile World, 2009).
Some systems present one side of the material to the detection system, while others position sensors on both sides. Scanning the fibres from both sides increases the efficiency of the detection process because it does not miss contaminants that are hidden behind tufts of cotton (Farber et al, 2010).
Barco is a technology company that designs and develops visualisation solutions for a variety of industries (Barco, 2011a). They contribute to the textile industry with their optical sensor technology, providing several products, which can be added to existing cotton yarn spinning plants for the detection of foreign materials, including CottonSorter, SliverWatch, and ABS.
For the blowroom, Barco offers CottonSorter, which is designed to be added after the bale opener. This machine detects foreign materials with “ultra-fast” CCD cameras and removes them from the cotton tufts with high-speed pneumonic guns (Barco, 2005).
The CottonSorter sends the cotton tufts through a transparent tunnel where the fibres are illuminated (16 high frequency light tubes) and four CCD line scan cameras (2,592 pixels each, 8-bit colour) observe from both the top-down and bottom-up. These images are processed and foreign fibres are identified on the basis of colour and size difference in comparison to cotton fibres. The machine judges the downstream speed of the fibre flow and ejects identified contaminants with the high speed air guns at the optimal moment (Barco, 2005).
The touch-screen user interface enables the operator to set tolerances. This allows for the machine to be more strict or lenient depending on the end-use of the cotton or other conditions. The machine also reports the size and colour of foreign fibres in case the plant manager has a need for this data (to evaluate cotton suppliers for example) (Barco, 2005).
Barco’s SliverWatch is designed to be installed at the creel of the first drafter or at the creel of the lapper. Upon detecting foreign materials among the cotton fibres, it will halt the draw frame or lapper, allowing for manual elimination by the operator (Barco, 2007).
SliverWatch passes the sliver through a sealed transparent guide, which is illuminated with either transmitters or light emitting diodes (LEDs). Receivers are positioned to absorb the amount of light in the sealed sensor. Foreign material will absorb a different amount of light than does cotton fibre (Figure 3), resulting in a different reading in the receiver. This is the function by which this system detects contaminant fibres. SliverWatch will also detect sliver breakage or the lack of a special coloured fibre if it is expected. This is the case with fancy yarn or cotton/black polyester heather yarn production (Barco, 2007).
Like CottonSorter, SliverGuard also allows the operator to decide on tolerance levels, which include minimum length and colour intensity. The sensors all connect to a central monitoring unit, which provides information to the operator, including the location of contaminants in need of removal during a machine stop (Barco, 2007).
Barco’s ABS is designed for open-end (rotor) spinning machines. Its purpose is to detect any foreign fibres that have made it all the way to the spinning stage of processing. It is said to be compatible with any open-end spinning machine regardless of the type of yarn clearer it has, and the system works on all yarn counts (Barco, 2011b).
The system is placed on the yarn withdrawal tube in the spinbox, as seen in Figure 4 below. It consists of LEDs and photoreceivers. Part of the tube is made transparent so the receivers can sense the total amount of light within the tube at any given time. Like SliverGuard, ABS receivers will notice the change in signal caused by the light absorption of the foreign material. Similar to the other Barco products, data from ABS receivers is collected and displayed, and tolerances can be set (Barco, 2011b).
The Trüetzschler Group is an international company involved in three areas of the textile industry. They provide card clothing services, spinning preparation machinery, and nonwoven production machinery (Truetzschler, 2011). Trüetzschler Spinning supplies technology for foreign matter separating at both the blowroom and carding levels of cotton yarn spinning.
Trüetzschler Spinning created technology that will detect and remove all types of foreign particles and fibres from cotton. They contend that in order to detect all of the various types of foreign contamination, numerous different techniques need to be employed (Truetzschler Spinning, 2011a).
This is the idea behind the modular system called the SECUROPROP, which features five modules. One of the modules is for dust removal, and another for maintaining continuous flow in the blowroom. The remaining three modules are designed to detect foreign material as described in Table 4. The individual plant can decide which modules to integrate into its SECUROPROP machine based on circumstance (Truetzschler Spinning, 2011a).
These modules are controlled using a touch screen interface similar to those on other Trüetzschler Spinning machinery (Truetzschler Spinning, 2011a). There is a teach-in procedure in which the machine learns the brightness values and colour components of the cotton being processed. These are then then used as reference values against which the foreign contaminants are compared. Once a foreign material is detected, it is removed with a newly developed air nozzle system. This ejection uses a 30 millisecond burst of compressed air (Farber et al, 2010).
While Trüetzschler Spinning hopes to detect most foreign materials in the blowroom with its SECUROPROP system, contaminants that slip through the cracks can be spotted with the NEPCONTROL TC-NCT system. As the name implies, this system is primarily for detecting neps. However, the electronic camera positioned under the take-off roll will also detect trash particles and seed coat fragments. It does this by filming the entire working width of the card, capturing 20 images every second (Truetzschler Spinning, 2011b).
Loepfe Brothers Ltd is a smaller company (approximately 160 employees) that concentrates on textile online quality control. To this end, they provide a variety of products for spinning and weaving that can be applied to pre-existing textile machinery (Loepfe Brothers Ltd, 2011a).
They were founded in 1955 and have been a part of the Barco group since 1994. Loepfe is the producer of the YarnMaster Spectra+, which is an optoelectronic yarn clearing system. This means that at the end of the ring spinning process, in the winding stage, yarn faults are cut out of the yarn and the two ends are spliced back together. YarnMaster Spectra+ is digital, micro-processor controlled, and modular. It detects an assortment of faults, including foreign fibres, neps, short thick places, long thick places, thin places, and splices. It is also an intelligent system, meaning it will selectively clear based on the end-use of the yarn. Therefore, if perfect yarn is not required, efficiency will not be hindered by an abundance of cuts and splices (Loepfe Brothers Ltd, 2011c).
Different models of the YarnMaster Spectra+ are available. Only the most advanced model, the YarnMaster Spectra+ 900, is capable of detecting “F-class” foreign materials including PP, hair, and jute. Loepfe has also developed a foreign fibre classification system through which the YarnMaster Spectra+ 900 is capable of assigning classes to any foreign material it detects (Loepfe Brothers Ltd, 2011e).
There are 64 fault classes in the Loepfe foreign matter standard, which is based on the YarnMaster classification chart. The classes are based on length, darkness, and fineness. Having these classes allows the operator to select which classes of fibres the machine will clear (Loepfe Brothers Ltd, 2011d).
The YarnMaster Zenit is the top-of-the-line offering from Loepfe for yarn clearing. It is capable of detecting contaminants that the Spectra+ is unable to, including transparent or fine contaminants, and is also effective at detecting foreign matter in coloured yarns (Loepfe Brothers Ltd, 2011f).
The YarnMaster Zenit FP model is the one capable of the most advanced detection. This is because of the P-Sensor, which uses a triboelectric measuring principle, in which the fibres exchange electrons with the sensor. Foreign fibres are detected because they result in a different voltage than does cotton when passed over the electrode. This is what makes the P-Sensor capable of finding fibres that are too transparent, fine, or cotton-coloured to be detected optically. With the P-Sensor combined with optical sensors, the YarnMaster Zenit FP system provides comprehensive yarn clearing (Loepfe Brothers Ltd, 2011b).
USTER Technologies AG offers systems and services for quality control in textiles from fibre to fabric, providing manufacturers with tools that help optimise quality and process efficiency (USTER, 2011). Among its systems, USTER offers a fibre measurement system called AFIS PRO 2, a yarn measurement system called TESTER 5, and a yarn clearing system called QUANTUM 3.
The USTER AFIS PRO 2 is an offline fibre measurement system. Samples of cotton are taken from the production line and tested. The results indicate the quality of product that is going through the factory. It separates the fibres and counts each and every fibre, nep, and trash particle to give the operator a very detailed assessment of the cotton at whichever point in the process the sample was taken from. While this system can find seed fragments and related trash, it does not detect foreign fibres such a polypropylene (USTER, 2007a).
The USTER QUANTUM 3 is an online yarn clearing system. It is capable of helping the operator to decide on the most efficient clearing limits given the quality of the fibre, the final fabric it is needed for, and how many cuts can be afforded (USTER, 2010).
The QUANTUM 3 comes complete with a technologically advanced foreign fibre identification system. The foreign matter sensor utilises multiple light sources to detect all coloured and shorter defects. It separates both foreign fibres and vegetable matter, and classifies them. This allows the more disturbing foreign fibres to be removed while vegetable matter can be left in the yarn, reducing the cost of cutting and splicing. Additionally, the QUANTUM 3 can be upgraded to have polypropylene clearing as well (USTER, 2010).
The USTER TESTER 5 is an offline yarn measurement system. It gives yarn manufacturers a detailed look at the quality of their product. The TESTER 5 measures everything from hairiness to diameter, but most relevantly, it detects and classifies foreign fibres. This is accomplished with the USTER FM sensor, which measures and classifies (Figure 5) any foreign matter that has made it all the way into the yarn (USTER, 2007b).
Researchers continue to look for more effective methods of interpreting the RGB (red, green, blue) measurements of basic colour image processing. In Jiao (2009), a new algorithm was created and tested for the detection of foreign matter using the same image gathering equipment already widely in use. The results were acceptable and met speed requirements; however the algorithm was not compared to other models.
Likewise, in Yang et al (2009), a new method of interpreting colour images is suggested. This is a four step process, and the idea is to better enhance the image and separate between objects and background. Ultimately, the researchers hope to provide a way to process images more effectively and efficiently. In Li et al (2010), multi-class support vector machines (MSVMs) were used to interpret the images gathered by pre-existing camera technology. MSVMs are classification methods based on statistical learning theory. The MSVMs tested in this study used three inputs gathered from the images: colour features, shape features, and texture features.
Two of the three MSVMs tested were considered successful in this experiment and further research was conducted on one in Yang et al (2011). This particular study focused on the speed in which images can be processed in the search for foreign fibres using the MSVM tested in the previous research. The researchers were satisfied that they met the requirements of the textile industry and would further hone the process in future research.
Much of the current research in the topic of foreign materials in cotton spinning is focused on the analysis of the images that can be gathered with current technology. However, there has been some exploration into methods not currently being utilised as well. For example, in He et al. (2008), the use of infrared absorption is explored. The key for any detection method is to measure the fibres on a characteristic in which foreign materials will differentiate themselves from cotton. This study works on the fact that cotton has a different infrared reflection spectrum than other materials (plastic thread and hair were used in the experiments). A neural network system was trained with the results of measuring all of these materials. Once trained, this system correctly identified the different materials, proving that infrared absorption is a viable option for future foreign material separators.
Thomasson & Sui (2000) attached sensors directly to a mechanical picker in the cotton fields at Mississippi State University. This allowed for trash content to be measured at the earliest point possible. Similarly, Schielack III et al (2009) created a system which would collect samples as they went through the duct of the harvester in order to take images and then returned the samples to the duct while saving the location of each sample. The images were taken four at a time, with each of the four being through a different filter and were taken in the visible and near infrared (NIR) regions of the light spectrum. NIR spectroscopy is particularly adept at detecting PET fibres in cotton because of the differences in the absorbance of NIR light between cotton and PET. It has been used in the past to determine the cotton content of blended fabric (Rodgers & Beck, 2009). While most products and research in this field focuses on detecting the foreign material somewhere along the spinning process, Farber et al (2010) showed the results of a trial done at an Indian roller gin. This test was done using the Trüetzschler Securoprop SP-FPU discussed earlier. The complete modular system was used, meaning that the colour module, PP module, and UV module were all utilised.
The detection process was expected to be less efficient at finding foreign materials that during the spinning process due to the reduced level of fibre opening at the gin. While this was true in the results, the system proved to be more efficient than expected. It was suggested that perhaps if this system, which spotted 68% of detectable contaminants, was used in combination with a similar system in the spinning mill, where it would typically find 75% of detectable contaminants, the result would be a total efficiency of 90% (Farber et al, 2010).
Research is also being done to reduce natural foreign matter picked up during the harvesting process. This relates to the topic of this paper because having less contaminants in harvesting would lead to the presence of less contaminants throughout the remainder of the processing. Sui et al (2010) determined that for each picking and ginning step of mechanisation that cotton goes through, foreign particles become smaller in size. The processes also increase neps, short fibre content, and the tightness of the particle-fibre attachment.
Researchers also used an imaging process to better understand the attachment mechanisms between contaminants and cotton fibres (Thomasson, Sui, Byler, & Barnes, 2008; Thomasson, Sui, Byler, Boykin, & Barnes, 2009). This was done under the assumption that a better understanding of these mechanisms would enable more efficient cleaning techniques that would cause less damage to the cotton fibres. The research found that the most common way the particles attach to the cotton is that cotton fibres stick into cracks in the foreign particles. However the results did not reveal any opportunities for novel cleaning methods.
The method with which the cotton is picked was also found to cause a difference in foreign matter. Faulkner et al (2008) compared picker and stripper harvesting systems regarding the effect each method has on cotton. Picker harvesters remove cotton from the boll using spindles whereas stripper harvesters employ brushes and bats. In the section of the research regarding foreign matter content, the spindle harvested cotton was determined to have much less contaminants. This leads to a better cotton grade and less subsequent cleaning work. Liu et al (2010) sought to improve upon high volume instrumentation (HVI) colour measurements by looking at cotton fibres in the visible, NIR, and UV ranges of light. This was done on baled cotton from the gin, before entering any spinning processes. The results pertaining to the visible trash content revealed that the system, which used the entire spectrum was more accurate than when only parts of the spectrum were used.
Measurement and control of foreign matter in cotton spinning will always be important because the existence of contaminants has a negative impact on both process and product quality. Foreign fibres in particular are easily camouflaged among cotton fibres. However, they later make their presence known in the form of end-breaks during processing, failures during fabric production, or unsightly faults in a dyed end-product.
There are two ways to address this issue. The first one is to stop the foreign matter from getting into the cotton in the first place. While there has been great strides made in this area, particularly with machine-picked cotton in places like the United States, it is entirely more difficult in places like Asia where older methods are still used.
The other avenue to take in regards to this situation is to detect the foreign material at some point before it makes it into the end-product. It makes the most logical and financial sense to detect and remove the contaminant as soon as possible so it will hinder a minimum of processing stages.
Technology exists and is available from several textile manufacturers that will detect foreign matter by means of a variety of sensors. The cost of this technology may preclude some yarn spinners from benefiting from such advancements. Also, depending on the end use of the yarn in question, it may be less important for certain manufacturers to avoid foreign matter at all costs.
In the end, it comes down to each individual situation. While strides have been made, there is still room for improvement in the measuring and controlling of foreign matter in the cotton yarn spinning industry.
Note: For detailed version of this article please refer the print version of The Indian Textile Journal April 2012 issue.
B J Hamilton
College of Textiles,
North Carolina State University,
2401 Research Dr,
Box 8301, Raleigh,
NC, 27695, USA.
K A Thoney
College of Textiles,
North Carolina State University,
2401 Research Dr,
Box 8301, Raleigh,
NC, 27695, USA.
College of Textiles,
North Carolina State University,
2401 Research Dr,
Box 8301, Raleigh,
NC, 27695, USA.