Intelligent yarn grading and fabric simulation computerised system

One of the problems in the Indian textile industry has been the absence of a method to quantitatively assign a grade when attempting to grade a yarn for appearance. To solve this problem, ATIRA has successfully designed and developed a low cost Intelligent Yarn Grading and Fabric Simulation Computerised System says C S Vora, M G Solanki and Kishori M Bhatt

Cotton yarn is one of the fastest growing segments in our country, both for domestic and textile export. The general appearance of yarn, especially cotton yarn with respect to unevenness, imperfections and random faults influences to a large extent the working of yarn on machines and the quality of end product. The inspection of yarn is very necessary at all times, especially with yarns to be used for value added fabrics, since an uneven yarn will affect the appearance of the fabric and consequently its commercial value. One of the problems in the textile industry has been the absence of a method to quantitatively assign a grade when attempting to grade a yarn for appearance. In most of the Indian mills, the yarn grading is traditionally performed by wrapping the yarns to be tested on an inspection board and the grading is solely based on the opinion of the technician after viewing the yarn on inspection board and comparing with American Society of Testing and Materials (ASTM) Standard photographs of specimen representing the appearance grades. It is fully subjective. However, there is no calculated basis for which a grade is assigned a number. The result obtained by this method may vary, takes more time, and it is highly human subjective.

Another problem in textile industry is to see quickly appearance of the fabric made from particular yarn. A sample of fabric is weaved on sample loom from a yarn and then after seeing the fabric appearance decision is taken for further process. It consumes lots of time.

Digital image processing is an alternative to the conventional method, and also it is widely being used in the textile industry for various applications. Same technology can be used to automatically grade the yarn electronically and to simulate the fabric made from the same yarn on computer system.

At present imported electronics system is available for automatic yarn gradation and fabric simulation, but due to its very high cost only few mills can afford to buy in India.

ATIRA has successfully designed and developed a low cost Intelligent Yarn Grading And Fabric Simulation Computerised System under the project sponsored by Department of Information Technology, MIT, Government of India and industry partner Viral Controls, Gandhinagar. The developed system inspects yarn electronically. A high resolution single line CCD digital camera scans yarn thickness and stores yarn diameter profile data for further analysis. It grades automatically the yarn using digital image processing technique and hence it eliminates human error. It also allows to grade the yarn for different width of fabric. It converts yarn image scanned by camera into virtual 3D yarn image. It converts virtually the yarn into fabric to enable one to see any defect which would otherwise be seen only after the actual weaving, which may be too late. The system allows standardisation of yarn grading method, highlights selected defects, allows visual inspection of various fault events and gives imperfection levels, simulates the fabric of different width and patterns and eliminates human error. It also provides various reports and stores results for future use. The software developed to see the appearance of 3D fabric on various 3D models like human, sofa, chair etc.

Principle of operation

Visual examination and grading of yarn have been traditionally performed by wrapping the yarns to be tested on an inspection rectangular or tapered board of color contrasting with the yarn. In this application a high resolution and high speed single line CCD digital camera is employed to scan the yarn running at constant speed. Yarn of certain length and width is canned and its diameter profile is stored in the memory for further processing.

Scanned yarn diameter data is processed for various yarn events. Yarn event is a section of yarn with diameter over or under a predetermined threshold level for a specific minimum length.

Every yarn event having a length of equal or larger than the minimum length are noted by the software. The yarn events are then grouped by their sizes. Each event will be assigned to a pre-selected bin of specific size. Depending on the various event counts, grade is given to the yarn.

System description

A drive which consists of an AC servo motor, servo drive, drive mechanism, rollers, guides, pulleys and motorised yarn tensioner has been successfully designed and developed to run yarn at very constant speed. Yarn speed and test time can be set from 8 to 200 metres/min. through computer. A Windows based software allow to set yarn speed and yarn test time, to start and stop motor etc.

The digital image processing technique has been used to capture the yarn profile. A high resolution line scan CCD digital camera has been used and interfaced with high end personal computer. It can capture 3.5 x 3.5 micron object at high speed. Basically variation in yarn diameter (thickness variation) is measured by camera with suitable light source, optics and yarn drive. The yarn is run at certain speed and yarn image of 1000 metre length is captured in step of 0.5 mm and stored it in memory.

After yarn test is over, stored yarn profile image is analysed. Three different software modules namely yarn grading, fabric simulation and 3D virtual models have been successfully designed and developed to analyse the stored yarn profile image. These modules are windows based and developed in VC ++ language.

Yarn Grading Module

This module finds number of events of different size and length. Every yarn event having a length of equal or larger than the minimum length are noted by the software. The yarn events are then grouped by their sizes. Each event will be assigned to a preselected bin of specific size. As per the number of different events, yarn is graded.

It also converts the yarn profile image into 3D yarn image for wrapping the yarn grading board electronically. It has been successfully developed and tested. Boards of different width and type can be wrapped using scanned yarn on computer.

Fabric Simulation Module

Software module for fabric simulation has been successfully developed and tested with scanned yarn data. Various parameters for weaving fabric like yarn count for weft and warp, yarn colour, weave type, fabric density, fabric width etc. can be set and accordingly simulate fabric can be seen on computer screen. It provides 3D view of simulate fabric and hard copy on high-resolution inkjet printer.

3D Virtual Model Module

This module has been specially designed and developed to see the real effect of simulated fabric woven by scanned yarn on various 3D models like human, sofa, chair etc. Various simulated as well as scanned can be applied on D models and its effect on models can be seen from different angles. It can also be seen under different lighting effects.

Evaluation and results

Total system has been successfully evaluated in the laboratory. Three different grades and counts of yarn packages were elected for evaluation of the system and to see the degree of correlation between yarn appearance results obtained on the system and obtained by visual inspection traditional method. These yarn samples were graded and supplied by one of the local mill. They were graded by mill on the basis of ASTM D2255-60. Each yarn sample was tested on the system and data captured were analysed.

The number of larger events increases by the lower grade. For example, Grade A has 151 for the bin size of 1 and 1 for bin size of 20, Grade B has 983 for bin size of 1,80 for bin size of 5, 13 for bin size 0,1 for bin size 15 and 1 for bin size 20. Grade C yarn package which is lowest in these three grades, obviously more events were counted. The system has been successfully demonstrated more than 30 mills at ATIRA. Total functioning and operation of the system were explained and also results obtained on the system were discussed with them in detail. There was very good response from them and they provided ATIRA with useful feedback.

Conclusions

The system is very easy to operate and gives quick result of yarn quality. Due to its low cost and simplicity in operation, many textile industries can afford to buy it.

(The authors are with ATIRA, Ahmedabad)