In this research, artificial neural networks, one of the most common branches of artificial
intelligence, were used to evaluate the diameter of the cotton yarn. The necessary data
were collected and practical tests were carried out. Then, an algori
thm for the artificial
neural network was established, which provides the possibility of determining the yarn
diameter from the input variables, represented by the count and twist yarn. Where after
the creation of many networks, one was selected which gave the lowest error rate.
research The aim of this is to find an explicit relationship between the strength of fibers, bundle and
yarns. The CV values of the strength are given as well. This provides a relatively simple method for
estimation of staple yarn strength, while a
voiding some difficulties of an empirical method. The effects of
important factors such as fiber length and thickness, twist and count of yarn are considered. The results
are verified by published data in Mechanical Engineering of Textile Industries and Their Technologies,
Faculty of Mechanical and Electrical Engineering, Damascus University. and also in Al-Kumasiah
company in Damascus city.
This research provides suggestions if we apply it it will help in estimating directly the yarn strength
without manufacturing and suitable testing of the yarn.
We have done many counts, strength and twist tests - which are considered the most important physical properties that determine yarns quality specification - on cotton yarns at the industrial researches& tests center. The researcher has depended on t
his tests results to find out formulas which match between physical properties of yarns, starting from common formulas then he has proved them by analysis of results that we got from graphic tests instruments, then we have figured out proportion constants between these physical properties from curves that we have got by tests results exemplification and we have shown if these constants describe common constants or new constants.
