Optimizing the yarn quality by design of experiment

Abstract

The manutactunng process, especially in a reputed industry like Padma Textile mlils Ltd (Group of Beximco), is equipped with the modem and sophisticated machllie and equipment However, there IS still room for optimizing the levels of processing parameters Three machines such as cardmg, ring frame & winding are such machines Experiments were conducted by the deslgn of experiment (DOE) approach to optimize some selected parameters with the aim of making greater amount of quality yam with the same amount of cotton, A terminology is used in D O.E to 'ndicate the target parameter and the higher the yam quality, the better is the situallOn for the industry. But there is a precondition of increasing the quality of yam must be achieved meetmg certain set condition. The condinon of the yam quality is measure of material utIlization Three machines are operated maintaining a range of set conditions, These conditions are vaned by the operator according to operation, the demand and capacity. To take into account the natuml variation in the response of the output, seveml experiments were conducted at the same set conditlons of the parameter. The average was taken as the esnmated of the true outputs. The quality standard was verified in each case PTML has to be very serious on the quality of its yam, Presently the rolll controls the process parameter accordmg to the manufacturer suggestion or by the expenence gained by tlte operating personnel. It ISstrongly felt that the yam qualrty can be improved by the process parameters thorough scieotific observanoo hke factorial experiments. All tests are conducted randomly to eliminate errors and considered mdividuals section to get accurate results, To avoid the varIation of data and complexity Tn the mdividuals section, each parameter has to be maintained within certain mnge and the combined effect of parameter determines the final yam quality, The individual section was nOl synchrolllzed because a test or series of test in which purposeful changes ""ere made to the input variables of a process or syslem so that onc may observe and identifY the reasons for change in the output. Carding machme is used to produce sliver and =penments were conducted to lind the mOSldeslfed combination of the levels of carding delivery speed, flal specd and cyhnder speed, These three factors were vaned into lwo levels and the,r influence was determined. The two levels of the factors are 80 and 1DO mlmin for delivery speed, 135 and i53m1mm for flat speed, 375 and 426rpmfor cylinder speed. All the lItree factors/parameters were found 10 have individual effect on the quality. Small interacllve effect of the parameters on lite qualrty was observed. A dehvery speed 80m/min, flal speed 135m/min and cylinder speed 426 rpm were found to increase the production S'gnificance tests were conducted by the analysis of varianCll, II was found that lite delivery speed, flat speed and cylinder speed. all three parameters had significant effcct on the quality. The interactions between cylinder speed & flat speed and cylinder speed & delivery speed have significance effect on the quality. In all cases the desired levels of quality standard was verified, Ring frame is performed to produce yam and experiments were conducted to find the most desired combination of the levels of ring speed and twist per inch, The level are varied WIth into three levels i.e. ring speed 18100,18600 and 19000 rpm and twist per Inch 25.18,25.48 and 25.78, Both the ring speed and twist per Inch were found to have main effect on the quahty, With the experimental limits of the factors, the mll! quality was observed for nng speed 19000 and the twist per inch 25.78. The most desirable conditi(}lI was observed at ring speed 18600 !pm and twIst per lOch 25.48. A significance test was illso conducted by the analysis of ~ariance, It was a significant interaction effect of the ring speed and twist per inch on the quality. Winding machine is used to produce cone and experiments was conducted to find out the most desired condition It may be mentioned here that each set condltion, experiment was rephcated for three times With three values standard deviation was calculated to cheek the variability Quality standard, a parameter to measure the retained material, was cheeked individually with very experiment to confirm the minimum acceptable level At wlllding speed 1450 mlmin, the variability was the lowest leading to the conslstency and customer satisfaction, Maximum quality standard was observed to occur at 1350m/min,