Near Infrared Quantitative Analysis Of Polyester Oil And Oligomers In Supercritical Waterless Dyeing And Removal Of Oligomer

Compared with the traditional water dyeing technology,supercritical dyeing technology has many advantages,such as no water pollution,low energy consumption,short dyeing time and so on.However,the polyester barrel yarn will have white barrel feet in supercritical dyeing,and it is easy to produce hairiness and yarn breakage in weaving.The main reasons are the precipitation of oligomers and the decrease of spinning oil content,which will have a great impact on fiber quality and subsequent weaving process.The traditional quantitative test method of spinning oil and oligomer takes a long time,uses more instruments and has poor test stability.While the near-infrared quantitative analysis technology has the advantages of accuracy,speed and no pollution.It has been applied in many fields and has become the standard detection technology in the industry.In order to accurately and quickly test the content of spinning oil and oligomer,so as to optimize the dyeing process and oligomer removal method,this paper selects the near-infrared modeling method,and selects the optimal modeling method by analyzing the model data and prediction results.The specific research contents are as follows:Firstly,the quantitative model of spinning oil is established.In this paper,UDY and DTY spinning oil samples are used for modeling,and the models are combined to test the content of different kinds of spinning oil.In order to improve the accuracy and stability of the model,different algorithms,normalization,baseline correction and band range are used.The best modeling method is optimized by comparing the R~2,prediction error and other data of the model.When selecting the modeling algorithm,it is found that the R~2and prediction errors of UDY model established by PLS1 are0.926 and 0.172 respectively,and the R~2and prediction errors of DTY model are0.841 and 0.1633 respectively,which are the best among several algorithms.When optimizing normalization and baseline correction,it is found that the most suitable for UDY model is the normalization of MSC and baseline correction of second derivative,with R~2of 0.976 and percentage prediction error of 7.9%;The pretreatment suitable for DTY model is the normalization of path length and the baseline correction of second derivative.The R~2and percentage prediction errors are 0.924 and 5.8%,and the accuracy of the model established by the pretreatment of MSC and second derivative is also good.However,UDY spinning oil model can not be predicted by using the pretreatment of path length and second derivative.In order to facilitate the combination of models,MSC and second derivative are used.When selecting the band range,it is found that the prediction error of the model established by using the full spectrum is the smallest.During external verification,it is found that the selection of samples will greatly affect the accuracy of the model.Properly eliminating abnormal samples can improve the R~2and prediction accuracy of the model,and the accuracy of the model will gradually improve with the increase of the total number of samples.Secondly,the oligomer precipitated by supercritical dyeing was analyzed and removed,and the oligomer model was established to optimize the removal method and reduce the damage of the removal process to the fiber.The oligomer was analyzed and characterized by testing the infrared spectrum,nuclear magnetic hydrogen spectrum,GPC chromatography and particle size of the oligomer without water dyeing.The analysis results showed that the white powder produced by water dyeing was ternary cyclic oligomer,its crystallinity was more than 90%and the average particle size was 2.2nm.The oligomers were removed by ultrasonic assistant method and alkali treatment method.The removal results were analyzed.It was found that the combination of J penetrant and oligomer remover could remove 95%of oligomers;When using 1g/L alkali solution for ultrasonic/alkali synergistic treatment,the remaining oligomer content is 0.264%,and when using 3g/L alkali for high temperature treatment alone,the remaining oligomer content is 0.285%,which proves that ultrasound can improve the removal of oligomers.However,ultrasonic treatment will reduce the crystallinity of the fiber,and alkali treatment will reduce the mechanical properties of the fiber.Rapid quantitative analysis of fiber oligomer is conducive to optimize and select the removal process of oligomer,improve the removal amount of oligomer and reduce the impact of treatment process on fiber.The oligomer model is established by using the near-infrared quantitative analysis method.The modeling process is similar to the spinning oil modeling process.Through the internal and external verification of R~2,error and other data,it is found that the model established by PLS1,path length,second derivative and full band method has the best effect,and the prediction percentage error is only 3.4%.