Integration of multivariate control charts and decision tree classifier to determine the faults of the quality characteristic(s) of a melt spinning machine used in polypropylene as-spun fiber manufacturing Part I: The application of the Taguchi method and principal component analysis in the processing parameter optimization of the melt spinning process

Abstract

Melt spinning is the most extensively used method of fabricating polymeric fibers in the textile industry. This series of studies aimed to construct an automatic abnormality diagnosis system for polypropylene (PP) as-spun fiber produced by the melt spinning process. Part I of this study aimed to construct the processing parameter optimization for the PP as-spun fiber produced by the melt spinning machine. The product quality resulting from the processing parameters of the melt spinning process included six control factors: extruder temperature, gear pump temperature, die-head temperature, rotational speed of extruder, rotational speed of gear pump, and take-up speed. The quality characteristics included fiber fineness, breaking strength, breaking elongation, and modulus of resilience. The quality data were derived from the experiments, the design of which were based on the orthogonal array of the Taguchi method in order to calculate the signal-to-noise ratio, analysis of variance, and confidence interval. Principal component analysis was then applied to eliminate the multi-correlation of the output responses and transform the correlated responses into principal components, to obtain multi-quality optimum processing parameters. These optimum parameters, including the extruder temperature (180°C), gear pump temperature (220°C), die-head temperature (240°C), the rotational speed of the extruder (7.5 rpm), the rotational speed of the gear pump (15 rpm), and take-up speed (700 rpm) would later be used to build a prediction of an abnormality diagnosis system for identification of fault processing parameters in a melt spinning machine in Part II of this study.