Sensitivity analysis of cotton fiber performance to spinning process settings by fusion of knowledge and data

Cotton fiber performance and spinning process settings have significant and complex effects on yarn quality. Matching and balancing cotton fiber performance and spinning processes is crucial for spinning mills to achieve cost reduction or product quality improvement. This paper proposes a computational model for the sensitivity of cotton fiber performance to spinning process settings. It involved collecting cotton fiber samples with varying properties from 40 different bales in China. These samples were spun into Ne 32 gray fabric yarns under specific spinning process settings, and their yarn quality were measured. Rough set theory was employed to extract data reasoning results of fiber performance on yarn quality, and spinning theory was employed to induce knowledge reasoning results. A comparison between the two was conducted, followed by an analysis and discussion of the sensitivity of various cotton fiber properties to spinning process settings. This study found that the sensitivity of fiber properties to the spinning process settings varies for different yarn quality indicators. For the six yarn quality indicators covered in the paper, upper half mean length, micronaire, fiber tenacity, and percentage of trash showed sensitivity to the process settings. The results can provide targeted recommendations for adjusting spinning process settings for the production of yarn with different quality requirements, enabling the process to fully match the cotton fiber performance and enhance the transfer efficiency from cotton fiber to yarns.