Intelligent optimization of flame-retardant/antistatic core-spun composite yarns via roving core-spinning and filament wrapping

To meet the growing demand for multifunctional protective textiles, flame-retardant/antistatic core-spun composite yarns were developed by combining roving core-spinning with filament wrapping. Nine yarn samples were systematically prepared by varying core twist (20, 25, 30 twists per meter (T/m)) and feeding spacing (1, 3, 5 mm), and their mechanical and functional properties were comprehensively evaluated. Optimal overall performance was achieved at a core twist of 25 T/m and a feeding spacing of 3 mm. Compared to the control (20 T/m core twist), the breaking tenacity of the optimal sample increased by 30.65%, and the coefficient of variation (CV) of yarn evenness decreased by 17.3%. Increasing feeding spacing from 1 to 3 mm improved breaking tenacity by 7.88%, reduced yarn evenness CV by 19.45%, and decreased short and long hairiness counts by 35.38% and 40.79%, respectively. All samples demonstrated excellent flame retardancy, while antistatic performance exhibited a nonlinear trend, peaking at a 3 mm feeding spacing. A backpropagation (BP) neural network model was established for further process optimization, predicting a global optimal combination of 26 T/m core twist and 2.9 mm feeding spacing. The response surface methodology subsequently validated the BP model, confirming the accuracy of these predicted parameters. Experimental validation using the BP-predicted parameters confirmed positive optimization across all performance indicators, fully verifying the model’s accuracy and reliability. This study offers a practical foundation for developing high-performance flame-retardant and antistatic composite yarns and advancing protective textile technologies.