A hybrid statistical framework for clearance optimization and burr height control in blanking of non-ferrous sheet metals

Abstract This study presents a hybrid Taguchi–Regression modelling framework for optimizing punch-die clearance in the sheet metal blanking process for Aluminium and Brass. By integrating the Taguchi design of experiments (DOE) with regression analysis, the study achieves precise control over burr height—one of the most critical quality indicators in blanking operations. Experiments were conducted using two input parameters: material thickness and punch-die clearance. The Taguchi L25 orthogonal array facilitated efficient experimental design, while regression modelling enabled predictive control over burr formation. Taguchi analysis revealed that a punch-die clearance of 5% coupled with a material thickness of 1.5 mm consistently resulted in the lowest burr height for both materials. Analysis of variance (ANOVA) confirmed clearance as the dominant factor, contributing 57.52% and 63.33% of the variation in burr height for Aluminium and Brass, respectively. The regression models demonstrated high accuracy, validated with maximum percentage errors of 13.38% for Aluminium and 8.29% for Brass. Predicted minimum burr heights were 0.0555 mm and 0.0525 mm for Aluminium and Brass, respectively. Signal-to-noise ratio and mean analyses further reinforced clearance as the key influence. The combined methodology significantly reduces reliance on trial-and-error methods, enhances dimensional accuracy, prolongs tool life, and contributes to sustainable manufacturing. This approach aligns with Industry 5.0 principles by fostering intelligent, eco-efficient production. Future research will focus on expanding the material dataset, implementing real-time sensor feedback, and integrating AI-driven adaptive learning to further advance smart and autonomous blanking systems. Graphical abstract