The optimal die semi-angle concept in wire drawing, examined using automatic optimization techniques

Coupling of evolutionary algorithms (EA) with meta-models (MM) is used to investigate the concept of the optimal die semi-angle in wire drawing. Traditional process design by minimization of the wire drawing force highlights an optimal die angle which increases with friction factor and reduction ratio. When wire drawing optimization is applied on the Latham and Cockcroft damage criterion, an optimal die semi-angle no longer appears: in this mono-objective optimization, the lowest industrially achievable die angle is recommended. Thanks to EA-MM coupling, multi-objective optimizations have been performed and the Pareto optimal front has been precisely plotted so as to find the best compromise. Simultaneous optimization of damage and wire drawing force suggests a refined vision of the optimal die semi-angle concept. Choosing a lower angle than the traditional optimum allows damage to be decreased without a significant increase of the drawing force. However, it is shown that a die semi-angle slightly above the optimum should be selected, for fear of friction drift; this explains the rather high traditional value of the angle.