Optimum Design of Microridge Deep Drawing Punch Using Regional Kriging Assisted Fuzzy Multiobjective Evolutionary Algorithm

Abstract

This paper introduces an optimized microridge-punch design for deep-drawing processes with improved formability and punch service life. Reducing the thickness of metal sheets results in a marked decrease in the microscale/mesoscale deep drawing rates. Prior studies have recognized that the drawing punch with a microridge design next to the punch nose lowers the thinning rate and improves the deep-drawing formability. However, the increased stripping force associated with the microridge design may result in deformation of the drawn cup and shorten the punch service life. Here, we present a parameter optimization of the microridge design, in which there is a tradeoff between multiple objectives in the deep-drawing process. We introduce a fuzzy inference to synthesize two design objectives, i.e., the thinning rate and stripping force, into single design fitness. We use Deform-2D software to simulate deep-drawing performance for design optimization. Since the simulation of design objectives is computationally expensive, we introduce a novel sequential surrogate-based optimization to improve the sampling and searching efficiencies. The proposed optimal microridge design outperforms the best design for both the thinning rate and stripping force of all the samples, as well as the design obtained by the Taguchi method based on the fuzzy design fitness. Compared with the Taguchi design, the proposed optimum design improves the fuzzy design fitness by 29.2%, whereby the stripping force is significantly reduced by 35.2%, with only a slight 2% increase in the thinning rate. The proposed design effectively ensures both the deep drawing formability and the microridge-punch service life.

Keywords

• Optimum design
• microridge-punch
• surrogate-based optimization
• kriging
• fuzzy assisted multiobjective evolutionary optimization