Abstract: Friction coefficient, blank holder force, drawbeads geometry and material properties in the actual process of Sheet metal forming all have a certain fluctuation, so they are uncertain. Traditional deterministic optimization design ignores uncertainties, which results in aggravating sheet metal forming performance fluctuation, reducing the product quality and qualification rate. Facing the polygon blank shape optimization problem which has a strong practical value, the uncertainties in the actual production process are further considered. Aiming at less information of uncertain data, non-linear interval number programming, function interval extension, response surface technology and multi-objective genetic algorithm are utilized to construct non-probabilistic robust optimization method based on interval analysis, and the practical examples demonstrate the validity and effectiveness of this method. This method greatly reduces the robust optimization requests for original data, reduces the mould design dependence on the designer’s experiences, advances the stability of product quality and decreases the rejection rate and the production cost.

Key words: sheet metal forming;polygon blank;uncertainty;interval analysis;robust optimization