Sheet metal components are ubiquitous, but options to produce them are relatively limited. Existing mass production processes such as deep drawing are inflexible, and consequently expensive for small batch, high-value production. In addition it takes time to produce and set up the required tooling. Furthermore, given the mechanics of these processes, half of all sheet metal never makes it into a product, having been cut off during manufacturing [1].There has been great interest in the invention and exploration of flexible processes for the past 30 years, aiming to automate the skills of the craftsmen of earlier times. However, this has proved challenging: automated design of toolpaths for spinning remains elusive [2]; incremental sheet forming is limited by poor accuracy and high residual stress [3]; other novel processes have not flourished outside laboratory conditions. Thorough reviews of flexible forming processes are given by Jeswiet et al. [4], Allwood and Utsunomiya [5] and Groche [6].
Innovations in the technology of flexible metal forming have to date arisen either from the product of laborious experimental study or out of isolated eureka moments. Is it possible to design processes automatically? Surveys in the literature can be used to map available processes and identify gaps. Similarly, the literature can inform some design choices, but it does not provide means to design future machines for particular forming requirements. This paper therefore asks whether, rather than relying on accidents of innovation, structured means could be developed to design new sheet metal forming processes to meet customers’ needs.