This work has presented a design strategy that can be employed for a variety of process constraints, and potentially to a wide array of challenging target parts. A general optimisation solver has been used in this work, but the method could be improved in future work with faster algorithms – see for example Raithatha The high variability in the direction of nodal forces in Fig. 11 is a consequence of the pure shear constraint imposed throughout this deformation. Future work will aim to reconcile the predicted forces to available tooling options, through iterative refinement of the process constraints to be more representative of the tractions that could be applied by real tools, as indicated in Fig. 1. This could include use of template forms of nodal force that could exist under a tool, and a clustering algorithm to classify regions of similar required forces.