Innovative metamodelling-based process design for manufacturing: an application to Incremental Sheet Forming

摘要

The use of metamodelling techniques in process design has become indispensable to perform faster solutions reducing time to market. This approach allows the implementation of decision support tools which are easier to use than the conventional numerical simulations. In this paper, a robust metamodelling technique has been designed and its feasibility has been validated for the crucial problem of localised thinning in sheet metal forming process. The proposed methodology is based on the innovative integration between the Design of Experimental statistical method and the Kriging one. This approach, in fact, allows to analyse contemporary the continuous and categorical factors and, as a consequence, to define a single tool for changing process conditions (i.e., material and product shape). To test the reliability of the mathematical approach, the same was performed for the case study of Incremental Sheet Forming, a process strongly affected by the not homogeneous distribution of the thickness. Taking advantage of this strategy, a wide experimental investigation has been performed to build the base of knowledge of the problem both for the metamodelling design and for the validation of the decision support tool; moreover, the experimental data were utilized to set and validate a numerical model, which was subsequently used to enrich the dataset. The proposed metamodel, suitably modified according to each process peculiarities, can be generally adapted for sheet thickness prediction.