Layer-to-Layer Norm-Optimal control of Incremental sheet forming with a Data-Driven model

Abstract

Incremental Sheet Forming (ISF) is a free form process for the manufacturing of sheet metal or polymer parts. In ISF, a tool is used to deform the sheet using a customized toolpath. ISF is ideal for low batch size applications, as it does not require geometry specific tooling. However, due to the lack of workpiece support throughout the forming process, ISF typically results in high geometric error when compared to other sheet forming processes such as deep drawing. Thus, in this work, a method to improve the accuracy of ISF by applying feedback control during the forming process is presented. Specifically, a norm-optimal control scheme was used to optimize the toolpath on each layer using a data-driven model which is created using a trial part. The results showed that the control algorithm implemented in this work was able to effectively improve geometric accuracy and resulted in a 50% reduction in the Mean Absolute Error when compared to the open loop. However, the controller was found to have large errors near the step-in region of each layer which were the result of model inaccuracies.