Control of a robotic arm: Application to on-surface 3D-printing

Abstract

3D-printing, also called Additive Manufacturing, is a rapidly developing technique in manufacturing. Nowadays 20% of the printed products is a ?nal product, the rest is used in Rapid Prototyping. It is claimed that in 2020, 50% of the printed output is a ?nal product. The overall goal of this research is to print 3D-structures on curved surfaces. The print technique used is called Drop-on-Demand and uses a print head with multiple nozzles. The technique enables accurate and fast 3D-prints, but needs a 6-DOF manipulator to be able to print on curved surfaces. For this purpose a 6-DOF robotic arm, the UR5, is used. In 2013 a working prototype is developed by C.J. Kruit. It was proven that a robotic arm can be bene?cial to Additive Manufacturing. Still as a ?rst prototype there is much room for improvements. The goal of this research is develop a control strategy to be able to print accurately on double curved surfaces. Model Predictive Control is used as a strategy to minimize the error of the print. In this research also an approach to acquire a local model of the UR5 is developed. Furthermore feasibility of the desired printing path and safety of using the UR5 are topics of this research.