Roboterbasierte multidirektionale additive Fertigung für das Lichtbogenschweißen mit Drahtzuführung

• Robotic multidirectional additive manufacturing for Wire Arc welding

Schmitz, Markus; Hüsing, Mathias (Thesis advisor); Reisgen, Uwe (Thesis advisor)

Aachen : RWTH Aachen University (2023)
Dissertation / PhD Thesis

Dissertation, Rheinisch-Westfälische Technische Hochschule Aachen, 2023

Abstract

Increasing complexity of components and at the same time increasing relevance of sustainable material usage motivate to near-net shape additive manufacturing of cost-intensive alloys. Wire-Arc Additive Manufacturing (WAAM) is also used in this context. The necessary process hardware (welding head with eccentric wire feed, ground cable, substrate plate and sensors) strongly limit the possible movement of the welding head. Rotation around the vertical axis of the welding head, which is necessary for eccentric wire feeding or suitable monitoring by sensors, cannot be enabled for complex paths. The drawbacks and challenges motivate the use of multidirectional additive manufacturing with pure object manipulation (MDAM). In this work, a process chain is developed that specifically ensures the robotic feasibility of a multidirectional welding process. The process chain depends on the specific robot and the position of the welding head in the robot's workspace. To be able to decide about the printable volume before the process starts and to optimize the welding head position on this basis, a corresponding method is developed, implemented and presented. It was shown that specific configurations of robot, substrate plate and welding head influence the printable volume. Not only the pure volume is relevant, but also the manipulability of the robot. Once the configuration has been determined, the process preparation, consisting of slicer, production planning, path planner and trajectory planner, can be carried out. The goal of path planning is to fill the slice with paths that are executable by the robot. The executability can be decomposed into the pure reachability of the necessary robot poses and the manipulability in the speed limits of the robot. To generate reachable paths, a Clustered Hamilton Path (CHP) search is used. This is based on a graph search. For this purpose, the developed path planning approximates and expands the polygon describing the slice. The polygon is then decomposed into convex partial polygons. For each partial polygon different infill structures can be generated. The result is a path that fills the slice with infills that are reachable for the specific robot. In addition to the pure reachability, process influences can also be incorporated via the weights in the graph. Furthermore, a trajectory optimization is developed in the context of this work, which generates executable trajectories tailored to the MDAM. An executability could be achieved by the local adaptation of the path velocity and the deviation from the given rotations. This adaptation of the kinematic parameters can be realized by simultaneous variation of the welding parameters. The developments of this work finally enable the execution of complex welding processes within the MDAM on the part of the robot.

Institutions

• Chair and Institute of Mechanism Theory, Machine Dynamics and Robotics [411910]