Geometry processing algorithms for robot-based additive manufacturing

Abstract

Robot-Based Additive Manufacturing (RBAM) is an Additive Manufacturing (AM) technology that combines a welding torch, a powder nozzle, or an extruder with robotic manipulators. RBAM is emerging as a solution to increase manufacturing flexibility and the build volume of standard Cartesian printing devices. Indeed, different from 3 degrees of freedom printers, the material can be deposited in multiple directions (multiaxial deposition) thanks to the dexterity of robots. So, it is possible to avoid the need for support structures while increasing the build volume. Also, it is possible to manufacture non-uniform layers where the thickness is not constant. Furthermore, non-planar layers can be realized, following the surface curvature to increase the surface finish and mechanical properties. Finally, non-uniform path planning can be implemented, adapting the infill according to layer shape. In this context, the process planning for RBAM becomes an aspect of primary importance. Process planning is defined as all the geometry processing steps involved in linking the 3D CAD geometry with the manufacturing stage, until the generation and the validation in virtual software environments of the instructions to control the path of a robot. In particular, multiaxial deposition, slicing strategies, and path planning steps must be appropriately implemented to obtain satisfactory outputs, avoiding printing failures. So, this work presents a framework for RBAM process planning, whose main goal is to exploit the possibilities that RBAM technology offers. Approaches for each step of the framework are presented, proposing solutions to take advantage of the flexibility of RBAM. First, a multiaxial volume deposition and a uniform planar slicing with layer infill algorithms have been developed. Then, non-uniform planar and cylindrical slicing approaches have been defined. The last algorithm regards a non-planar slicing method. These algorithms have been developed and tested using Rhinoceros®, Grasshopper® and RoboDK®. These procedures are crucial steps to fully implement the RBAM process planning. Finally, experiments were performed with wire arc additive manufacturing technology, proposing a procedure to analyze the bead geometry and extract parameters for the proposed algorithms.