Computational Challenges and Solutions in Multi-Axis 3D Printing

Unlike three-axis 3D printing, the additional rotations in multi-axis means that printed slices are no longer constrained to a plane. By escaping this constraint, better toolpaths are possible which can improve surface quality, improve part strength, and avoid the necessity of support structures. Thankfully, the majority of hardware needed for multi-axis printing is already in place thanks to CNC five-axis mills and robotics, but there are still some significant challenges on the software side. In this talk, three broad computational challenges are presented and solutions are offered. Firstly, nonplanar toolpaths may be specifically tailored to meet design specifications, but there is currently no way to convey this design information between relevant manufacturing entities. G-code is not a viable option here since the exact manufacturing equipment may not be known. As a result, an international working group has been developing an extension in 3MF to capture these toolpath details.

The second challenge is about the optimal creation of layers since one is no longer bound by parallel planar slices. Many current efforts in the literature simply conform to a nonplanar worksurface, but arbitrarily curved layers can be defined throughout a model to optimize the entire printed geometry or meet goals related to loading conditions. A solution involving quadric surfaces is presented as a potential solution to this problem.

Once layers are defined, the third challenge is in defining the perimeter and infill paths on these curved layers. Conventionally, such paths are determined by 2D polygonal operations, but these will produce inaccuracies on 3D surfaces resulting in gaps or over-deposition. A summary of current approaches to this problem will be presented and a novel iterative approach is presented that minimizes nozzle interruptions while maintaining proper deposition distances.