Research@AT: #3DPrinting3

3D printers typically operate layer by layer. They build up the geometry based on thin layers of material that are applied sequentially from the bottom up. This means that the final product is an anisotropic component that exhibits different mechanical properties within a layer and between layers. For some components, it can be advantageous to align the layers with the component’s principal stress. This approach is used in many composite materials, such as CFRP or GFRP components, which utilize internal fabrics that are clamped into a mold and coated with epoxy resins.

Attempting to apply this technology to 3D printing would mean giving each layer not just a two-dimensional shape, but a three-dimensional surface. Standard 3D printers do not support this type of control, whereas 6DoF robots do. The first step in exploring this idea is therefore to develop algorithms that control a 6DoF robot using G-code, the standard input format for 3D printers.

These are some test results showing a virtual industrial robot controlled by G-code generated by a slicer, as is typically used for 3D printers. This involves a huge amount of data, and the robot control algorithms must be able to process it properly. The main result here is that the industrial robot is controlled by MATLAB but it still runs in a real-time execution environment. I call it Rapid-Robot-Prototyping.

Author: Prof. Dr.-Ing. E. Engels

Figures: E. Engels

Back to News & More