Adaptive algorithms for NC guided polishing of complex surfaces

Thanks to new algorithms for generating NC tools, with automated polishing it is soon possible to take a preferred polish direction and the condition of the surface into account.

Highly precise components often require a final polishing so as to achieve the surface quality requested. A great deal of research is recently being conducted into replacing traditional manual polishing with an automated variant, usually performed by a robot. Parallel tracks, distributed equally over the surface, are currently generated to guide the robot. However, this method fails to take any preferred polishing direction, nor the current condition of the surface, into account. Nevertheless, new algorithms for generating NC tool tracks do offer this possibility. 

The new algorithms are based on the Hilbert curve. This mathematical model generates a continuous curve in such a way that a certain space or surface is filled entirely. The figure below provides a number of examples of Hilbert curves with differing thickness. If you go down the different lines yourself, you will notice you cross the surface in different directions, which benefits the polishing.

Researchers have integrated the Hilbery model within the generation of tool-tracks for NC guided polishing, allowing patterns with differing density on the surface to be combined. This makes it possible to cross surfaces requiring more polishing several times, while less critical sections are passed over quickly. Vectors were also added to the model, so that the user is able to influence the polishing direction. The figure below presents the comparison between a standard (a) and the new tool tracks (b).

Although the new algorithm is currently still in the research phase, lab tests are yielding extremely positive results: double curved surfaces with roughness under 20 nm. 

(Source: International Journal of Advanced Manufacturing Technology (2015))

If you would like to read more about creating high-quality surface finish, register and receive our second white paper on precision production!

Sign up to receive the white papers 

Tags: