There are two good reasons to use industrially used diamond grinding wheels as efficiently as possible: their material value in itself – such a wheel costs up to 1000 Euros - and the costs incurred by the time required for dressing, because each removal of the worn layer and sharpening of the diamonds takes about 20 minutes; during this time the wheel and the machine are not available for grinding and therefore also not for the production process.
Until now, the machine operator has usually decided by means of a "thumb test" when it is time to dress a grinding wheel and then remove 200 micrometers each in the standard process. With a total grinding wheel coating thickness of 10 to 20 millimeters, this is a process that is possible about 50 to 100 times for one grinding wheel.
Maikel Strug, an engineering scientist at the Institute of Production Engineering and Machine Tools (IFW) at the Hannover Centre for Production Technology, has been providing new insights and new possibilities for making grinding smarter since the beginning of 2018 as part of the "Intelligent Grinding" research project. "There was a similar attempt 20 years ago," he explains, "but it failed at the limits of data processing. Today it is possible to stand next to the machine with a laptop and evaluate live what data the laser scan generates.
One challenge for the laser scans is the reflections of the diamond grains - another challenge is posed in particular by the grinding wheels with ceramic bond. In contrast, grinding wheels bonded in synthetic resin or metal or hybrid are comparatively easy to pick up and interpret.
Strug is working together with Walter Maschinenbau GmbH on the new measuring system integrated into the machine tool. The heart of their system is a laser triangulation sensor. The measuring system itself is already in place. But by the end of the project - it will be funded by the German Research Foundation until the end of 2020 - the machine should be able to judge completely independently. It must generate 3D surface parameters of the grinding wheel from the laser scan data and learn to read and interpret them in such a way that an objective recommendation for the machine operator can be derived from them: Does it have to be dressed? and if so, how many micrometers deep? The final part of the project is therefore to optimise self-learning.
The interest of grinding wheel and machine tool manufacturers is already enormous. Strug announces that he and his research colleagues will continue to present their state of affairs on many occasions throughout the year and will also be available for presentations in small groups - the entire measuring system is modular and portable.
Note to the editors: For further information, please contact Maikel Strug, Institute of Production Engineering and Machine Tools, at +49 511 762 18338 or by e-mail at strug@ifw.uni-hannover.de.
