Identify components without marking

© BMWK
Marking-free component identification of milling tools

Since 2020, research has been conducted at the Institute of Production Engineering and Machine Tools (IFW) together with several project partners as part of an AiF grant on marker-free component identification. The goal of the project was to transfer a method from the Collaborative Research Center (SFB) 653 to industrial series production. "A major aspect of the research was focused on the metrological recording of the surface under conditions close to production" explains project member Hendrik Voelker from the IFW. With the end of the research project, proof of functionality could be provided to industrial partners.

For years, the German mechanical and plant engineering industry has been fighting against returns caused by faulty production and against product counterfeiting by third parties. In this context, the monetary damage caused by plagiarism in 2020 amounted to 7.6 billion euros for German companies alone. To track components and thus optimize manufacturing, companies often use active markings (e.g., QR code, RFID tag, etc.) applied to the surface. Two major disadvantages arise in this process. On the one hand, at least one additional process step is necessary in manufacturing and, on the other hand, the markings have insufficient counterfeit protection.

Your surface can do more!

The project at the IFW aims to counteract these disadvantages with the method developed in SFB 653. The innovative method of mark-free component identification makes use of the stochastic features that arise during machining. By measuring the surface with a high-resolution industrial camera, the features can be extracted from a 2D profile. In this process, the 2D profile is transferred into a local spectrum by the continuous wavelet transform (CWT). Then, the positions of the frequencies with the highest match are identified from the local spectrum and defined as features. "The arrangement of the features is so unique that millions of components produced in the same way can be distinguished" adds Hendrik Voelker. In SFB 653, the method was successfully validated experimentally using 500 ground workpieces. In the course of the project at the IFW, the method could now be validated for other manufacturing processes.

Successful identification of milled and turned surfaces

To achieve the project goal, the first step was to qualify the use of an industrial camera for surface measurement. The industrial camera is highly cost-effective due to low acquisition costs, a short cycle time and is easy to integrate into the production process. In addition to surface grinding, the industrial camera was also used to qualify external cylindrical grinding, external longitudinal turning, flank milling and face milling for mark-free component identification. Due to the use of a defined cutting edge, turned and milled components tend to exhibit a lower proportion of stochastic influencing factors and features than ground workpieces. Nevertheless, as for grinding processes, an identification reliability of 1020 must be achieved. For this purpose, the profile cross-section must be adapted for identification. The profile cross section determines which surface features are used for identification.

In order to determine the robustness of mark-free component identification, the influence of mechanical wear and corrosion was determined during the project at the IFW. For this purpose, in addition to experimental test series, the influence of e.g. scratches or contact wear on the surface was determined. Subsequently, the characterization was used to simulate the percentage change of a 2D profile. The simulated results, as well as the experimental test series, could show that identification was still possible despite a surface change between 35 - 45%. "An enormous robustness of the method against external influences could be observed. Nevertheless, it is a good idea not to use functional surfaces for identification" notes Hendrik Voelker.

Successful validation of markerless component identification in industry

"Overall, we are seeing a great deal of interest from industry in the research project," notes Hendrik Voelker. For this reason, the applicability of mark-free component identification in series production was validated in cooperation with several companies. In the process, the method was successfully put to robust use. "Through the project, the requirements for the measurement technology could be researched, which showed that already existing measurement technology at the partners can be used, so that nothing stands in the way of putting the method into operation" says Hendrik Voelker. The results showed that robustness in production depends on the measurement equipment. It is imperative for the commissioning of the method to find a robust setup, which provides repeated measurements with the same quality.

After the successful completion of the project, the industrialization will be continued in different cooperations. Should you also be interested in this or in further information, Hendrik Voelker from the Institute of Production Engineering and Machine Tools will be happy to assist you at +49 511 762 8078 or by e-mail at voelker@ifw.uni-hannover.de.