Surface quality check for bearing balls

08 February 2021

The Center of Device Development at the Fraunhofer Institute for Silicate Research has developed a fully automated checking system for bearing balls.

High-precision ball bearings need to meet special tribological requirements to ensure their long-term, fault-free performance in drive technology – for example in gearboxes without lubricants. Currently, ceramic ball bearings offer the best technical properties for such applications.

Production of the bearing components, and the bearing balls in particular, requires high precision and process reliability. Even the slightest deviation from the optimum shape or the desired surface quality can cause bearing damage and failures.

To minimise risk, the Center of Device Development (CeDeD) at the Fraunhofer Institute for Silicate Research (ISC) has created a fully automated, robot-based inspection system for one bearing ball manufacturer.

The Ceramic Ball Check System (CBCS) has been in operation since the beginning of 2019 and its performance largely depends on the built-in machine vision systems.

The developers at the Fraunhofer Institute split the complete bearing ball check into two separate units – In the first step, the components’ roundness is checked in accordance with the standards, followed by quality control of the ball surfaces in the second step. The exact feeding of the bearing balls to the respective testing module, the passing-on of defect-free balls and the ejection of faulty products is carried out by a robot. The required testing speed of the system resulted from the preceding manufacturing process: the developers’ goal for the entire system was 2,000 balls per hour. Besides the requirement for speed, the properties of the ceramic balls posed further challenges for the machine vision system.

The surfaces to be measured are highly reflective and require a special illumination system that minimises the reflections in order to solve the various measurement tasks. In addition, the system had to be flexible, as various standardised ball sizes – with diameters of between 3 to 10mm – had to be tested with it. These were very demanding requirements for a fully automated quality inspection system.

The design team understood that the reliable inspection of bearing balls at the required speed would only be possible with the help of a machine vision system: “In the past, we have implemented around 50 systems in seven different application areas, each one tailored to the customers’ needs and in use all over the world,” explained Dr. Andreas Diegeler, head of CeDeD. “Machine vision has always been a core element of our measuring systems”. 

Despite the in-house expertise in plant engineering with integrated image processing and in-house software development, Diegeler did not consider himself and his team as being imaging experts – instead they rely on STEMMER IMAGING.

"At the beginning of this project, we had done intensive research to find out which providers had the required service portfolio," said Diegeler. "Our starting point was the special illumination we needed because of the bearing balls' reflections. The manufacturer of the appropriate lighting technology for this project was a STEMMER IMAGING partner, so we were able to obtain all the vision system components from a single source".

The vision experts were able to offer the special lenses used, the cameras and the software for evaluating the recorded images as a package that was precisely tailored to meet specific requirements, and they were also able to provide support in programming the system. "STEMMER IMAGING's machine vision knowledge was particularly evident in the required combination of the two measurement modules," said Diegeler.
 
Diegeler sees promising opportunities for the future, particularly in the enhancement of the Common Vision Blox software library to include artificial intelligence. The system is able to three-dimensionally measure with an accuracy of 0.3 µm. To achieve this level of accuracy, high-precision camera systems with a resolution of 1 µm are used.

Deviations from the standardised roundness and various surface defects – such as scratches, breaks, dents or colour deviations – must be reliably detected. According to Diegeler, one of the major challenges is that the system works with high precision and is able to detect any deviation: "The system's cleanliness is vital because otherwise even the tiniest grain of dust could be identified as a defect. This means that the bearing ball production must be optimised in this respect too".

The test routine automatically evaluates deviations from the specified standard values. If the specified tolerances for ball shape or surface quality are exceeded, the defective balls are sorted out accordingly. This enables quality controls and conclusions to be drawn about any process errors during ongoing production. Here, the integration of vision components into the system via OPC-UA was a prerequisite for digitalised production and allows for adjustments to be made for individual tasks.


Contact Details and Archive...

Related Articles...

Print this page | E-mail this page