Crack detection during bending straightening

July 14, 2020 by

Simple, clear and effectiv

Simple and clear user interface including pattern recognition with the ability to filter pseudo cracks (false positives). Within a few weeks the QASS-Straightening Team has created a new software for automatic crack detection during straightening using our new PenGUI, our programmable and extensible graphical user interface. The wishes of the users have long been directed towards a simplified operability of our measurement technology.
Up to now, however, this has been opposed by the simultaneous compulsion for more complex data analysis. The team has now elegantly combined both requirements.

Under the hood

The QASS pattern recognition works under the hood and is therefore able to learn all the working noises of the machine and to distinguish them from real crack signals in the best possible way. On the surface we show the straightening sequence, a simplified overview of the pattern search, but above all the detection statistics for cracks and a simplified shift counter.
In addition, “specialists” or particularly interested parties can display a list of the last defective components and directly display the detected signals with all QASS crack detection information.
The team has also contributed a new page for simplified sensor replacement. This page shows the structure-borne sound coupling when the screw-on torque of the sensor changes.
A curve of the sensor coupling and a red/green bar graph display make it easy to correctly screw on the sensor. The team is now working to implement the results of the signal clustering, which we have optimized for bending straightening, on a 4th page.

We have already gained experience with our new optimization concept.

The Optimizer4D equipped with pattern recognition software collects data in the straightening process for about 2 weeks (while the recognition works with the previous sensitivity). Afterwards we subject the found “crack signals” to a clustering process and get the different noise types of the straightening process. In the course of optimization, noise types are then stored in the pattern recognition system, which must immediately recognize them when they occur and no longer mistakenly interpret them as crack signals.

Share this post