Monitoring of slow turning roller bearings in the paper industry
The monitoring of slow turning roller bearings is a special challenge. The signal strength for classic methods such as vibration measurement is very often insufficient to obtain a qualified statement about the condition. Despite the use of vibration measurement, unplanned failures therefore occur time and again in operational practice with slow turning roller bearings. The combination of ultrasound / acoustic emissions with artificial intelligence is an established and very efficient alternative.
Interlinked plants pose a special challenge in the paper industry. Due to the size of the rolling bearings and the time-consuming replacement, a preventive maintenance strategy (= replacement of roller bearings based on operating hours/time) is very cost-intensive. A condition-based maintenance strategy, or predictive strategy, can save considerable costs. In this case study, roller bearings were monitored using the Ultrasound Condition Monitoring System from Senzoro, which evaluates acoustic emissions / ultrasound in the frequency range from 20 kHz up to 700 kHz with artificial intelligence.
Illustration of the analysis
The artificial intelligence (AI) of the Ultrasound Condition Monitoring System was trained on the basis of more than 20,000 roller bearings. Over a period of several years, many hundreds of plant data were collected and stored in an AI-compliant manner. The remaining lifetime of the roller bearing in the case study was declared as less than three months and the customer decided to replace it, which is why "before" and "after" data is also available.
Comparison "before" and "after“
The Ultrasound Condition Monitoring System uses more than 400 different factors in the time and frequency signal to assess roller bearings. The "classic view" on images of time signal, bearing frequencies and frequency spectrum is not part of the analysis. The images of the time signal are therefore only used for illustration purposes to show the sensitivity of the system.
The high sensitivity of ultrasound also enables the safe monitoring of slow turning bearings. The combination with AI allows automatic and therefore efficient evaluation without expert knowledge.
Author: Dipl. Ing. Mag. Markus Loinig, Senzoro GmbH
Email: markus@senzoro.com
