Fraunhofer Institute for Ceramic Technologies and Systems IKTS
Fraunhofer Institute for Ceramic Technologies and Systems IKTS

Acoustic diagnostics – Defect detection | Signal evaluation | Quality assurance

Wear monitoring

Moving parts in machines and plants are subject to high loads. Failure can lead to downtime, costly repairs, and production losses.

Early detection of defects in wear parts, such as rollers, bearings, valves, and gears can be very beneficial from an economic standpoint. In safety-sensitive areas, such as chemical plants, failure of these parts can have devastating consequences. For example, valves that cannot be closed properly and hence allow liquid to continue flowing unimpededly must be replaced before failure occurs.

A mere good-or-bad distinction is not sufficient. Wear monitoring solutions developed at Fraunhofer IKTS offer a new approach: with the help of service life analysis, the residual life of a part can be determined. This provides an indication of when the part in question will need to be replaced.

Walzen
© iStock: morenosoppelsa

Rollers

 

Rollers in spinning machines, such as are used in the textile industry, undergo significant wear. Mechanical defects, such as cracks, friction-induced damage, and spalls, occur frequently.

State-of-the-art acoustic diagnostic methods help detect impending failure of rollers and other machine parts. They are based on typical signatures in the vibration spectra of the respective functional components. These patterns are generated by mechanical defects and can be determined through time-frequency analysis.

The method developed at Fraunhofer IKTS evaluates these signatures. Failure can hence be detected in advance and appropriate preventive measures introduced.

© iStock: ultrapro

Valves

Mechanical changes arising due to progressing age and wear affect the switching noise emitted by a valve. This noise has a pronounced spectral and temporal structure and forms a good basis for condition monitoring.

Pattern recognition methods are hence ideal for classifying these noises. With knowledge of the behavior over various lifetimes for known valves, the remaining lifetime can be determined for unknown valves. This allows impending failure to be identified and prevented. Conversely, unnecessary replacement of properly functioning valves can be avoided and maintenance costs thereby lowered.

© iStock: loonger

Hydraulic test stands

 

Until now, hydraulic test stands need to undergo maintenance at regular intervals, whether faults are present or not. Fraunhofer IKTS has developed a system for condition-dependent maintenance that obviates this costly practice.

The monitoring strategy, which is realized via structure-borne sound sensors, targets especially wear-prone hydraulic components, such as servo-driven valves. Through the oil flow, acoustic signals characterizing the hydraulic processes are generated with a very wide range of frequencies. Changes in the conditions of components through which the oil flows lead to changes in the signal characteristics. From the envelopes of the time signals and the amplitude range of the signals, features can be extracted and used for condition monitoring of the components. In this way, the necessity of maintenance measures can be assessed.