Pressemitteilung 126/21 - 20.12.2021

Christmas knowledge: Of bells and acoustic emissions

Christmas would be a silent holiday without the ringing of bells − but would you have guessed that there is a connection between bells and measuring acoustic emissions in the field of materials research?

“Jingle bells, jingle bells, jingle all the way.“ – who doesn’t know the lyrics to this famous Christmas song? In order for a bell to actually jingle all the way without fail, it is important that its material structure is flawless. Bell founders test this by conducting a knock test. They check the sound by striking in different locations and deduce whether cracks or other irregularities have formed during the production process. A very similar procedure is used in materials research– there it is called acoustic emission measurement and takes place in the ultrasound spectrum.

Of smalles vibrations and earhthquakes

So how does acoustic emission measurement work? “The researchers attach sensors to a structure or a material and record even the smalles sound waves that occur for example if cracks appear in it − even if they only measure micrometres“, describes Prof. Dr. Markus Sause from the University of Augsburg, worldwide one of the leading experts in the field of acoustic emission measurement. It can be imagined like an earthquake, only much smaller in scale, with nearly impossible to discern waves in the ultrasonic range. “The sounds tell us, what amount of stress causes a crack and where it occurs. Companies use this for instance to test the quality of products like hydrogen tanks“, explains Sause.


The graphic shows the application of acoustic emission measurement in materials research using the example of a pressure vessel made of fibre-reinforced composites: When damage occurs, for example during a pressure test, the resulting sound is detected by sensors. By means of triangulation and the use of AI, the exact location can be determined and the type of fracture process can also be determined. Both are only possible to a very limited extent without AI due to the geometry of the container and the structure of the material. © University of Augsburg

Tracking down the causes with artificial intelligence

Currently researchers at the University of Augsburg are trying to find out how artificial intelligence can be used to better interpret the recorded sound waves: “With structures that consist of fibre-reinforced composites or that have a distinctive shape – like bells – , it often is difficult to interpret the signals correctly“, he explains. The problem with fibre-reinforced composites or special geometries is that sound waves do not spread evenly in them. “A bell has many curvatures and various wall thicknesses, and fibre-reinforced composites consist of a mix of reinforcing fibres and a matrix material holding them together. Both conduct sound waves at different speeds. Local deviations in fibre orientation or fibre volume also quickly bring the accuracy of analytical calculations to its limits“, Sause explains.

The researchers therefore use atificial intelligence methods on the one hand to enable them to precisely determine the sound source in any geometry and on the other hand to identify the exact nature of the sound source. This is particulalry important if such sensor technology is to be used for permanent monitoring of structures (so-called Structural Health Monitoring), such as an aircraft landing gear or aircraft wing. “This requires the highest possible level of realiability of these systems. We are currently studying this in the European network ODIN - Optimized Design for Inspection“, says Sause.

Scientific Contact

Director AI production network
Mechanical Engineering

Media Contact

Corina Härning
Stellvertretende Pressesprecherin
Stabsstelle Kommunikation & Marketing
  • Phone: + 49 821 598-2098
  • Email:
  • Room 3002a