If material cracking occurs, an energy accumulated in the material is released as an elastic wave. The wave propagates from the source through the specimen body. On the specimen surface, the wave may be detected by a piezo-transducer as a burst of an electric signal. However, a lot of physical phenomena produce signals, e.g. friction or electrical noise. There exists a defined AE event and its parameters. The AE event is registered if the measured signal crosses over a defined threshold level and is considered as terminated if the burst signal disappears. The common problem of advanced inspection method with rich instrumentation is the correct data interpretation resulting in a reliably conclusion statement.
Acoustic emission analyser is a measuring device with extremely high sensitivity, time resolution and data flow capacity demands. As the AE method is used for inspections of large structures for a long time, new techniques of data processing are necessary to design for specific environment of the aircraft industry. Small dimensions of structural part lead to multiple reflections of single wave and therefore an event should be detected carefully in the measured signal. For practical usage, a “death time” period is defined as the period after one event has ended and another event is not registered. Real events occurred in this period or during the first event duration are missed. An advanced algorithm should be used if the missed events are necessary to be detected, e.g. event sample analyzing, or artificial neural network design. Another aspect of thin-walled structure is coming, as the measured event shape is often deformed by wave dispersion along the travel between the emission source and sensor location. The source identifications based on the frequency spectra analysis are often useless for monitoring of an airframe parts. However, some more robust ways are still available, for example an adaptation of a simple parameter of AE rate or correlation of AE data and actual load level.