Reconstruction of Wave Forms for Parameter Correction and Pattern Recognition for Acoustic Emission Testing on Pressure Vessels

The aim of this project is to improve the analysis of acoustic emission testing on cylindrical pressure vessels. The analysis of acoustic emission tests on pressure vessels is currently based on signal parameters. Those signal parameters are extracted from the wave form as it is measured by the sensor which is the closest to the source. This approach comes with the drawback that changes to the wave form occurring prior to arrival of the wave at the sensor, potentially can influence the analysis.  In this project, a new analysis technique is elaborated which makes use of all available information to determine the waveform at the source. The new technique should show improved resistance against influences by extraneous noise or irregularities on the propagation path in order to be industrially applicable. Since the result of the new analysis technique is a corrected waveform, corrected signal parameters can be extracted. This will make the analysis technique compatible with currently used evaluation criteria. After the new technique had been developed, this project will investigate how frequently encountered irregularities on the path between source and sensor can be recognized. Furthermore, the project involves a validation study of the new analysis technique on pressure vessels in industrial environment. Finally, the analysis technique provides an improved basis for pattern recognition, which can be a basis for new applications of acoustic emission.