Automated fatigue crack detection for structural health monitoring of metallic aerospace structures
Lead Research Organisation:
Cardiff University
Department Name: Sch of Engineering
Abstract
Structural health monitoring (SHM) offers the ability to monitor structures, in real time, throughout their service lives to ensure they are safe for operation. Such technology has the potential to provide significant savings in maintenance and repair costs whilst increasing safety. For this reason there is currently a significant technology pull through for SHM techniques by the aerospace, energy, transport and infrastructure sectors.
Acoustic emission (AE) is a passive SHM technique that is able to globally monitor large structures in real-time by detecting small amounts of energy that are released when damage grows in a structure. However, like many SHM techniques acoustic emission requires operator interpretation of data and can be less reliable in complex structures, which is seen as the greatest barrier to industrial implementation. This problem is made all the more challenging by the presence of uncertainty which results from measurement noise, lack of system knowledge, variability in operating conditions, etc.
The applicant will develop excellent skills in signal processing, multi-variate statistics, novelty detection and uncertainty analysis. The aim of the work is to develop data processing methodologies that allow the automated detection of fatigue cracking in aluminium hangers. This will include techniques for accurate location of damage, the separation of extraneous noise from fracture signals, novelty detection and feature extraction. The developed techniques will be supported and validated by an extensive experimental fatigue testing programme, where the applicant will gain expertise in acoustic emission monitoring and experimental testing. The developed skills will also be widely applicable to the monitoring of a range of structures outside the initial aerospace focus.
Supporting Documentation.
Acoustic emission (AE) is a passive SHM technique that is able to globally monitor large structures in real-time by detecting small amounts of energy that are released when damage grows in a structure. However, like many SHM techniques acoustic emission requires operator interpretation of data and can be less reliable in complex structures, which is seen as the greatest barrier to industrial implementation. This problem is made all the more challenging by the presence of uncertainty which results from measurement noise, lack of system knowledge, variability in operating conditions, etc.
The applicant will develop excellent skills in signal processing, multi-variate statistics, novelty detection and uncertainty analysis. The aim of the work is to develop data processing methodologies that allow the automated detection of fatigue cracking in aluminium hangers. This will include techniques for accurate location of damage, the separation of extraneous noise from fracture signals, novelty detection and feature extraction. The developed techniques will be supported and validated by an extensive experimental fatigue testing programme, where the applicant will gain expertise in acoustic emission monitoring and experimental testing. The developed skills will also be widely applicable to the monitoring of a range of structures outside the initial aerospace focus.
Supporting Documentation.
Organisations
Studentship Projects
| Project Reference | Relationship | Related To | Start | End | Student Name |
|---|---|---|---|---|---|
| EP/N509449/1 | 01/10/2016 | 30/09/2021 | |||
| 1943540 | Studentship | EP/N509449/1 | 01/10/2017 | 31/03/2021 | Frederick Purcell |
| Description | A novel technique has been developed to estimate thickness maps, an image of a structure with about 1mm per pixel showing thickness of metallic structures. This makes geometric features and defects such as corrosion defects, stiffener dis-bonding, etc. visible and quantifiable. The technique is non-contact using laser laser ultrasound to measure multi-frequency ultrasonic waves in structure. This allows parts with both greater thickness and thickness rangers to be examined accurately. This technique has also been broadened to be applied to composite structures which are significantly more complex in nature as wave velocity and nature varies depending on fibre direction and the nature of the composite. Using automated image analysis a technique was developed to identify the ultrasonic wave characteristics which out prior knowledge of the material. Areas of different thickness can then be identified showing hard to identify defects such as delamination. |
| Exploitation Route | EPSRC grant application in conjunction with the University of Sheffield and the University of Strathclyde, the university of Luxembourg and Hull University. The project has support letters from Airbus, Siemens, Polytec and the National Physical Laboratory. |
| Sectors | Aerospace, Defence and Marine,Energy |
| Title | WaveNumber |
| Description | A python module with a broad range of wavenumber (spacial frequency) analysis tools for the purpose of damage detection. Can use a wide range of data (from modelled data to that taken form a Doppler Laser vibrometer) |
| Type Of Technology | Software |
| Year Produced | 2020 |
| Impact | As of yet not distributed but has so far allowed for a publication to be realised and a new quantitative non destructive technique to be developed. The large scale of data sets meant other analysis methods would be cumbersome or not possible to use |