SELF-SCAN · Neural Net based defect detection system using LRU technology for aircraft structure Monitoring

This project will develop an integrated system to monitor the condition of aircraft components, using integrated transducer arrays for improved long range ultrasonic testing (LRUT) optimised to maximise UT wave-defect interaction in order to boost sensitivity. The project will: •Improve the defect detection capabilities of guided waves by generating / selecting wavemodes on the basis of optimised wave-defect interaction, rather than selecting one non-dispersive mode facilitating visual signal interpretation, as is the current practise. •Make use of Neural Nets for data interpretation and defect classification. Neural Nets are, in a monitoring type system, ideally suited to detect minute changes in signals, caused by defect initiation and subsequent growth, and separate them from changes in signal caused by other factors. •Develop and validate novel flexible MFC transducers / magnetostrictive transducers suitable to be bonded to / integrated into aircraft components to form LRU sensor arrays enabling detection, localisation and sizing of flaws. •Development of Focusing thechniques such as Time reversal focusing and Time delay focusing in complex materials used for aircraft component manufucturing. •Develop, train and validate the Neural Net defect detection and classification system using LRU technology for aircraft components Monitoring. •Develop a central software program with high-level functions comprising data collection, signal processing, data analysis and representation, information storage and user interface. Additional software will be developed to enable focusing of LRU to identifiy significant potential failure sources. •Undertake modular integrations of the sensors/transducers, signal processing and software functionalities to develop the prototypes and demonstrate its the capability to monitor , to reduce the maintenance costs and increase the safety of aircraft components.