Structural health monitoring based on principal component analysis: damage detection, localization and classification

Structural Health Monitoring is an area that its main objective is the verification of the state or the health of the structures in order to ensure proper performance and maintenance cost savings using nondestructive tests. Currently, in the CoDAlab group it is used an active piezoelectric system which involves the use of piezoelectric transducers that are attached to the surface of the structure in order to apply vibrational excitations and collect dynamic responses at different points. As pattern recognition technique, Principal Component Analysis is used to perform the analysis, built a base-line model of the structure without damage and, subsequently to compare the data of the current structure under test. Different indices are calculated to determine how different is the structure under test. Using these indices, it is possible to detect, classify and locate the damage by means of the contribution of each sensor to each index. These methodologies are tested using two different structures, one aircraft turbine blade and one aluminium plate, which were instrumented with seven and four Piezoelectric transducer discs respectively. Seven simulated damages were made in the aircraft turbine blade and four real damages in the aluminium plate.