Premium
Structural Health Monitoring of an Annular Component using a Statistical Approach
Author(s) -
Mustapha F.,
Manson G.,
Pierce S. G.,
Worden K.
Publication year - 2005
Publication title -
strain
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.477
H-Index - 47
eISSN - 1475-1305
pISSN - 0039-2103
DOI - 10.1111/j.1475-1305.2005.00207.x
Subject(s) - structural health monitoring , aerospace , transducer , novelty detection , outlier , principal component analysis , component (thermodynamics) , acoustics , structural engineering , cylinder , ultrasonic sensor , computer science , engineering , artificial intelligence , mechanical engineering , novelty , aerospace engineering , physics , philosophy , theology , thermodynamics
For aerospace components there is undoubtedly a critical need to detect incipient damage in the structure, as any microscopic crack or defect can potentially lead to catastrophic failure and loss of human life. This paper investigates the scattering of an ultrasonic‐guided wave into a hollow cylinder‐like structure, under both damaged and undamaged conditions. Hollow cylinder structures are widely used not only in aerospace components but also in other engineering applications. The wave was sequentially transmitted and captured by means of a ‘real‐time data‐acquisition system’ combined with integrated disc‐shaped piezoceramic transducers. The integration of the tested structure and the transducers formed a structural health monitoring system. Wave responses were recorded from both of the structural conditions for the purpose of damage identification using a novelty detection method called ‘outlier analysis’. The principal component analysis method of reducing the dimensionality of the feature space is also presented in this paper, with its main aim being to visualise how the data sets behave as a function of the structural conditions.