Quantitative imaging algorithms for in situ damage characterisation using guided waves

Chan, E 2015, Quantitative imaging algorithms for in situ damage characterisation using guided waves, Doctor of Philosophy (PhD), Aerospace, Mechanical and Manufacturing Engineering, RMIT University.


Document type: Thesis
Collection: Theses

Attached Files
Name Description MIMEType Size
CHAN.pdf Thesis Click to show the corresponding preview/stream application/pdf;... 12.40MB
Title Quantitative imaging algorithms for in situ damage characterisation using guided waves
Author(s) Chan, E
Year 2015
Abstract Guided wave based structural health monitoring has attracted tremendous attention in recent times from industry groups and research communities alike. This new approach holds great promises to supplement conventional quantitative non-destructive evaluation methods for high-value structural assests with in-situ techniques in the future.

The aim of this research is to develop and validate quantitative imaging algorithms to determine the size and severity of structural damage from guided wave response measured using an array of active sensors. Three new imaging algorithms have been developed and validated viz. modified diffraction tomography, modified beamforming and modified time-reversal for digital reconstruction and characterisation of in-plane damage in plate-like structures using guided waves. The proposed algorithms treat damage as a flexural inhomogeneity within the framework of Mindlin plate theory. By applying the Born approximation theory, inverse solutions of the damage intensity can be derived using measured response as input. The imaging algorithms are applicable to early corrosion damage in metallic structures and impact damage in fibre-laminated plates, where the damage can be approximated as weak scatterers. Unlike existing imaging strategies using guided waves, the new algorithms lead to reconstruction integrals that have a very similar mathematical structure as that pertinent to far-field methods. Furthermore, the new algorithms consider the multi-static data matrix and Green’s function of the structure to provide damage characterisation in the near-field and compensation for multi-path wave interactions.

The imaging algorithms are extensively tested using numerical data and finite element simulated data for four appropriate indicators of imaging performance viz. reconstruction quality, prediction of damage geometry as well as an estimate of damage size and severity.

In the numerical studies, analytical solutions are used to generate the data matrix and Green’s function. Initially, the scatterers are simulated as a variation in wavespeed or refractive index during the process of formulating the algorithms. However, for testing of the algorithms using plate-waves, the scatterers are simulated as localised reductions in material thickness and inertia to model corrosion damage and delamination damage, respectively. The results show exceptional quality of reconstruction for a single scatterer, even when the scatterer and sensors are in close proximity. Additionally, the time-reversal based imaging algorithms exhibited the least susceptibility to noise in the data matrix. This result is significant in practice for locating the damage, along with providing an approximation to damage shape and size.

Validation of the new imaging algorithms using synthetic data (multi-static scattering matrix and Green’s function) generated by the finite element analyses reveal that the imaging algorithms are capable of good estimation of damage shape, size and severity for a variety of cases including multi-site scatterers, arbitrary shaped damage and plate configurations featuring multi-path interactions. Amongst the new algorithms, the modified time-reversal formula provided the best imaging results for complex structures, even in the presence of strong multi-path wave interactions between the structural features and damage. Apart from these results, an approach that allows accurate reconstruction of the damage using minimal number of sensors is developed and implemented.
Degree Doctor of Philosophy (PhD)
Institution RMIT University
School, Department or Centre Aerospace, Mechanical and Manufacturing Engineering
Keyword(s) Guided waves
Imaging
Diffraction tomography
Time-reversal
Beamforming
Versions
Version Filter Type
Access Statistics: 190 Abstract Views, 187 File Downloads  -  Detailed Statistics
Created: Mon, 16 Nov 2015, 10:36:02 EST by Denise Paciocco
© 2014 RMIT Research Repository • Powered by Fez SoftwareContact us