Structural Health Monitoring with Laser Doppler Imaging of Ultrasonic Guided Waves
This thesis deals with ultrasonic guided wave based non-destructive inspection and health monitoring of aircraft structural components using a non-contact sensing scheme based on laser Doppler scanning imaging technique. Ultrasonic guided waves are elastic waves which travel relatively long distance and are guided by the geometry of thin structures with minimal loss in amplitude. Experiments are conducted using piezoelectric actuator and a 3D scanning Laser Doppler Vibrometer (LDV). The piezoelectric actuators bonded on the structure excite ultrasonic guided waves. The laser beam from the LDV senses the surface vibration based on optical interference of the laser beam. The optimal frequency range is selected based on the wave dispersion characteristics, wave mode strength, thickness of the test specimen and size of the defects that need to be determined. Analytical models are developed to determine these relations and signal calibration. For defect detection, schemes for both baseline-free and baseline-dependent algorithms are developed. The baseline-dependent algorithms use amplitude, phase and Time-of-Flight (TOF) information to detect defects. Experiments are conducted on large structural panel with inaccessible defects. The measurements are obtained from sparse sensing points which are in form of either compact or a sparse distributed array on the accessible side of the panel. Triangulation technique along with TOF information is used to localize the defects. Sparse measurements are used to reconstruct virtual spatio-temporal wavefield using a pixel-based image reconstruction technique. The schemes allow fast inspection of large and complex structural parts without structural disassembly. In a baseline-free detection process, the schemes that are developed include a spatio-temporal wavefield imaging scheme, a spectral power flow scheme, a local wavelength and mode conversion strength-based scheme. The spatio-temporal ultrasonic wavefield reconstructed using scanning LDV is further used to understand the wave scattering behavior in the presence of defects. Further, the spectral power flow-based scheme is developed that can estimate the location and size parameters of defect simultaneously. This scheme uses an invariance property of the guided wavefield to detect defects. A spatial power flow map is further derived from this scheme to eliminate the background wavefield and highlight the defect. These schemes are tested on a metallic structural panel having sub-wavelength defects such as notch and pitting corrosion. In a composite structural component, a delamination is first identified using the spatio-temporal ultrasonic wavefield imaging scheme. Further, the delamination parameters are estimated using local wavelength-based scheme involving dynamics of sub-laminate wise wave splitting in the delamination region. The technique is tested on a laminated composite and a co-cured co-bonded composite T-joint. For fast inspection of structural components, a sparse sensing scheme with a circular array of sensing points is chosen. Using this scheme, along with the estimated mode conversion strength due to wave scattering, the problem of classifying different defects are explored. Toward the end of the thesis, a reliability analysis is carried out to establish the newly developed inspection technique in terms of sensitivity and probability of detection. Environmental and measurement uncertainties are introduced by adding noise in the excitation signal. Probabilistic detection curves for varying defect parameters are obtained from the experimental data. Future potential of the developed technique in the aircraft maintenance, repair and overhaul industry is highlighted.