This study addresses the application and enhancement of laser ultrasonic imaging for damage monitoring. The enhancement is based on applying an optimal filter known as the Kalman filter to the wave propagation data so that the influence of an anomaly ...
This study addresses the application and enhancement of laser ultrasonic imaging for damage monitoring. The enhancement is based on applying an optimal filter known as the Kalman filter to the wave propagation data so that the influence of an anomaly on the wave due to random noise can be improved.
The study will constitute: the development of the optimal filter for the laser-generated wave propagation, the application of the filter to the real data for validation purposes, and finally, the case study performed on the specific case of damage, which for this case is plate corrosion dimension sizing. The development of the method was performed in two stages. The first stage developed the filtering algorithm for a waveform representing a single laser scan point. In doing so, an analytical approach was used to formulate an ultrasonic wave equation. The state-space equations are then computed from it. A difficulty had arisen in matching the waveform from the analytical equation with the experimental counterpart. The wave reflections from edges that are part of the experimental waveform are not accounted for in the analytical formulation. The discrepancy between the two waveforms was dealt with by proposing polar Fourier transform-based removal of the edge reflection. The second development stage aimed to remove the first stage's drawbacks. These drawbacks are the requirement of removing the edge reflection and a high computation time due to separate filtering of the waveforms at each laser-scanning point. In order to address these issues, the second stage/phase considered the fusion of a finite element simulated wavefield with the experimental wavefield, the means of fusion being the Kalman filter. The matrices in the state-space equation were derived from the Finite element software. The input vector components of the state space equation were also extracted from the simulation results. Apart from the wave quality enhancement-driven application of the Kalman filter, novel damage detection/quantification methods are also developed for the case study. The plate corrosion dimension sizing and data-driven dispersion curves (experimental dispersion curves) were used to quantify the depth of the corrosion damage.