In guided wave (GW)-based structural health monitoring (SHM), ultrasonic elastic waves are used to detect damages in structures by comparing the acquired signals with those acquired before defect formation. Making the SHM system automatic, especially for similar structures, such as turbine blades, is rather challenging. The high sensitivity of GWs to environmental and operational conditions, the variabilities due to sensor positioning, sensor coupling, and material variability in composites limit the baseline application. This work presents a machine learning (ML)-based damage detection method using aggregated baselines independent of their damaged states to enhance the generalization capability of ML algorithms by considering similar structures’ variabilities. The methodology relies on feature extraction from raw GW signals and training classification algorithms (e.g., kernel machines, ensemble methods, and neural networks). Two experimental data sets on composite panels are used. The first experimental data set of 45 composite panels is used to validate the approach by considering the aforementioned inter-specimen variabilities. Half of the 45 panels provide pristine data, and the rest provide damaged data so that the same sample is present in the training or test set but never in both. High classification performance is obtained, demonstrating that the classifier has successfully learned to recognize defect signatures despite the influence of the variabilities linked to the multiple instrumented specimens. The second experimental data set of 1 composite panel with temperature variation is used. Good classification performance is obtained without using baseline correction methods.
Accurate defect characterization is desirable in the ultrasonic nondestructive evaluation as it can provide quantitative information about the defect type and geometry. For defect characterization using ultrasonic arrays, high-resolution images can provide the size and type information if a defect is relatively large. However, the performance of image-based characterization becomes poor for small defects that are comparable to the wavelength. An alternative approach is to extract the far-field scattering coefficient matrix from the array data and use it for characterization. Defect characterization can be performed based on a scattering matrix database that consists of the scattering matrices of idealized defects with varying parameters. In this article, the problem of characterizing small surface-breaking notches is studied using two different approaches. The first approach is based on the introduction of a general coherent noise model, and it performs characterization within the Bayesian framework. The second approach relies on a supervised machine learning (ML) schema based on a scattering matrix database, which is used as the training set to fit the ML model exploited for the characterization task. It is shown that convolutional neural networks (CNNs) can achieve the best characterization accuracy among the considered ML approaches, and they give similar characterization uncertainty to that of the Bayesian approach if a notch is favorably oriented. The performance of both approaches varied for unfavorably oriented notches, and the ML approach tends to give results with higher variance and lower biases.