The aim of this work is to achieve reliable nondestructive evaluation (NDE) of adhesively bonded aerospace components by developing novel multidimensional data fusion techniques, which would combine the information obtained by ultrasonic and X-ray NDE methods. Separately, both NDE techniques have their advantages and limitations. The integration of data obtained from pulse echo immersion ultrasound testing and radiography holds immense potential to help improve the reliability of non-destructive evaluation. In this study, distinctive features obtained from single techniques, traditional ultrasonic pulse echo testing, and radiography, as well as fused images, were investigated and the suitability of these distinctive features and fusion techniques for improving the probability of defect detection was evaluated. For this purpose, aluminum single lap joints with brass inclusions were analyzed using ultrasound pulse echo and radiography techniques. The distinctive features were extracted from the data obtained, and images of features obtained by both techniques were fused together. Different combinations of features and fusion algorithms were investigated, considering the desire to automate data evaluation in the future.

Ultrasonic non-destructive evaluation in pulse-echo mode is used for the inspection of single-lap aluminum adhesive joints, which contain interface defects in bonding area. The aim of the research is to increase the probability of defect detection in addition to ensuring that the defect sizes are accurately estimated. To achieve this, this study explores additional ultrasonic features (not only amplitude) that could provide more accurate information about the quality of the structure and the presence of interface defects. In this work, two types of interface defects, namely inclusions and delaminations, were studied based on the extracted ultrasonic features in order to evaluate the expected feasibility of defect detection and the evaluation of its performance. In addition, an analysis of multiple interface reflections, which have been proved to improve detection in our previous works, was applied along with the extraction of various ultrasonic features, since it can increase the probability of defect detection. The ultrasonic features with the best performance for each defect type were identified and a comparative analysis was carried out, showing that it is more challenging to size inclusion-type defects compared to delaminations. The best performance is observed for the features such as peak-to-peak amplitude, ratio coefficients, absolute energy, absolute time of flight, mean value of the amplitude, standard deviation value, and variation coefficient for both types of defects. The maximum relative error of the defect size compared to the real one for these features is 16.9% for inclusions and 3.6% for delaminations, with minimum errors of 11.4% and 2.2%, respectively. In addition, it was determined that analysis of the data from repetitive reflections from the sample interface, namely, the aluminum-adhesive second and third reflections, that these contribute to an increase in the probability of defect detection.

Adhesively bonded dissimilar materials have attracted high interest in the aerospace and automotive industries due to their ability to provide superior structural characteristics and reduce the weight for energy savings. This work focuses on the improvement of disbond-type defect detectability using the immersion pulse-echo ultrasonic technique and an advanced post-processing algorithm. Despite the extensive work done for investigation, it is still challenging to locate such defects in dissimilar material joints due to the large differences in the properties of metals and composites as well as the multi-layered structure of the component. The objective of this work is to improve the detectability of defects in adhesively bonded aluminum and carbon fiber-reinforced plastic (CFRP) by the development of an advanced post-processing algorithm. It was determined that an analysis of multiple reflections has a high potential to improve detectability according to results received by inspection simulations and the evaluation of boundary characteristics. The impact of a highly influential parameter such as the sample curvature can be eliminated by the alignment of arrival time of signals reflected from the sample. The processing algorithm for the improvement of disbond detectability was developed based on time alignment followed by selection of the time intervals with a significant amplitude change of the signals reflected from defective and defect-free areas and shows significant improvement of disbond detectability.

This study presents an ultrasonic non-destructive method with convolutional neural networks (CNN) used for the detection of interface defects in adhesively bonded dissimilar structures. Adhesive bonding, as the weakest part of such structures, is prone to defects, making their detection challenging due to various factors, including surface curvature, which causes amplitude variations. Conventional non-destructive methods and processing algorithms may be insufficient to enhance detectability, as some influential factors cannot be fully eliminated. Even after aligning signals reflected from the sample surface and interface, in some cases, due to non-parallel interfaces, persistent amplitude variations remain, significantly affecting defect detectability. To address this problem, a proposed method that integrates ultrasonic NDT and CNN, and which is able to recognize complex patterns and non-linear relationships, is developed in this work. Traditional ultrasonic pulse-echo testing was performed on adhesive structures to collect experimental data and generate C-scan images, covering the time gate from the first interface reflection to the time point where the reflections were attenuated. Two classes of datasets, representing defective and defect-free areas, were fed into the neural network. One subset of the dataset was used for model training, while another subset was used for model validation. Additionally, data collected from a different sample during an independent experiment were used to evaluate the generalization and performance of the neural network. The results demonstrated that the integration of a CNN enabled high prediction accuracy and automation of the analysis process, enhancing efficiency and reliability in detecting interface defects.