In the current state of the art of process industries/manufacturing technologies, computer-instrumented and computer-controlled autonomous techniques are necessary for damage diagnosis and prognosis in operating machinery. From this perspective, the paper addresses the issue of fatigue damage that is one of the most commonly encountered sources of degradation in polycrystalline-alloy structures of machinery components. It is possible to conduct in-situ detection & classification of damage as well as an assessment of the remaining service life through ultrasonic measurements of material degradation and their computer-based analysis. In this paper, tools of machine learning (e.g., convolutional neural networks (CNNs)) are applied to synergistic combinations of ultrasonic measurements and images from a confocal microscope (Alicona) to detect and evaluate the risk of fatigue damage. The database of the confocal microscope has been used to calibrate the ultrasonic database and to provide the ground truth for fatigue damage assessment. The results show that both the ultrasonic data and confocal microscope images are capable of classifying the fatigue damage into their respective classes with considerably high accuracy. However, the ultrasonic CNN model yields better accuracy than the confocal microscope CNN model by almost 9%.

The global objective of this study was to investigate the best features of the surface topography for fatigue damage detection and classification. The presence of the stress concentration in valleys of the surface topography causes a grain slip and a crack initiation at the surface of the machined structure and finally leads to fatigue failures. Therefore, the surface topography has a major influence on the fatigue strength of the machined structure. An optical confocal measurement system (Alicona) was applied to measure the surface topography parameters. These parameters are the arithmetical mean height S a , the root-mean-square height S q , the maximum peak height S p , the maximum valley depth S v , the maximum height S z , and ten-points height S z 1 0 . In this paper, feature selection using the Pearson correlation method was adopted to select the best surface textures that provide best the neural network (NN) model performance.The proposed NN models have been trained using the scaled conjugate-gradient back-propagation method. Results showed that the best surface topography parameters were S a , S v , S 1 0 Z , S z , where the NN model can detect and classify the damage with an accuracy of up to ∼94 .4%.

This paper develops a data-driven autonomous method for detection of fatigue damage and classification of the associated damage risk in mechanical structures, based on ultrasonic signal energy. The underlying concept is built upon attenuation of the signal and stability of the attenuation process. The attenuation provides pertinent information for damage quantification, whereas the stability represents resistance towards the fatigue damage growth. The proposed neural network (NN) model has been trained using the scaled conjugate-gradient back-propagation method. The NN model is capable of damage detection and damage classification into five classes of increasing risk. The Daubechies wavelet transform has been used to reduce the noisy pattern of the ultrasonic signal energy by using the associated approximation coefficients. The results show that the proposed method of approximation signal energy can detect and classify the damage with an accuracy of up to ∼ 9 8 . 5 % .