Thermal Images and Temperature Matrices for the State Assessment of Rolling Bearings

The assessment of the health state of rolling bearings is important for supporting the decision-making concerning their maintenance and operation. Even if different techniques and methods have been utilized, approaches based on Infrared Termography (IRT) have not been sufficiently explored. In this paper, the potential that IRT may have for classifying the severity of failures of rolling bearings is investigated. This study compares different approaches for analyzing thermal data woth the purpose of detecting outer-race defects in rolling bearings. Specifically, the study considers thermal-based analysis (TBA) using temperature matrices, intensity-based analysis (IBA) using thermal images, and combinations of these two methods. These approaches are evaluated based on their ability to classify the severity of the defects, building on promising results previously obtained in the field of Infrared Breast Thermography. An accuracy and F1-score exceeding 90% were achieved by combining the temperature matrix with the thermal image using pseudo colors and processing them with the VGG deep learning algorithm. These outcomes indicate the potential of IRT in assessing the condition of rolling bearings. It should be noted that while this work has explored the use of IRT for the classification of the health state of rolling bearings by running different operative conditions and taking thermal images from various angles and distances, further experiments are needed to fully validate the effectiveness of this approach.