Crack detection in rotating shafts

Abstract

Detecting cracks in rotating shafts is crucial for ensuring the reliability and safety of various mechanical systems. This final year project presents a novel approach to crack detection in rotating shafts using advanced signal processing techniques and machine learning algorithms. The proposed method leverages vibration signals obtained from sensors attached to the shaft to identify and localize the presence of cracks. Initially, the project involves the acquisition of vibration data from a laboratory-scale test rig simulating real-world operating conditions. Signal processing techniques such as Fourier analysis and wavelet transform are applied to extract relevant features from the vibration signals. These features are then utilized to train machine learning models, including support vector machines (SVM) and artificial neural networks (ANN), for crack detection. The effectiveness of the developed crack detection system is evaluated through extensive experimental validation on the test rig. The results demonstrate the capability of the proposed method to accurately detect and localize cracks in rotating shafts, even in the presence of noise and varying operating conditions. Furthermore, the project explores the integration of the crack detection system into existing condition monitoring frameworks for proactive maintenance of industrial machinery. Overall, this final year project contributes to the advancement of fault diagnosis techniques for rotating machinery, with potential applications in industries such as manufacturing, power generation, and transportation.