Crack detection in a rotary shaft analytical and experimental analyses: A review

Abstract

Detection of depth of cracks and its positions is the broad region of investigator over the universe. The article discussed a critical review of recent research done on crack identification and localization in rotary machine shaft using experimental modal analysis. Crack detection and localization in rotor shaft is very essential in various engineering applications. An enormous vibration in a rotary shaft causes premature failures of machine components. It is required to recognize the damage in time to avoid catastrophic failure of rotary machine shaft. Experimental investigation is used to observe their vibration characteristics like natural frequency of vibration, damping and mode shapes of structures. Natural frequency of rotary shaft can be obtained using the modal analysis technique. Also trained machine learning (ML) based artificial neural networks (ANN) can be applied to detect crack depth and its location. From the review by different research papers, it is observed that a lot of the research has been done on rotary machine shafts with open transverse cracks. Excessive speed and frequency for identified cracks position as well as cracks depth can be used to train an ANN. By using ML based artificial neural networks, location of cracks, and its depth identified precisely.