Crack detection in shafts of rotating machinery using active sensing with an external excitation on a bearing

Because several power plants have undergone burst shafts with catastrophic consequences, methods for detecting cracks in the shafts of rotating machinery are of great interest to the research community. Cracked shafts represent a very small percentage of the vibration issues in machinery (less than 2%) compared with other faults such as misalignment and mass imbalance, which are responsible for 85% of these issues. However, if a cracked shaft is not detected early enough, it can jeopardize the safety of operators and result in high costs for replacement. A method sensitive enough to detect cracks in the shafts of rotating machinery is investigated in this paper. It is shown that changes in the natural frequencies and/or mode shapes are not sensitive enough indicators to detect transverse cracks at the mid-span of shafts in rotating machinery. An exhaustive numerical analysis using a finite element model of a simple machine is conducted to consider many different measurements of the machinery vibratory responses that most clearly detect cracking. It is shown that external excitations applied on a bearing enhance the sensitivity of vibration measurements to cracking. Experimental tests are also used to validate several of the numerical simulation findings.