Bearing element integrity analysis using dimensional analysis and response surface method (RSM) for fault diagnosis of rotor bearing system

• Main Authors: Rakate, G. N., Jadhav, P. V., Patil, V. R., Barawade, R. A., Shinde, P. V., Gondkar, V. S.
• Format: Conference Proceeding
• Language: English
• Subjects: Amplitudes; Bearings; Catastrophic failure analysis; Clearances; Condition monitoring; Defects; Dimensional analysis; Dynamic response; Excitation; Fault detection; Fault diagnosis; Fault minimization; Integrity; Mathematical models; Numerical prediction; Parameters; Power plants; Regression models; Response surface methodology; Rotor-bearing systems; Unbalance; Vibration measurement; Vibration monitoring
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0136995

Vibration measurements are commonly used to diagnose bearings in order to monitor the integrity of various bearing components. The aim is to either prevent catastrophic failure or use condition monitoring techniques. As the failure of bearing will cause the breakdown of power plant, it leads to huge loss and directly affects on the performance of plant. Even a very small fault in bearing significantly affects on performance of system. Hence it is important to identify and diagnose these faults to minimize the vibration levels which affect the performance of the system. In the available literature, the dynamic analysis of bearings focused only on localized defects and distributed defects in bearings. These studies have not given attention to the integrity of different components in bearing due to external sources of excitation. the objective of this work to develop the mathematical model to predict the vibration level and defect frequencies of defective bearing and correlate this with integrity of bearing elements such as bearing clearance, unbalance etc. also to study the dependency of the input parameters on response parameters and interactions. Here, a horizontal bearing system is investigated to study the integrity of various elements in bearing mathematically and experimentally. The two sources of excitation are disc unbalance, and bearing clearance has been considered. A generalized dimensional formulation with a dimensional approach with Buckingham’s pi theorem, which considers the effect of both discs unbalance and bearing clearance, is developed. The novel scheme is proposed integrating dimensional approach with Response Surface Method (RSM) that correlates predicted response and response obtained experimentally. The combined use of RSM was used to examine the interactive effects of three parameters such as disc unbalance, bearing clearance, and speed. The impact of internal radial clearance, disc unbalance and speed on the dynamic response analysis system is studied. It found that results obtained through constructed model lie within the range of 5%, which ensures the developed model is fit for evaluating the response. The percentage of error of developed regression model is found within of± 10%. Maximum vibration amplitude was observed at shaft speed of 1300 rpm and unbalances mass of 125 gm of unbalance mass with bearing clearance of 0.018mm and 0.035 mm. Minimum amplitude of vibration was obtained at radial load of 250N; unbalance mass of 25gm