Damage Diagnosis of Cardan Shafts in Mobile Mining Machines using Vibration Analysis

Abstract

The ability of prediction of machine condition may become even more crucial in mobile mining machines considering their instantly-varying conditions. The field of early damage diagnosis for mobile machinery is gaining increasing attention in modern and predictive maintenance strategies. In general, vibration-based machine condition monitoring methods offer the possibility to detect and localize the damages in mechanical systems at their early stages. In vibration monitoring, the vibration signals are captured via accelerometers, and are processed in the time and frequency domains in signal processing packages to extract diagnostic information. However, in real-world mobile applications, vibration analysis alone may not reveal the diagnostic information due to extreme transient and nonstationary operating conditions. In this regard, additional metadata available on the machine’s CAN bus system, can improve the reliability of vibration analysis for mobile applications. The current study aims to investigate the usage of metadata as additional information to the vibration analysis on cardan shaft of a test wheel loader in field conditions. Two various damage types, namely, misalignment and bearing clearance, both on cardan shaft, are investigated in some detail. The experimental results show how vibration analysis together with metadata processing can identify the state of the machine even in harsh operating conditions.