Lab-Scale Cable Shovel Model with Potential for Fault Detection Applications

Abstract

Cable shovels are mostly used as primary equipment in large-scale surface mining operations. A single unexpected component failure could cause unforeseen costly shutdowns, reduced productivity and may even pose a great danger to on-site personnel. The ideal maintenance strategy is to reveal incipient faults in advance that the repair schedule and equipment have enough time to get prepared. Many challenges will occur if conducting the fault detection experiments on the real mining shovels. In this research, a lab-scale shovel model with the potential for fault detection is proposed.

With the knowledge of the benefits of physical modelling and full-size cable shovel structures and fundamental functions, a lab-scale shovel model is designed and fabricated based on the Komatsu P & H 4100XPC cable shovel with a scaled-down factor of 15. The shovel model has validated that model has a similar working mechanism and performance as the full-size mining shovel. Besides, the shovel model is able to simulate the fundamental operation processes by a robust control system based on EtherCAT technology.

Vibration analysis is an effective method for fault detection based on the fact that the vibration patterns under normal and abnormal conditions are different and the variation is the indicator of the presence of faults. A PC-based data acquisition system is designed to obtain the vibration signal of the mechanical systems. Two artificial failure modes with different time-varying behaviours are seeded to the shovel model. In order to extract the fault signatures from the signal, various signal processing techniques in time, frequency and time-frequency domains are used to distinguish the faulty components from the healthy ones and some techniques are also able to predict the scale of faulty conditions.

A lab-scale cable shovel model with time-varying behaviours provides a platform for fault detection that can be also applied to other areas such as soil-tool interaction, equipment health monitoring and autonomous excavation.