Condition Monitoring and Fault Detection of Longwall Shearer

Improvement in equipment availability is recognised as a key factor to maximise returns from Australian longwall mining assets.  Previous analysis of LW equipment downtime identified the shearer as a significant contributor to this situation.  This project set out to build and field test a complete system of data logging and analysis hardware and software to monitor bearing vibration at multiple locations on shearer ranging arms and haulage drives and the loads applied to the ranging arms.  The system is based around the acquisition of raw vibration and strain gauge data from underground equipment and the transfer and storage on the surface at sufficiently high rates to allow both “real time” and subsequent analysis of data.  The intent being to provide both operational warnings of pending bearing failure, and the vibration information and externally applied loading data for subsequent machine design improvement or failure investigation.

A literature search and some signal processing work was carried out during the project to assess the feasibility of detecting incipient bearing failure during periods of normal production by the shearer.  The conclusion drawn is that the detection of incipient bearing failure from analysis of vibration data collected while the machine is subjected to extremely variable and random loading is not practical.  It is also suggested that the benefit sought by continuous condition monitoring during production may be met for shearers by recognising the frequently occurring periods when the machine momentarily operates at constant speed with no load and applying traditional condition monitoring techniques.

Despite operational difficulties preventing the final field trials, the project has delivered a rugged, high speed, multi-channel, data logger suitable for mounting in flameproof enclosures in the underground environment and the associated software required to send and receive large amounts of data over the mine network.  The project has also delivered a graphical interface for the subsequent investigation of the immense array of data generated by the logger.

A future opportunity exists to integrate the hardware and software generated during the project with traditional condition monitoring techniques to deliver a virtual continuous condition monitoring system for underground mining equipment.