Open Cut » Maintenance & Equipment
The ability to increase the accuracy and effectiveness of monitoring the condition of mobile machinery has been a long held vision of the mining industry. Traditional condition monitoring techniques of mobile machinery typically require the interruption of normal operations, often resulting in relatively long time intervals during which the condition of critical components is unknown, leaving opportunity for faults to rapidly deteriorate and potentially triggering catastrophic failures. While in recent years the ever increasing instrumentation of the plant in the form of on-board sensors has permitted more regular monitoring of some components, it still requires the software and hardware infrastructure to handle and process the data; for operations with dozens or hundreds of plants this may result in the need for significant infrastructure, as well as increased maintenance costs for this array of on-board sensors and computers.
Laser doppler vibrometers are capable of collecting acoustic data from a target surface at a distance with great sensitivity and for a wide frequency response range. This technology has been implemented in the manufacturing industry for quality assurance, however it may be able to be used in the mining industry to improve the current condition monitoring of mobile machinery.
This project aimed to provide foundational research into the feasibility of using laser vibrometers to collect acoustic signals for condition monitoring purposes. Several laboratory tests were carried out to explore the capabilities of the laser vibrometer unit and to characterise its performance under different conditions. These tests resulted in the derivation of a feasible operational envelope for a non-contact laser sensing station for workshop use.
Alongside these tests a software suite was developed for the pre-processing, filtering and post-processing of the data collected with the laser unit. This software suite (also referred to as the signal processing pipeline) aided by mathematical models derived by the team, allowed for the analysis of the acoustic data collected from two small-scale final drive units, with different levels of wear. The results from the data analysis showed a marked difference in the spectra obtained from the two units and processed with Mining3's software suite. Condition monitoring specific techniques were implemented to characterise these differences seen in the spectra, resulting in the quantification of the difference in wear between these two final drive units.
Finally, a field test was carried out using Mining3's CAT777B haul truck as the target. This test aimed to validate the operational envelope derived from laboratory tests and also to compare the performance of the laser vibrometer against an accelerometer, firmly mounted to the final drive unit casing. The results showed that the laser was able to collect acoustic signals that contained the expected fundamental frequencies and also that the acoustic signals collected are comparable to that recorded with the accelerometer on-board the truck.
Feasibility of using laser vibrometers for condition monitoring purposes was shown and a suitable operational envelope for a remote acoustic sensing station for workshop use was derived, which will allow for the validation of the technology through an extended field trial.
Another significant outcome of this project is the software suite developed, which is capable of pre-processing, filtering and post-processing acoustic data to produce clean spectra, critical for the development of condition monitoring algorithms of potentially many other components beyond the final drive unit, which was used as a demonstrator in this project.