Smart Condition Monitoring of Worm Gearboxes

Worm gearboxes are commonly used in many various fields of industrial applications such as escalators, presses, conveyors etc. However, heavy industries face important problems about this type of gears due to undetected failures. The vibration signal of a gearbox carries the signature of the fault. Hence, vibration measurement and graphical representation plays an important role for analysis of physical conditions. Although there are many options for vibration measurement systems, cost effective and portable microcontrollers based monitoring system is a good option. In this study, smart monitoring system for worm gearboxes is investigated. Data acquisition system, various vibration analysis techniques, fault diagnosis and visualization is developed via Arm Cortex M4 microcontroller. It is shown that by analyzing the vibration signal using signal processing algorithms including time synchronous average (TSA), fast fourier transform (FFT) and statistical metrics; early fault detection of the worm gearbox is possible. To classify unknown gear conditions, the single metric of Mahalanobis distances are calculated, utilizing the maximum probability estimates. From the experimental results, the most suitable indicator for fault diagnosis of worm gearboxes is determined.