CONSTRUCTION OF FAULT EVOLUTION MAPS FOR DAMAGE LEVEL ASSESSMENT OF RAILWAY BEARINGS

Abstract

The diagnosis of the state of machinery is becoming increasingly important in a more competitive industrial sector. Therefore, maintenance strategies are essential to ensure the correct and continuous operation of equipment. In the last decade, the application of condition based monitoring (CBM) has increased significantly in the detection of bearing faults. CBM is based on the characterisation of equipment behaviour and focuses on preventing a machine from failing. Therefore, fault characterisation and the construction of health indicators is a fundamental issue in CBM. This study evaluates the performance of the diagnostic methodology based on the construction of contour maps from envelope spectra. This allows the construction of visual maps that characterise the evolution of the amplitude of a fault frequency and its harmonics. They also serve as a reference to establish health indicators of future equipment conditions. The results suggest that the methodology holds promise for bearing condition classification. Furthermore, the use of contour maps is a visually intuitive tool for the analyst.