Use of fleet data and seeded fault testing to enhance diagnostic operations and provide the basis for safe prognostics of a helicopter bearing

摘要

A Condition-Based Maintenance (CBM) program calls for transitioning from time based part replacements to performing maintenance upon evidence of need. For the U.S. Army's CBM+ plan this entails eliminating the use "time before overhaul" (TBO) definitions currently driving vehicle component maintenance schedules. Although Health and Usage Monitoring Systems (HUMS) and Digital Source Collectors (DSC) have the potential to support this goal, their ability to diagnose component faults early is limited, and implementation of prognostics is rare. Key causes of these limitations include (i) the sensitivity of sensors and condition indicators to signal noise, specific fault modes, and variations in environmental and operating conditions; (ii) the fleet-wide performance of diagnostic processes with mostly empirical condition indicators; (iii) uncertainties in damage or wear detection and progression; (iv) the inherent risk of implementing prognostics on degrading or faulty components; and (v) the challenges involved with validation and verification in deployed systems. This paper describes a methodology to overcome these limitations using an integrated set of diagnostic enhancement algorithms and the implementation of a dual prognostics approach. This architecture was developed and demonstrated using the case of a bearing in the drive train of H-60 helicopters in service for the U.S. Army, which is a critical component and has had a relatively high incidence of replacements and faults, some of which have involved contaminated grease and corrosion, of concern for potential rapid degradation and failure driven by loss of lubricity and corrosion fatigue mechanisms. This paper analyzes the case and fault modes of the bearing and describes tools developed for the improved use of monitoring data, enhanced diagnostics, and the implementation of prognostics in support of a transition from TBO-based maintenance decisions to condition based maintenance, illustrating capabilities with examples using available fleet data and seeded fault tests of the helicopter bearing.