Downloaded from https://www.everyspec.com ADS-79D-HDBK to refine a CI are outlined in Appendix D. Guidelines to determine the appropriate Sample Size for replacing current maintenance practices are outlined in Appendix I. 5.9.1 Enhancing current maintenance with CBM on legacy aircraft. Validated and unvalidated CI/HI algorithms may be utilized for both data gathering and aircraft maintenance diagnostics/ prognostics while legacy maintenance practices remain in place. Data gathering permits time to adjust maintenance alert levels built into the algorithms (CI/HI refinement) while not degrading any aspect of continued airworthiness associated with the legacy maintenance of the aircraft.
In fact, adding sensors and associated algorithms to the aircraft legacy maintenance practices can actually increase the reliability of the overall aircraft system provided the sensor hardware reliability is not mission or flight critical and does not cause unscheduled maintenance impacting aircraft readiness. Note also, the sensors and associated algorithms may be applied to focus on specific component failure modes versus all failure modes of complex components (e.g. transmissions and engines).
Since baseline risk is not degraded by the sensors, validated as well as unvalidated algorithms may be employed to enhance current maintenance practices and develop new diagnostic/prognostic maintenance. Advantages to this approach are:
relatively low initial cost approach that does not require a large number of sample specimens (see Appendix D) to demonstrate algorithm reliability for each failure mode; ability to enhance maintenance by focusing on specific failure modes versus all failure modes of complex components; relatively short timeframe involved prior to field implementation due to the reduction in test requirements; and the ability to gather data during normal aircraft operation to facilitate verification/validation efforts. Disadvantages to this approach are:
number of false positives indications with unvalidated algorithms along with the associated maintenance increases; and the determination of return on investment may only be conducted after field data is collected, the tear down analyses confirm sensor indications, and the diagnostics/prognostics are matured. 5.9.2 Modifying or replacing overhaul intervals on legacy aircraft.
Prior to considering modifications to or replacing legacy aircraft maintenance, it is important to understand what initial specification design, maintenance, and reliability requirements were placed on the legacy platform as well as the engineering rigor utilized to verify, validate, and establish legacy maintenance practices. Consequently, any maintenance modifications or replacement should be validated as good as, or better than, legacy maintenance practices. For US Army aircraft propulsion and drive systems this involves:
User requirements for aircraft usage, maintenance, and reliability; bearing B10 analyses; bearing endurance testing; lubrication shelf life analyses; gear tooth bending fatigue life analyses; gear tooth contact fatigue life analyses; thousands of hours of endurance testing; engine component life testing/analyses; wear rate analyses to ensure component reparability; and TBO/On-Condition establishment. Therefore, any CBM system implemented to modify or replace legacy maintenance practices should undergo similar analytical and testing rigor.
In the case of vibration monitoring, CBM algorithms implemented to accurately depict actual hardware condition and replace current maintenance practices should be required to be validated. CBM algorithm validation will require both faulted and unfaulted components. The statistically significant sample size (see Appendix I) for faulted and unfaulted components should take into account the required confidence and reliability guidelines within this document.
In addition to understanding legacy requirements to verify and validate legacy maintenance practices, it is important to note that TBO interval extensions are generally limited by the repair limits and calculated fatigue lives of components within a system under consideration for maintenance modification. An exception to the fatigue life limit is to employ CBM monitoring if the fatigue failure mode is detectable utilizing a validated detection system and will not result in the failure mode progressing or manifesting into a failed state within 2 data download intervals of the monitoring system. 25