ADS-79D-HDBK Downloaded from https://www.everyspec.com At this point, the question arises, what if the health algorithm doesn‘t accurately diagnose the component? The fall-back here is regularly scheduled maintenance. If maintenance calls for the part to be removed, it should be removed even if the diagnostics algorithm doesn‘t say it should. If the part is removed due to maintenance, check whether the algorithm‘s diagnosis is on par with maintenance. If, however, the part is removed based on the health algorithm‘s recommendation, then check the part against the maintenance manual to ensure that the removal was necessary. There is no time limit for in-service validation.
Refer to Appendix D for more details on algorithm validation. Seeded/Natural Fault Test Validation: This is a viable option for health credit validation. To start off, validation tests need to be designed by the supplier and the tests need to be approved by AED. These tests can be either a seeded fault test or a natural fault test. A seeded fault test is where a fault is integrated into the target component and the component is used for the test. In a natural fault test, the component used for the test must have been removed previously from the aircraft due to a fault. Once the test methodology is determined and the tests approved, it is time to perform the validation.
The validation process also has two options: Bench Test or On-Aircraft Test. Upon successful validation, the candidate system can acquire Maintenance Credit for the component. A document about the validation process will have to be generated to close out the process. For more details on seeded fault testing, reference Appendix J. K.5.2 Mechanical Health-based Prognostics V&V. The determination of Remaining Useful Life (RUL) of a component based on its mechanical health is the definition of Mechanical Health-based Prognostics. This approach is a viable option for a system attempting to achieve prognostics for mechanical components.
It is to be noted that for a system to conduct prognostics, the candidate system must have at least one mature or validated HI for the target component. This section will provide a detailed view of the processes of Verification and Validation for acquiring prognostics for a component using the Mechanical Health approach. K.5.2.1 Mechanical Health-based Prognostics Verification Process. In this section, a detailed view of the Mechanical Health-based Prognostics Verification process (See Figure K-8 for flow chart) is provided. This process also discusses certain development steps needed to have a successful verification.
The entry criterion for starting the prognostics verification process is that the candidate system must have at least one mature HI established for the target component. This is to ensure that even if the prognostics fails to function accurately, maintenance still can be performed based on health diagnosis. 206