ADS-79D-HDBK Downloaded from https://www.everyspec.com b. This method does not specify a minimum number of faulted items; rather, it requires that a minimum number of healthy items are able to exceed the desired CI threshold while still remaining healthy. c. This method can also be used to compute a minimum required sample size when one or more faults (true yellow or true red) are encountered in the tested items. d. Finally, it may be possible to develop similar methods for other distributions, such as the lognormal distribution.
I.5 SUMMARY It is the goal of this appendix to further provoke thoughts and encourage others to perpetuate research in maintenance credits for extending rotorcraft drive/engine systems time on wing with, and without, CBM. Further research, documentation, and experience need to be accrued to achieve the full benefits of both extending component service time on wing and CBM. The examples within this appendix provide practical implementation methods to meet requirements for maintenance credits intended to modify or replace legacy inspections or TBOs. When selecting examples, important considerations are: a.
the engineering rigor utilized to establish original maintenance on legacy rotorcraft prior to pursuing maintenance modification/replacement methods; b. what is at stake when attempting to modify/replace legacy maintenance practices on legacy rotorcraft components. c. technical variables surrounding a specific form of vibratory CBM monitoring device. Metrics for monitoring should be handled carefully so as to promote the objective for TBO extensions or paths to On Condition. Variables with Vibratory CBM can involve: data ski slopes indicative of bad accelerometers, wiring, or amplifiers; noise which may mask or simulate fault signals; and harmonics which may register a false fault alert. d.
sample sizes necessary to validate a specific vibratory CI/HI for modifying/replacing legacy rotorcraft maintenance. Sample size calculations are based on assumptions that should be tested for validity. e. CI data following a Weibull distribution may be capable of reducing a minimum sample size required to demonstrate CI reliability at a desired level of confidence. f. continuous field data assessment is necessary to ensure future faults follow the same distribution established during the initial sample size evaluation. 174