ADS-79D-HDBK Downloaded from https://www.everyspec.com For the on-aircraft test, a specimen should be chosen that has already reached, or is very close to reaching, the desired recognizable fault condition by the HUMS. This will allow for a reasonable amount of normal operation to progress the fault to a detectable level and provide data for again evaluating the condition or health indicator‘s ability to measure fault progression. Because a faulted, or compromised, component is being introduced into the aircraft, the on- aircraft testing should be conducted with ample consideration given to vehicle and operator safety (e.g. ground run only).
In fact, the earlier laboratory testing should provide as accurate an estimate as possible to the remaining safe-life using legacy maintenance methods for the specimen prior to installation on the aircraft. This will provide some continued airworthiness assurance the test article will not progress to a component failure during the on aircraft test. However, to provide meaningful results the test should obtain fault and indicator data over the full range of aircraft operating regimes when possible. Therefore, the aircraft testing should normally be performed as part of an experimental flight with a trained test pilot unless circumstances justify otherwise. J.4.4 Step 4: Follow-on efforts.
Pending the conclusions and results found in the bench and on-aircraft test reports, an AWR or Safety Message may be generated to alert the fleet as to any changes required in a fielded CBM system, and to include any CBM maintenance manual changes. Depending on the purpose and intended goals of the SFT, this information can range from the introduction of a new condition or health indicator; retirement of an existing, prescribed condition or health indicator; or change in threshold value of an indicator for inspection or replacement of a part.
J.5 DETAIL GUIDANCE To further illustrate and provide detail guidance in executing the seeded fault test process, example references are cited in Section J.2, Applicable Documents, of this appendix. While these examples do not specifically utilize all steps of the Figure J-1 process, these references are good examples of where, following a rigorous experimental process, SFT led to obtaining a conclusion as to the effectiveness of a CBM application.
In the NASA / US Army Research Laboratory study on crack detection reference sited in Section J.2, thirteen vibration-based diagnostic metrics were compared for their ability to detect tooth fracture and progression to tooth separation in a spiral bevel pinion of a Bell OH-58 main rotor gearbox. The specific fault condition under test was identified, and the test specimen was prepared by manually placing a notch into the fillet region of one spiral bevel pinion tooth using electro-discharge machining (ultimately, trial and error determined the minimum notch size used to induce the intended fault).
The test specimen was installed in an OH-58 main transmission and mounted in a Helicopter Transmission Test Stand at NASA‘s Glenn Research Center. Bench testing commenced with the pinion operated at the design speed and at various percentages of maximum design torque, with the overall goal of the testing to initiate a crack in the pinion at the lowest possible torque. Three metric indicators proved sensitive enough to detect the damage while not being overly sensitive to torque fluctuations. The other diagnostic metrics either could not reliably detect the fault condition or were too noisy in their indications to be used as a viable field solution. 182