Originally Posted by
AnFI
riff raff
just a hypothetical question
How many gearboxes would you have to run and for how long under what conditions to demonstrate a catestrophic in service failure rate of < 10^-9 per hour?
what is the acceptable catestrophic failure rate for a gearbox? is there one?
AnFI,
Sorry for the tardy response.
The answer to your first question is that for single path flight critical systems/components, an example of the FAA threshold for catastrophic failure events is "extremely improbable" as defined in
AC 25.1309-1A section 7d. Compliance with this requirement is normally demonstrated entirely by analysis, and the acceptable approach is described in section 8d of the document noted. To demonstrate by test that a complete main gearbox system has a catastrophic failure rate below the "extremely improbable" threshold of 1x10^-9/flight-hour, you would need to test a statistically relevant number of type-confoming gearboxes to their design TBO. And then calculate the failure rate based on the number of catastrophic failures and total accumulated test hours.
One thing you need to consider is the huge number of test hours that might be required to produce a valid result for a very high reliability rate. And each gearbox build would only provide maybe 2000 hrs of test, so you would need to test a very large number of gearboxes. There is some work being done on methods for accelerated life testing,
like this example.
Regarding the speed/load conditions used for this type of test procedure they are defined by a "mission profile", which is a representation of the speed/load/time intervals during a typical flight cycle. You can read more about reliability testing in
MIL-HDBK-781.