Achieving Six-Nine's Reliability Using an Advanced Fatigue Reliability Assessment Model

The idea that the reliability of helicopter safe-life components could be statistically quantified, as opposed to estimating that the chance of failure was "extremely remote", has been around for more than 20 years. Originally the work was done via tedious Monte-Carlo simulations that used a lot of (primitive) computer time. This was not practical for routine fatigue reliability evaluations. But now a suite of essentially analytical solutions - the Fatigue Reliability Assessment Model or FRAM - has been developed at Sikorsky. The FRAM provides numerically efficient and accurate solutions to easily determine fatigue reliability for comparison to the "six-nines" standard (i.e., less than one component in one million will fail within its established safe life). While this evaluation can only consider those statistical elements of fatigue that are known in advance, it can still serve as a standard for the component’s basic fatigue substantiation. This is the first step in the establishment of a new methodology that will provide reliability-based component retirement times. It also provides a means to evaluate the relative contributions and sensitivity of the strength, loads, and usage elements of fatigue reliability, so that emphasis can be placed on protecting the component from critical deviations in the original statistical characteristics assumed. Sensitivity to the occurrence rate of one specific flight condition would be an example. This paper reviews the advancement of statistical modeling and analysis in achieving six-nine's fatigue reliability design for dynamic components. It also presents fundamentals of advanced probabilistic methodology and its application potential to establish more rational fatigue reliability designs. An evaluation of this capability is presented here, based on the fatigue substantiation of a main rotor shaft with measured flight loads and a simplified usage spectrum.