A Tool for Probabilistic Damage Tolerance of Hole Features in Turbine Engine Rotors

Abstract

Over the past two decades, the Federal Aviation Administration (FAA) and the aircraft engine industry (organized through the Rotor Integrity Sub-Committee (RISC) of the Aerospace Industries Association) have been developing enhanced life management methods to address the rare but significant threats posed by undetected material or manufacturing anomalies in high-energy rotating components of gas turbine engines. This collaborative effort has led to the release of several FAA advisory circulars providing guidance for the use of probabilistic damage tolerance methods as a supplement to traditional safe-life methods. The most recent such document is Advisory Circular (AC) 33.70-2 on “Damage Tolerance of Hole Features in High-Energy Turbine Rotors.” In parallel with this effort, the FAA has also been funding research and development activities to develop the technology and tools necessary to implement the new methods, including a series of grants led by Southwest Research Institute® (SwRI®). The most significant outcome of these grants is a probabilistic damage tolerance computer code called DARWIN® (Design Assessment of Reliability With INspection). DARWIN integrates finite element models and stress analysis results, fracture mechanics models, material anomaly data, probability of crack detection, and uncertain inspection schedules with a user-friendly graphical user interface (GUI) to determine the probability of fracture of a rotor disk as a function of operating cycles with and without inspection. This paper provides an overview of new DARWIN models and features that directly support implementation of the new AC on hole features. The paper also simultaneously provides an overview of the AC methodology itself. Component geometry and stresses are addressed through an interface with commercial three-dimensional finite element (FE) models, including management of multiple load steps and multiple missions. Calculations of fatigue crack growth (FCG) life employ a unique interface with the FE models, sophisticated new stress intensity factor solutions for typical crack geometries at holes, shakedown modules, a menu of common FCG equations, and algorithms to address the effects of varying temperatures on crack growth rates. The primary random variables are based on the default anomaly distributions and probability-of-detection (POD) curves provided directly in the AC. Fracture risk is computed on a per-feature basis using one of several available computational methods including importance sampling, response surface, and Monte Carlo simulation. The approach is illustrated for risk prediction of a representative gas turbine engine disk. The results can be used to gain a better understanding of the AC and how the problem is solved using the probabilistic damage tolerance framework provided in DARWIN.