Calculation methodology for turbine rotor fragment containment

Abstract

The design process of gas turbine engines is complex, subjected to numerous regulations imposed by national and international authorities and driven by economical requirements in a competitive market. Within the design process, predictive analytical methods are used to determine the minimum case thickness necessary to contain the high-energy fragments that may be released by a compressor or turbine rotor failure. The current methodologies often lead to inaccurate results. This study proposes a new analytical method to determine the minimum turbine case thickness required for the containment of a rotor fragment, based on investigation results from three distinct research areas: gas turbine rotor fragment containment, terminal ballistics, and finite element analysis. The general agreement in the industry regarding fragment containment is that all the translational kinetic energy of the fragment is exchanged with the casing, the impact area is proportional with the blade dimensions, the fragment is considered not deformable, and the following blades contribution to the impact is negligible. The present research shows that a fraction of the translational kinetic energy of the fragment is exchanged with the casing and reveals that the deformation of the blade fragment and its interaction with the following blades are not negligible. In addition, it proves that contrary to the legacy modeling, the blade fragment airfoil does not significantly affect the casing, as the main damage is created by the bulky section (the shroud, the blade platform, or its root). Finally, this study highlights key differences between modeling the containment of a shrouded versus a shroudless blade fragment, aspects not considered in the current literature. Some researchers, when characterizing the terminal ballistic model, assume the damage as resulting from shearing of the casing, while others assume tension and/or bending as primary failure modes. This thesis proposes a new methodology based on defining the threshold between the two main modes of failure, shearing and membrane stretching, as an established function of the fragment geometry and kinematic conditions. The proposed methodology has been validated with results obtained during containment tests on two Pratt&Whitney engines deemed representative for large gas turbines: JT3D and JT8D. The casing material (A-286) has been characterized using the state-of-art material model definition (Johnson-Cook). The material curves have been calibrated with results obtained from tests performed at various strain rates and temperatures. The damage factor was determined from another series of tests as a surface function of different triaxialities and Lode parameters. In summary, the present work proposes a new, refined analytical method to determine the minimum thickness required for a gas turbine case to successfully contain a single blade fragment released from the rotor. This improved model of the impact phenomenon leads to accurate results in the preliminary design phases, which translate subsequently in significant weight reduction.

Date	10 Jul 2022
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	Henri Champliaud (Supervisor)