Description |
-
Abstract: Finite element analysis is used extensively in the aircraft turbine engine industry to predict stresses to calculate low cycle fatigue (LCF) life. An accurate prediction of stresses is especially important to a specific subset of engine hardware that is defined by the FAA as life-limited parts (LLP's). LLP's include rotor and major structural parts such as disks, spacers, hubs, shafts, h... read moreigh pressure casings, and non-redundant mount components. A failure of an LLP can lead to a potentially catastrophic event due to non-containment of high energy debris, uncontrolled fire, or a complete inability to shut the engine down. Under-predicted stress can cause the life limits to be set too high, which is a safety hazard. Over-predicted stress can cause the life limits to be set too low, which adds cost due to the need to replace expensive engine hardware more frequently. High fidelity stress analysis is necessary to appropriately set LCF life limits. One common engine feature analyzed with 3D stress analysis is a rotor bolted joint. Geometrical features associated with bolted joints such as holes, fillets, and scallops cause stress concentrations. Often the life limiting feature in a rotor LLP is a geometrical feature in close proximity to the joint. Unfortunately, the detailed stress analysis associated with accurately predicting stress in the joint is costly and time consuming. Analysis assumptions that can simplify the effort, yet still produce accurate results, would be valuable to the industry. The focus of this study is on the bolt-nut interface modeling assumptions associated with a rotor bolted joint stress analysis for LCF predictions. A 3D finite element model of an actual aircraft engine rotor bolted joint is created. A series of eleven cases are analyzed and compared to investigate how the thread modeling assumptions affect the calculated life in the mated rotor LLP hardware. Walker-adjusted alternating stress, σ0,alt, is used to measure the life impact. The impact is limited to the edges of the two critical features closest to the bolt-nut interface. The results demonstrate that factors such as the thread mesh density, elastic versus elastic-plastic bolt/nut material properties, and the inclusion of the helical thread shape have only minor impact. The inclusion of contact elements at the interface instead of couples has a moderate impact of 1.1 to 1.2 ksi. When couples are used, the placement of the first couple is critical, impacting the results by 1.1 to 2.6 ksi. Also, when couples are used to join the interface, the explicit inclusion of the thread shape has only 0.5 to 0.6 ksi impact.
Thesis (M.S.)--Tufts University, 2012.
Submitted to the Dept. of Mechanical Engineering.
Advisor: Anil Saigal.
Committee: Thomas James, and David Kelpe.
Keyword: Mechanical engineering.read less
|
This object is in collection