S. K. Srivatsa, J. R. Groh, GE Aircraft Engines, Cincinnati, OH; T. Limbo, Smiths Aerospace Components, Ashville, NC

Summary:

Extrude + isoforged powder alloy René 104 was initially developed for thick-section, critical rotating aircraft engine turbine disks subject to elevated temperature for extended periods during operation. Dwell fatigue and metallurgical stability benefits versus current fine grain Ni-base alloys also make René 104 a viable candidate to improve the durability of non-critical turbine engine components under severe operating conditions. However, machining distortions, especially of thin parts, made from René 104 need to be characterized. This paper presents the results of finite element modeling which was successfully utilized to abate the risk of unacceptable distortion during the machining of thin section turbine hardware.

2-D machining distortion analysis was performed using the commercial software DEFORM. Residual stresses and strains from the modeling of prior heat treat operations, and the material removal sequence for each machining operation, were used as input to the machining model. A sensitivity analysis of the finished part location within the heat treat envelope and the differences in the two pieces (stacked 2–high) obtained from within the heat treat shape were investigated. Modeling was carried out until the operation where three-dimensional features such as bolt holes and scallops are machined.

Three finished aft cooling plates were manufactured to confirm producibility, quantify material distortion, and to provide hardware for follow-on engine test. Dimensional inspection was performed at a series of 5 radial locations, every 45 degrees around the circumference for each of three pieces. Component distortion during turning operations was significantly greater than that experienced for the same part made of another alloy (René 95). However, all three pre-forms yielded acceptable component quality and dimensions without either rework or case record activity. Measured distortion was in general less than that predicted by the model. However, modeling predictions did raise operator awareness, identified and abated processing risks and resulted in successful first-time machining.