Published Online: 31 July 2008
Page Count: 12
Golden, Patrick J.
Air Force Research Laboratory, Wright-Patterson AFB, OH
University of Dayton Research Institute, Dayton, OH
The University of Texas, Austin, TX
(Received 4 December 2007; accepted 30 June 2008)
The objective of this work was to evaluate life prediction methodologies involving fretting fatigue of turbine engine materials with advanced surface treatments. Fretting fatigue tests were performed on Ti-6Al-4V dovetail specimens with and without advanced surface treatments. These tests were representative of the conditions found in a turbine engine blade to disk attachment. Laser shock processing and low plasticity burnishing have been shown to produce deep compressive residual stresses with relatively little cold work. Testing showed these advanced surface treatments improved fretting fatigue strength by approximately 50 %. In addition to advanced surface treatments, several specimens were also coated with diamond-like carbon applied through a nonline-of-sight process capable of coating small dovetail slots in an engine disk. Testing with this coating alone and combined with advanced surface treatments also significantly improved fretting fatigue strength due to a decreased coefficient of friction along with the compressive residual stresses. This work presents a mechanics based lifing analysis of these tests that takes into account the local plasticity and the redistribution of residual stresses due to the contact loading. The use of superposition of the residual stresses into the contact stress analysis results in unconservative crack growth life predictions. Finite element analyses were conducted to predict the redistribution of residual stresses due to the contact loading. The redistributed residual stresses were used to make improved crack growth life predictions when possible. The results showed very little redistribution of residual stresses for the advanced surface treatments, however, a significant change in shot peened residual stress gradients was predicted.
Paper ID: JAI101610