Fretting Fatigue of Nickel Based Superalloys at Elevated Temperature

Abstract

Experiments are conducted to characterize the fretting fatigue behavior of advanced nickel based alloys common to turbine disk and blade applications. The experiments address the effects of different materials, loading types, and surface treatments. Block loading experiments show that minor cycles reduce fretting fatigue life. Accurate friction coefficient data necessary to develop fretting fatigue life prediction is also reported. Background Fretting refers to the small scale relative tangential motion that occurs at the edge of nominally clamped contacts subjected to oscillatory loading. The formation and propagation of cracks in the presence of fretting induced stresses is known as fretting fatigue. Blade/disk attachments in gas turbine engines are susceptible to fretting fatigue. This paper describes experiments that characterize fretting fatigue in advanced nickel based alloys subjected to contact stresses similar to those that occur at blade/disk contacts. The experimental apparatus used for these experiments has been described elsewhere. 1 A schematic defining the applied loads is given in Figure 1. The experiments are designed to address the influence of different loading types, surface treatments, contact geometries, and contact materials on the fretting fatigue performance of advanced nickel based alloys. An initial set of experiments establishes baseline fretting fatigue life parameters. The baseline cyclic experiments are used to determine the applied load levels to produce a nominal fretting fatigue life. The desired fretting fatigue life is 100,000 cycles to failure. The normal load is held constant for the eight tests, and the applied bulk stress amplitude is varied to produce the desired life. These baseline experiments are for the contact of polycrystalline nickel alloys. The specimen material is Rene’ 95 and the pad material is a directionally solidified Rene’ 80. The contact pad geometry is flat with rounded edges depicted in Figure 2. Both the specimen and pad were shot peened. The baseline results are shown in Figure 3. Note, one experiment went to runout at approximately one million cycles. This runout experiment had the standard axial load ratio of approximately 0, but the shear load ratio was -3. Table 1 lists the experiments, with the appropriately measured experimental loads. The measured and controlled loads are identified separately as the tangential load (Q) is the difference between the loads measured by the top and bottom load cells (Figure 1). The final 3 experiments (test numbers 6-8) given in Table 2 are approximate repeats of the nominal loading profile for the Rene’ 95/Rene’ 80 flat shot-peened experiments. These baseline results form the basis of comparison to assess the effect of shot peening and material on fretting fatigue life. The difference between the loads given in Tables 1 and 2 is the inclusion of the prestress (as measured by the bottom load cell) that is applied to the specimen before application of the normal load.