Progression of fretting fatigue damage in Ti–6Al–4V

Abstract

An investigation was conducted to explore the nature of fretting fatigue damage in the stages prior to crack formation. In the unique experimental apparatus employed in this study, where total slip never occurs, several locations on each test specimen exist where cracks can develop due to local contact conditions. Under the test conditions used, not all of the sites had cracks upon test completion. This study evaluated the condition of non-cracked sites on several fretted specimens in an effort to identify differences between these and sites where small cracks were observed. A single test condition of 620 MPa average applied static clamping stress and 250 MPa applied axial fatigue stress for R = 0.5 was selected, which corresponds to a fretting fatigue life of 10 7 cycles based on prior work. For specimens tested to 10 6 cycles, or 10% of life, several destructive and non-destructive characterization methods were chosen: scanning electron microscopy (SEM), residual stress measurement and transmission electron microscopy (TEM). Each site at which crack nucleation could be expected was inspected in the SEM and was then characterized using surface X-ray diffraction to quantify the residual stresses field near that location. Then TEM foils were cut from one area on a specimen with tiny cracks and dislocation densities were observed. A novel technique was used which permitted TEM samples to be obtained from regions in close proximity on the original specimen. Comparisons were made between as-received (AR) and stress-relief annealed (SRA) specimens, on which the stress-relief was applied prior to fretting fatigue testing. SEM inspection was useful for qualitative analysis of wear debris and identification of cracks as small as 20 μm, but was unable to provide quantitative data on the level of fretting fatigue damage beyond crack size. Although differences were noted in the residual stresses for the SRA versus the AR specimens, no residual stress peaks were noted in the edge of contact regions where cracks would eventually develop. TEM observations in the vicinity of the crack nucleation region showed that the dislocation structure decayed rapidly into the specimen thickness. The cause of the dislocations was attributed to plastic deformation caused by the clamping stresses.