Fretting fatigue crack initiation in titanium alloy, Ti−6Al−4V, was investigated experimentally and analytically by using finite element analysis (FEA). Various types of fretting pads were used in order to determine the effects of contact geometries. Crack initiation location and crack angle orientation along the contact surface were determined by using microscopy. Finite element analysis was used in order to obtain stress state for the experimental conditions used during fretting fatigue tests. These were then used in order to investigate several critical plane based multiaxial fatigue parameters. These parameters were evaluated based on their ability to predict crack initiation location, crack orientation angle along the contact surface and the number of cycles to fretting fatigue crack initiation independent of geometry of fretting pad. These predictions were compared with their experimental counterparts in order to characterize the role of normal and shear stresses on fretting fatigue crack initiation. From these comparisons, fretting fatigue crack initiation mechanism in the tested titanium alloy appears to be governed by shear stress on the critical plane. However, normal stress on the critical plane also seems to play a role in fretting fatigue life. At present, the individual contributions/importance of shear and normal stresses in the crack initiation appears to be unclear; however, it is clear that any critical plane describing fretting fatigue crack initiation behaviour independent of geometry needs to include components of both shear and normal stresses.

The effects of shot-peening on the fretting fatigue behavior of titanium alloy, Ti-6Al-4V were investigated. Specimens were shot-peened as per AMS 2432 standard. X-ray diffraction analysis measured a maximum compressive stress of 800 MPa at the specimen surface, which reduced to zero at a depth of 188 μm. The compensatory residual tensile stress in the specimen was estimated using a curve fitting technique, the maximum value of which was found to be 260 MPa at a depth of 255 μm. Fretting fatigue tests were conducted at room temperature at a cyclic frequency of 200 Hz. Scanning electron microscopy of the shot-peened fretting fatigue specimens showed that the crack initiated at a point below the contact surface, the depth of which was in the range of 200–300 μm. Finite element analysis of the fretting fatigue specimens was also conducted. Fatigue life diagrams were established for the fretting fatigue specimens with and without shot-peening, and were compared to those under the plain fatigue condition, i.e. without fretting. Shot-peening improved the fretting fatigue life of Ti-6Al-4V; furthermore, it moved the crack initiation site from the fretting contact region to a region inside the specimen. Moreover, stress analysis showed that the fatigue failure of shot-peened specimens was caused by the compensatory tensile residual stress.

No abstract

A fretting fatigue crack initiation mechanism (number of cycles, location and orientation angle) using critical plane based parameters has been addressed by several researchers. There are several process variables that can affect these parameters and thereby the prediction of fretting fatigue crack initiation behaviour. Effects of two such parameters, viz. process volume and the coefficient of friction, were investigated in this work. Fretting fatigue experiments with a titanium alloy were conducted with different contact pad geometries. Finite element analysis (FEA) was used to obtain a stress state in specimens for the experimental conditions used during fretting fatigue tests. Analysis was carried out for two values of the coefficient of friction, thereby providing a framework for calculation of several critical plane based multiaxial fatigue parameters for different process volumes. A program was developed to compute these multiaxial fatigue parameters from the FEA data for different values of process variables. It was observed that parameters for cylindrical pad geometries with no singularity-type behaviour were inversely proportional to the size of process volume and directly proportional to the coefficient of friction. There was no change in the predicted orientation of the primary crack for this geometry, due to variations in these process variables. Parameters for flat-pad geometries with behaviour approaching that of a singularity were also inversely proportional to the size of process volume, but the coefficient of friction had a minimal effect on their values. Predicted orientation of the primary crack for these geometries changed slightly when the process volume increased from that of a grain size of the tested material to a larger size, and then did not change with the increase of process volume size. Overall, the effect of these process variables on the critical plane based parameters was similar in all five contact geometries used in this study, when the scatter in fatigue data is kept in mind. Finally, the modified shear stress range parameter satisfactorily predicted the crack initiation location, orientation angle and number of cycles to fretting fatigue crack initiation independent of the contact geometry for a given process volume size and coefficient of friction.

No abstract

No abstract

No abstract

Fretting fatigue is complex phenomenon which depends on various factors. The present study focuses on the contact geometry dependence of fretting fatigue crack initiation behaviour by combining tests and their analysis. Various geometries were investigated which ranged from cylindrical shape of different radii to essentially flat including flat with rounded edges in Ti alloy Ti-6Al-4V. The measured fretting fatigue life data were compared using stress range, effective stress and a critical plane based parameter, modified shear stress range (MSSR) for the eight pad geometries where first two are the variations of the applied cycling stress on the substrate while the third one is the combination of normal and shear stresses on the critical plane in the contact region. Fatigue life relationships from all contact geometries as well as from the plain fatigue (i.e. without fretting) were within a scatter band as commonly seen in the fatigue tests. Further, MSSR predicted the crack initiation location and its orientation at the contact surface, which were in good agreement with their experimental counterparts. Fretting fatigue crack initiation behaviour is thus governed by a combination of shear and normal stresses on the critical plane in the tested material. A methodology to design against the fretting induced crack initiation damage can be thus developed from the approach of this study using the plain fatigue data only, especially in the high cycle fatigue regime which is primarily governed by the threshold considerations.