Crack initiation under fretting conditions: experimental and numerical evidence for a dominant mechanism

This work investigates the crack-initiation mechanism under fretting conditions. Experimental and numerical investigations were carried out on a fretting pad model with flat-on-flat contact conditions using 34CrNiMo6 + QT steel. The investigations comprised two series of experiments: under fretting-fatigue loading and under plain fretting loading. All experiments were conducted with a nominal contact pressure of p = 40 M P a and a slip-amplitude of s a = 5 μ m . The slip-amplitude was measured and controlled in real-time from the fretting loops using a new method. In the fretting-fatigue tests, the fretting-fatigue limit was first determined. Subsequently, several fretting-fatigue tests (slightly above the fretting-fatigue limit) and plain fretting tests were stopped after defined load cycle numbers. The specimens were examined metallographically to determine crack-initiation lifetimes and crack-growth behaviour. The crack-initiation lifetimes hardly differed between the test series and were about 5000 load cycles. The crack lengths for short cracks were also similar at comparable load cycles. In all cases, cracks initiated near the centre of the contact rather than at the contact edges, where the highest stresses are expected if Coulombic friction is assumed. Three effects were evaluated in the numerical simulations: Coulombic friction, local form fitting, and local material bonding due to adhesion. The latter two assumptions resulted from microscopic analysis of the fretting scars, which showed conspicuous spots. Crack-initiation position, lifetime, angle, and fretting-loop shape were evaluated using the Smith–Watson–Topper (SWT) and Fatemi–Socie (FS) parameters and linear-elastic fracture–mechanical approaches. Most indicators pointed to a local adhesion spot as the predominant factor.