Rolling contact fatigue tests of JIS-SUJ2 steel were performed using plate specimens having a small drill hole with various diameters and depths. In all the tests, fatigue crack initiated at the edge near the bottom of the hole just after starting the fatigue test, and then propagated by shear mode. Even in the unbroken specimens tested up to N = 2×108 cycles, a short fatigue crack was found at the edge. Fatigue life, Nf, plotted against the maximum contact pressure, qmax, was greatly varied according to the diameter and depth of the hole. The effect of the depth at crack initiation on the fatigue life Nf was uniquely characterized by using the nominal shear stress amplitude, τa, instead of qmax. Further, based on the consideration of the rolling contact fatigue strength as a small shear-mode crack problem, the fatigue life data were plotted using the novel parameter, τa/(√area)-1/6, where the √area is a projected area of the hole. Consequently, all the fatigue life data were successfully fitted on a unified line irrespective of the diameter and depth of the hole, i.e. a defect size dependence in the rolling contact fatigue strength was manifested in a small crack regime. Moreover, the fatigue life data for smooth specimens fractured from an internal non-metallic inclusion were also in accordance with the data for the drill holes in the plot of Nf versus τa/(√area)-1/6. The result infers that a lower limit of the rolling contact fatigue strength of bearings with various dimensions can be evaluated by means of the following two dominant parameters: (i) maximum inclusion size predicted by the statistics of extremes in a given control volume and (ii) the maximum value of nominal shear stress amplitude produced under the rolling contact.
