Issue 48

C. Santus, Frattura ed Integrità Strutturale, 48 (2019) 442-450; DOI: 10.3221/IGF-ESIS.48.42 444 [34]. Though type I and type II cracks where not differentiated, they discussed several anomalies between the experimental orientation of the crack and the predictions, and an intermediate experimental crack angle, with respect to the SWT (tensile) and the FS or MWCM (shear) criteria, was observed in one case. The normal stress component for the correction of both FS and MWCM criteria can play a role in distinguishing between the two equal maximum shear stress amplitude directions, at an angular distance of 90°. This is of great interest for this kind of analysis, especially considering that the crack orientation outward, which can be predicted as an effect of the normal stress correction, is very rarely observed. In this paper the fretting fatigue tests, which we previously presented in Bertini and Santus [22], are reconsidered and the initial crack orientation, evident by means of Scanning Electron Microscope (SEM) observations, is discussed by comparing these approaches proposed in the literature. The main interest of these tests is that the contact pressure fluctuates with the bulk bending load, which causes the tensile stress to be less predominant than in the usual test rigs where the pressure remains (almost) constant during the fretting cycle. Consequently, the observed cracks are clearly oriented with a shallow, or even very shallow, initiation angle towards the inside of the contact thus obviously type I (shear) cracks. E XPERIMENTAL TESTS he setup of the fretting experimental tests investigated in this paper is shown in Fig. 1 (a), following our previous work: Bertini and Santus [22]. The specimen shaft is manufactured in aluminium alloy 7075-T6, with a conical end which is then shrink-fitted into a steel hub, allowing the connection shrinkage intensity to be controlled with an adjusting nut. The alternating load was applied at the other side, with a hydraulic actuator, and for all the tests the load ratio of this remote force was R = -1. The fretting crack nucleated at the edge of the “flat and rounded” contact, which can be assumed as (locally) plane strain. Figure 1 : (a) Fretting test setup for shrink-fitted shaft-hub connection, geometry details and (b) contact pressure fluctuation at the fretting hot-spot during the loading cycle. T