PSI - Issue 75

Tomáš Karas et al. / Procedia Structural Integrity 75 (2025) 150–157 T. Karas et al. / Structural Integrity Procedia 00 (2025) 000–000

156

7

stress is involved ( R = 0), the estimate for both pad geometries sits around 10% below the measured FL. This relatively small dispersion of estimate can be extremely advantageous and definitely deserves further investigation, unfortunately the amount of experimental data for fretting fatigue under R = 0 loading is sparse and in combination with thermography measurement practically non-existent.

3.3. FEA, Multiaxial Fatigue Criteria and Fractography Results

As mentioned in Section 2.5, the FEA and post-processing results can be divided into three categories: local FI values as the primary output from the selected criterion and gradient-corrected FI values using the Line and Point methods of TCD. The FI values obtained from the fatigue criteria primarily indicate regions with an increased likelihood of crack initiation and provide a measure of the local loading severity. Both the Dang Van and Papuga QCP criteria correctly identify the exposed contact edges as the critical area. In this region, the FI values slightly ex ceed 1.0, indicating conservativeness of prediction and therefore the need for further gradient correction using TCD. The gradient corrected FI by LM and PM gives similar results, with LM being less conservative and showing higher values of FI at the critical distance. Results of the LM are shown in Fig. 5, so the focus can be placed more on the di ff erence between the criteria and the geometries of the pads. In addition, fractography images were used to validate the predicted fracture behaviour. The main point of interest is the initiation process, where two regimes were observed: multiple point initiation along the whole edge of contact and single point initiation at the edge near the sample corner. For the single contact pad with r = 50mm and R = 0 both selected criteria predicted a shift between these two regimes with decreasing σ a . For higher stress amplitudes, the initiation is scattered across the whole edge, while with lower amplitudes the initiation unifies to a single point and propagates asymmetrically. This is in agreement with the fractography showing that both criteria are applicable. For the other case (r = 1mmand R = 0), the same shift between two regimes is not predicted by the criteria and, in addition, both distinguishably disagree on the initiation position. While the fractography shows similar behaviour as for the other pad. With this in mind, we can look at the range of PM results and see that the di ff erence between the maximum and minimum values is considerably small. Therefore, in agreement with multiple point initiation, but further comparison with fractography is necessary to decide which criterion is more precise.

Fig. 6. Dang Van and Papuga QCP fatigue indexes corrected by Line method compared with fractography.