PSI - Issue 51

Michael Horvath et al. / Procedia Structural Integrity 51 (2023) 95–101 M. Horvath et al. / Structural Integrity Procedia 00 (2022) 000–000

98 4

Fig. 2. Calibration of the critical distance L M as a function of the number of cycles to failure N f .

Fig. 3. Validation of the TCD-framework utilizing generalized S-N curves of notched specimens (2  = 135°).

Table 1 shows the average deviations of the numerically derived results in regard to the experimental data.

Table 1. Average deviations between numerical and experimental results for the investigated load ratios R = -1, R = 0 and R = 0.5.

N f < 10

f < 5∙10

7

5

N

Avg. Deviation  [%]

Avg. Deviation  [%]

R [ - ]

Avg. Deviation N f [%]

Avg. Deviation N f [%]

-1

2.3 3.5

-20.2 -28.1 -69.2

0.9 2.1

-5.1 -11.1 -61.2

0

0.5

18.2

18.1

Considering the results at load ratios of R = -1 and R = 0, the calculated datapoints show only a small deviation from the corresponding S-N curves. Only the results for R = 0.5 represent a more conservative fatigue assessment. Due to the high upper loads, elastic-plastic mean stress relaxation occurs, resulting in an effective load ratio of R eff less than 0.5, which cannot be considered properly by the applied linear-elastic methodology. Hence, by invoking elastic-plastic material behaviour into the numerical routine, prediction accuracy may improve significantly. However, in terms of stress range, the averaged deviations of less than twenty percent reflect the typical accuracy found in literature for fatigue strength of notched components invoking the TCD (Chiandussi and Rossetto 2005, Taylor et al. 2000), confirming the sound applicability of the assessment method in the medium-cycle fatigue regime.