PSI - Issue 68

S.R. Raghuraman et al. / Procedia Structural Integrity 68 (2025) 769–775 S.R. Raghuraman et al. / Structural Integrity Procedia 00 (2025) 000–000

772

4

Fig. 2: (a) Trend S-N curve calculated according to StressLife including the primary loading stress amplitude σ a, PL ; (b) material response of CATs at a stress amplitude of 720 MPa; (c) definition of the fatigue stage for primary loading experiments as a function of normalised lifetime based on the change in temperature The results of these tests are shown in terms of cyclic deformation curves based on the change in temperature in Fig. 2 (b). Since the change in temperature consists of elastic as well as plastic portions, Equation (1) is used in order to calculate ΔT , where T 1 , T 2 and T 3 , denote the temperature measured at the centre of the gauge length and the shafts of the specimen respectively. (1) The material response exhibits a long initial phase with almost constant values concerning the change in temperature (up to approximately 70 % of the number of cycles to failure N f ) for all CATs illustrated in Fig. 2 (b), before a significant increase can be observed as a result of cyclic softening processes. To ensure a reliable damage assessment on the specimen surface, selected specimens were primarily loaded up to a defined fatigue stage, characterised by sufficient damage induction. This requires a consideration of the material-related scattering of the fatigue lifetime to prevent premature specimen failure during primary loading. To determine the fatigue stages, the material responses were represented as a function of the normalised cycle numbers. A continuous increase regarding the change in temperature can demonstrably be observed in Fig. 2 (c) after an initial increase of approximately 60 % from a lifetime of approximately 75 %. Based on these findings the extent of primary loading was defined up to 75 % of fatigue life for the primary loading stress amplitude of σ a, PL = 720 MPa. The damage state can furthermore be quantified using magnetic methods such as magnetic Barkhausen noise (MBN). The results of the MBN measurements of the primarily loaded (PL) specimens including a comparison with thermographic data are visualised in Fig. 3. The initial condition (IC) serves as reference. Fig 3 (a) depicts the so called butterfly curve of measurements in IC in contrast to the PL condition. According to Schneider et al. (2012) materials that exhibit a pronounced cyclic softening behaviour and are therefore magnetically softer, tend to have a higher effective voltage ΔV RMS as material response to the applied magnetic field H . In this case, ΔV RMS corresponding to the specimens in PL condition ( ΔV RMS, PL ; red) is comparatively higher than in the IC specimens ( ΔV RMS, IC ; black),