PSI - Issue 57

David Mellé et al. / Procedia Structural Integrity 57 (2024) 61–72 David Melle´ / Structural Integrity Procedia 00 (2023) 000–000

70

10

Figure 11 shows the same Kitagawa-Takahashi diagram as Figure 5, with the previously identified El-Haddad and Topper model and material strength variation, from critical features from all surface states. Three 30 min etched specimen surface features sizes populations were also drawn on this graph to illustrate the following demonstration. The scatter on the stress values for a given specimen is due to the stress variation on the gauge surface (due to longitudinal and thickness stress gradients).

Fig. 11. Kitagawa-Takahashi diagram including El-Haddad and Topper model, the material strength variation identified on critical points and three 30 min chemically etched coupons full surface features populations.

It is easy to understand that for a coupon with large surface features and a low strength material surrounding them, the failure will occur quickly after the largest features enter the critical area (grey area). Only a few surface features are then critical and the killer surface feature will be among the largest if not the largest one. Out of the 9 coupons with 30 min etched surfaces in Figure 9, two exhibit this configuration (coupon 1 and coupon 8). The surface feature population of coupon 8 is shown in Figure 11 as an illustration. An increased strength in the material surrounding the largest features leads to an increase in the stress level needed to achieve failure. Even with large surface features. More features may then enter the critical area but the probability that the killer-feature is among the largest ones is still high. Coupon 2 is the only coupon in the 30 min etched batch which follows this logic. Its surface feature population is shown in Figure 11. Finally, for coupons exhibiting either low surface features maximal sizes (coupon 3, coupon 4 or coupon 7) or / and whose largest surface feature surrounding material exhibits a high material strength (coupon 5, coupon 6 and coupon 9), the stress level needed to achieve failure is so high that almost the whole surface features population is in the critical area. The killer-feature may then be significantly smaller than the largest surface features. To illustrate this, the coupon 3 surface feature population is shown in Figure 11. This explains why the killer-feature does not have the highest size or the highest ∆ K value. This explanation is obviously highly dependent on the scatter in the inherent strength (defect free strength) of the material.

4. Conclusion

In this work, an experimental fatigue testing campaign has been conducted on flat bending coupons, additively manufactured out of the Ti-6Al-4V alloy. Fatigue test were conducted at a stress ratio of R = 0 . 1 and the e ff ect of chemical etching was investigated. The surface topography of the gauge surface of all coupons was scanned and the