Issue 48

L. Reis et alii, Frattura ed Integrità Strutturale, 48 (2019) 318-331; DOI: 10.3221/IGF-ESIS.48.31 330 ER1 ER2 ENR FSm Findley 0º 0º 0º 21º B-M 0º 0º 0º 21º S-W-T 0º 0º 0º 24º F-S 0º 0º 0º 20º LIU I 0º 0º 0º 20º LIU II 90º/0º/+90º 90º/0º/+90º 90º/0º/+90º -70º/+20º Measured 0º -22º/0º/5º 0º 0º Table 4: Comparison of the measured crack plane angle with estimates. For the ER1 and ENR loading cases, all models correctly predicted the crack plane for all stress levels. For the ER2 sequence, two of the specimens showed distinct crack plane angles. In all the criteria, in each plane, the damage parameter showed the same value for loadings ER1, ER2 and ENR, which share the same loading path. This behaviour shows that critical plane models do not consider the load sequence. For the FSm loading case, all the specimens fractured at 0º, and none of the selected criteria was able to predict that crack plane angle. C ONCLUSIONS n this work, one non-random and three random sequences were applied to specimens of the 42CrMo4 steel under different stress levels. Fatigue life from the ER1 and the ER2 loading sequences were considerably shorter than the lives obtained with the non-random equivalent loading sequence, ENR. The randomness introduced in the loading path, shared by these three loading sequences, significantly reduced fatigue life. For all the loading cases fatigue life was estimated by the SSF model using Miner and Morrows damage accumulation rule and using as cycle counting methods the Rainflow rule and vcc method. The difference between Miner and Morrow’s rules is negligible in comparison with Wang- Brown in all loading cases. There is only a big difference between the SFF models when the cycle counting method is changed, having almost opposite behaviors, the Rainflow tends underestimate life and vcc tends to overestimate, but both are able to predict fatigue life under the boundary life factor 3, which is commonly used in fatigue life correlation. Damage accumulation, at time of fracture, was not only estimated with the SSF model but also with Wang-Brown’s model (W-B). Wang-Brown’s underestimated the damage of the ENR, ER1 and ER2 loading sequences, as oppose the damage overestimation shown in the FSm loading. The SSF model also showed a tendency to underestimate damage, although, in a smaller scale, compared to W-B. Even though the SSF criterion, at its present state, can’t account the mean stress effect, it showed good results when estimating fatigue life with non-proportional random loadings with variable stress amplitudes and mean stress, performing better than the W-B model. However, additional studies are needed to confirm, or not, the verified trend according to which the SSF equivalent stress tends to be lower than the value it should be under random loadings with variable amplitude. A CKNOWLEDGEMENTS This work was supported by FCT, through IDMEC, under LAETA, project UID/EMS/50022/2019. R EFERENCES [1] Cui, W. (2002). A state-of-the-art review on fatigue life prediction methods for metal structures, Journal of Marine Science and Technology, 7(1), pp. 43-56. DOI: 10.1007/s007730200012. [2] Campbell, G. S., Lahey, R. (1984). A survey of serious aircraft accidents involving fatigue fracture, International Journal of Fatigue, 6(1), pp. 25-30. DOI: 10.1016/0142-1123(84)90005-7. [3] Gates, N., Fatemi, A. (2016). Multiaxial variable amplitude fatigue life analysis including notch effects, International Journal of Fatigue, 91(2), pp. 337-351. DOI: 10.1016/j.ijfatigue.2015.12.011. [4] Freitas, M., Reis, L., Meggiolaro, M. A., de Castro, J.T.P. (2016). Comparison between SSF and Critical-Plane models to predict fatigue lives under multiaxial proportional load histories, Frattura ed Integrità Strutturale, 38, pp. 121-127. DOI: 10.3221/IGF-ESIS.38.16. I