PSI - Issue 13

Fidan Smaili et al. / Procedia Structural Integrity 13 (2018) 1347–1352 Fidan Smaili, Tomaž Vuherer / Structural Integrity Procedia 00 ( 2018) 000 – 000

1350

4

a

b Fig. 5. Result of instrumented Charpy tests; the CG HAZ microstructure; (b) the FG HAZ microstructure

Specimens of FG HAZ and CG HAZ with a dimensions 10×10×55 mm were used fatigue crack growth tests Results of the tests are presented and plotted in Figure 6. Before starting the experiment, special thin foil crack gauge was attached on one side of the specimen. It was used for measuring the crack propagation during the loading. Testing were performed by bending loading, where the load ratio (R) was 0.1. The initial 2 mm long pre-crack was prepared with the low ∆K. The crack threshold was obtained step by step by reducing ∆K from initial ∆ K, which was used for pre-crack creation.

Fig. 6. Results of fatigue crack growth tests; the CG HAZ microstructure; (b) the FG HAZ microstructure

The threshold for long crack fatigue propagation in CG HAZ microstructure is 10.50 MPam 0.5 while the threshold for long crack fatigue propagation in FG HAZ is 4.33 MPam 0.5 , what is almost 2.5 lower threshold. These mean that small crack in FG HAZ can easily initiate and propagate from the defect, which can be present in FG HAZ microstructure, because the grain size is relatively small. Furthermore, the grain boundary do not represent strong barrier at the early crack growth when the initiated small crack is still physically small (few grains long). When the small crack initiate in FG HAZ because of the small crystal grains boundary is so weak microstructurally barrier that cannot stop crack propagation and it propagate until crack become long crack. Initiated microstructural small crack (smaller grain or just one grain or few grains in size) in CG HAZ is at the beginning retard with strong microstructural barrier which it is grain boundary. In some cases these barriers are so strong that the microstructurally small cracks can be stopped in their propagation if the loading rate is not too high. Some non-propagation small crack can appear during testing before crack is reaching the threshold for long crack fatigue propagation because it is relatively so high. Fatigue crack growth of the long cracks is at the beginning lower in CG HAZ microstructure in comparison to FG HAZ microstructure, but when the crack is long enough or loaded with higher load (  K ) crack growth become faster in CG HAZ where the microstructure has lower fracture toughness. Few different fatigue tests with artificial defect in CG HAZ and FG HAZ microstructure were tested in order to confirm before mentioned statements. Specimens had the same stress concentration like in real weld ( K t = 1.74). Fatigue specimens are presented in Figure 7.