PSI - Issue 66

Nur Mohamed Dhansay et al. / Procedia Structural Integrity 66 (2024) 87–101 Author name / Structural Integrity Procedia 00 (2025) 000–000

99 13

For LPBF Ti-6Al-4V, when annealing below the β -transus, the PBG boundaries are retained, but some morphological change may occur and increased texture due to transformation. In terms of grain morphology, both CM and LPBF produced Ti-6Al-4V can obtain the same morphology i.e. lamellar, bi-modal and equiaxed, and typically with the exception of PBG. However, due to the initial microstructures being different, applying the same heat treatments will result in some morphological differences between the two. Regarding the PBG, CM processes are able to obtain columnar PBG through thermomechanical processes above the β -transus. However, through the BOR controlling mechanisms, the PBG produced via the LPBF process will have a greater influence on the α lath morphology than the CM post-processing techniques. Considering that Δ K th is a crack propagation method, crack initiation has been approximated to occur at ~ 10 -9 m/cycle (Campbell, (2008)). Therefore, Δ K th in the global “closure-free” regime provides us with CGRs which can provide more insight into crack initiation. In the investigation by Bantounas et al (Bantounas et al., (2009)), cross rolled and unidirectionally rolled forged bars were considered for their fatigue behaviour. It was observed that the α grain orientation influenced the fatigue performance i.e. unfavourably orientated grains cause weaker fatigue performance (exacerbate crack initiation) (Bantounas et al., (2009)). More specifically, when the crystallographic orientation is favoured for basal slip, the shortest fatigue life is achieved and when orientated away from basal slip, the longest fatigue life is achieved. Given that near-threshold FCGRs are argued to be representative of crack initiation growth rates, it stands to reason that the ZX orientation’s crystallographic orientation is favoured more for crack initiation or faceted fracture compared to the remaining orientations and therefore has the lowest Δ K th . Similarly, in the work of Nalla et al (Nalla et al., (2002)) where at large enough R-ratios produced similar Δ K th for different microstructures, we observe similar behaviour for the ZX and XZ orientations bi-modal condition. This suggests that when fatigue behaviour is considered from a global “closure-free” perspective, LPBF and CM Ti-6Al-4V may have similar behaviour. 5. Conclusions The current investigation considered the near-threshold FCGRs of LPBF produced Ti-6Al-4V in its AF and SR condition compared to a bi-modal condition. The following can be concluded:  The bi-modal condition showed to have larger α -grains and the presence of β -phase than the LPBF conditions. This resulted in rougher fracture surfaces and more tortuous crack paths due to larger grain size, larger facets and α / β interfaces. These mechanisms are more influential in the global “closure-affected” regions.  The global “closure-free” region was also affected by the grain size and presence of β -phase. However, the mechanisms which improve the near-threshold FCGRs of bi-modal condition are related to plastic flow abilities of the bi-modal condition over the LPBF conditions.  For the LPBF conditions, a Δ K * th of ≈ 1.6 ± 0.2 MPa.m 0.5 and K * max ≈ 3 MPa.m 0.5 . This improved to Δ K * th of ≈ 2.7 MPa.m 0.5 in the ZX and XZ orientation and Δ K * th ≈ 3.5 MPa.m 0.5 for the XY orientation. For all three orientations a K * max ≈ 6 MPa.m 0.5 was achieved.  The ZX orientation in the bi-modal condition showed the most facet regions on the fracture surface an and for all R-ratios had the lowest Δ K th than the remaining orientations. The XZ orientation obtained a comparable Δ K th as the ZX orientation, but at a higher R-ratio. The XY orientation had the highest Δ K th and is likely to achieve a comparable Δ K th to the remaining orientations at higher R-ratios.  The LPBF bi-modal condition was shown to have larger, but comparable Δ K th to that of CM produced Ti 6Al-4V. Acknowledgements The authors acknowledge the powder analysis data from the Central University of Technology, the support received from the Stellenbosch Technology Centre and gratefully the funding from the Department of Science and Innovation through the Collaborative Program for Additive Manufacturing (CPAM).