Fig. 1. Schematic illustrations of (a) heat treatment routes for the LMD-fabricated TA19 alloys, (b) SEN samples and (c) sample sampling direction (Z-X). Here, LD: loading direction and ASD: applied stress direction.

Fig. 2. OM images of columnar PBGs for the (a) as-built and heat-treated samples with different conditions of (b) 940SAA, (c) 970SAA, (d) 1030SAA and (e) 1030SFA.

Fig. 3. OM and SEM (upper right corner illustration) images of microstructures for (a, b) as-built and heat-treated samples with different conditions of (c) 940SAA, (d) 970SAA, (e) 1030SAA and (f) 1030SFA.

Fig. 4. (a) FCG rate (da/dN) versus stress intensity factor range (ΔK) with double logarithm case, (b) relation between the FCG thresholds (ΔKth) and the microstructure characteristic scale for the as-built and heat-treated samples, (c) near-threshold crack growth paths for the 940SAA sample and (d) near-threshold FCG behavior on the fracture for the 940SAA sample.

Fig. 5. Paris law parameters n and c of the LMD-fabricated TA19 alloys in this study and the forged Ti6242 alloys in the literature [40], [41], [42].

Fig. 6. (a) Angle (γ) of the growth direction of columnar PBG deviating from build direction (Z) and deflection angle (θij) of fatigue cracking path at columnar PBG boundary of the 1030SAA sample and (b) the crack deflection angle (ψij) at lamellar colony boundary in the 1030SFA sample (A-G are the symbols marking lamellar colony).

Fig. 7. Schematic illustrations of (a) the coordinate system for the AM system (X-Y-Z), (b) the crystal coordinate system (x[100]-y[010]-z[001]) of β grain, (c) the three-dimensional plate-like α lath and (d) the angle (φ) between the trace direction of the α lath and applied stress direction (ASD) on the observation surface (XOZ).

Fig. 8. Schematic illustrations of the long axis trace direction of different α variants in the three columnar PBGs containing the crack path for the 1030SAA sample.

Fig. 9. Theoretically calculated and experimentally measured angles (φ) and their angle difference (Δφ) of different α laths in the adjacent columnar PBGs of 1030SAA samples and the angle (γ) of the growth directions for the columnar PBGs deviating from the build direction (Z).

Fig. 10. Statistical distribution of experimentally measured φ of α laths in different heat-treated samples (The specific data are from Fig. 8 and Fig. S5.).

Fig. 11. Deflection angles θij or ψij at columnar PBG boundaries or colony boundaries of heat-treated samples.

Fig. 12. LCM micrographs of fracture roughness for the (a) as-built and heat-treated samples with different conditions of (b) 940SAA, (c) 970SAA, (d) 1030SAA and (e) 1030SFA, (f) roughness values of different samples.

Fig. 13. Model diagrams of (a) the relationships between the angle (γ) of columnar PBGs deviating from build direction (Z) and the deflection angle (θij) of fatigue crack at columnar PBG boundary and the interaction between fatigue crack and α laths with different φ, as well as (b) the relationships between the angle (γ) and the angle (φ) of α laths.

Fig. 14. (a) Relation between the angle (γ) and the angle (φ), (b) schematic illustration of the best combination of angles (γ) of several adjacent columnar PBGs for high FCG resistance.

Fig. 15. Schematic diagram of (a) the scanning strategy and (b) the relationship between the proposed AM fabrication process parameters, molten pool geometry and columnar PBG growth direction.