| Abstract Scope |
Introduction:
Additive manufacturing (AM) of Nb-bearing nickel superalloys promotes the formation of various precipitates, such as carbides and intermetallic phases, due to elemental segregation during solidification. After subsequent exposure to high temperature, solidified microstructures are susceptible to the accelerated precipitation of detrimental δ-phase due to Nb enrichment. Conditions that promote δ-phase precipitation are generally avoided because the characteristic needle or platelet morphologies can serve as crack initiation sites and reduce ductility, limiting high temperature applications. The volume fraction of δ-phase that forms at high temperature is directly linked to the amount excess Nb available in solution. The distribution of Nb between the matrix, solidification products, and δ-phase can be tailored through variations in initial alloy chemistry. When significant amounts of nitrogen are present in AM Inconel® 625, various Nb-containing nitrides form that are not commonly observed in nickel superalloys. The types of nitrides that form depend on the relative mass fractions of minor alloying elements and can significantly alter the amount of δ-phase that precipitates upon exposure to high temperature.
Experimental Procedures:
Two Inconel® 625 alloys with different Fe, Si, and Ti mass fractions were used to build multi-pass, L-shaped wall structures by laser-based directed energy deposition AM. The materials were deposited using a 2000 W laser power, 10.6 mm/s scan speed, 4 mm beam diameter, and approximately 0.25 g/s powder flow rate. Selected portions of the structures underwent hot isostatic pressing (HIP) at approximately (1163 ± 25) °C for 4 h at 101 MPa to reduce porosity and homogenize elemental segregation. Samples from both materials in the as-deposited and HIP conditions were subjected to isothermal heat treatment at 870 °C in a tube furnace for times ranging from 0.5 h to 1000 h. After air cooling to room temperature, the samples were sectioned using a low-speed diamond saw and surfaces were metallographically ground and polished to a 0.05 μm finish using colloidal silica. Precipitates were revealed by electrolytic etching with 10 % chromic acid at 3.5 V for approximately 5 s.
Precipitates in the as-deposited and HIP conditions were identified by high-energy synchrotron X-ray diffraction using beam line 11-ID-B at the Advance Photon Source of Argonne National Laboratory. Diffraction patterns were generated with a beam energy of 58.59 keV and 0.2113 Å wavelength, and the individual peaks were identified based on powder diffraction files from the NIST Inorganic Crystal Structure Database. Focused ion beam milling was used to prepare TEM samples, where the crystal structures and chemical compositions of individual precipitates were analyzed. After isothermal heat treatment, precipitate phase fractions were quantified by the point count method using images acquired from scanning electron microscopy. The spatial distribution of chemical composition was tracked by energy dispersive spectroscopy for different heat treatment times.
Results & Discussion:
In the Inconel® 625 with relatively low silicon (0.05 %) and high titanium (0.21 %) mass fractions, Nb- and Ti-rich MN type nitrides are identified in the as-deposited microstructures. The low volume fraction ((1.1 ± 0.2) %) of MN nitrides leaves much of the Nb in the alloy in solid solution, where the average Nb mass fractions in the interdendritic regions and dendrite cores measure (6.19 ± 0.84) % and (2.02 ± 0.27) %, respectively. After HIP, the elemental segregation of Nb is homogenized to a value of approximately to a mass fraction of (3.57 ± 0.12) %, and the volume fraction of MN nitrides remained largely unchanged. However, the nitrides became more enriched in Ti after HIP. When subjected to isothermal heat treatment at 870 °C, a δ-phase volume fraction of approximately 10 % was measured in both the as-deposited and HIP conditions after 1000 h.
In contrast, Z-phase (NbCrN) and η-nitrides (M6N) form in the alloy with high silicon (0.39 %) and low titanium (0.03 %) mass fractions, and the overall volume fraction of nitrides in the as-deposited condition is (2.3 ± 0.4) %. The spatial variations in Nb are reduced to (3.49 ± 0.40) % in the interdendritic regions and (1.73 ± 0.12) % in the dendritie cores due to consumption by nitrides. After HIP, elemental segregation is eliminated and the nitrides persist. During heat treatment, the Z-phase dissolves and the volume fraction of η-nitrides increases to (4.6 ± 0.6) %. With much of the excess Nb being consumed by η-nitrides, the total volume fraction of δ-phase remains below approximately 2 % for the duration of the 1000 h heat treatment. For comparison, wrought Inconel® 625 plate formed approximately 6 % δ-phase after the same heat treatment.
Conclusions:
The type and phase fraction of nitrides resulting from small changes in AM Inconel® 625 alloy composition influence δ-phase formation during isothermal heat treatment at 870 °C. The root cause controlling the volume fraction of δ-phase that forms is the amount of niobium available in the matrix. Alloy compositions favoring the formation of a large fraction of Nb-containing precipitates reduce the source of Nb necessary for δ-phase formation. For an alloy composition that promotes Ti-rich MN nitride formation, enough Nb is available in solution to form approximately 10 % δ-phase after 1000 h of heat treatment. However, the alloy containing Z-phase and η-nitrides consumes much of the available Nb, leading to a relatively low δ-phase of less than 2 % after the same time. The results show that control of initial alloy composition provides a means for outperforming wrought Inconel® 625, which contains relatively low fraction of M(C,N) carbonitrides and η-carbides and forms about 6 % δ-phase. |