The results of SEM inspections have confirmed a high sphericity and particle size of the commercial H-X superalloy powders that were used as the batch materials (Fig. 2a, b). As declared by the producers, the powder showed an average particle size below 15 µm. The powders were found to be free of shape defects (e.g. satellites). Spherical carbamide particles (Fig. 2c) combined with the applied processing allowed replicating their shape in the final porous samples. Finally, highly porous H-X sinters were successfully produced (Fig. 2d).
3.1. Macro- and microstructure of porous H-X alloy
The results of non-destructive CT analyses (see example in Fig. 3) revealed homogeneous distribution of porosity within the produced sinters. As revealed by the reconstructed models, even for a porosity as high as 70%, the sinters showed a good integrity and a well-defined porous structure. The shape and size of pores replicate the morphology of used carbamide particles. The results of metallographic analyses (Fig. 4) revealed: (i) an average grain size of ~ 120 µm: (ii) an existence of some internal microporosity (Fig. 4a); and (iii) a presence of skeleton-like precipitates at grain boundaries (GBs). More detailed analyses by simultaneous SEM/EDS/EBSD method (Fig. 4b-d) allowed recognizing these structural features as Cr-rich M23C6 and Mo-rich M6C carbides. The phase identification of GB precipitates is in line with results reported for additively manufactured (AM) H-X alloy [7], [8].
On the other hand ,a skeleton-like morphology of the GB carbides points towards a discontinous precipitation (DP) as the main governing reaction. The basic feature of this phenomenon is a “lamellar, transformation product behind a GB advancing into a supersaturated matrix” [9]. Over the years, the presence of DP-like products in heat treated nickel based superalloys has been documented and widely discussed by many authors [10]. Furthermore, we have also recently documented a presence of analogous structural features in another Ni-Fe-Cr-based alloy subjected to non-equilibrium oversaturation followed by aging treatments [11], [12].
For the sake of discussing the results of our present work, the model recently proposed by Atrazhev et al. [13] might be adopted. The authors have proposed that GBs mobility is the key factor governing the formation of either GB serration under low GBs mobility conditions or skeleton-type GBs structures (similar to these observed in the present work) when the mobility of grain boundaries is high. It is reasonable to assume that very high temperature applied during the final processing step (T = 1300°C = 0.96Tm), supports both high GBs mobility and diffusion kinetics. Therefore, it is proposed that the formation of skeleton-like M23C6/M6C products is driven by a local segregation of Cr to GBs areas combined with an effective grain growth (i.e. a migration of GBs) inside the matrix of produced porous sinters and a high chemical affinity of Mo to carbon. Furthermore, organic materials used during the fabrication process (paraffin wax and carbamide) can easily serve as the carbon source supporting the DP reaction.
3.2. Room temperature mechanical properties
Compressive stress-strain curves obtained for H-X samples with porosities of 50, 60 and 70%, are shown in Fig. 5, while values of quantitative parameters are listed in Table 1. The reference Hastelloy-X sample (marked as 0%) produced according to the same powder metallurgy-based procedure as described in the Section 2.1, but without introducing porosity formers, was used for the sake of comparison.
The results obtained for the reference H-X alloy sample show a good agreement with these recently reported in the literature for the alloy processed by additive manufacturing techniques [7], [14], [15]. However, it should be noted, that literature data regarding mechanical response of the AMed H-X alloy shows rather high scattering (Yield strength of 290–690 MPa; Ultimate Strength of 560–1060 MPa), as many variants of the processing and/or heat treatment, are applied in various laboratories.
On the other hand, it is worth noting that values available in the literature regarding mechanical properties of the Hastelloy-X alloy are mostly limited to those produce in tensile tests on additively manufactured specimens, and are rather highly scattered.
Table 1
The results of compression tests carried out on porous Hastelloy-X superalloy samples with 50–70% porosity (bulk alloy sample, marked as 0% was used as reference) and a comparison to reported literature data.
| Porosity [%] | Yield Strength (0.2) [MPa] | Ultimate Strength [MPa] | Total strain [%] | Compressive stress [MPa] |
| at 5% strain | at 10% strain | at 25% strain | at 50% strain |
| 0 | 251 | 847 | 19 | 445 | 583 | n.a | n.a |
| 50 | n.a | 1965 | 74 | 96 | 116 | 195 | 511 |
| 60 | n.a | 1144 | 73 | 82 | 100 | 163 | 375 |
| 70 | n.a | 1043 | 74 | 69 | 75 | 115 | 280 |
The results obtained for porous Hastelloy-X samples show typical effects of introducing a high volumetric content of porosity. With increasing porosity content in the Hastelloy-X alloy, the obtained curves became more and more “flattened”, i.e. a noticeable decrease of compressive strain was observed over the wide strain range. After reaching yielding point at a low level of ~ 100 MPa, the porous samples underwent a plastic deformation in two stages: (i) under a near linear-like course associated with a densification of sinters through deforming and breaking individual walls; and (ii) in non-linear regime characterized by a more intensive strain hardening. This destruction mechanism has been reported for example in the case of highly porous Fe-Al intermetallics [16]. It is found that a general strain hardening coefficient (expressed as a local slope of the compressive curve) decreases with the increment in the volumetric content of pores. This effect seems to be reasonable, as more pores in the sample volume, means a smaller load bearing cross-section area. Analogous mechanical behavior (and a similar shape of compressive curves and stress/strain values) has been previously reported by Unver et al. for porous 625 Ni superalloy [6].