Solution Treatment—T6

Several researchers [34,38,65] have focused on the T6 heat treatment, consisting of a solution treatment (450–550 ◦C) [38,58] followed by quenching and artificial aging (150–180 ◦C) [65]. Both Li et al. [38] and Gu et al. [58] solution treated SLM produced AlSi10Mg material at 450, 500, and 550 ◦C for 2 h, followed by water quenching. In both cases, results showed that the microstructure became coarser with increasing solution temperatures (see Figure 9). After solution treatment at 450 ◦C for 2 h (see Figure 9d), Si precipitates were small (<1 μm) and uniformly distributed in the Al matrix. When the temperature was further increased to 500 and 550 ◦C (see Figure 9e,f), particle coalescence and Oswald ripening occurred, resulting in a further increase in particle size (2–4 μm) [38]. The volume fraction of Si increased, proving that Si precipitates during heat treatment [58]. In the work of Li et al. [38], half of the specimens additionally underwent artificial aging at 180 ◦C for 12 h after the solution treatment. Further coarsening occurred with the addition of the artificial aging procedure (see Figure 9g–i), despite the temperature being seemingly too low to induce Si coarsening [38,55]. This is in contrast to the work of Aboulkhair et al. [65], where a solution treatment at 520 ◦C was applied for 4 h. The addition of 6 h of artificial aging at 160 ◦C did not seem to alter particle size or density.

**Figure 9.** Secondary electron micrographs showing the microstructure of as-built and heat-treated AM AlSi10Mg. (**<sup>a</sup>**–**<sup>c</sup>**) Represent the microstructure of the as-built samples in different zones and at different magnifications. (**d**–**i**) Represent the microstructure of the samples after different heat treatments: (**d**) 450 ◦C for 2 h; (**e**) 500 ◦C for 2 h; (**f**) 550 ◦C for 2 h; (**g**) 450 ◦C for 2 h + 180 ◦C for 12 h; (**h**) 500 ◦C for 2 h + 180 ◦C for 12 h; (**i**) 550 ◦C for 2 h + 180 ◦C for 12 h. (Adapted from reference [38]).

#### Artificial Aging Heat Treatments

In most studies, artificial aging (AA) heat treatments are performed after a solution treatment (e.g., T6 heat treatment). Kempen et al. [23] noted that this is not ideal for SLM parts as it undoes the fine microstructure giving the superior mechanical properties. The effect of AA heat treatment on the microstructure of as produced SLM material was studied by Rubben et al. [54]. No significant microstructure evolution was observed compared to the untreated specimens. An explanation for this can be found in the works of Fiocchi et al. [69] and Rafieazad et al. [55]. The temperature of the heat treatment is significantly below the peak temperature (273 ◦C [55]) of precipitation, coarsening, and spheroidization of Si particles.

#### 3.2.2. Effect on Corrosion

#### Stress Release Heat Treatments

Several studies have focused on the effect of stress-relieving heat treatments on corrosion behaviour. Cabrini et al. [52] studied the effect of stress release for 2 h at 200, 300, 400, and 500 ◦C. They performed intergranular corrosion tests in 30 g/<sup>L</sup> NaCl + 10 mL/L of HCl at room temperature for 24 h. Another study by Rubben et al. [54] focused on the effect of stress release for 2 h at 250 and 300 ◦C. To achieve this, they performed open circuit potential, potentiodynamic polarization, and immersion experiments in 0.1 M NaCl.

On untreated specimens, selective corrosion attacks occur at the melt pool borders [52,54]. Galvanic coupling between α-Al and Si causes the dissolution of Al (as already explained above). Selective corrosion attacks at MPBs were still detected after performing stress release up to 300 ◦C for 2 h [52,54]. This is despite significant microstructure evolution, with the microstructure changing from a continuous silicon network to separate Si precipitates, with larger precipitates formed at the MPBs. Revilla et al. [32] and Rubben et al. [54] performed SKPFM measurements on specimens before [32,54] and after stress release at 300 ◦C for 2 h [54]. For both cases, measurements showed Volta potential di fferences between the primary aluminium and the more noble Si, with larger Volta potential di fferences at the melt pool borders compared to the interior of the melt pools. This indicated that the larger microstructure found at the MPBs is more prone to galvanic corrosion, explaining the selective attack at the MPBs before and after heat treatment.

While preferential corrosion at MPBs was noticed for these types of specimens, Rubben et al. [54] illustrated that the underlying mechanism may change depending on the temperature of stress release. This was linked to a corresponding change in microstructure (as explained in the previous section). When the Si network is still intact (which is the case for untreated specimens, and heat-treated at 200 ◦C), the corrosion is mostly superficial with microcrack formation at the heat-a ffected zone next to the MPBs. This allows corrosion to locally spread deeper in the material. When the Si network has broken up (for instance after stress release at 300 ◦C for 2 h), the spread of corrosion is no longer stopped by the presence of an Si network, leading to a more broadly penetrating attack around MPBs. This behaviour is summarized in Figure 10. Depending on how discontinuous the silicon network is, both mechanisms may be active at the same time [54].

**Figure 10.** Schematic diagram representing the evolution of the corrosion attack for two separate cases: (**top**) When a connected Si network is present, as is the case for as-built AM Al-Si specimens, and (**bottom**) when the Si is broken up in separate precipitates, as is the case for heat-treated AM Al-Si parts at relatively high temperatures (~300 ◦C or higher). The illustrations portray the corrosion process seen from a cross-sectional perspective, with the top side of the images representing the surface exposed to the corrosive medium. The Al phase is represented with green, while black portrays the Si phase. For as-built as well as heat-treated specimens, the corrosion initiates at the melt pool borders (MPB), due to the higher driving force for galvanic corrosion in these regions. For as-built specimens corrosion spreads superficially, accompanied by the formation of micro-cracks along the heat-a ffected zones. On the other hand, a dip and relatively wide penetration of the corrosion attack characterizes heat-treated specimens due to the presence of separate Si precipitates. (Adapted from reference [54]).

After stress release at 400 or 500 ◦C for 2 h, no more selective attacks at MPBs were detected after intergranular corrosion tests [52]. Instead, more general corrosion morphologies were noticed. This was linked to a significant modification of the microstructure, showing an α-Al matrix and rounded

precipitates of Si. This leads to the disappearance of the characteristic melt pool macrostructure. As there is no longer a larger microstructure at the MPBs, no preferential attack occurs in those cases.

#### Heating of Building Platform

Intergranular corrosion tests were performed by Cabrini et al. [52] on specimens fabricated with a building platform temperature of either 35 or 100 ◦C. Specimens were immersed for 24 h in 30 g/<sup>L</sup> NaCl + 10 mL/L of HCl at room temperature. The corrosion morphology showed a superficial attack, with a deeper attack for specimens built at 35 ◦C. This change can be attributed to either the small change in microstructure (see the previous section) or to the reduction of thermal stresses. The latter is supported by the works of Revilla et al. [32,33,54], where it was postulated that the presence of residual stresses combined with the exposure of the silicon network after corrosion might lead to the development of micro-cracks, allowing the spread of corrosion further in-depth.
