| Abstract Scope |
Introduction:
Submicron inclusions have been reported in the microstructures of duplex stainless steel alloys fabricated through additive processes. These inclusions have positive and negative impacts on microstructures and properties. For example, these inclusions serve as heterogenous nucleation sites for the formation of fine intragranular austenite resulting in tensile properties equivalent to wrought materials. However, these submicron inclusions also significantly degrade low temperature Charpy impact properties, which is an important property used for duplex stainless steel part qualification in the oil and gas industries. While these submicron inclusions have significant impact on microstructures and properties, a detailed understanding of their formation is not well established. An in-depth knowledge of these inclusions is required to control the formation in the microstructures of duplex stainless steel alloys fabricated through additive manufacturing processes.
Experimental:
A series of lean (UNS S32101), standard (UNS S32205), and super (UNS S32507) DSS powder feedstocks were used to fabricate builds using a DED-L AM process. Samples for microstructural evaluation were extracted from the transverse cross section at the centerline of the builds and were mounted and ground using a SiC grinding paper up to P4000 ISO grit size then polished using 3µm diamond followed by a 1 µm diamond slurry. Samples used for optical microscopy were etched using a KOH electrolytic etch (25 g KOH, 50 mL distilled water, 5 VDC for 5 sec) in order to reveal the austenite/ferrite microstructure and to distinguish between the different austenite morphologies in images captured using an Epiphot inverted microscope. To obtain high resolution microstructural images and to analyze the samples for elemental segregation, samples were prepared using the same procedures with the addition a final vibratory polish of with 0.05 µm colloidal silica for 12 hours and analyzed using a Helios NanoLab 660 scanning electron microscope (SEM) with X-MaxN (EDS), and NordlysMax2(EBSD) detectors were used.
Results and discussion:
Nearly 1000 inclusions in the as-deposited and post-process (hot isostatically pressed) (HIP) conditions were measured using a scanning electron microscope (SEM). While the median inclusion size remains about the same, the frequency decreases when comparing the as-deposited to the post-process conditions. This decrease in inclusion frequency suggests the post-process HIP treatment is reducing the number of inclusions. This change is further supported by SEM-EDS data, which showed the chemistry of the inclusion changed from rich in silicon to depleted in silicon after post-process HIP. Transmission electron microscopy (TEM) was used to study these inclusions at higher resolutions. In the as-deposited condition two types of inclusions can be observed, one rich in Cr, Mn, and O, and the other rich in Mn, Si, and O. In the post-processed HIP condition, again two inclusions are observed, one rich in Cr, Mn, and O, while the other rich in Mn and S.
Computational thermodynamics was used to identify the stable phases predicted at equilibrium. Results showed that a spinel phase (MnCr2O4) was predicted. Because the additive manufacturing process deviated from equilibrium conditions, the metastable phases were also evaluated. This was accomplished by suppressing the stable spinel phase and rerunning the analysis. The results revealed that a rhodonite phase (Mn5Si5O15) was a metastable phase.
Using the computational thermodynamics, the chemistry of the phases can be predicted. This predicted chemistry was then compared to the measured chemistry collected using STEM-EDS. From this analysis it is likely the phases in the as-deposited and post-processed HIP conditions are a combination of spinel MnCr2O4, rhodonite Mn5Si5O15, and MnS. Selected area diffraction was used to verify these phases, spinel MnCr2O4 has a face-center cubic crystal structure and the lattice parameter was measured to be 8.47A. This is nearly an exact match to the value of 8.44A which has been reported in the literature.
It appears there is a combination of both stable and metastable phases within the duplex stainless steel microstructures. In the as-deposited condition, stable spinel and metastable rhodonite are observed, while in the post-processed HIP condition, both stable spinel and MnS are observed. Both the spinel and rhodonite inclusions show high-temperature stability and coexistence with the liquid phase. The rapid solidification and cooling rates associated with the DED-L process (103 K/sec), it is not unusual to find metastable phases in the microstructure. Additionally, during the DED-L process the build is at higher temperatures for longer periods due to the reheating from the deposition of subsequent layers and passes which likely permits the formation of the stable MnCr2O4 phase. After fabrication, the builds were subjected to a post-process HIP treatment at a temperature of 1170°C and pressure of 145 MPa for 3 hours. This HIP treatment permitted the microstructure to reach a state more similar to thermodynamic equilibrium, which contributes to the presence of stable MnCr2O4 and MnS phases observed in the microstructure. The metastable Mn5Si5O15 phase, observed in the as-deposited condition, appears to be dissolved during post-processed HIP.
Conclusions:
A mixture of spinel MnCr2O4, rhodonite Mn5Si5O15, and MnS inclusions have been observed in the lean (UNS S32101), standard (UNS S32205), and super (UNS S32507) duplex stainless steel microstructures fabricated using laser-based directed energy deposition additive manufacturing techniques. These inclusions have been observed to have both positive and negative impacts on mechanical properties. The primary conclusions from this study are as follows:
• In the as-deposited condition, stable spinel MnCr2O4 and a metastable rhodonite Mn5Si5O15 inclusions are observed. While after the post-process HIP treatment, the system reaches a state closer to thermodynamic equilibrium and the metastable rhodonite inclusions appear to dissolve. Stable spinel MnCr2O4 and MnS are observed in the post-process HIP condition.
• The inclusions appear to have no negative impacts on uniaxial tensile properties. After post-process HIP, when alloy chemistry is within specification, wrought tensile properties are achievable.
• The presence of inclusions likely degrade the Charpy impact toughness of duplex stainless steels. Impact toughness values were significantly lower than wrought products. |