Microstructural and Nanostructural Evolution in Splat Quenched Stainless Steels

This paper will show the impact that rapid solidification has on the microstructure and nanostructure of a wide range of stainless steels. There is now strong interest in using fusion-based additive manufacturing (AM) methods for the production of complex components made of austenitic stainless steels such as 304 and 316. In AM techniques, such as laser powder bed fusion (LPBF), the cooling rates and solidification rates can be quite rapid with cooling rates estimated to be in the range of 106-107˚K/s. It is well known from the welding literature that the solidification microstructures of austenitic stainless steels evolve as a function of Cr and Ni content and solidification rate. Lippold has provided a map that connects weld microstructure phase content as a function of Cr/Ni equivalency ratio and solidification rate. [1] When the solidification rates reach those appropriate for LPBF, the current knowledge predicts what might happen, but the data has not yet been generated to test these predictions. To address this knowledge gap, we have generated a series of >20 stainless steel alloys in the Fe-Cr-Ni-Mo system and have used the two-piston splat quenching (SQ) process to produce the rapidly solidified materials. The SQ process is able to generate cooling rates as high as 108K/s using only small (<1 g) amounts of material, thus allowing for the efficient exploration of alloy space for rapid solidification. In this study, we have used 316L stainless steel as the base alloy (nominal composition according to ASTM-A240). The Cr to Ni ratio was systematically varied following the WRC-1992 definition of Cr/Ni equivalency [2], while the Mo, Mn, and Si contents were fixed. Arc melting was used to produce the range of alloys. Multiple spat quenches were produced for each alloy. The resultant samples (approximately 25 mm in diameter and 70 µm thick) were cross sectioned and mounted in thermoset epoxy. Standard metallographic techniques were combined with vibratory polishing to prepare the cross-sectional samples for electron backscatter diffraction and other SEM-based microscopy. FIB milling was employed to prepare TEM specimens oriented such that the foil was across identified solidification cells. EBSD phase maps clearly show that the stainless steel changes from fully austenite to fully ferrite as the Cr/Ni ratio is increased (Figure 1). More interestingly, the range of compositions over which the primary ferrite has massively transformed to austenite is much wider than what was predicted in the welding research of Lippold [1]. It should be noted that this massively transformed austenite phase is only clear upon the combination optical microscopy on etched samples with crystallographic phase information. STEM-HAADF images show a similar cellular structure (cell diameter 200-400 nm) across the explored Cr/Ni ratio range. Interestingly, small Mn-Si rich oxide precipitates were observed decorating all of the cell boundaries (Figure 2). These precipitates have also been reported in the solidification microstructures of austenitic stainless steels produced by LPBF. [3] STEM-EDX shows segregation of Cr, Ni, and Mo to the cell boundaries. The degree of segregation was not sensitive to the Cr/Ni ratio, but it did change with the primary solidification phase. Cr and Mo segregated to cell boundaries for primary austenite solidification, while Ni segregated to cell boundaries for primary ferrite solidification. [4] EBSD phase map. Blue represents the austenite/FCC phase; while red represents the ferrite/BCC phase. The alloy number and Cr/Ni equivalency ratio are indicated above, A, B, and C. A. STEM-HAADF image of splat quench microstructure. The red line is draw across two cell boundaries. B. Ratio of integrated x-ray peak intensities versus iron intensity across the cell boundaries indicated in A. (Note: Kα lines used for each element.)