Abstract Scope |
Austenitic stainless steels have a combination of good weldability, mechanical, and corrosion properties, making them suitable candidates for additive manufacturing (AM) and use in Army ground vehicle systems. During AM processes, the 316L feedstock material undergoes melting followed by fast solidification and partial remelting of previous build layers, resulting in unique final microstructures and textures throughout AM builds. In this study, the solidification behavior and crystallographic texture of AM 316L produced using a laser-wire directed energy deposition (DED-LB) process was investigated and delta-ferrite morphology was utilized to determine the solidification pathway of AM 316L.
Variations in shielding gas set-up and build type, single versus multi-track walls, resulted in differences in primary solidification mode and final microstructure morphology among the DED-LB builds. Three delta-ferrite morphologies were formed in the as-built microstructure as a result of variations in the solidification sequence for which distinct orientation relationships and crystallographic texture between delta-ferrite and austenite developed. Skeletal and interdendritic delta-ferrite exhibited a (001)delta // (001)gamma relationship with austenite and a {001} solidification texture aligned with the build direction was observed for both austenite and delta-ferrite. Lathy delta-ferrite exhibited a Kurdjimov-Sachs orientation relationship due to austenite nucleation in the solid state.
The build type and shielding gas set-up directly impacted the solidification conditions and element vaporization of the build process, resulting in the range of final AM 316L microstructures. Primary ferrite solidification is predicted based on feedstock material but variations in solidification mode from primary ferrite to primary austenite occurred due to increased solidification velocity at the top of the melt pool stabilizing the austenite dendrite tip over the delta-ferrite dendrite tip as well as due to compositional variations between builds. The initial feedstock composition cannot accurately represent all the solidification mode variations of AM 316L indicating the need for the as-built wall composition.
A dendrite growth model was paired with Rosenthal’s analytical heat transfer model for the development of solidification maps, which were used to predict solidification microstructure morphology, size, and shift in solidification mode. The solidification maps were assessed against experimental dendrite arm spacing data for which there was good agreement between the experimental data, the dendrite growth model, and the solidification path generated from Rosenthal’s analytical heat transfer model. |