Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Different processing methods are utilized in the production of titanium (Ti) parts. In recent years, additive manufacturing has gained a lot attention due to its inherent flexibility and freedom in design. Therefore, during the production of Ti parts the material is subjected to a broad range of cooling rates. During the cooling down phase, titanium will experience a solid-state phase transformation going from the β to α phase. The cooling rate will have an important impact on the nucleation and growth during this solid-state phase transformation and will greatly influence the final microstructure, and therefore the properties of the part. However, the relation between the cooling rate and the formation of the microstructure in pure Ti has not been clearly understood yet. In scope of this problem, a phase-field simulation model has been developed by Feyen and Verbeeck, which is able to simulate homogeneous nucleation and growth of the α phase in the parent β phase under constant cooling rate. This thesis aimed to validate and improve the existing model through simulations and experiments. The model was extended by redefining the temperature profile through heat equations, to include cooling based on convective heat transfer with the environment and latent heat production upon phase transformation. As a result, the rate of phase transformation was found to be controlled by the heat flux to the environment. Furthermore to improve the material-specific input parameters for the phase-field model, the interfacial energy and mobility were investigated through experiments. Differential thermal analysis (DTA) of extra pure Ti revealed the undercooling required for phase transformation to be very low. Therefore, the interfacial energy upon nucleation was determined to be lower than the one used in the model and close to zero, indicating a high degree of coherency of the α nucleus upon nucleation. Confocal scanning laser microscopy (CSLM) was used to observe in-situ the phase transformation and measure the interfacial velocity of the growing α phase. In this way, an estimation for the interfacial mobility was made which was found to be an order of magnitude larger than the one used for the model. Finally, Electron backscatter diffraction (EBSD) was performed on samples which were subjected to different heat treatments i.e., furnace cooling, air cooling and water quenching, to validate the simulated microstructures for different cooling rates. The improved model simulates larger grains, which is also closer to the grain size in the EBSD cross-sections, differing by only one order of magnitude, compared to provided model which differs by two order of magnitudes, for the air cooled and water quenched samples. A qualitative match of the grain size was obtained for the microstructure of furnace cooled sample.

Choose an application

• Reference Manager
• EndNote
• RefWorks (Direct export to RefWorks)

Laser-Powder Bed Fusion (L-PBF) manufactured 316L stainless steel has a peculiar subgrain microstructure with cellular structures comprised of a network of dislocations and a nanophase dispersion of oxide inclusions in the as-built material. This microstructure provides enhanced mechanical properties and corrosion resistance. However, it also adds a significant amount of residual stresses and microstructural inhomogeneity, resulting in variability in material properties. Many studies have been performed to characterize the cellular structure, but none are focused on studying the evolution of the nanophase oxide dispersion and its effect on grain growth. \ A typical feature of the microstructure in L-PBF 316L stainless steel is a bimodal particle size distribution of inclusions, which has not been modelled yet for any material system. Secondly, coarsening of second-phase particles in grain growth simulations has not been considered yet. In this thesis, a computational framework supported by experimental data to predict the evolution of the average grain size is presented, considering the bimodal dispersion and coarsening of inclusions. The starting point for the simulations is assumed as a fully recrystallized equiaxed microstructure. \ The framework is thoroughly evaluated for 316L stainless steel. The recrystallization time for L-PBF 316L stainless at 1200 $^circ$C is determined as 1h. All the inclusions are identified using Scanning Electron microscopy and identified using EDS (energy dispersive X-ray spectroscopy). Rhodonite was found as a meta-stable phase present after recrystallization until 2 hours of heat treatment at 1200 $^circ$C. Further increasing the heat treatment temperature, Rhodonite was transforming to $SiO_2$. The particle size distribution and grain size are calculated and used as an initial condition for the grain growth simulations using the phase field method. No significant coarsening of particles and grain growth was observed. The understanding of the experimental observations is built with thermodynamic simulations, and grain size evolution is computed with phase field simulations. \ The existing framework is currently constrained by the limited resolution of nanophase oxides (0.1 $mu m$) and grains ((10 $mu m$)) within the same mesh, which subsequently limits the statistics required for grain growth simulations. Despite this limitation, the framework enables a comprehensive analysis of the material's grain growth phenomenon.