Phase Field Modeling of Additively Manufactured Ti-Al-V-Fe alloy

Abstract

Over the past three decades, Additive Manufacturing (AM) has emerged as a widely utilized method for fabricating titanium components, offering notable advantages such as reduced materials waste and enhanced flexibility in geometry design. However, the widespread adoption of AM is hindered by the challenge of ensuring consistent product quality, necessitating a comprehensive understanding of the process-structure-property relationship. Modeling approaches serve as invaluable tools in bridging this gap, providing insights into the AM process while significantly reducing time and experimental costs. Among these modeling techniques, Phase Field (PF) modeling has garnered significant attention as an effective approach for simulating microstructure evolution during AM processing. Within this thesis, I present a comprehensive study utilizing a multi-component phase-field method to investigate the AM process, with a specific focus on Ti-1Al-8V-5Fe (wt\%), a cost-effective Beta-Ti alloy. This thesis encompasses the development, verification, applications and validations of the multi-component PF model, advancing our understanding of AM and its application to Beta-Ti alloys.

First, a multi-component phase equilibrium prediction method was developed for use in the multi-component phase field models. The prediction of phase equilibrium is a main time-consuming section of a multi-component PF model. To improve the computational efficiency, a new convex-based method for temperature dependent phase equilibrium prediction was proposed in this study. To show the accuracy of the convex hull method, 1-D phase field simulations utilizing the Convex-hull method were performed under isothermal and continuous cooling conditions. The 1-D simulation results were compared against Thermo-Calc calculations, which shows that a high accuracy of equilibrium prediction is achieved at a single and multiple temperatures.

Second, the implementation of the multi-component PF model was further verified via performing a benchmark analysis on different 2-dimensional models of solidification in multi-component alloys. Specifically, the multi-component PF model and two pseudo-binary PF models were applied on the isothermal and directional solidification of the Ti-185 alloys. The results showed that a very good similarity in microstructure was achieved between the three phase field models during both isothermal and directional solidification. The results demonstrate the usefulness of different PF modelling approaches and highlight cases where a full multi-component model is needed.

Third, the multi-component PF model was applied on the Laser Powder Bed Fusion (LPBF) process. In this study, two large-scale PF simulations were performed to simulate the microstructure evolution during LPBF process using pre-alloyed powder and blended elemental powder. This work aims to provide a clear picture of the in situ alloying process and improve our fundamental understanding of the competitive growth phenomenon. The results show that evenly distributed finer columnar grains formed while using pre-alloyed powder. For the case of using elemental powders, the results indicate that full alloying is difficult to achieve during the LPBF printing process; this incomplete alloying greatly influences the dendrite morphology and solute distribution.

Finally, the multi-component PF model was applied to the Wire Arc Additive Manufacturing (WAAM) process and to construct the solidification process map of the Ti-185 alloy. The solidification process map was compared with experimental results for the model validation. The process map of the Ti-185 alloy was constructed via performing a series of simulations with constant temperature gradient and solidification rate. The process map shows that a CET transition occurs at the top of the sample, this trend shows an excellent agreement with experimentally characterization.