Abstract: The precipitation of titanium oxides and nitrides during steel solidification plays a critical role in microstructure control, yet their accurate prediction remains a long-standing challenge. While classical models such as the Goto and Goto + precipitation consumption (GPC) model approaches have provided valuable insights into solute microsegregation and inclusion formation, they are typically applied to single-type predictions and therefore leave room for further improvement in describing the concurrent evolution of multiple Ti-bearing phases. To advance this understanding, this study developed an Optimized Coupled Prediction (OCP) model that integrated solute microsegregation, thermodynamics-driven competitive precipitation, and diffusion-controlled growth. By incorporating a Gibbs free energy-based competition module, the model quantitatively predicts the coexistence and transformation of Ti₃O₅ oxides and Ti₂O₃-TiN complex inclusions under varying metallurgical conditions. Validation against scanning/transmission electron microscopy and literature data demonstrated that the OCP model successfully reproduced observed inclusion morphologies and achieved higher accuracy in size prediction, with the lowest root mean square error (0.478 µm) and the highest statistical consistency (Wilcoxon p=0.569) among compared models. This work established a more comprehensive framework for understanding multiphase inclusion evolution, offering both theoretical insight and practical guidance for inclusion engineering in advanced steelmaking.

Key words: Inclusion engineering, Titanium oxide, Prediction model, Microsegregation, Precipitation