This study develops an advanced multiphysics modeling framework to investigate the Powder Bed Fusion (PBF) additive manufacturing process for Inconel 625, with a focus on melt pool dynamics, dendritic solidification, and crack formation. By integrating the Allen-Cahn phase-field approach with an elastoplastic J2 plasticity model within the Multiphysics Object-Oriented Simulation Environment (MOOSE), the simulations employ adaptive mesh refinement (AMR) to precisely resolve the evolving solidification front and laser-powder interactions [1]. The framework establishes a critical link between macroscale PBF process parameters and microscale solidification behavior, enabling the prediction of grain boundary segregation and liquation cracking. A novel multiscale methodology facilitates automated data transfer from thermal process simulations to phase-field models, incorporating temperature gradients and cooling rates to simulate dendrite growth in a representative Ni–Nb–Al ternary alloy system. Experimental validation through single-track laser melting experiments demonstrates good agreement between predicted and observed melt pool geometries under varying process conditions. Furthermore, phase-field simulations reveal microstructural features such as solute segregation and cellular dendrite formation at melt pool boundaries, supported by the first melt pool-scale phase-field model. This work provides new insights into defect formation mechanisms, microstructure evolution, and process-structure-property relationships in metal additive manufacturing, offering a robust computational tool for process optimization and quality control [2].