Mixing-Driven Defects and Composition Evolution in Multi-Material Metal Additive Manufacturing

Multi-material laser powder bed fusion (LPBF) enables the fabrication of components with spatially tailored compositions, but reliable process design remains largely limited by an incomplete understanding of melt pool interactions, mixing-induced instabilities, and defect formation. In this work, we combine operando synchrotron X-ray radiography with high-fidelity multi-material multiphysics modeling to resolve, for the first time, the real-time coupling between inter-material mixing, keyhole instability, and pore formation. We show that material mixing fundamentally alters the stability landscape of the keyhole, giving rise to composition-driven shape transitions and mixing-induced collapse events that generate large pores. Meanwhile, microscale gas entrapment during forced convection of dissimilar materials produces a distinct class of fine pores. Quantitative mapping of transient composition fields reveals how local alloying gradients govern defect trajectories and melt pool transport. This work establishes a validated mechanistic framework for multi-material LPBF and provides a quantitative basis for process optimization toward defect-controlled, compositionally graded metal additive manufacturing.