Dataset for: Limits of homogeneous-continuum modelling in directed energy deposition of immiscible Fe–Cu alloys: dual-axis validation via neutron diffraction and digital image correlation

Directed energy deposition of immiscible Fe–Cu functionally graded materials is limited by severe residual stresses and uncertain strength, yet conventional homogeneous continuum models are seldom verified against multi-axis experimental data. A layer-resolved finite-element framework of an immiscible 50:50 AISI 316L/UNS C61800 co-deposition is validated across two axes against reference datasets: neutron diffraction for directional residual stresses and digital image correlation for tensile response. The two axes converge on an identical physical threshold of the homogeneous J2 continuum. The framework reproduces qualitative signs, anisotropy, and localization topologies. A parametric convergence audit reveals that deceptive coarse-mesh agreement arises from mutually compensating discretization errors. Quantitatively, the converged framework provides a yield-limited upper bound, overestimating the longitudinal residual stress plateau by a systematic factor of 1.95 (–682 versus –350 MPa). Symmetrically, integrating a strain-averaged fracture criterion predicts the structural limit load (Fpred = 8.20 kN) within +1.8% error. The apparent localized modulus and skewed displacement partition are diagnosed as measurement system artefacts induced by porous and frame compliance. Bounding the continuum framework traces the discrepancy to specific two-phase microstructural and porous features rather than empirical parameters, establishing a baseline for future phase-resolved and ultrasound-assisted modeling.