Residual stress and multiscale microstructural evolution in steels processed by laser powder bed fusion

Laser powder bed fusion (LPBF) additive manufacturing can produce metallic components with complex geometries, but residual stress induced by non-equilibrium thermal histories still limits dimensional stability, crack resistance, fatigue performance, and service reliability. Thermal gradients and constrained cooling are central to residual stress formation, but they cannot fully explain how residual stress develops in steels that undergo solid-state phase transformations. This review compares residual stress and multiscale structures in non-transforming and phase-transforming steels, with these terms used operationally for LPBF residual stress analysis. In the non-transforming regime, no major build-cycle solid-state transformation strain is required to interpret the as-built residual stress state, whereas in phase-transforming steels thermal contraction interacts with transformation strain. The comparison is organized around LPBF thermal history, stress generation mechanisms, phase transformation, and measurement conditions used to report residual stress. In this review, non-transforming steels, represented by predominantly austenitic LPBF 316L under the conditions considered here, are used as the thermal-stress-dominated reference case, in which melt pool gradients, scan history, heat accumulation, build plate constraints, and post-processing treatments control stress evolution without requiring a major contribution from martensitic transformation strain. Phase-transforming steels, such as H13 hot-work tool steel, add martensitic or bainitic transformation, retained austenite, intrinsic tempering, carbide evolution, and preheating-dependent phase evolution to this pathway. These metallurgical processes can reduce, redistribute, or locally change the sign of the measured stress. Reported residual stress values should therefore be interpreted together with measurement depth, gauge volume, stress component, build state, phase selection, and stress scale. Further progress depends on transformation-aware process design, phase-resolved three-dimensional characterization, and models that link LPBF process history to the measured residual stress state.