Process window, microstructure evolution and hardness response of single-track Fe–Mn–Si–Cr–Ni shape memory alloy coatings fabricated on CrNi3MoVA gun steel by laser cladding

Fe-Mn-Si-Cr-Ni shape memory alloy coatings were fabricated on CrNi3MoVA gun steel by single-track laser cladding to establish the process-morphology-microstructure-hardness relationship. An L25(5 3 ) orthogonal design was employed to evaluate the effects of laser power ( P ), scanning speed ( V ) and powder feed rate ( f ) on track geometry, dilution behavior, defect formation and phase constitution. A reduced regression model was developed for the dilution rate, revealing that dilution was governed by the competition between substrate penetration and powder deposition. Deposition-dominated tracks exhibited low dilution below 20%, whereas penetration-dominated tracks promoted substrate remelting and increased dilution above 40%. Defect evolution was strongly dependent on dilution and line energy. Lack-of-fusion pores and unmelted particles were prevalent in coatings produced at low line energy, while bottom-localized defects were gradually suppressed with increasing line energy. Cracking occurred only under the combined condition of low line energy, high V and severe lack-of-fusion defects. Hardness did not show a monotonic dependence on any single processing parameter. Instead, a P × V - f interaction map divided the samples into four processing regions, among which low f combined with intermediate P × V produced the highest hardness of 474 HV. The coating hardness was primarily controlled by nature and spatial distribution of the FCC-typed austenite and BCC phase rather than its fraction alone. In the low-dilution condition, coarse low-strain ferritic BCC phase and austenitic FCC phase segregation resulted in moderate hardness. The highest hardness originated from a fine α′-martensite-dominated dual-phase structure with dispersed retained austenite, whereas over-dilution led to a coarse austenite-dominated structure and significant softening.