Scanner-based laser welding of additively manufactured titanium lattice structures—melt pool characteristics and joint performance control

Additively manufactured lattice structures exhibit significant potential in aerospace and insulation structures, yet their wider industrial implementation is limited by intricate geometries, prolonged production cycles, and high-demanding equipment performance. Laser additive-welding integrated manufacturing offers a practical route to assemble small AM lattice modules into large, flight-relevant structures, but reliably joining open-cell titanium lattice nodes remains challenging because slender struts and multi-surface exposure impose atypical heat-extraction pathways that destabilize the melt pool. Fundamental, mechanism-based guidance linking laser parameters to melt-pool dynamics, microstructure, and joint performance is still scarce. We address this gap with a geometry-aware, scanner-based circular-path welding approach for LPBF-fabricated Ti600 lattices, in which the scan path itself—rather than robot motion—sets local energy density and dwell, enabling precise control of both the magnitude and spatial distribution of heat input. By systematically varying laser power (500–700 W) and scan radius (80–160 μm) at fixed speed/defocus, we build a processing–structure–property map that correlates melt-pool morphology with porosity (CT), α′-martensitic evolution and misorientation (EBSD), hardness, and node-level tensile response. We show that lattice-specific heat extraction—conduction along slender struts combined with convection/radiation to air—controls melt-pool residence and solidification; under an intermediate heat-input window exemplified by a 120 μm path radius with 600–700 W, the pool stabilizes and continuous fusion is achieved, with CT-measured porosity of approximately 0.3–0.9 %, EBSD-resolved α′ packets characterized by grain sizes of ∼2.27–3.31 μm, and the highest node-level strength (maximum tensile force 86.03 N at 700 W, 120 μm). Insufficient input causes lack-of-fusion brittleness, whereas excessive dwell prolongs the liquid lifetime, promotes pore coalescence, and degrades strength; critically, path planning at the trajectory level allows us to tune azimuthal energy sweep and local dwell so as to balance penetration at the node crown with heat conduction into strut roots, suppressing lack-of-fusion at low input and hot-spot-induced porosity at high input. Using the scan path as a tunable parameter yields a mechanism-based process window and predictive maps in terms of energy density and dwell, enabling repeatable, reliable laser additive-welding of porous metallic architectures for large-scale aerospace lattice assembly.