Beyond expert intuition: evaluating a vector‑adaptive scan strategy to enhance geometric fidelity in hard‑to‑print Ti‑6Al‑4V thin‑walled structures

The fabrication of metallic thin-walled structures (TWS) via laser-based powder bed fusion of metals (PBF-LB/M) presents significant challenges in maintaining geometric fidelity and minimizing defects, both of which are strongly influenced by the laser scan strategy. This study introduces and empirically benchmarks a novel vector-adaptive scan strategy (Free Float) against conventional methods for producing dense TWS. The adaptive strategy dynamically modulates hatch vector number, length, and orientation in response to local geometry—specifically wall thickness and inclination angle—to control scanned area per layer and actively mitigate deleterious thermal gradients. A full-factorial design comprising 300 titanium alloy (Ti-6Al-4V) specimens, spanning ten wall thicknesses (0.1–1.0 mm), ten inclination angles (0°–90°), and three scan strategies were executed. Results demonstrate that the vector-adaptive strategy enhanced dimensional accuracy, particularly for critical shallow overhangs (10°–30°). It reduced average dimensional deviation by 29.7–30.6% compared to a standard strategy and by 3.5–3.7% relative to surface-optimized strategy with down-skin parameter adaptation. Specimens with wall thickness ≥ 0.2 mm achieved high structural integrity, with relative densities exceeding 98.65%. This work establishes a quantitative benchmark for adaptive scanning in PBF-LB/M, confirming its potential for fabricating high-integrity lightweight components.