Numerical analysis of rotating scans’ effect on surface roughness in laser-powder bed fusion

Abstract

The rotating scanning strategy is a critical factor for top surface roughness in laser-powder bed fusion (L-PBF), yet the underlying mechanism remains elusive. This study introduces an integrated L-PBF model, encompassing powder packing and molten pool fluid dynamics analysis, to investigate how surface roughness accumulates across layers with non-rotating scans but is mitigated with rotating scans. To address computational complexity and prediction errors associated with multi-layer simulations, the model utilizes an experimentally built part as substrate and simulates single-layer depositions with varying scanning rotation angles. Additionally, an as-machined surface representing ideal flat substrate is included for comparative analysis. Results indicate that on as-built surfaces featuring alternating peak-valley patterns, both powder packing structure and fusion zone shape are location-dependent. To effectively characterize the packing structure on such non-flat surfaces, an effective packing density is proposed. Specifically, peak locations exhibit a thinner powder layer, a 14.5% increase in effective packing density, and a 7% decrease in penetration depth compared to valley locations. The circulation flow, driven by recoil pressure and Marangoni force, is responsible for the reduced surface roughness with rotating scans. Specifically, the circulation flow intersects with multiple peaks and valleys, effectively promoting material redistribution among these tracks and disrupting the peak-valley patterns. Conversely, the peak-valley patterns are preserved and may propagate to subsequent layers with non-rotating scans.