Layer-wise adaptive control of thin wall geometries in laser-powder direct energy deposition

Direct Energy Deposition (DED) is an Additive Manufacturing technology that enables the manufacturing of metallic parts by melting the material as it is deposited. By stacking single passes over subsequent layers, DED can produce thin-wall structures with high aspect ratios, with potential applications in aerospace and prosthetics. However, the reliable fabrication of thin walls is challenged by heat accumulation and speed reductions at sharp corners, leading to local variations in the deposition rate. Static part programs rely on passive self-stabilization, requiring extensive optimization experiments to find the optimal deposition strategy that deposits the target part height, avoiding local lacks of material or protrusions. While active feedback loops can adjust DED process parameters, their effectiveness in thin-wall structures with sharp corners is limited by temporal real-time constraints, since deviations at a given location can only be corrected after the laser returns to that position in a subsequent pass. This study presents a robust layer-wise adaptive control method for manufacturing thin walls using DED that monitors and optimizes the local stand-off distance. Our approach introduces three key innovations: high-speed online stand-off measurement, a robust layer-wise proportional control system, and an adaptive part program generator. Extensive experiments on triangular thin walls demonstrate that our method significantly reduces height deviations due to protrusions on the corners compared to passive self-stabilization while maintaining a high average layer thickness. Moreover, our control approach proved resilient to variations in DED process parameters, nozzle occlusion, and applies to several thin-wall geometries described by closed contours, enabling the fabrication of complex structures for aerospace and prosthetic applications.