Factor screening for adaptive powder feeding in Laser-Beam Directed Energy Deposition

Abstract Laser beam powder directed energy deposition (DED-LB/Powder) is a promising additive manufacturing (AM) technology for large parts, repairs, and multi-material builds, yet many process parameters govern part quality. This study investigates how these parameters affect powder delivery and deposition height for AISI 316L. A moving control-volume model is coupled with a factor-screening design to relate powder feed settings to travel speed and to assess four responses: height, density, porosity, and microhardness. Within the tested window, final height is primarily governed by the Z -increment (Δ Z ), with additional contributions from energy density ( E ), overlap efficiency ( O e ), and κ . The approach is evaluated through a design of experiments that demonstrates sensitivity to selected factors and their impact on specimen responses. Boundary conditions in the building chamber are monitored with a custom system. A full-factorial plan with replication, randomization, and blocking is used, and analysis of covariance (ANCOVA) identifies which factors and noise sources affect each response. The results map factor effects and interactions, provide simple response surfaces for height control without real-time feedback, and indicate that density and hardness are relatively insensitive within the tested ranges, whereas porosity requires dedicated analysis. The study also highlights non-significant factors that can be deprioritized in subsequent optimization and robustness studies.