The optimization of process parameters in powder Directed Energy Deposition (DED) is essential for achieving consistent, high-quality bead geometries, which directly influence the performance and structural integrity of fabricated components. As a subset of additive manufacturing (AM), the DED process, also referred to as laser metal deposition (LMD), enables precise, layer-by-layer material deposition, making it highly suitable for complex geometries and part repair applications. Critical parameters, such as the laser power, feed rate, powder mass flow, and substrate temperature govern the deposition process, impacting the bead height, width, contact angle, and dilution. Inconsistent control over these variables can lead to defects, such as poor bonding, dimensional inaccuracies, and material weaknesses, ultimately compromising the final product. This paper investigates the effects of various process parameters, specifically the substrate temperature, on bead track geometry in DED processes for stainless steel (1.4404). A specialized experimental setup, integrated within a DED machine, facilitates the controlled thermal conditioning of sample sheets. Using Design of Experiments (DoE) methods, individual bead marks are generated and analyzed to assess geometric characteristics. Regression models, including both linear and quadratic approaches, are constructed to predict machine parameters for achieving the desired bead geometry at different substrate temperatures. Validation experiments confirm the accuracy and reliability of the models, particularly in predicting the bead height, bead width, and contact angle across a broad range of substrate temperatures. However, the models demonstrated limitations in accurately predicting dilution, indicating the need for further refinement. Despite some deviations in measured values, successful fabrication is achieved, demonstrating robust bonding between the bead and substrate. The developed models offer insights into optimizing DED process parameters to achieve desired bead characteristics, advancing the precision and reliability of additive manufacturing technology. Future work will focus on refining the regression models to improve predictions, particularly for dilution, and further investigate non-linear interactions between process variables.
Read full abstract