A Deep Learning Strategy to Control Bead Morphologies of Arc-Laser Hybrid Direct Deposition with In-situ Micro-Rolling
摘要
Wire direct energy deposition shows excellent prospects for industrial application, as it improves building efficiency with high quality. The hybrid deposition with micro-rolling (HDMR) technique, which builds parts under the micro-rolling, provides the built part with high performance comparable to the forging counterpart. Wire arc-laser hybrid direct energy deposition with in-situ micro-rolling (AL-DED-MR) technique combines two different heats with the subsequent rolling force, resulting in difficulties in controlling bead morphologies due to a complex thermal-mechanical coupling effect. This paper proposed a strategy to build a model between process parameters of AL-DED-MR and specific bead morphology, which provides a new approach to improve printing accuracy and efficiency for AL-DED-MR.