Intelligent control of additively manufactured beads on the basis of target morphologies and thermal field  

作　　者：Xuhui YANG Rui LI Kelong HU Aidong SUN Xiaoyao MA Mingxin LIU Mingtian WANG Runsheng LI Gang ZHAO Wenzheng ZHAI Hao SONG Zili LI Haiou ZHANG 

机构地区：[1]School of Mechanical Science and Engineering,Huazhong University of Science and Technology,Wuhan 430074,China [2]School of Software Engineering,Huazhong University of Science and Technology,Wuhan 430074,China [3]College of Mechanical and Electronic Engineering,China University of Petroleum,Qingdao 266580,China [4]Hubei Key Laboratory of Mechanical Transmission and Manufacturing Engineering,Wuhan University of Science and Technology,Wuhan 430065,China

出　　处：《Frontiers of Mechanical Engineering》2024年第6期87-98,共12页机械工程前沿(英文版)

基　　金：supported by the National Natural Science Foundation of China(Grant No.52175169);the Open Fund of Hubei Key Laboratory of Mechanical Transmission and Manufacturing Engineering at Wuhan University of Science and Technology,China(Grant No.MTMEOF2023A01);the Hubei Province Science and Technology Innovation Talent Plan,China(Grant No.2023DJC053);the Hubei Technological Innovation Special Fund,China(Grant No.2023BAA019-4).

摘　　要：Hybrid deposition with microrolling is a promising arc-based direct energy deposition technique to rapidly build complex parts,whose performance is comparable to that of their wrought counterparts.Complex forming conditions and bead morphologies pose difficulties in controlling the morphologies from a single weld bead to built part profiles,and these difficulties hinder the widespread application of the technique.Here,a model that can automatically generate optimal process parameters on the basis of the infrared image,including thermal information and the point cloud information of the target weld bead,is developed.Results show that the errors in critical parameters,namely,feed speed,travel speed,and rolling force,are below 0.4%,0.9%,and 2%,respectively,indicating that the proposed technique outperforms the compared methods.Furthermore,validation reveals that the actual depositing bead is similar(deviation below 0.05 mm)to the target bead.The proposed strategy provides an effective foundation for dynamic path planning and can considerably improve printing efficiency and precision.

关 键 词：wire arc additive manufacturing deep learning thermal field point cloud stainless steel 

分 类 号：TG148[一般工业技术—材料科学与工程]