Study On Infrared Visual Intelligent Diagnosis System For Defects In CMT Additive Manufacturing Process

With the rapid development of arc additive manufacturing technology,CMT(Cold metal transfer)as a new manufacturing technology has attracted more and more attention and research in recent years.However,in view of the inherent defects of "repeated thermal cycle" in the manufacturing process of metal additive,inherent defects such as thermal stress caused by directional heat dissipation and deviation accumulation of morphology and size,the development of its shape-controlling technology is an important means to improve the forming performance and quality of additive manufacturing parts.Therefore,this paper established a CMT arc additive manufacturing process detection system based on infrared vision,by deeply studying the infrared image and temperature distribution data of defects of additive parts,the temperature field correlation rule of normal and defect samples and the infrared image feature information were proved,and the defect self-identification and intelligent analysis algorithm was designed accordingly,finally providing a shape control method for the manufacturing process of CMT arc additive parts.The main research conclusions of this paper include:(1)Set up a set of successful CMT arc material system,and selects the FLIR A320 and GC86 G match the two infrared thermal imager,based on the characteristics of CMT technology design and assembly fixtures,filter,thermal imager enclosure,successfully developed the infrared monitoring system hardware part of the form a complete set,the effective implementation of CMT arc increasing material process of infrared image and temperature data collection.(2)Use the CMT arc additive system to successfully produce the samples without defects;Four kinds of typical problem samples,such as surface pores or holes,spatter,collapse,hump and uneven shape of welding seams,were also produced.Infrared images and temperature data of each process were collected by the infrared monitoring system.At the same time,the accuracy of temperature data is analyzed by thermocouple experiment.(3)The temperature field correlation rules of four types of problem samples and normal samples in the additive process were obtained,and the detailed features of infrared images corresponding to each type of problem were summarized,which laid a theoretical foundation for the design of defect recognition algorithm.(4)By using MATLAB functions and image processing toolbox,corresponding defect identification algorithms are designed according to the corresponding characteristics of defects,which can realize the identification of four types of typical problems,such as surface pores or holes,spatter,collapse,hump and uneven shape of welding seams.(5)CMT arc additive "temperature-defect" intelligent analysis system software was developed by using C++ language,which can realize real-time acquisition and display of infrared images and temperature data in additive manufacturing process,identify the 4 typical problems in additive manufacturing process,and record the whole process of additive.