【作者】 刘锦；

【导师】 洪波；

【作者基本信息】 湘潭大学 ， 材料科学与工程， 2021， 硕士

【摘要】 通过高能量密度的等离子电弧对焊丝或粉末进行加热熔化,然后将熔化的焊材层层堆敷进而熔敷出所设计的构件的方法被称为等离子电弧增材制造技术。该方法现阶段主要应用于尺寸大、形貌复杂的构件。由于等离子电弧的能量密度高且集中,其熔敷过程容易导致熔池溢流现象的出现,因此,如何减少熔滴溢流进而减小熔敷层坍塌的几率是等离子电弧增材制造是一个急需解决的问题。在熔滴溢流后导致增材制造层形貌可能导致后续的堆积过程与原本计划相差较大,得到的工件形貌尺寸与原本设计的模型差异较大,因此,提高增材制造熔敷层的精度是本文解决的另一个问题。本文首先设计磁控等离子电弧传感器来减少熔敷层热量累积导致的熔池溢流的现象,然后建立了磁控等离子电弧增材制造系统,对单道多层等离子电弧增材制造成型信息进行检测及自适应控制展开了研究。(1)本文通过对摆动等离子电弧及熔池的受力分析,并综合磁控摆动电弧方法与等离子电弧增材制造方法设计了磁控等离子电弧焊枪,在此基础上搭建磁控等离子电弧增材制造平台;通过磁控控制电弧在增材制造熔敷层上摆动,有效的减少了沉积层表面累积的热量,减少了熔池中液态金属的凝固时间。从而减少了等离子电弧增材制造过程中由于热量累积导致熔池溢流现象。(2)本文对磁控等离子电弧增材制造层形貌检测方法进行设计与研究,通过建立弧长与各参数的数学模型分析得到了检测熔敷层形貌尺寸的参数,主要以增材制造时的熔敷电流、熔敷电压、励磁电流频率以及焊枪高度检测的输入量,然后进行了大量的工艺实验,获取了焊接电流、焊接电压、励磁电流频率、焊枪高度以及增材制造层形貌尺寸数据。(3)接下来对实验数据实施归一化计算,成功构建BP神经网络熔敷层高度及层宽预测模型,针对这一模型进行训练及验证处理,进而确定了4-13-2的BP神经网络结构,并建立了神经网络预测模型;通过对神经网络模型的训练,实现了增材制造过程中熔敷层形貌尺寸的预测。(4)针对磁控等离子电弧增材制造技术的特点,设计了磁控等离子电弧增材制造熔敷尺寸控制器,通过该控制器对增材制造熔敷层形貌尺寸进行实时调控,进而实现等离子电弧增材制造熔敷层的精确成形。(5)最后,在本文所搭建的平台上通过本文所述的方法制造相关构件,实验结果表明,需制作的构件的形貌规则、成形精度高,无明显缺陷发生。因此采用本文所设计的磁控摆动等离子电弧增材制造形貌检测及自适应控制系统,可以提高等离子电弧增材制造的成形精度。更多还原

【Abstract】 Welding wire or powder is heated and melted by plasma arc with high energy density,and then the melted welding material is stacked layer by layer to deposit the designed component,which is called plasma arc additive manufacturing technology.At present,this method is mainly applied to the components with large size and complex morphology.Due to the high energy density and concentration of plasma arc,the deposition process is easy to lead to the appearance of molten pool overflow phenomenon.Therefore,how to reduce the droplet overflow and reduce the probability of deposition layer collapse is an urgent problem for plasma arc additive manufacturing.After the droplet overflow,the morphology of additive manufacturing layer may lead to a large difference between the subsequent stacking process and the original plan,and the morphology and size of the workpiece obtained are quite different from the original design model.Therefore,improving the accuracy of additive manufacturing coating is another problem to be solved in this paper.In this paper,firstly,a magnetron plasma arc sensor is designed to reduce the phenomenon of molten pool overflow caused by the accumulation of heat in the cladding layer.Then,a magnetron plasma arc additive manufacturing system is established to detect the forming information of single pass and multi-layer plasma arc additive manufacturing and develop adaptive control.(1)In this paper,the magnetron plasma arc welding gun is designed based on the force analysis of swing plasma arc and molten pool,and the magnetron swing arc method and plasma arc additive manufacturing method;By controlling the arc swing on the deposited layer of additive manufacturing by magnetic control,the accumulated heat on the surface of the deposited layer is effectively reduced,and the solidification time of liquid metal in the molten pool is reduced.Thus,the phenomenon of molten pool overflow caused by heat accumulation in the manufacturing process of plasma arc additive is reduced.(2)In this paper,the design and research of the method for detecting the morphology of the coating in the manufacturing process of magnetron plasma arc additive are carried out.By establishing the mathematical model of arc length and various parameters,the parameters for detecting the morphology and size of the coating are obtained,which are mainly based on the input of the deposition current,deposition voltage,excitation current frequency and torch height,The data of welding current,welding voltage,exciting current frequency,welding torch height and additive manufacturing layer morphology and size were obtained.(3)next the first mock exam is carried out on the experimental data.The prediction model of the height and width of the BP coating is successfully constructed.The training model is trained and verified,and the 4-13-2 neural network structure of BP is established.Through the training of neural network model,the prediction of cladding morphology and size in additive manufacturing process is realized.(4)According to the characteristics of magnetron plasma arc additive manufacturing technology,a deposition size controller for magnetron plasma arc additive manufacturing is designed.Through the controller,the morphology and size of the deposition layer for additive manufacturing can be adjusted in real time,and then the precise forming of the deposition layer for plasma arc additive manufacturing can be realized.(5)Finally,the relevant components are manufactured by the method described in this paper on the platform built in this paper.The experimental results show that the components to be manufactured have regular morphology,high forming accuracy and no obvious defects.Therefore,the shape detection and adaptive control system designed in this paper can improve the forming accuracy of plasma arc additive manufacturing.更多还原