| Action bracket,as an important part of the tank action device and suspension device,has the diversified shape characteristics and complex processing.The production quality of the action bracket is related to the safety and service life of the whole vehicle.The production method of action bracket is gradually changing from the traditional to intelligence,with the dual drive of the “14th Five-Year Plan” and the “Made in China2025”.The discrete intelligent production method as an important solution to break through the traditional production uniformity technology bottleneck,which has datadriven intelligent sensing and collaborative control capability.However,the main technical problem limiting discrete intelligent production is the process equipment not flexibility.In addition,the action bracket involves a large number of face milling processes with high accuracy requirements,and traditional contact surface roughness measurement does not detect quality problems immediately in production processes.Selecting appropriate machining parameters and tool parameters can only be based on trial and error and manual experience,which undoubtedly wastes production resources.This paper aims to address the above technical challenges.Under the support of the“Major Science and Technology Innovation Engineering Projects of Shandong Province”and “Science and Technology SMEs Innovation Capacity Improvement Project of Shandong Province”,conducted research on the design and quality control of the tank action bracket intelligent production line.The tank action bracket intelligent production line and flexible process equipment is designed,the surface roughness prediction model is established,and the surface roughness prediction system is developed.The main studies are as follows:(1)The current status of domestic and international research on intelligent production line and surface roughness prediction model is analyzed.The research content and technical route of this paper are determined by combining the major needs of discrete intelligent production and quality control of tank action brackets.(2)The intelligent production line and main systems for tank action brackets are designed.The tank action bracket intelligent production line layout and main functions is designed,based on the analysis of the action bracket production process and the design requirements.The movable fixture process system,material conveying system and process equipment system of the tank action bracket intelligent production line are designed,based on the shape characteristics and process of the action bracket.The main structural reliability of the system equipment is analyzed.(3)A face milling surface roughness prediction model of tank action bracket is proposed,the profile forming mechanism is revealed,the influence of machining parameters and tool parameters on surface roughness is investigated,and the accuracy of the model is experimentally vererified.The profile-forming mechanism is analyzed based on kinematics and geometry.A face milling surface roughness prediction model is proposed,taking into consideration the influences of insert back cutting and stepover ratio.The 2D profile and 3D topography of different locations is reconstructed by numerical analysis.The influence of the machining parameters(feed per tooth,overlap rate)and tool parameters(corner radius,minor cutting edge angle)on surface roughness was investigated separately using single factor analysis,and the reason in terms of contour forming mechanism is analyzed.The experiment of milling aerospace aluminum alloy7075 is suggested to verify the model,and the Z-Map prediction model is introduced for comparison.(4)A surface roughness prediction model considering the tool runout is proposed,the profile forming mechanism under the influence of tool runout is revealed,the influence law of tool runout on surface roughness,and the accuracy of the model is experimentally vererified.The profile forming geometry and kinematics under the influence of tool runout are analyzed.A surface roughness prediction model considering the tool runout is proposed based on the face milling surface roughness prediction model.The 2D profile and 3D topography of same locations is reconstructed under different tool runout by numerical analysis.The influence of different tool runout on surface roughness is investigated,and the causes from the perspective of profile forming is analyzed.The experiment of milling ZG32 Mn Mo is suggested to verify the model,and the face milling prediction model is introduced for comparison.(5)A dynamic surface roughness prediction model based on convolutional neural networks is proposed,the profile forming mechanism under the influence of dynamic factors is revealed,and the accuracy of the model is experimentally vererified.The influence of tool vibration,elastic recovery and tool wear on surface roughness during machining are analyzed.The equation between the above factors and surface roughness is established based on the cutting force.The cutting data sets is collected by milling ZG32 Mn Mo.A surface roughness prediction model based on convolutional neural network is proposed,with cutting force as the input signal and the difference in height between the theoretical profile and the real profile as the output signal.Experimental was conducted to analyze the accuracy of the CNN prediction model,and the prediction model of tool runout is introduced for comparison.(6)Surface roughness prediction system developed to enable quality control of tank action mount production.The surface roughness prediction system is developed based on prediction model of runout and CNN prediction model,including surface roughness offline predict and online monitoring.The main interface,functions and procedures of the surface roughness prediction system are designed,and analyze the usage of the system.A surface roughness visualization platform is built by interfacing the surface roughness prediction system with MES system of the tank action bracket intelligent production line. |
PDF Full Text Request