【作者】 王鑫；

【导师】 徐雷； 任清川；

【作者基本信息】 四川大学 ， 机械工程（专业学位）， 2021， 硕士

【摘要】 航空机载铝合金共形件因具有比重小、比强度高、与载体共形而不破坏载体的机械结构等特点而被广泛应用于航天航空领域。但由于其刚度低、壁薄、弹性模量大,具有复杂的曲面特征,在铣削加工中应力情况复杂,导致很难达到理想的表面完整性要求,在特殊情况下甚至会造成零件报废。为此,研究铝合金共形件铣削过程中残余应力和表面粗糙度分布情况,进行铣削参数优化,同时研究铣削过程中表面完整性的检测方法,对于提高零件表面完整性和生产效率具有重要的指导意义和应用价值。本文以5A06铝合金共形件为研究对象,在深入分析铝合金三维铣削加工机理的基础上,建立了多道铣削有限元仿真模型,应用仿真分析的方法对铝合金共形件已加工表面粗糙度和残余应力的分布规律进行了研究,并基于灰色关联度和主成分分析进行了铣削加工参数优化的研究,同时提出了基于神经网络的表面完整性检测方法,最后开发了一个铝合金共形件铣削加工表面完整性优化软件。主要研究内容如下:(1)铝合金共形件多道铣削三维有限元模型构建。在阐述了金属切削加工机理等相关理论基础上,建立了多道铣削加工的三维螺旋刃铣刀有限元模型及材料为5A06铝合金的工件模型。利用该模型进行加工模拟,分析了已加工表面表面粗糙度和残余应力的分布情况。为铝合金共形件表面完整性的优化与研究建立基础。(2)铝合金共形件表面完整性检测方法。表面粗糙度和表面残余应力是共形件表面完整性的很重要的评价指标。该方法通过快速傅里叶变换和小波包分析获取切削过程中的切削力和噪音数据的时域、频域图,构建了一种以Resnet为基础的多通道输入神经网络,获取时、频域图与表面粗糙度及表面残余应力之间映射关系,通过神经网络通过识别时域、频域图,实现对铝合金共形件的表面粗糙度和表面残余应力检测。实验结果表明,该方法检测准确率为94.97%,具有一定的有效性和应用价值。(3)基于表面完整性的铣削参数优化。在共形件铣削有限元仿真模型的基础上,应用灰色关联度和主成分分析方法构建了铣削参数与表面完整性之间的数学模型。以该数学模型为基础,求解出最优铣削参数组合并研究了铣削参数对表面完整性的影响趋势。最后通过正交设计实验法验证了该模型的有效性。(4)铝合金共形件表面完整性优化软件的开发。本文在上述研究基础上,以多通道输入的神经网络为核心,使用Python语言和Pyqt开发了一个软件。该系统能过实际刀具和工件几何参数构建铣削有限元模型,并以铣削过程仿真的结果为依据建立铣削参数与表面粗糙度及表面残余应力间的数学模型,求解该模型即可确定最优铣削参数组合。同时该系统还实现了根据铣削过程中的噪音和切削力,检测共形件的表面完整性的功能,为优化铝合金共形件的表面完整性提供便利。更多还原

【Abstract】 Aluminum alloy aircraft conformal parts are widely used in the field of aerospace because of their low specific gravity,high specific strength,conformal with the carrier without damaging the mechanical structure of the carrier.However,due to its low stiffness,thin wall,large elastic modulus and complex surface characteristics,it is difficult to achieve the ideal surface integrity requirements in milling because of the complex stress conditions,and even parts scrapping might happen in special cases.Therefore,it is of great significance and application value to study the distribution of residual stress and surface roughness in the milling process of aluminum alloy conformal parts,optimize the milling parameters,and study the detection method of surface integrity in the milling process.In this paper,5A06 aluminum alloy conformal part is taken as the research object.Based on the in-depth analysis of the three-dimensional milling mechanism of aluminum alloy,a multi-channel milling finite element simulation model is established.The distribution of machined surface roughness and residual stress of aluminum alloy conformal part is studied by using the simulation analysis method,At the same time,a method of surface integrity detection based on neural network is proposed.Finally,a prototype system of milling surface integrity optimization for aluminum alloy conformal parts is developed.The main research contents are as follows:(1)The three-dimensional finite element model of aluminum alloy conformal part multi-channel milling was built.On the basis of expounding the metal cutting mechanism and other related theories,the three-dimensional finite element model of spiral edge milling cutter and the workpiece model of 5A06 aluminum alloy are established.The surface roughness and residual stress distribution of machined surface were analyzed by using the model.It is the foundation for the optimization and research of surface integrity of aluminum alloy conformal parts.(2)Surface integrity testing method of aluminum alloy conformal parts.Surface roughness and surface residual stress are very important evaluation indexes for surface integrity of conformal parts.In this method,the time-domain and frequency-domain diagrams of cutting force and noise data in the cutting process are obtained by fast Fourier transform and wavelet packet analysis,and a multi-channel input neural network based on RESNET is constructed to obtain the mapping relationship between time-domain and frequency-domain diagrams and surface roughness and surface residual stress,The surface roughness and surface residual stress of aluminum alloy conformal parts can be detected.The experimental results show that the detection accuracy of this method is 94.97%,which has certain effectiveness and application value.(3)Milling parameters optimization based on surface integrity.Based on the finite element simulation model of conformal part milling,the mathematical model between milling parameters and surface integrity is established by using grey correlation degree and principal component analysis.Based on the mathematical model,the optimal combination of milling parameters is obtained,and the influence trend of milling parameters on surface integrity is studied.Finally,the effectiveness of the model is verified by orthogonal design experiment.(4)Development of surface integrity optimization prototype system for aluminum alloy conformal parts.Based on the above research,this paper develops a prototype system with multi-channel input neural network as the core,using Python language and pyqt.The system can build the milling finite element model through the actual tool and workpiece geometric parameters,and establish the mathematical model between milling parameters and surface roughness and surface residual stress based on the results of milling process simulation.The optimal milling parameter combination can be determined by solving the model.At the same time,the system also realizes the function of detecting the surface integrity of conformal parts according to the noise and cutting force in the milling process,which provides convenience for optimizing the surface integrity of aluminum alloy conformal parts.更多还原