随着《中国制造2025》宏伟蓝图的展开，现代装备制造业对切削加工的精度、效率、成本、安全性和环境友好性提出了更高的要求。高性能铣削已成为现代高端装备制造的一种共性关键技术。铣削过程动力学建模和稳定性分析是高性能铣削的重要理论基础，为保证工件加工精度、表面质量和提高加工效率提供了关键支撑。 传统的铣削过程动力学建模与稳定性分析方法本质上是一种针对静态工况的建模与分析方法，不能准确反映实际切削工况下铣削过程的真实动力学行为及其对应的稳定性叶瓣图的动态变化情况，导致铣削稳定性叶瓣图失真。随着以高速铣削、精密铣削、强力铣削等为代表的高性能铣削的兴起，传统方法存在的不足逐渐凸显，所引发的切削颤振给实际生产造成了极大的损害。 本文针对传统方法存在的不足，围绕准确性和高效性两大主题深入研究适用于真实切削工况的铣削过程动力学建模与稳定性动态分析方法，称为切削态铣削过程动力学建模与稳定性动态分析方法。主要研究内容包括： 研究了切削态铣削过程动力学建模与稳定性动态分析方法的基本原理，明确了前者的关键是切削态铣削工艺系统动力学的准确建模与高效分析，构建了刀具-刀柄悬伸/主轴-刀柄凸缘二子结构刚性耦合研究模型；明确了后者的关键是分析算法的准确性、高效性与通用性，研究了分析算法的判稳原理，构建了研究工作的整体框架。 基于Timoshenko梁和旋转Timoshenko梁理论，分别建立了刀具-刀柄悬伸子结构中低速和高速工况动力学模型，考虑了剪切变形、截面转动惯量和陀螺效应对子结构动力学特性的影响，相较传统Euler-Bernoulli梁模型更准确地反映了子结构的真实动力学行为。针对所建模型提出一种子结构动力学谱元分析方法，将谱方法、有限元法和动力刚度矩阵法相结合，构造频率相关的动力形函数插值形式的谱单元分别用于中低速和高速工况子结构动力学分析，仿真结果证明所提方法相较传统有限元法提高了分析的精度和效率，为切削态铣削工艺系统刀尖动力学建模提供了关键支撑。 提出一种切削态铣削工艺系统刀尖动力学建模方法。该方法基于切削态刀尖动力学质量正则化模态振型不变性原理，考虑了主轴转速和接触条件的变化对切削态铣削工艺系统刀尖动力学行为的影响，将动柔度耦合子结构分析所得的空转态刀尖模态留数与谐激励型运行模态分析获取的切削态铣削工艺系统固有频率、阻尼比相结合建立了切削态铣削工艺系统刀尖动力学模型。实验证明该方法解决了传统铣削过程动力学建模方法中无法获取真实切削工况下铣削工艺系统刀尖动力学特性的问题，为建立切削态铣削过程动力学模型提供了关键结构参数。 提出一种用于切削态铣削过程稳定性动态分析的径向浸入率通用型时域谱元法。该方法基于周期系数微分方程Floquet稳定性理论，针对切削态铣削过程动力学模型将基于Legendre-Gauss-Lobatto节点的高阶时域谱元与积分方程数值解法相结合提高了状态向量离散动态映射矩阵的计算效率，解决了小径向浸入率铣削的强不连续性导致的谱元收敛率降低的问题。仿真实验表明所提方法显著提高了稳定性分析的精度、效率与通用性，实现了切削态铣削过程稳定性叶瓣图的动态更新。 综合上述研究成果，给出一种切削态铣削过程动力学建模与稳定性动态分析的完整工程应用流程。应用于CY-VMC850三轴立式铣削加工中心的生产加工过程，通过切削实验验证了本文所提方法的准确性、高效性和实用性。 本文研究成果进一步丰富和发展了铣削过程动力学建模和稳定性分析的理论与技术，对于实现高性能铣削加工具有重要的理论意义和工程价值。 关键词：铣削过程；切削态动力学建模；稳定性动态分析；谱元法；径向浸入率通用型时域谱元法

 "Made in China 2025" strategy makes more demands on cutting process about precision, efficiency, cost, safety and environmental protection. High-performance milling becomes a common key technology in high-end equipment manufacturing. Dynamics and stability of milling process are important theoretical foundations for high-performance milling which provide key support for ensuring machining precision, surface quality and improving productivity. Traditional dynamics modeling and stability analysis of milling process are based on static condition. It failed to present the actual milling process dynamics and the changes of stability lobes diagram (SLD). High-performance milling has become more popular, such as high-speed milling, precision milling, heavy milling, which highlights the disadvantages of traditional method. The distortion of stability lobes diagram leads to chatter and is harmful to manufacture. Considering the defects of the traditional method, operational dynamics modeling and stability dynamic analysis of milling process are comprehensively studied on two topics: precision and efficiency. The main research contents include: The principles of operational dynamics modeling and stability dynamic analysis of milling process are studied. The key of the former is operational dynamics of milling system. The rigid coupling model of tool-toolholder overhang and spindle—toolholder flange is proposed. The key of the latter is the precision, efficiency and universality of analysis algorithm. The stability criteria are studied. The research framework has been built. Based on the theory of Timoshenko beam and spinning Timoshenko beam,the substructure dynamics of tool-toolholder overhang are modeled respectively for medium or low speed condition and high speed condition. The effects of shear deformation, section moment of inertia and gyroscopic effect are considered, so that the model is more precise than the traditional Euler-Bernoulli beam model. A spectral element method (SEM) is proposed for analysis of the substructure dynamic model. The SEM combines spectral method, finite element method (FEM) and dynamic stiffness matrix method to build spectral elements with frequency-dependent dynamic shape functions. The simulation shows that the SEM is more precise and effective than traditional FEM. The proposed model and analysis method provide the key support for modeling operational tool tip dynamics. The method for modeling operational tool tip dynamics is proposed. Based on the principle that tool tip normal modal shapes are identical in operational and idle conditions, it combines the idle tool tip modal residues with operational natural frequencies and damping ratios to model operational tool tip dynamics. The effects of spindle speed and contact condition in operational conditions are considered. The idle tool tip modal residues are obtained by receptance coupling substructure analysis (RCSA). The natural frequencies and damping ratios are identified by operational modal analysis (OMA) with harmonic excitation. The results of experiments show that the proposed method enables obtaining operational tool tip dynamics model which provides the key structure parameters for operational milling process dynamics. The universal radial immersion temporal spectral element method for milling stability dynamic analysis is presented. Based on Floquet theory, it combines the high-order temporal spectral element with Legendre-Gauss-Lobatto (LGL) nodes and the numerical methods for integral equations to improve the computational efficiency of discrete dynamic mapping matrix and handle the low radial immersion milling with strong discontinuity. The results of simulation show that the presented method has significantly higher computational efficiency than the well-known semi-discretization method without loss of numerical precision, so that it enables the dynamic update of operational milling process SLDs. Based on the above research results, a complete engineering application of operational milling process dynamics modeling and stability dynamic analysis is presented. It is applied to CY-VMC850 three axis vertical milling machining center. The precision, efficiency and practicability of the proposed method are verified by experimental results. The work develops the theory and technology of milling process dynamics modeling and stability analysis. It has important theoretical significance and engineering value for high-performance milling. Key words: milling process; operational dynamics modeling; stability dynamic analysis; spectral element method; universal radial immersion temporal spectral element method