上一篇:有限元方法问答 13:热传导 · 系列总目录 · 下一篇:有限元方法问答 15:加权残值法

L15a — Vibration & Dynamics(振动与动力学)🟡 样卷未考


Q114 🟡 多自由度运动方程

EN: The equation of motion for a multi-DOF structural system is:

Mu¨+Cu˙+Ku=f(t)\boxed{\mathbf{M}\ddot{\mathbf{u}} + \mathbf{C}\dot{\mathbf{u}} + \mathbf{K}\mathbf{u} = \mathbf{f}(t)}

Physical meaning: Inertia forces + Damping forces + Elastic forces = Applied forces

  • M\mathbf{M} — mass matrix (from kinetic energy)
  • C\mathbf{C} — damping matrix
  • K\mathbf{K} — stiffness matrix
  • u\mathbf{u}, u˙\dot{\mathbf{u}}, u¨\ddot{\mathbf{u}} — displacement, velocity, acceleration

中文: 动力方程:Mu¨+Cu˙+Ku=f(t)\mathbf{M}\ddot{\mathbf{u}}+\mathbf{C}\dot{\mathbf{u}}+\mathbf{K}\mathbf{u}=\mathbf{f}(t)。惯性力+阻尼力+弹性力=外力。M是质量矩阵,C是阻尼矩阵,K是刚度矩阵。这是结构动力学的基本方程。


Q115 🟡 无阻尼自由振动

EN: For undamped free vibration (f=0\mathbf{f} = \mathbf{0}, C=0\mathbf{C} = \mathbf{0}):

Mu¨+Ku=0\mathbf{M}\ddot{\mathbf{u}} + \mathbf{K}\mathbf{u} = \mathbf{0}

Assuming harmonic motion u(t)=ϕsin(ωt)\mathbf{u}(t) = \boldsymbol{\phi}\sin(\omega t), we get the generalized eigenvalue problem:

(Kω2M)ϕ=0\boxed{(\mathbf{K} - \omega^2\mathbf{M})\boldsymbol{\phi} = \mathbf{0}}

For non-trivial solutions: det(Kω2M)=0\det(\mathbf{K} - \omega^2\mathbf{M}) = 0. This yields nn eigenvalues ωi2\omega_i^2 (squared natural frequencies) and nn eigenvectors ϕi\boldsymbol{\phi}_i (mode shapes), where nn is the number of DOFs.

中文: 无阻尼自由振动 → 广义特征值问题 (Kω2M)ϕ=0(\mathbf{K}-\omega^2\mathbf{M})\boldsymbol{\phi}=0。n个DOF就有n个固有频率和n个模态。固有频率通过 det(Kω2M)=0\det(\mathbf{K}-\omega^2\mathbf{M})=0 求解。


Q116 🔴 集中质量矩阵 vs 一致质量矩阵

EN: Two types of mass matrices:

Lumped mass matrix: Mass is concentrated at nodes → diagonal matrix. For a bar element: M=ρAL2[1001]\mathbf{M} = \frac{\rho AL}{2}\begin{bmatrix}1&0\\0&1\end{bmatrix}. Simple, computationally efficient.

Consistent mass matrix: Derived from the same shape functions as the stiffness matrix: M=VρNTNdV\mathbf{M} = \int_V \rho\mathbf{N}^T\mathbf{N}\,dV. For a bar element: M=ρAL6[2112]\mathbf{M} = \frac{\rho AL}{6}\begin{bmatrix}2&1\\1&2\end{bmatrix}. More accurate for a given mesh, but non-diagonal.

The consistent mass matrix follows the same formulation philosophy as the stiffness matrix (hence “consistent”).

中文: 集中质量矩阵 = 对角矩阵(质量集中在节点),计算效率高;一致质量矩阵 = ρNTNdV\int\rho\mathbf{N}^T\mathbf{N}dV(与刚度矩阵使用相同的形函数),更准确但非对角。杆单元一致质量矩阵 = ρAL6[2112]\frac{\rho AL}{6}\begin{bmatrix}2&1\\1&2\end{bmatrix}


Q117 🟡 固有频率公式

EN: For a single-DOF system, the circular natural frequency is:

ω=km[rad/s]\omega = \sqrt{\frac{k}{m}} \quad \text{[rad/s]}

The cyclic frequency: f=ω2πf = \frac{\omega}{2\pi} [Hz].

For a fixed bar: ω1=πLEρ\omega_1 = \frac{\pi}{L}\sqrt{\frac{E}{\rho}} (first axial natural frequency).

For MDOF systems: ωi=λi\omega_i = \sqrt{\lambda_i} where λi\lambda_i are eigenvalues of Kϕ=λMϕ\mathbf{K}\boldsymbol{\phi} = \lambda\mathbf{M}\boldsymbol{\phi}.

中文: 单DOF固有频率 ω=k/m\omega=\sqrt{k/m}(单位rad/s);频率 f=ω/(2π)f=\omega/(2\pi)(Hz)。多DOF系统:ωi\omega_i 通过特征值求解。


Q118 🔴 低阶模态比高阶模态更准确

EN: In FEM dynamic analysis, lower modes (smaller ωi\omega_i) are more accurate than higher modes. This is because:

  • Lower modes have longer wavelengths, requiring fewer elements per wavelength for accurate representation
  • Higher modes have shorter wavelengths, and the FE discretization cannot capture them well unless the mesh is very fine
  • A rule of thumb: at least 6-10 elements per wavelength are needed for acceptable accuracy

For a given mesh, only the lowest n/3n/3 to n/2n/2 modes are typically considered reliable (nn = number of DOFs).

中文: 低阶模态(小频率)比高阶模态更准确。因为低阶模态波长大,有限元离散容易捕捉;高阶模态波长短,需要极密网格才能准确。经验法则:每个波长至少6-10个单元。给定网格下,通常只有最低的1/3到1/2阶模态可靠。


Q119 🔴 刚体模态(Rigid Body Modes)

EN: Rigid body modes correspond to ω=0\omega = 0. These represent motions that produce no strain energy — pure translation or rotation of the unconstrained (or insufficiently constrained) structure. In a free-free vibration analysis:

  • A 2D structure has 3 rigid body modes (2 translations + 1 rotation)
  • A 3D structure has 6 rigid body modes (3 translations + 3 rotations)

Practical use: Rigid body modes can be used to check the FE model — unexpected zero-frequency modes indicate mechanisms, missing constraints, or disconnected parts.

中文: 刚体模态对应 ω=0\omega=0(无应变能的运动)。2D有3个刚体模态(2平动+1转动),3D有6个。可用于检查模型——多余的零频率说明有机构、缺少约束或连接错误。


Q120 🔴 模态位移的大小无物理意义

EN: In normal mode analysis (eigenvalue analysis), the magnitude of the eigenvector has NO physical meaning — only the shape (relative proportions between DOFs) matters. Eigenvectors can be arbitrarily scaled. Common normalization: ϕiTMϕi=1\boldsymbol{\phi}_i^T\mathbf{M}\boldsymbol{\phi}_i = 1 (mass normalization) or set the maximum component to 1.

中文: 模态分析中,位移大小无物理意义,只有形状(各DOF的相对比例)有意义。可以任意缩放。常见归一化:质量归一化 ϕiTMϕi=1\boldsymbol{\phi}_i^T\mathbf{M}\boldsymbol{\phi}_i=1,或最大分量为1。


Q121 🟡 模态正交性

EN: Mode shapes satisfy orthogonality with respect to M\mathbf{M} and K\mathbf{K}:

ϕiTMϕj=0,ϕiTKϕj=0for ij\boldsymbol{\phi}_i^T \mathbf{M} \boldsymbol{\phi}_j = 0, \quad \boldsymbol{\phi}_i^T \mathbf{K} \boldsymbol{\phi}_j = 0 \quad \text{for } i \neq j

Mass-normalized: ϕiTMϕi=1\boldsymbol{\phi}_i^T \mathbf{M} \boldsymbol{\phi}_i = 1 and ϕiTKϕi=ωi2\boldsymbol{\phi}_i^T \mathbf{K} \boldsymbol{\phi}_i = \omega_i^2.

Orthogonality enables modal superposition — the response can be expressed as a linear combination of mode shapes.

中文: 不同模态关于M和K正交。正交性使得模态叠加法成为可能(系统响应 = 各模态贡献的线性组合)。这是模态分析的理论基础。


Q122 🟡 阻尼比和Rayleigh阻尼

EN: Damping ratio: ζ=c/cc\zeta = c/c_c, where cc=2km=2mωc_c = 2\sqrt{km} = 2m\omega is the critical damping. Structural damping: 0<ζ<0.150 < \zeta < 0.15, typically 1–5%. For small damping: ωdω\omega_d \approx \omega.

Rayleigh (proportional) damping: C=αM+βK\mathbf{C} = \alpha\mathbf{M} + \beta\mathbf{K}, where α\alpha and β\beta are determined from: ζi=α2ωi+βωi2\zeta_i = \frac{\alpha}{2\omega_i} + \frac{\beta\omega_i}{2}. This form preserves modal orthogonality and is convenient for FEM implementation.

中文: 阻尼比 ζ=c/cc\zeta=c/c_c,结构阻尼约1-5%,小阻尼下 ωdω\omega_d\approx\omega。Rayleigh阻尼 C=αM+βK\mathbf{C}=\alpha\mathbf{M}+\beta\mathbf{K} 保持了模态正交性,是FEM中最常用的阻尼模型。



上一篇:有限元方法问答 13:热传导 · 系列总目录 · 下一篇:有限元方法问答 15:加权残值法