UNIT 1: FUNDAMENTALS OF VIBRATION & SINGLE-DEGREE-OF-FREEDOM SYSTEMS
1.0 INTRODUCTION & BASIC CONCEPTS
1.1 Importance and Study of Vibrations
-
Causes: Unbalanced forces, friction, elastic deformation, wind, seismic activity.
-
Effects: Fatigue failure, noise, discomfort, malfunction, structural failure.
-
Applications: Vibratory feeders, compactors, ultrasonic cleaners, musical instruments, earthquake-resistant design.
-
Goal: Understand, predict, control (reduce or utilize) vibrations.
1.2 Basic Elements of a Vibratory System
| Element | Symbol | Function | Idealization |
|---|---|---|---|
| Mass/Inertia | $m$ or $J$ | Stores kinetic energy | Lumped mass or rotary inertia |
| Spring | $k$ (linear) or $$\displaystyle k_t $$ (torsional) | Stores potential energy, provides restoring force | Linear elastic, stiffness constant |
| Damper | $c$ (viscous) | Dissipates energy (heat) | Force $\propto$ velocity: $$\displaystyle F_d = c\dot{x} $$ |
1.3 Degree of Freedom (DOF)
-
Definition: Minimum number of independent coordinates required to define the system's configuration at any instant.
-
Types:
-
Translational DOF: Linear motion (e.g., $x$).
-
Rotational DOF: Angular motion (e.g., $\theta$).
-
-
Examples:
-
1-DOF: Simple spring-mass-damper, pendulum (small angle).
-
2-DOF: Two masses on three springs, car suspension (heave & pitch).
-
Multi-DOF: Building frames, multi-mass shafts.
-
[!TIP] Exam Focus: Be able to identify DOF from diagrams. A rigid body in plane has 3 DOF (2 trans + 1 rotation).
1.4 Modeling: Lumped vs. Distributed Parameters
-
Lumped Parameter: Mass, stiffness, damping concentrated at discrete points. Used for SDOF & MDOF. Governing ODEs.
-
Distributed Parameter: Properties vary continuously (e.g., beam, rod). Governing PDEs. Approximated by lumped models.
1.5 Classification of Vibrations
| Basis | Types | Key Characteristics |
|---|---|---|
| Energy Input | Free | Initial energy only; response decays (damped) or persists (undamped). |
| Forced | Continuous external force; steady-state + transient response. | |
| Damping | Undamped | No energy loss; perpetual oscillation. |
| Damped | Energy dissipation; amplitude decays. | |
| Linearity | Linear | Superposition holds; $$\displaystyle m\ddot{x}+c\dot{x}+kx=F(t) $$. |
| Non-linear | Superposition fails; e.g., Coulomb friction, large deflection. | |
| Force Nature | Deterministic | Precisely known (harmonic, periodic). |
| Random | Probabilistic description (e.g., road roughness, wind). |
1.6 Vibration Analysis Approach
-
Modeling: Idealize physical system → mathematical model (lumped parameters).
-
Equation of Motion (EOM): Derive using Newton's laws or energy methods (Lagrange).
-
Solution: Solve ODE/PDE (analytical/numerical) for displacement $x(t)$.
-
Interpretation: Extract natural frequency, damping ratio, mode shapes, response amplitude.
2.0 UNDAMPED SINGLE-DEGREE-OF-FREEDOM (SDOF) SYSTEMS
2.1 EOM for Undamped Spring-Mass System
System: Mass $m$, spring stiffness $k$, no damping, force $F(t)$.
DiagramCANVAS: Simple horizontal spring-mass system with coordinate x from equilibrium
Derivation (Newton's 2nd Law):
$$\sum F = m\ddot{x} \quad \Rightarrow \quad -kx + F(t) = m\ddot{x}$$
Standard Form:
$$m\ddot{x} + kx = F(t)$$
For free vibration ($$\displaystyle F=0 $$):
$$\boxed{m\ddot{x} + kx = 0}$$
2.2 Simple Harmonic Motion (SHM)
-
Mathematical Definition: Solution to $$\displaystyle \ddot{x} + \omega_n^2 x = 0 $$ is $$\displaystyle x(t) = C_1 \sin(\omega_n t) + C_2 \cos(\omega_n t) $$.
-
Natural Frequency ($$\displaystyle \omega_n $$):
$$\omega_n = \sqrt{\frac{k}{m}} \quad \text{(rad/s)}$$
- Time Period ($$\displaystyle T_n $$):
$$T_n = \frac{2\pi}{\omega_n} = 2\pi\sqrt{\frac{m}{k}} \quad \text{(s)}$$
-
Vector Representation:
DiagramCANVAS: Phasor diagram showing rotating vector for displacement x, velocity (90° lead), acceleration (180° lead)-
Displacement: $$\displaystyle x = X \sin(\omega_n t) $$ or $$\displaystyle X \cos(\omega_n t) $$.
-
Velocity: $$\displaystyle \dot{x} = \omega_n X \cos(\omega_n t) $$ (leads $x$ by $$\displaystyle 90^\circ $$).
-
Acceleration: $$\displaystyle \ddot{x} = -\omega_n^2 X \sin(\omega_n t) $$ (lags $x$ by $$\displaystyle 180^\circ $$).
-
2.3 Energy Method (Rayleigh's Method)
- Principle: For undamped, conservative systems, max KE = max PE.
$$\left( \frac{1}{2}m\dot{x}_{max}^2 \right)_{t=0} = \left( \frac{1}{2}kx_{max}^2 \right)_{t=\frac{T_n}{4}}$$
- Natural Frequency Estimate:
$$\omega_n^2 \approx \frac{k_{eq}}{m} \quad \text{where} \quad k_{eq} \text{ is stiffness "seen" by mass.}$$
> [!CAUTION] Rayleigh's method gives **upper bound** for $$\displaystyle \omega_n $$ if the assumed mode shape is exact; otherwise an approximation.
2.4 Natural Frequency with Multiple Springs
-
Equivalent Stiffness:
-
Parallel: $$\displaystyle k_{eq} = k_1 + k_2 + ... $$
-
Series: $$\displaystyle \frac{1}{k_{eq}} = \frac{1}{k_1} + \frac{1}{k_2} + ... $$
-
-
Complex Systems: Reduce stepwise to single $$\displaystyle k_{eq} $$.
2.5 Natural Frequency: Rotational & Pendulum Systems
- Torsional System: $$\displaystyle J\ddot{\theta} + k_t\theta = 0 $$
$$\omega_n = \sqrt{\frac{k_t}{J}}$$
> DiagramCANVAS: Torsional system with disk of inertia J, torsional spring kt
- Simple Pendulum (small $\theta$):
$$\omega_n = \sqrt{\frac{g}{L}}$$
- Compound Pendulum:
$$\omega_n = \sqrt{\frac{mgh}{J_O}} \quad \text{where } h \text{ is distance from pivot to CG.}$$
2.6 Natural Frequency: Rigid Body Motion
-
Example: Mass on inclined plane with spring.
DiagramCANVAS: Mass on frictionless incline angle α, spring parallel to incline-
Static equilibrium: $$\displaystyle mg\sin\alpha = k\delta_{st} $$.
-
EOM about equilibrium: $$\displaystyle m\ddot{x} + kx = 0 $$ (same as horizontal!).
-
$$\displaystyle \omega_n = \sqrt{k/m} $$ independent of gravity/incline angle.
-
3.0 DAMPED SINGLE-DEGREE-OF-FREEDOM SYSTEMS
3.1 Types of Damping
| Type | Force Law | Characteristics | Common in |
|---|---|---|---|
| Viscous | $$\displaystyle F_d = c\dot{x} $$ | Linear, frequency-independent, easy math. | Hydraulic dampers, lubricated parts. HIGHLY FREQUENT |
| Coulomb (Dry Friction) | $$\displaystyle F_d = \mu N \cdot \text{sgn}(\dot{x}) $$ | Constant magnitude, non-linear, stick-slip. | Brakes, unlubricated joints. |
| Structural/Hysteresis | $$\displaystyle F_d = \beta \dot{x} $$ (complex $k$) | Energy loss per cycle $\propto$ displacement^2. | Materials (metal fatigue, rubber). |
| Solid Damping | $$\displaystyle F_d = b\dot{x} + d\dot{x}^3 $$ | Combination of viscous & non-linear. | Some polymers at large amplitudes. |
3.2 EOM for Viscously Damped System
$$m\ddot{x} + c\dot{x} + kx = F(t)$$
Free vibration ($$\displaystyle F=0 $$):
$$m\ddot{x} + c\dot{x} + kx = 0 \quad \text{or} \quad \ddot{x} + 2\zeta\omega_n\dot{x} + \omega_n^2 x = 0$$
where:
-
$$\displaystyle \omega_n = \sqrt{k/m} $$ (undamped natural frequency)
-
Critical Damping Constant: $$\displaystyle c_c = 2\sqrt{km} = 2m\omega_n $$
-
Damping Ratio: $$\displaystyle \zeta = \frac{c}{c_c} $$ (dimensionless, key parameter)
3.3 Solution for Free Vibration (Viscous)
Assume solution $$\displaystyle x = e^{st} $$. Substituting gives characteristic equation:
$$ms^2 + cs + k = 0 \quad \Rightarrow \quad s^2 + 2\zeta\omega_n s + \omega_n^2 = 0$$
Roots: $$\displaystyle s_{1,2} = -\zeta\omega_n \pm \omega_n\sqrt{\zeta^2 - 1} $$
| Damping Case | Condition | Roots $s$ | Response $x(t)$ | Behavior |
|---|---|---|---|---|
| Underdamped | $$\displaystyle \zeta < 1 $$ | $$\displaystyle -\zeta\omega_n \pm i\omega_d $$ | $$\displaystyle e^{-\zeta\omega_n t}(C_1\sin\omega_d t + C_2\cos\omega_d t) $$ | Oscillatory, amplitude decays. |
| Critically Damped | $$\displaystyle \zeta = 1 $$ | $$\displaystyle -\omega_n $$ (repeated) | $$\displaystyle (C_1 + C_2 t)e^{-\omega_n t} $$ | Fastest return to equilibrium without oscillation. |
| Overdamped | $$\displaystyle \zeta > 1 $$ | Two real, negative | $$\displaystyle C_1e^{s_1 t} + C_2e^{s_2 t} $$ | Slow return, no oscillation. |
Damped Natural Frequency ($$\displaystyle \zeta < 1 $$):
$$\omega_d = \omega_n\sqrt{1 - \zeta^2}$$
3.4 Logarithmic Decrement ($\delta$)
- Definition: Natural log of ratio of successive amplitudes $n$ cycles apart.
$$\delta = \ln\left(\frac{x(t)}{x(t+nT_d)}\right) \approx \frac{2\pi\zeta}{\sqrt{1-\zeta^2}} \quad \text{(for $$\displaystyle \zeta < 0.2 $$)}$$
> DiagramCANVAS: Decaying envelope of underdamped response showing peaks at $$\displaystyle x_1, x_2, ... $$
- Exact Formula:
$$\boxed{\delta = \frac{2\pi\zeta}{\sqrt{1-\zeta^2}}}$$
-
For small $\zeta$ ($$\displaystyle <0.2 $$): $\delta \approx 2\pi\zeta$.
-
Measurement: $$\displaystyle \zeta = \frac{\delta}{\sqrt{4\pi^2 + \delta^2}} $$.
[!TIP] Exam Focus: Interconversion between $c$, $$\displaystyle c_c $$, $\zeta$, $\delta$ is VERY FREQUENT. Remember: $$\displaystyle \zeta = c/c_c $$, $\delta \approx 2\pi\zeta$ for light damping.
3.5 Measurement of Damping
-
Logarithmic Decrement Method: Measure $\delta$ from free decay response → compute $\zeta$.
-
Half-Power Bandwidth Method (Forced Vibration): From MF curve, measure bandwidth $\Delta\omega$ at $$\displaystyle MF = 1/\sqrt{2} $$ of peak. $$\displaystyle \zeta \approx \frac{\Delta\omega}{2\omega_n} $$.
3.6 Comparison: Underdamped vs. Coulomb Damping
| Feature | Viscous Damping | Coulomb Damping |
|---|---|---|
| Force Law | $$\displaystyle F_d = c\dot{x} $$ | $$\displaystyle F_d = \mu N \cdot \text{sgn}(\dot{x}) $$ |
| Amplitude Decay | Exponential: $$\displaystyle x \propto e^{-\zeta\omega_n t} $$ | Linear: $$\displaystyle x = x_0 - \frac{4\mu N}{m\omega_n}t $$ per half-cycle |
| Frequency | Slightly less than $$\displaystyle \omega_n $$: $$\displaystyle \omega_d = \omega_n\sqrt{1-\zeta^2} $$ | Independent of amplitude: $$\displaystyle \omega_n' \approx \omega_n\sqrt{1 - \frac{4\mu N}{\pi m\omega_n x_0}} $$ |
| Math | Linear, solvable analytically. | Non-linear; requires approximate or numerical methods. |
3.7 Special Cases & Interconversion
Given any two of: $c$, $$\displaystyle c_c $$, $\zeta$, $\delta$ → find others.
-
$$\displaystyle \zeta = c / c_c = c / (2\sqrt{km}) $$
-
$$\displaystyle \delta = \frac{2\pi\zeta}{\sqrt{1-\zeta^2}} $$
-
$$\displaystyle c = \zeta \cdot 2\sqrt{km} $$
4.0 FORCED VIBRATION OF SDOF SYSTEMS
4.1 Forcing Functions
-
Harmonic: $$\displaystyle F(t) = F_0 \sin \omega t $$ (most common).
-
Periodic: Can be expressed as Fourier series (sum of harmonics).
-
Arbitrary: Solved via convolution/Laplace transform.
4.2 EOM
$$m\ddot{x} + c\dot{x} + kx = F_0 \sin \omega t$$
4.3 Steady-State Solution (Particular Solution)
Method: Complex Notation (Phasors).
Assume complex force: $$\displaystyle \tilde{F} = F_0 e^{i\omega t} $$ and complex response: $$\displaystyle \tilde{x} = X e^{i\omega t} $$.
Substitute into EOM (with $$\displaystyle d/dt \rightarrow i\omega $$):
$$(-\omega^2 m + i\omega c + k)\tilde{x} = \tilde{F}$$
$$\tilde{x} = \frac{F_0}{k - m\omega^2 + i c\omega} = \frac{F_0/k}{1 - r^2 + i(2\zeta r)}$$
where $$\displaystyle r = \omega / \omega_n $$ (frequency ratio).
Magnitude (Magnification Factor, MF):
$$X = \frac{F_0/k}{\sqrt{(1-r^2)^2 + (2\zeta r)^2}}$$
$$\boxed{M = \frac{X}{X_{st}} = \frac{1}{\sqrt{(1-r^2)^2 + (2\zeta r)^2}}}$$
where $$\displaystyle X_{st} = F_0/k $$ (static deflection).
Phase Angle ($\phi$): Phase of displacement lag behind force.
$$\tan\phi = \frac{2\zeta r}{1 - r^2}$$
DiagramCANVAS: MF vs r curve for different ζ, showing resonance peak and phase transition from 0° to 180°
4.4 Analysis of MF Curve
-
Effect of Damping ($\zeta$):
-
Decreases peak magnification $$\displaystyle M_{max} $$.
-
Widens bandwidth (frequency range where $$\displaystyle M > 1 $$).
-
Eliminates infinite resonance at $$\displaystyle r=1 $$ for $$\displaystyle \zeta > 0 $$.
-
-
Resonance: Occurs near $$\displaystyle r=1 $$. For light damping ($$\displaystyle \zeta < 0.2 $$), resonant frequency:
$$r_{max} \approx \sqrt{1 - 2\zeta^2} \quad \Rightarrow \quad \omega_{max} \approx \omega_n\sqrt{1 - 2\zeta^2}$$
$$\displaystyle M_{max} \approx \frac{1}{2\zeta} $$.
-
Regions:
-
$r \ll 1$ (Rigid Body): $M \approx 1$, $$\displaystyle \phi \approx 0^\circ $$.
-
$r \gg 1$ (High Freq/Isolation): $$\displaystyle M \approx 1/r^2 $$, $$\displaystyle \phi \approx 180^\circ $$.
-
4.5 Total Response
$$x(t) = x_c(t) + x_p(t)$$
-
$$\displaystyle x_c(t) $$: Complementary (Transient) solution of homogeneous equation (free vibration). Decays with time if $$\displaystyle \zeta > 0 $$.
-
$$\displaystyle x_p(t) $$: Particular (Steady-State) solution: $$\displaystyle x_p(t) = X \sin(\omega t - \phi) $$.
-
Given ICs: $$\displaystyle x(0)=x_0 $$, $$\displaystyle \dot{x}(0)=v_0 $$, determine $$\displaystyle C_1, C_2 $$ in $$\displaystyle x_c(t) $$.
[!TIP] Exam Focus: "Find total response" problems are FREQUENT. Compute $$\displaystyle \omega_n $$, $\zeta$, $X$, $\phi$, then apply ICs to $$\displaystyle x_c + x_p $$ at $$\displaystyle t=0 $$.
4.6 Forced Vibration with Base Excitation
System: Base motion $y(t)$, mass relative displacement $$\displaystyle z = x - y $$.
DiagramCANVAS: Mass-spring-damper on moving base y(t), relative coordinate z
EOM (relative to base):
$$m\ddot{z} + c\dot{z} + kz = -m\ddot{y}$$
If $$\displaystyle y = Y \sin \omega t $$, then RHS is $$\displaystyle -m\omega^2 Y \sin \omega t $$.
- Displacement Transmissibility ($$\displaystyle T_d $$): Ratio of transmitted base amplitude to machine amplitude.
$$T_d = \frac{X}{Y} = \frac{r^2}{\sqrt{(1-r^2)^2 + (2\zeta r)^2}}$$
- Force Transmissibility ($$\displaystyle T_f $$): Ratio of transmitted force to static force.
$$T_f = \frac{\sqrt{(kX)^2 + (c\omega X)^2}}{kY} = \sqrt{1 + (2\zeta r)^2} \cdot T_d$$
Vibration Isolation:
-
Goal: Reduce transmitted force/acceleration to foundation ($$\displaystyle T_f < 1 $$).
-
Condition: Requires $$\displaystyle r > \sqrt{2} $$ and some damping ($$\displaystyle \zeta > 0 $$).
-
Design: Use soft springs (low $$\displaystyle \omega_n $$) to achieve $$\displaystyle r = \omega/\omega_n > \sqrt{2} $$ for operating frequency $\omega$.
5.0 APPLICATIONS OF SDOF FORCED VIBRATION
5.1 Vibration Isolation Design
-
Given: Machine weight $$\displaystyle W = mg $$, operating speed $\omega$, desired transmissibility $$\displaystyle T_f $$.
-
Steps:
-
Find required $$\displaystyle r = \omega/\omega_n $$ from $$\displaystyle T_f $$ formula (often iterative).
-
Compute $$\displaystyle \omega_n = \omega / r $$.
-
Stiffness: $$\displaystyle k = m\omega_n^2 $$.
-
Damping: Choose $\zeta$ (typically 0.05–0.2) → $$\displaystyle c = \zeta \cdot 2\sqrt{km} $$.
-
5.2 Rotating Unbalance
-
Model: Mass $$\displaystyle m_u $$ rotating at radius $e$ at speed $\omega$ → harmonic force $$\displaystyle F_0 = m_u e \omega^2 $$.
-
Steady-State Amplitude:
$$X = \frac{m_u e r^2}{\sqrt{(1-r^2)^2 + (2\zeta r)^2}}$$
where $$\displaystyle r = \omega / \omega_n $$, $$\displaystyle \omega_n = \sqrt{k/(m+m_u)} $$.
- Critical Speed: Speed at which $$\displaystyle r=1 $$ → amplitude large unless damped.
5.3 Speed-Dependent Excitation (Vehicle on Sinusoidal Road)
-
Model: Base excitation $$\displaystyle y = Y \sin(\frac{2\pi}{\lambda} v t) = Y \sin \omega t $$, where $$\displaystyle \omega = 2\pi v / \lambda $$.
-
Amplitude of Vibration (relative to base):
$$Z = Y \cdot \frac{r^2}{\sqrt{(1-r^2)^2 + (2\zeta r)^2}}$$
- Critical Speed: $$\displaystyle v_{cr} = \frac{\lambda \omega_n}{2\pi} $$ (when $$\displaystyle r=1 $$).
5.4 Recoil Mechanisms (Gun Recoil)
-
Goal: Limit recoil distance $$\displaystyle x_{max} $$ and time.
-
Model: Gun mass $M$, recoil spring $k$, damper $c$. Initial recoil velocity $$\displaystyle v_0 $$.
-
Design for Critical Damping ($$\displaystyle \zeta=1 $$):
-
$$\displaystyle c_c = 2\sqrt{kM} $$.
-
Maximum recoil: $$\displaystyle x_{max} = v_0 / \omega_n $$ (since critically damped, no overshoot).
-
Given $$\displaystyle x_{max} $$ and $$\displaystyle v_0 $$ → $$\displaystyle \omega_n = v_0 / x_{max} $$ → $$\displaystyle k = M \omega_n^2 $$.
-
5.5 Whirling of Shafts
-
Definition: Lateral vibration of a rotating shaft due to unbalance, causing the shaft axis to whirl (rotate in a curved path).
-
Critical Speed ($$\displaystyle N_c $$): Rotational speed at which $\omega$ (rotational) = $$\displaystyle \omega_n $$ (natural frequency of transverse vibration).
-
Proof: For a simple rotor on massless shaft:
-
Unbalance force: $$\displaystyle F = m e \omega^2 $$.
-
EOM: $$\displaystyle m\ddot{x} + kx = m e \omega^2 \sin \omega t $$.
-
Steady-state amplitude: $$\displaystyle X = \frac{m e \omega^2 / k}{1 - (\omega/\omega_n)^2} $$.
-
As $$\displaystyle \omega \to \omega_n $$, $X \to \infty$ (theoretically). This speed is critical.
-
-
Factors Affecting $$\displaystyle N_c $$: Shaft length, diameter, material ($E$), support stiffness, rotor mass & eccentricity.
6.0 MULTI-DEGREE-OF-FREEDOM (MDOF) SYSTEMS - INTRODUCTION
6.1 Modeling & Coordinates
-
DOF: Number of independent coordinates needed.
-
Coordinates: Cartesian ($$\displaystyle x_1, x_2 $$), angular ($$\displaystyle \theta_1, \theta_2 $$), or modal (principal coordinates).
6.2 Formulation of EOM
-
Newton's 2nd Law (FBD): Draw free body diagram for each mass → write $$\displaystyle \sum F = m_i\ddot{x}_i $$. FREQUENT IN EXAMS.
DiagramCANVAS: 2-DOF spring-mass system with masses m1, m2, springs k1, k2, k3, coordinates x1, x2 -
Rayleigh's Energy Method: For approximate $$\displaystyle \omega_n $$ (usually fundamental). Assume mode shape → equate max KE & PE.
6.3 Matrix Form
$$\boxed{[M]\ddot{\mathbf{x}} + [C]\dot{\mathbf{x}} + [K]\mathbf{x} = \mathbf{F}(t)}$$
where:
-
$[M]$: Mass matrix (symmetric, positive definite).
-
$[K]$: Stiffness matrix (symmetric, positive semi-definite).
-
$[C]$: Damping matrix (often proportional: $$\displaystyle [C] = \alpha[M] + \beta[K] $$).
-
$\mathbf{x}, \mathbf{F}$: Displacement & force vectors.
6.4 Free Vibration of Undamped MDOF
EOM: $$\displaystyle [M]\ddot{\mathbf{x}} + [K]\mathbf{x} = \mathbf{0} $$.
Assume harmonic solution: $$\displaystyle \mathbf{x}(t) = \boldsymbol{\phi} \sin(\omega t + \phi) $$.
Substitute → Eigenvalue Problem:
$$\left([K] - \omega^2[M]\right)\boldsymbol{\phi} = \mathbf{0}$$
-
Natural Frequencies ($$\displaystyle \omega_i $$): Roots of characteristic equation $$\displaystyle \det([K] - \omega^2[M]) = 0 $$. For $n$-DOF, $n$ eigenvalues $$\displaystyle \omega_1^2 \le \omega_2^2 \le ... \le \omega_n^2 $$.
-
Mode Shapes ($$\displaystyle \boldsymbol{\phi}^{(i)} $$): Corresponding eigenvectors (shape of vibration at $$\displaystyle \omega_i $$). Defined up to a scale factor. Often mass-normalized: $$\displaystyle {\boldsymbol{\phi}^{(i)}}^T[M]\boldsymbol{\phi}^{(i)} = 1 $$.
-
Nodes: Points in a mode shape with zero displacement.
-
Orthogonality Properties:
-
Mass Orthogonality: $$\displaystyle {\boldsymbol{\phi}^{(i)}}^T[M]\boldsymbol{\phi}^{(j)} = 0 $$ for $i \neq j$.
-
Stiffness Orthogonality: $$\displaystyle {\boldsymbol{\phi}^{(i)}}^T[K]\boldsymbol{\phi}^{(j)} = 0 $$ for $i \neq j$.
-
Generalized Orthogonality: $$\displaystyle {\boldsymbol{\phi}^{(i)}}^T([K] - \omega_j^2[M])\boldsymbol{\phi}^{(j)} = 0 $$ for $i \neq j$.
-
Solution Methods:
-
Determinant (Characteristic Equation): For small $n$ (2-3 DOF). Solve $$\displaystyle \det([K] - \omega^2[M]) = 0 $$ for $$\displaystyle \omega_i $$, then find $$\displaystyle \boldsymbol{\phi}^{(i)} $$ from $$\displaystyle ([K] - \omega_i^2[M])\boldsymbol{\phi} = \mathbf{0} $$.
-
Matrix Iteration (Power Method): Finds fundamental ($$\displaystyle \omega_1 $$) mode shape. Start with guess vector $$\displaystyle \mathbf{x}_0 $$, compute $$\displaystyle \mathbf{x}_1 = [K]^{-1}[M]\mathbf{x}_0 $$, normalize, repeat → converges to $$\displaystyle \boldsymbol{\phi}^{(1)} $$.
-
Jacobi Method: For all eigenvalues/eigenvectors (conceptual).
6.5 Free Vibration Response via Mode Superposition
- Assumption: Total response is linear combination of normal modes:
$$\mathbf{x}(t) = \sum_{i=1}^{n} q_i(t) \boldsymbol{\phi}^{(i)}$$
where $$\displaystyle q_i(t) $$ are **generalized (modal) coordinates**.
- **Substitute into EOM:**利用 orthogonality → uncoupled equations:
$$\ddot{q}_i + \omega_i^2 q_i = 0 \quad \text{(undamped)}$$
Solution: $$\displaystyle q_i(t) = A_i \sin(\omega_i t) + B_i \cos(\omega_i t) $$.
- Determine $$\displaystyle A_i, B_i $$ from ICs: Expand initial conditions $$\displaystyle \mathbf{x}_0, \dot{\mathbf{x}}_0 $$ in terms of mode shapes:
$$\mathbf{x}_0 = \sum_{i=1}^{n} q_i(0) \boldsymbol{\phi}^{(i)} = [\boldsymbol{\Phi}]\mathbf{q}_0$$
where $$\displaystyle [\boldsymbol{\Phi}] = [\boldsymbol{\phi}^{(1)} \boldsymbol{\phi}^{(2)} ... \boldsymbol{\phi}^{(n)}] $$.
Multiply by $$\displaystyle {\boldsymbol{\phi}^{(j)}}^T[M] $$ → $$\displaystyle q_i(0) = \frac{{\boldsymbol{\phi}^{(i)}}^T[M]\mathbf{x}_0}{{\boldsymbol{\phi}^{(i)}}^T[M]\boldsymbol{\phi}^{(i)}} $$ (if mass-normalized, denominator=1).
6.6 Principal (Modal) Coordinates
-
Definition: Coordinates $$\displaystyle q_i(t) $$ that decouple the equations of motion. They represent the amplitude of each normal mode.
-
Determination: $$\displaystyle \mathbf{q} = [\boldsymbol{\Phi}]^{-1}\mathbf{x} $$ if $[\boldsymbol{\Phi}]$ is invertible (non-defective systems). For orthogonal modes, $$\displaystyle [\boldsymbol{\Phi}]^T[M][\boldsymbol{\Phi}] = [I] $$ (mass-normalized) → $$\displaystyle \mathbf{q} = [\boldsymbol{\Phi}]^T[M]\mathbf{x} $$.
-
Significance: In principal coordinates, each DOF vibrates independently at its own natural frequency.
7.0 SPECIAL TOPICS & NOISE FUNDAMENTALS
7.1 Fourier Series Expansion
-
Purpose: Represent periodic, non-sinusoidal forcing function $F(t)$ as sum of sinusoids.
-
Fourier Series (General):
$$F(t) = a_0 + \sum_{n=1}^{\infty} \left( a_n \cos n\omega_0 t + b_n \sin n\omega_0 t \right)$$
where $$\displaystyle \omega_0 = 2\pi/T $$ (fundamental frequency).
-
For Impact Force (e.g., forging hammer): Often represented by half-range sine/cosine series depending on symmetry.
DiagramCANVAS: Periodic impulse force F(t) with period T, showing its Fourier series approximation -
Response: System responds at each harmonic $$\displaystyle n\omega_0 $$ with amplitude $$\displaystyle X_n = \frac{F_n / k}{\sqrt{(1-r_n^2)^2 + (2\zeta r_n)^2}} $$, where $$\displaystyle r_n = n\omega_0 / \omega_n $$.
-
Total Steady-State Response: $$\displaystyle x(t) = \sum_{n=1}^{\infty} X_n \sin(n\omega_0 t - \phi_n) $$.
7.2 Torsional Vibration Systems
-
Elements: Rotational inertia $J$ (kg·m²), torsional stiffness $$\displaystyle k_t $$ (N·m/rad).
-
EOM for 2-DOF Torsional System:
DiagramCANVAS: Two disks with inertias J1, J2 connected by shaft with stiffness kt1, kt2, fixed at one end
$$J_1\ddot{\theta}_1 + k_{t1}(\theta_1 - \theta_2) = 0$$
$$J_2\ddot{\theta}_2 + k_{t1}(\theta_2 - \theta_1) + k_{t2}\theta_2 = 0$$
-
Matrix Form: $$\displaystyle [J]\ddot{\boldsymbol{\theta}} + [K_t]\boldsymbol{\theta} = \mathbf{0} $$.
-
Natural Frequencies: Solve $$\displaystyle \det([K_t] - \omega^2[J]) = 0 $$.
-
Mode Shapes: Displacement ratios $$\displaystyle \theta_1/\theta_2 $$ at each $$\displaystyle \omega_i $$.
7.3 Noise Fundamentals (Applied Vibration)
Sound Pressure Level (SPL)
- Definition: Logarithmic measure of sound pressure relative to reference.
$$L_p = 10 \log_{10}\left(\frac{p^2}{p_{ref}^2}\right) = 20 \log_{10}\left(\frac{p}{p_{ref}}\right) \quad \text{(dB)}$$
where $$\displaystyle p_{ref} = 20\ \mu\text{Pa} $$ (threshold of human hearing at 1 kHz).
-
Key Points:
-
0 dB SPL does not mean no sound; it means sound pressure equals $$\displaystyle p_{ref} $$.
-
6 dB increase ≈ doubling of sound pressure.
-
10 dB increase ≈ perceived doubling of loudness.
-
Sound Power Level (SWL)
- Definition: Logarithmic measure of total acoustic power radiated.
$$L_W = 10 \log_{10}\left(\frac{W}{W_{ref}}\right) \quad \text{(dB)}$$
where $$\displaystyle W_{ref} = 10^{-12}\ \text{W} $$.
- Difference from SPL: SPL is location-dependent (distance from source), SWL is source property (total power output). SWL is fixed for a machine; SPL varies with distance.
Inverse Square Law
-
For a point source in free field, sound intensity $$\displaystyle I \propto 1/r^2 $$, sound pressure $p \propto 1/r$.
-
Doubling distance: SPL decreases by 6 dB.
$$\Delta L_p = 20 \log_{10}\left(\frac{r_2}{r_1}\right)$$
If $$\displaystyle r_2 = 2r_1 $$, $$\displaystyle \Delta L_p = 20 \log_{10}(2) \approx 6\ \text{dB} $$.
Octave Band Analysis
-
Purpose: Break complex noise into frequency bands to identify dominant frequencies (source diagnosis).
-
Octave Band: Frequency range where upper limit = 2 × lower limit. Center frequencies: 31.5, 63, 125, 250, 500, 1000, 2000, 4000, 8000 Hz.
-
1/3 Octave Bands: More precise; each octave split into 3 bands (e.g., 1000 Hz octave → 800, 1000, 1250 Hz bands).
-
Output: SPL in each band → noise spectrum (plot of dB vs. band center freq).
Human Response & Hearing Conservation
-
Audible Range: 20 Hz – 20 kHz (decreases with age).
-
Damage Risk Criteria (Permissible Exposure Limits - PELs):
-
Based on A-weighting (dBA) to mimic human ear sensitivity.
-
OSHA/NIOSH standards: e.g., 90 dBA/8 hr (OSHA), 85 dBA/8 hr (NIOSH) with 3 dB exchange rate (NIOSH) or 5 dB (OSHA).
-
Exposure Time: Halved for every 3–5 dB increase above threshold.
-
-
Precautions & Remedies:
-
Source Control: Balance rotors, use quieter processes, isolate vibrations.
-
Path Control: Enclosures, barriers, acoustic lining, vibration isolation mounts.
-
Receiver Protection: Earplugs, earmuffs, administrative controls (shift rotation).
-
[!TIP] Exam Focus: Know SPL formula, reference pressure, 6 dB/doubling distance rule, octave band purpose, and basic hearing conservation (dBA, exposure time trade-off).