35. Vibrations
## Solving Vibration Problems: A Systematic Approach
Vibrations are oscillatory motions about an equilibrium point, common in mechanical structures, vehicles, and electronic systems. Solving a vibration problem means determining how a system responds to initial conditions or external forces over time. The process follows a structured path: idealization, mathematical modeling, solving the equation of motion, and interpreting results.
### Step 1: Idealize the Physical System
Real systems are complex. You must create a simplified model that captures essential vibratory behavior. Common idealizations include:
- **Lumped parameter models**: Assume mass is concentrated at points, springs are massless and linear, and dampers provide viscous force proportional to velocity.
- **Degrees of freedom (DOF)** : The minimum number of independent coordinates needed to describe motion. A mass on a spring is single-DOF; a beam vibrating in multiple shapes requires many DOFs.
For a typical single-DOF system, you identify:
- Mass ( m ) (stores kinetic energy)
- Spring stiffness ( k ) (stores potential energy)
- Damper coefficient ( c ) (dissipates energy)
- External force ( F(t) ) (if any)
### Step 2: Derive the Equation of Motion
Using Newton’s second law or energy methods (Lagrange’s equation), you obtain a differential equation. For a forced, damped single-DOF system:
[
mddot{x} + cdot{x} + kx = F(t)
]
Here ( x(t) ) is displacement from equilibrium. This is a linear, second-order ODE with constant coefficients – the bedrock of vibration analysis.
### Step 3: Solve the Homogeneous Equation (Free Vibration)
Set ( F(t)=0 ) to study natural behavior. Assume a solution ( x(t)=e^{lambda t} ), leading to the characteristic equation:
[
mlambda^2 + clambda + k = 0
]
Solutions depend on the damping ratio ( zeta = frac{c}{2sqrt{mk}} ):
- **Underdamped** (( zeta < 1 )): Oscillatory decay. Solution:
[
x(t)=e^{-zetaomega_n t}left(Acos(omega_d t)+Bsin(omega_d t)right)
]
where ( omega_n=sqrt{k/m} ) (natural frequency) and ( omega_d=omega_nsqrt{1-zeta^2} ). Constants A, B come from initial displacement ( x_0 ) and velocity ( v_0 ).
- **Critically damped** (( zeta = 1 )): Fastest return to zero without oscillation.
- **Overdamped** (( zeta > 1 )): Slow, non-oscillatory decay.
### Step 4: Solve the Non-Homogeneous Equation (Forced Vibration)
For harmonic force ( F(t)=F_0cos(omega t) ), the total solution is homogeneous + particular. The particular solution is:
[
x_p(t)=Xcos(omega t - phi)
]
where amplitude ( X = frac{F_0/k}{sqrt{(1-r^2)^2+(2zeta r)^2}} ) and phase ( phi = tan^{-1}left(frac{2zeta r}{1-r^2}right) ), with frequency ratio ( r=omega/omega_n ).
**Resonance** occurs when ( rapprox1 ), causing large amplitudes if damping is low. In practice, design avoids operating near resonance.
For non-periodic forces (step, impulse, arbitrary), use **Duhamel’s integral** (convolution) or Laplace transforms. The impulse response function ( h(t)=frac{1}{momega_d}e^{-zetaomega_n t}sin(omega_d t) ) lets you compute response to any force via convolution.
### Step 5: Apply Initial Conditions
After finding the general solution (homogeneous + particular), use initial displacement and velocity to determine the unknown constants. This yields the complete time history ( x(t) ).
### Step 6: Interpret Results and Extract Key Metrics
Solving gives ( x(t) ), but engineering questions require specific metrics:
- **Natural frequencies** ( omega_n ): Avoid matching excitation frequencies.
- **Damping ratio** ( zeta ): Measure decay rate. From logarithmic decrement: ( zeta = frac{delta}{sqrt{4pi^2+delta^2}} ), where ( delta=ln(x_n/x_{n+1}) ).
- **Transmissibility**: For vibration isolation, ratio of transmitted force to input force. Isolators work when ( omega/omega_n > sqrt{2} ).
- **Frequency response function (FRF)** : Plot ( X/F_0 ) vs. frequency – peaks indicate natural frequencies and damping.
### Advanced Considerations
- **Multi-DOF systems**: Lead to coupled ODEs, solved via matrix methods. Write ( [M]{ddot{x}}+[C]{dot{x}}+[K]{x}={F(t)} ). Find natural frequencies by solving ( det(K-omega^2M)=0 ). Normal modes decouple the system.
- **Continuous systems**: Beams, strings, rods have infinite DOFs. Solve partial differential equations (wave equation) with boundary conditions. Solutions are series of mode shapes multiplied by time-dependent modal coordinates.
- **Nonlinear vibrations**: If stiffness or damping depends on displacement (e.g., large deflections), the ODE becomes nonlinear. Use perturbation methods, numerical integration (Runge-Kutta), or phase plane analysis.
### Practical Problem-Solving Workflow
1. **Draw a free-body diagram** – include all inertia, spring, damper, and external forces.
2. **Apply Newton’s laws** (or Lagrange’s equation) to derive the ODE.
3. **Identify system parameters** (m, c, k) from geometry and material properties. For example, a cantilever beam’s stiffness ( k=3EI/L^3 ).
4. **Classify the problem**: Free or forced? Damped or undamped? Harmonic or arbitrary force?
5. **Select solution method**:
- Free, underdamped → standard exponential/trig solution.
- Harmonic force → use frequency response formula.
- Transient force → Laplace or convolution.
- Multi-DOF → modal analysis or numerical integration (e.g., Newmark’s method).
6. **Apply initial/boundary conditions** to find constants.
7. **Compute engineering outputs**: maximum displacement (for clearance), maximum acceleration (for comfort), force transmitted to support (for isolation), stress (for fatigue life).
8. **Validate**: Does response decay? Does resonance occur near predicted ( omega_n )? Are amplitudes physically reasonable?
### Common Pitfalls to Avoid
- Forgetting units – stiffness in N/m, mass in kg, frequency in rad/s (convert Hz by ( omega=2pi f )).
- Confusing static deflection with dynamic amplification.
- Ignoring damping when it’s small but crucial near resonance.
- Misidentifying DOFs – e.g., a rotating machine may have both translational and rocking modes.
- Using linear models when clear nonlinearity exists (loose joints, large rotations).
### Conclusion
Solving vibrations is a blend of physics, mathematics, and engineering judgment. Start by idealizing the system, derive the governing ODE, solve analytically where possible (free, harmonic, or impulse excitation), and use numerical methods for complex cases. Always relate the mathematical solution back to physical parameters: natural frequency, damping ratio, and mode shapes. Mastery comes from practice – analyze a simple mass-spring, then a car suspension, then a building during an earthquake. With this systematic approach, you can predict and control vibration in any mechanical system.