
E12 Linear Physical Systems Analysis
April 9, 2026
Warning
This lecture, due to technical issues, is best viewed in the ‘slides’ format available by clicking the RevealJS button on the right. You can also click here.


Recall that overdamped means \(\zeta > 1\) where \[\zeta = \frac{b}{2\sqrt{m k}}\]

\[ \begin{aligned} m &= 50 \\ b &= 100.5 \\ k &= 1 \end{aligned} \]
Differential equation of the system \[50 \ddot{x} + 100.5 \dot{x} + x = f(t)\]
Transfer function: \[T(s) = \frac{X(s)}{F(s)} = \frac{1}{50s^2 + 100.5s+1}\]
Using the roots of the denominator, we can re-write as \[T(s) = \frac{X(s)}{F(s)} = \frac{1}{(100s+1)(0.5s+1)}\]
The (magnitude) Bode Plot needs the complex number \(\boxed{T(i\omega)}\) \[\left| T(i\omega) \right| = \left| \frac{1}{(100 i \omega +1)(0.5 i \omega + 1)}\right| = \frac{1}{\sqrt{\left( (100 \omega)^2 + 1 \right)\left( (0.5 \omega)^2 + 1 \right)}}\]


Suppose \(m = 2, b = 14, k = 20\). Recall that \(\zeta = \frac{b}{2\sqrt{m k}}\).
Is this system overdamped? Yes! \(\zeta = \frac{14}{2 \sqrt{2 \times 20}} \approx 1.1\)
Given the diff. eq. \(\quad 2 \ddot{x} + 14 \dot{x} + 20 x = f(t), \quad\) write the transfer function for this system in the form \(\frac{...}{(\tau_1 s + 1)(\tau_2 s + 1)}\)
\(T(s) = \frac{X}{F} = \frac{1}{2 s^2 + 14s +20}\) \(= \frac{1}{2(s+2)(s+5)}\) \(= \frac{0.05}{(0.5s+1)(0.2s+1)}\)
So the corner frequencies of this system are \(1/\tau_1 = 1/0.5 = 2\) and \(1/\tau_2 = 1/0.2 = 5\)

Setting \(m = k = 1\) and varying only \(b\) in the range such that \(\zeta >= 1\):





\(\displaystyle T(s) = \frac{X(s)}{F(s)} = \frac{1}{m s^2 + b s + k}\)
\(\displaystyle T(s) = \frac{k X(s)}{F(s)} = \frac{k }{m s^2 + b s + k}\) \(= \frac{k/k}{(m/k) s^2 + (b/k) s + (k/k)}\)
\(\displaystyle T(s) = \frac{1}{(m/k) s^2 + (b/k) s + 1} = \frac{1}{\left(\frac{s}{\omega_n}\right)^2 + 2 \zeta (s/\omega_n) + 1}\)
\(\displaystyle T(s) = \frac{k X(s)}{F(s)} = \frac{\omega_n^2}{s^2 + 2 \zeta \omega_n s + \omega_n^2}\)
\(T(i\omega) = \frac{\omega_n^2}{(i \omega)^2 + 2 \zeta \omega_n (i \omega) + \omega_n^2}\) \(\phantom{T(i\omega) }= \frac{\omega_n^2}{- \omega^2 + \omega_n^2 + 2 \zeta \omega_n \omega i}\)
\(T(i\omega) = \frac{\omega_n^2}{- \left( \frac{\omega}{\omega_n} \right)^2 \cdot \omega_n^2 + \omega_n^2 + 2 \zeta \omega_n \left( \frac{\omega}{\omega_n} \right) \cdot \omega_n i }\)
\(\phantom{T(i\omega)} = \frac{\omega_n^2}{\omega_n^2 \left(1 - r^2 \right) + 2 \zeta r \omega_n^2 i}\) \(= \frac{1}{\left(1 - r^2 \right) + 2 \zeta r i}\)
The magnitude of this complex number is \(\frac{1}{\sqrt{(1-r)^2 + 4 \zeta^2 r^2}}\)
The (magnitude) Bode plot for a 2nd order system \(T(s) = \frac{\omega_n^2}{s^2 + 2 \zeta \omega_n s + \omega_n^2}\) can therefore be represented as \[|T(i \omega)| = \frac{1}{\sqrt{(1-r^2)^2 + 4 \zeta^2 r^2}}\]










Let’s zoom in to the Bode plot at small damping ratios.









Three frequencies inherent to a second-order system.
The undamped natural frequency \[\omega_n = \sqrt{k/m}\] at which the system would hypothetically oscillate if there were no damping.
The damped natural frequency \[\omega_d = \omega_n \sqrt{1-\zeta^2}\] at which an underdamped system actually oscillates.
The resonant frequency \[\omega_r = \omega_n \sqrt{1-2\zeta^2}\] at which, if we provide an input to the system, the system’s oscillations will grow with time.

Predict the frequency with which the spring-mass system from Tuesday’s class should be oscillated to achieve resonance. Give your answer in Hertz.
Note that each mass was 200 grams, and the spring is listed as having a spring constant of \(0.2~\mathrm{lbf}/\mathrm{in}\).
E12 • Spring 2026 • Lecture 22 • April 9, 2026