Flow Lab

Syllabus / Module 5: Second-Order Linear ODEs & Oscillations

Lesson 5.6

Forced Oscillations & Resonance

Beats, resonance, frequency response curves, phase lag, Q factor — and the Tacoma Narrows story (and what really happened).

Written byClaude Sonnet

🌊
Why this matters for simulation: Flutter — wind feeding energy into a wing's vibration — is a fluid–structure resonance that can destroy aircraft.

By the end of this lesson you can…

  • Derive the steady-state amplitude $A(\omega)$ and phase lag $\varphi(\omega)$ of a driven damped oscillator from $1/P(i\omega)$.
  • Locate the amplitude peak at $\omega_0\sqrt{1 - 2\zeta^2}$ and separate it from $\omega_0$ and $\omega_d$.
  • Compute $Q$ and the half-power bandwidth, and judge when $\Delta\omega \approx 2\zeta\omega_0$ is accurate.
  • Derive beats by sum-to-product, and see $t\sin\omega_0 t$ as an endless beat that damping caps on the time scale $1/(\zeta\omega_0)$.
  • Explain why Tacoma Narrows was flutter and not simple resonance.
On this page
·
  1. The driven oscillator
  2. The steady state, from the transfer function
  3. Where is the peak? Three different frequencies
  4. Quality factor and bandwidth
  5. Beats: the first picture of resonance
  6. Damping caps the growth
  7. Energy: power in, power out
  8. The same ω inside the integrator
  9. Tacoma Narrows: flutter, not resonance (from memory)
  10. Vortex shedding and lock-in (from memory)
  11. An honest limit: the spring is not linear
  12. Lab: curve, peak, Q, beats, ring-up and power (blocks A to F)
  13. Lab: step size, aliasing, a cubic spring, flutter and Strouhal (blocks G to K)
  14. Exercise: compute the resonance curve
  15. Problems
  16. Key Takeaways

A washing machine at spin-up, a wing in a gust, a cable in a crosswind: each is the damped oscillator of 5.3 Mechanical Vibrations: Spring–Mass–Damper pushed periodically. 5.4 Nonhomogeneous Equations: Undetermined Coefficients supplied the machinery (G=1/P(iω)G = 1/P(i\omega), the modification rule). The designer’s question: at which frequency does a push of fixed size do the most harm, how much, and how fast does it build? Each answer is checked against an integrator.

▶ Video Transient and steady state, the amplitude and phase curves, beats, growth at resonance, and why a bridge can fail without resonancecaptions only
Transcript
  1. A mass, spring and damper pushed by F zero cosine omega t: mass point four, stiffness ten, damping point four, force point four. So omega zero is five, zeta point one. Pushed at omega four from rest, a free part dies like e to the minus point five t, leaving the steady state.
  2. Divide the force by P of i omega. At r equals point eight, P over omega zero squared is point three six plus point one six i. The gain is one over its length, about two point five four; the lag is its angle, about point four two.
  3. Sweep the drive frequency. Each dot is an amplitude measured in a full simulation; the curve is the formula. Far below resonance the gain is one. At omega zero it is five, one over two zeta. Far above, it falls.
  4. Zoom in: three frequencies are easy to confuse. The forced amplitude peaks at about four point nine five, below the free ringing frequency, about four point nine seven, below omega zero. Damper loss grows with frequency, dragging the peak down. Peak gain: about five point oh two five.
  5. The lag rises from zero to pi, and is pi over two at omega zero for every damping. There spring and inertia cancel, so the damper alone opposes the push: c x prime equals F. Far above, inertia wins and x opposes F.
  6. Q is the gain at resonance, one over two zeta: five. The half-power width is where A squared falls to half its value at r equals one. For zeta point one it is about point two oh four, and two zeta is two percent off; at zeta point three, twenty-one percent.
  7. With no damping, a drive off resonance gives beats: a fast carrier in a slow envelope, by sum to product. First nodes: about fifteen point seven seconds at omega four point six, sixty-two point eight at four point nine. As omega reaches omega zero, the beat never ends.
  8. With damping, growth starts on the undamped line and levels off at point two metres: the envelope is point two times one minus e to the minus point five t. Measured against it, the ratio is about point eight six at two seconds, about one by six. Ninety-five percent first arrives at six point five six four.
  9. The force delivers F times velocity; the damper burns c times velocity squared; averages match. At omega four, about point oh three three watts. At omega zero velocity is in phase with the force: the curves coincide, and the mean is F zero squared over two c, point two watts. Stored over lost energy gives Q, five.
  10. Accuracy depends on h times omega zero. Below point five, doubling N cuts the RK4 error about sixteen times; at three point nine, it explodes. And a solver sees only samples: at h point two, a drive at about thirty-six point four looks like a drive at five, giving amplitude about point one seven, not point zero zero zero seven seven.
  11. From memory, Tacoma Narrows was flutter, not resonance: the wind force depends on the motion and can make damping negative, so the toy grows like e to the point one t with no periodic force. Vortex shedding is genuine forcing: St times U over D, fifty hertz here.
  12. An honest limit: the spring is not linear. With a cubic term the peak bends and the answer depends on history. At omega five the amplitude is about point four nine, not one. At five point four, the same push gives about point two six or about point eight.
  13. To recap: gain and lag; three frequencies; Q sets width and fill time; beats become capped growth; power balances; flutter is not resonance. Next: higher-order linear equations.

The driven oscillator

Divide mx′′+cx′+kx=F0cos⁡ωtmx'' + cx' + kx = F_0\cos\omega t by mm and use ω02=k/m\omega_0^2 = k/m, 2ζω0=c/m2\zeta\omega_0 = c/m:

x′′+2ζω0 x′+ω02 x=F0mcos⁡ωt.x'' + 2\zeta\omega_0\,x' + \omega_0^2\,x = \frac{F_0}{m}\cos\omega t .

One system runs through: m=0.4m = 0.4 kg, k=10k = 10 N/m, c=0.4c = 0.4 N s/m, F0=0.4F_0 = 0.4 N. So ω0=5\omega_0 = 5 rad/s, ζ=0.1\zeta = 0.1, F0/m=1F_0/m = 1, and the static deflection F0/k=0.04F_0/k = 0.04 m is the yardstick: the gain is A/(F0/k)A/(F_0/k).

By 5.1 Second-Order Linear Equations: Structure, x=xp+xhx = x_p + x_h: the free part xhx_h carries the initial conditions and dies like e−ζω0te^{-\zeta\omega_0 t}, leaving the steady state xpx_p.

The steady state, from the transfer function

With P(s)=s2+2ζω0s+ω02P(s) = s^2 + 2\zeta\omega_0 s + \omega_0^2 we have P(iω)=ω02[(1−r2)+2iζr]P(i\omega) = \omega_0^2\bigl[(1 - r^2) + 2i\zeta r\bigr] with r=ω/ω0r = \omega/\omega_0. The rule yp=F∣P∣cos⁡(ωt−φ)y_p = \tfrac{F}{|P|}\cos(\omega t - \varphi), φ=arg⁡P\varphi = \arg P, gives

 A(ω)=F0/k(1−r2)2+(2ζr)2,φ(ω)=atan2⁡(2ζr, 1−r2) \boxed{\,A(\omega) = \frac{F_0/k}{\sqrt{(1 - r^2)^2 + (2\zeta r)^2}}, \qquad \varphi(\omega) = \operatorname{atan2}\bigl(2\zeta r,\ 1 - r^2\bigr)\,}

since (F0/m)/ω02=F0/k(F_0/m)/\omega_0^2 = F_0/k. The imaginary part 2ζr2\zeta r is never negative, so φ\varphi runs from 00 to π\pi; a plain arctangent would hide the jump past r=1r = 1, where the real part changes sign.

  • Slow push, r≪1r \ll 1. The spring balances the force: x≈(F0/k)cos⁡ωtx \approx (F_0/k)\cos\omega t, gain near 11, lag near 00.
  • At r=1r = 1. Spring force and inertia cancel (ω02−ω2=0\omega_0^2 - \omega^2 = 0), so only the damper opposes the push: cx′=Fcx' = F. Then x′x' is in phase with FF, xx lags by π/2\pi/2 for every ζ\zeta, and the gain is 1/(2ζ)1/(2\zeta).
  • Fast push, r≫1r \gg 1. Inertia wins, mx′′≈Fmx'' \approx F, so x≈−F/(mω2)x \approx -F/(m\omega^2) opposes the force (lag near π\pi) and the gain falls like 1/r21/r^2.

Block A of the lab integrates the ODE with RK45 and fits acos⁡ωt+bsin⁡ωta\cos\omega t + b\sin\omega t to the tail:

rrω\omegagain, formulagain, measuredlag, formulalag, measured
0.40.4221.1851141.1851141.1851141.1851140.0949520.0949520.0949520.094952
0.80.8442.5383652.5383652.5383652.5383650.4182240.4182240.4182240.418224
1.01.0555.0000005.0000005.0000005.0000001.5707961.5707961.5707961.570796
1.11.15.55.53.2879803.2879803.2879803.2879802.3329432.3329432.3329432.332943
2.02.010100.3304090.3304090.3304090.3304093.0090413.0090413.0090413.009041

Matching the cos⁡\cos and sin⁡\sin parts for xp=acos⁡ωt+bsin⁡ωtx_p = a\cos\omega t + b\sin\omega t gives a 2×22\times2 system:

(pq−qp)(ab)=(F0/m0),p=ω02−ω2,  q=2ζω0ω.\begin{pmatrix} p & q \\ -q & p \end{pmatrix}\begin{pmatrix} a \\ b \end{pmatrix} = \begin{pmatrix} F_0/m \\ 0 \end{pmatrix}, \qquad p = \omega_0^2 - \omega^2,\ \ q = 2\zeta\omega_0\omega .

The matrix is ∣P∣|P| times a rotation by −φ-\varphi, so solving it shrinks the force vector by ∣P∣|P| and turns it by +φ+\varphi: (a,b)=A(cos⁡φ,sin⁡φ)(a, b) = A(\cos\varphi, \sin\varphi).

◆ Interactive Sweep the drive frequency through resonance and watch the measured amplitude and lag land on the formulas; switch to a single frequency (transient plus steady state) or to beats

Loading simulation…

In sweep mode each dot is measured from one full drive cycle; the dashed curves are the formulas. Slow the sweep and the dots sit on the curve; speed it up and they lag, since the oscillator needs about 1/(ζω0)1/(\zeta\omega_0) to adjust. In single-frequency mode xx is the steady state plus a fading transient.

Where is the peak? Three different frequencies

Minimise D(r)=(1−r2)2+4ζ2r2D(r) = (1 - r^2)^2 + 4\zeta^2r^2. With u=r2u = r^2, D=u2−2(1−2ζ2)u+1D = u^2 - 2(1 - 2\zeta^2)u + 1 has its minimum at u=1−2ζ2u = 1 - 2\zeta^2, a legal r2r^2 only for ζ<1/2\zeta \lt 1/\sqrt2. Then

ωpeak=ω01−2ζ2,Dmin⁡=4ζ2(1−ζ2),Amax⁡=F0/k2ζ1−ζ2.\omega_{\text{peak}} = \omega_0\sqrt{1 - 2\zeta^2}, \qquad D_{\min} = 4\zeta^2(1 - \zeta^2), \qquad A_{\max} = \frac{F_0/k}{2\zeta\sqrt{1 - \zeta^2}} .

For ζ≥1/2=0.7071\zeta \ge 1/\sqrt2 = 0.7071 the minimum is at r=0r = 0 and the curve only falls. Three frequencies are easy to confuse: ω0\omega_0, where the lag is π/2\pi/2; ωd=ω01−ζ2\omega_d = \omega_0\sqrt{1 - \zeta^2}, where the free oscillator rings; and ωpeak\omega_{\text{peak}}, where the forced amplitude is largest. Since 1−2ζ2<1−ζ2<11 - 2\zeta^2 \lt 1 - \zeta^2 \lt 1, ωpeak<ωd<ω0\omega_{\text{peak}} \lt \omega_d \lt \omega_0. The damper term 4ζ2r24\zeta^2r^2 grows with rr and penalises the high side, so the peak sits low. Block B finds the peak from the formula and as the argmax of a fine grid (in brackets):

ζ\zetaωpeak\omega_{\text{peak}}ωd\omega_dpeak gaingain at ω0\omega_0excess
0.10.14.949754.94975 [4.949704.94970]4.974944.974945.025195.025195.000005.000000.504%0.504\%
0.30.34.527694.52769 [4.527754.52775]4.769704.769701.747141.747141.666671.666674.828%4.828\%
0.50.53.535533.53553 [3.535503.53550]4.330134.330131.154701.154701.000001.0000015.470%15.470\%
0.70.70.707110.70711 [0.707100.70710]3.570713.570711.000201.000200.714290.7142940.028%40.028\%
0.80.8none3.000003.000001.000001.000000.625000.62500n/a

At ζ=0.1\zeta = 0.1 the peak is 0.5%0.5\% above the resonant gain and 1.0%1.0\% below ω0\omega_0; at ζ=0.5\zeta = 0.5 the gain at ω0\omega_0 has fallen to the static value; at ζ=0.7\zeta = 0.7 the peak is 0.02%0.02\% above the static deflection and resonance has effectively disappeared.

Quality factor and bandwidth

Define Q=A(ω0)/(F0/k)=1/(2ζ)Q = A(\omega_0)/(F_0/k) = 1/(2\zeta), the resonant gain. It is also exactly 2π×2\pi\times(stored energy)/(energy lost per cycle) at steady-state resonance (derived below); in a free ring-down the same ratio needs 4πζ≪14\pi\zeta \ll 1 (5.3 Mechanical Vibrations: Spring–Mass–Damper).

The half-power bandwidth is the width where A2A^2 (stored energy) is at least half its value at r=1r = 1: (1−r2)2+4ζ2r2=8ζ2(1 - r^2)^2 + 4\zeta^2r^2 = 8\zeta^2, a quadratic in u=r2u = r^2 with roots

r±2=1−2ζ2±2ζ1+ζ2,Δω=ω0(r+−r−)≈2ζω0=ω0Q.r_\pm^2 = 1 - 2\zeta^2 \pm 2\zeta\sqrt{1 + \zeta^2}, \qquad \Delta\omega = \omega_0(r_+ - r_-) \approx 2\zeta\omega_0 = \frac{\omega_0}{Q}.

The lower root exists only for ζ<1/(22)=0.3536\zeta \lt 1/(2\sqrt2) = 0.3536. Block C gives the exact width and the shortcut:

ζ\zetar−r_-r+r_+exact r+−r−r_+ - r_-2ζ2\zetashortcut error
0.050.050.9459780.9459781.0464821.0464820.1005040.1005040.10.1−0.50%-0.50\%
0.10.10.8826110.8826111.0867371.0867370.2041260.2041260.20.2−2.02%-2.02\%
0.30.30.4399790.4399791.2026711.2026710.7626920.7626920.60.6−21.33%-21.33\%

The shortcut is good to 2%2\% for ζ≤0.1\zeta \le 0.1 (Q≥5Q \ge 5) and poor at 0.30.3, because the half-power points are not equidistant from r=1r = 1. For absorbed power (derived below), proportional to (2ζr)2/[(1−r2)2+(2ζr)2](2\zeta r)^2/[(1 - r^2)^2 + (2\zeta r)^2], the half points solve 1−r2=±2ζr1 - r^2 = \pm2\zeta r, so r=1+ζ2∓ζr = \sqrt{1 + \zeta^2} \mp \zeta and the width is exactly 2ζ2\zeta; block C’s “power width” equals 2ζ2\zeta to six digits for all five ζ\zeta it tries (0.010.01 to 0.30.3).

Beats: the first picture of resonance

Drive an undamped oscillator from rest, ω≠ω0\omega \ne \omega_0: x′′+ω02x=fcos⁡ωtx'' + \omega_0^2x = f\cos\omega t, x(0)=x′(0)=0x(0) = x'(0) = 0, f=F0/mf = F_0/m. The particular solution is fω02−ω2cos⁡ωt\frac{f}{\omega_0^2 - \omega^2}\cos\omega t; subtracting fω02−ω2cos⁡ω0t\frac{f}{\omega_0^2 - \omega^2}\cos\omega_0 t makes x(0)=0x(0) = 0 and keeps x′(0)=0x'(0) = 0. By cos⁡A−cos⁡B=2sin⁡A+B2sin⁡B−A2\cos A - \cos B = 2\sin\frac{A + B}{2}\sin\frac{B - A}{2},

x(t)=fω02−ω2(cos⁡ωt−cos⁡ω0t)=2fω02−ω2 sin⁡(ω0−ω)t2 sin⁡(ω0+ω)t2.x(t) = \frac{f}{\omega_0^2 - \omega^2}\bigl(\cos\omega t - \cos\omega_0 t\bigr) = \frac{2f}{\omega_0^2 - \omega^2}\,\sin\frac{(\omega_0 - \omega)t}{2}\,\sin\frac{(\omega_0 + \omega)t}{2}.

A fast carrier at (ω0+ω)/2(\omega_0 + \omega)/2 sits inside an envelope of height 2f/∣ω02−ω2∣2f/|\omega_0^2 - \omega^2|. The sine factor oscillates at ∣ω0−ω∣/2|\omega_0 - \omega|/2, but the envelope is its absolute value, so the swell repeats every 2π/∣ω0−ω∣2\pi/|\omega_0 - \omega|. Mechanism: the driven oscillation and the free one that x(0)=0x(0) = 0 forces into existence start in opposition, drift in phase at rate ω0−ω\omega_0 - \omega, add to twice the driven amplitude after half a beat and cancel after a full one.

Block D integrates ζ=0\zeta = 0 and finds the envelope nodes (where x2+(x′/carrier)2x^2 + (x'/\text{carrier})^2 vanishes):

ω\omegaω0−ω\omega_0 - \omegacarrierfirst node2π/(ω0−ω)2\pi/(\omega_0 - \omega)max⁡∣x∣\max\lvert x\rvertbound 2f/∣ω02−ω2∣2f/\lvert\omega_0^2 - \omega^2\rvert
4.64.60.40.44.84.815.70815.70815.70815.7080.519720.519720.520830.52083
4.94.90.10.14.954.9562.83262.83262.83262.8322.020202.020202.020202.02020

The bound needs a carrier peak exactly at the envelope crest: true for 4.94.9, missed by 0.2%0.2\% for 4.64.6. As ω→ω0\omega \to \omega_0, 2sin⁡(Δt/2)/Δ→t2\sin(\Delta t/2)/\Delta \to t and

x→tsin⁡ω0t2ω0,x \to \frac{t\sin\omega_0 t}{2\omega_0},

the pure-resonance solution of 5.4 Nonhomogeneous Equations: Undetermined Coefficients: a beat that never ends. Here t/(2ω0)=t/10t/(2\omega_0) = t/10, and block D reads max⁡∣x∣=3.98987\max|x| = 3.98987 up to t=40t = 40 against the line value 44.

Damping caps the growth

For ζ>0\zeta \gt 0 and ω=ω0\omega = \omega_0 the steady state is f2ζω02sin⁡ω0t\frac{f}{2\zeta\omega_0^2}\sin\omega_0 t. Adding e−ζω0t(c1cos⁡ωdt+c2sin⁡ωdt)e^{-\zeta\omega_0 t}(c_1\cos\omega_dt + c_2\sin\omega_dt) and imposing x(0)=x′(0)=0x(0) = x'(0) = 0 gives c1=0c_1 = 0, c2ωd=−f/(2ζω0)c_2\omega_d = -f/(2\zeta\omega_0):

x(t)=f2ζω02[sin⁡ω0t−ω0ωde−ζω0tsin⁡ωdt].x(t) = \frac{f}{2\zeta\omega_0^2}\Bigl[\sin\omega_0 t - \frac{\omega_0}{\omega_d}e^{-\zeta\omega_0 t}\sin\omega_d t\Bigr].

Two nearly equal sinusoids subtract (a beat cut short by the decay), so for light damping the envelope is

Aenv(t)≈f2ζω02(1−e−ζω0t),A_{\text{env}}(t) \approx \frac{f}{2\zeta\omega_0^2}\bigl(1 - e^{-\zeta\omega_0 t}\bigr),

which starts as ft/(2ω0)ft/(2\omega_0), the undamped line, and levels off at the steady amplitude. The time scale is 1/(ζω0)=21/(\zeta\omega_0) = 2 s, which is 1/(2πζ)=Q/π=1.591/(2\pi\zeta) = Q/\pi = 1.59 cycles per ee-fold. Block E compares the largest ∣x∣|x| in each drive cycle with AenvA_{\text{env}}:

ttmeasuredenveloperatio
220.1085650.1085650.1264240.1264240.85870.8587
440.1644230.1644230.1729330.1729330.95080.9508
660.1899500.1899500.1900430.1900430.99950.9995
12120.1994240.1994240.1995040.1995040.99960.9996

The ratio is within about 0.1%0.1\% from t=6t = 6 on; at t=2t = 2 it is 14%14\% low because the amplitude is still growing fast during the cycle. The time to 95%95\% is ln⁡20/(ζω0)=5.991\ln20/(\zeta\omega_0) = 5.991 s; the first time ∣x∣\lvert x\rvert reaches 0.190.19 is 6.5646.564 s, later by less than the half-period 0.6280.628 s between peaks.

A high QQ therefore means a tall, narrow resonance that is slow to fill and to ring down, and a sweep faster than about (ζω0)2(\zeta\omega_0)^2 per second measures an oscillator that has not had time to respond.

Energy: power in, power out

In the steady state x=Acos⁡(ωt−φ)x = A\cos(\omega t - \varphi) and x′=−Aωsin⁡(ωt−φ)x' = -A\omega\sin(\omega t - \varphi). The force delivers Fx′Fx'. Because cos⁡ωt sin⁡(ωt−φ)\cos\omega t\,\sin(\omega t - \varphi) averages to −12sin⁡φ-\tfrac12\sin\varphi over a period,

⟨Fx′⟩=12F0Aωsin⁡φ,⟨cx′2⟩=12cω2A2.\langle Fx'\rangle = \tfrac12F_0A\omega\sin\varphi, \qquad \langle cx'^2\rangle = \tfrac12 c\omega^2A^2 .

They are equal: E=12mx′2+12kx2E = \tfrac12mx'^2 + \tfrac12kx^2 obeys E′=x′(F−cx′)E' = x'(F - cx'), which is periodic in the steady state, so its average is zero. With A=F0/(k−mω2)2+(cω)2A = F_0/\sqrt{(k - m\omega^2)^2 + (c\omega)^2},

Pabs=cω2A22=F022c⋅(cω)2(k−mω2)2+(cω)2≤F022c,P_{\text{abs}} = \frac{c\omega^2A^2}{2} = \frac{F_0^2}{2c}\cdot\frac{(c\omega)^2}{(k - m\omega^2)^2 + (c\omega)^2} \le \frac{F_0^2}{2c},

with equality exactly at ω0\omega_0. The mechanism is phase: at ω0\omega_0 the velocity is in phase with the force, which always pushes along the motion. Block F averages the simulated velocity over ten drive periods:

ω\omegaAA (m)φ\varphimean Fx′Fx' (W)mean cx′2cx'^2 (W)12F0Aωsin⁡φ\tfrac12F_0A\omega\sin\varphi12cω2A2\tfrac12c\omega^2A^2
440.1015350.1015350.4182240.4182240.0329890.0329890.0329900.0329900.0329900.0329900.0329900.032990
550.2000000.2000001.5707961.5707960.1999990.1999990.1999990.1999990.2000000.2000000.2000000.200000

and F02/(2c)=0.16/0.8=0.2F_0^2/(2c) = 0.16/0.8 = 0.2 W. At resonance the stored energy is constant, E=12kA2E = \tfrac12kA^2, and the loss per cycle is 0.2⋅2π/ω0=πcω0A20.2\cdot 2\pi/\omega_0 = \pi c\omega_0A^2, so 2πE/ΔE=k/(cω0)=1/(2ζ)2\pi E/\Delta E = k/(c\omega_0) = 1/(2\zeta) exactly; block F prints 5.0000005.000000.

The same ω inside the integrator

A simulation must resolve two frequencies: the drive ω\omega and the oscillator’s own ω0\omega_0, which the transient excites and which sets the stability limit. Block G integrates ζ=0.1\zeta = 0.1 over [0,40][0, 40] with RK4, NN steps per drive period (error: largest deviation from RK45 over the steady amplitude):

NNω=5\omega = 5: hω0h\omega_0, errorω=2\omega = 2: hω0h\omega_0, error
441.5711.571, 3.64×10−13.64\times10^{-1}3.9273.927, 2.60×10392.60\times10^{39}
880.7850.785, 3.09×10−23.09\times10^{-2}1.9631.963, 3.61×10−13.61\times10^{-1}
16160.3930.393, 2.00×10−32.00\times10^{-3}0.9820.982, 2.90×10−22.90\times10^{-2}
32320.1960.196, 1.25×10−41.25\times10^{-4}0.4910.491, 1.94×10−31.94\times10^{-3}
64640.0980.098, 7.76×10−67.76\times10^{-6}0.2450.245, 1.19×10−41.19\times10^{-4}

Once hω0<0.5h\omega_0 \lt 0.5 each doubling of NN cuts the error by about 1616 (4.1 Local Error, Global Error & Order). Steps per drive period is the wrong yardstick: at ω=2\omega = 2 the same NN gives a 2.52.5 times larger hh. At N=4N = 4, hω0=3.927h\omega_0 = 3.927 is past the RK4 stability limit (2.62.6 to 2.92.9 in the table of 5.3 Mechanical Vibrations: Spring–Mass–Damper) and the run explodes. Forward Euler with 10241024 steps per period reaches 3.18×10−23.18\times10^{-2}, what RK4 gets with N=8N = 8: 10241024 against 8×4=328\times4 = 32 force evaluations.

Aliasing. A solver reads the forcing only at its stage times, and cosines differing in frequency by 2π/h2\pi/h agree at every tn=nht_n = nh. Block H takes h=0.2h = 0.2 and a drive at ωa=5+2π/h=36.41593\omega_a = 5 + 2\pi/h = 36.41593, far above resonance, where the true response amplitude is 0.0007680.000768 m. Symplectic Euler (4.7 Energy & Symplectic Integrators) reads cos⁡ωatn=cos⁡5tn\cos\omega_at_n = \cos5t_n: its numbers differ from the ω=5\omega = 5 run by at most 2.7×10−142.7\times10^{-14}, and its amplitude 0.165960.16596 is a resonance that does not exist. RK4 also reads midpoints, aliasing differently but still wrongly: 0.068180.06818, about 8989 times the truth; at h=0.001h = 0.001 it gives 0.0007680.000768. Seeing the forcing at all needs h<π/ωa=0.0863h \lt \pi/\omega_a = 0.0863 (Nyquist).

Tacoma Narrows: flutter, not resonance (from memory)

The Tacoma Narrows Bridge opened in July 1940 and failed on 7 November 1940 in a steady wind (from memory; I give no figures I could not verify). The textbook story, that wind vortices matched the bridge’s natural frequency, is not what the later analyses I remember support: a steady wind has no periodic force at that frequency, and the deck showed a growing torsional oscillation.

The mechanism is a different equation. In resonance the force is given and independent of xx; in aeroelastic flutter the aerodynamic force depends on the motion itself, on twist and its rate. The part proportional to the rate acts like damping, and above a critical wind speed it has the wrong sign. Moved to the left side it gives x′′+2ζeffω0x′+ω02x=0x'' + 2\zeta_{\text{eff}}\omega_0x' + \omega_0^2x = 0 with ζeff=ζ−κ(U)\zeta_{\text{eff}} = \zeta - \kappa(U), κ\kappa growing with wind speed UU. When ζeff<0\zeta_{\text{eff}} \lt 0 the roots have positive real part: negative damping, exponential growth with no periodic force.

Block J is a toy, not Tacoma data: ζeff=−0.02\zeta_{\text{eff}} = -0.02, ω0=5\omega_0 = 5, x(0)=0.01x(0) = 0.01. The envelope grows like e0.1te^{0.1t} (ee-fold time 1010 s), by e6=403.43e^6 = 403.43 in 6060 s, and the largest ∣x∣|x| for t∈[50,60]t \in [50, 60] is 3.91593.9159. Resonance is capped at QQ times the static deflection; this has no linear cap.

Vortex shedding and lock-in (from memory)

Steady flow past a circular cylinder of diameter DD sheds vortices alternately at f=St U/Df = \mathrm{St}\,U/D, with Strouhal number St≈0.2\mathrm{St} \approx 0.2 over a wide range of Reynolds numbers (a rule of thumb from memory). Each vortex pushes the cylinder sideways, so the lateral force oscillates at ff: for a stiff structure this is this lesson’s forced oscillator. Block K: D=0.04D = 0.04 m, U=10U = 10 m/s give 50.050.0 Hz, and a 1212 Hz cable mode is excited at U=fnD/0.2=2.40U = f_nD/0.2 = 2.40 m/s.

Once the cylinder moves appreciably, the shedding (from memory) stays at the structure’s frequency over a range of UU instead of following St U/D\mathrm{St}\,U/D: lock-in, where the forcing depends on the response.

An honest limit: the spring is not linear

The curve assumed a Hooke spring, but real structures stiffen or soften with amplitude. Add a cubic term, x′′+2ζω0x′+ω02x+βx3=cos⁡ωtx'' + 2\zeta\omega_0x' + \omega_0^2x + \beta x^3 = \cos\omega t (a Duffing oscillator). For amplitude AA the effective frequency rises to about ω2≈ω02+34βA2\omega^2 \approx \omega_0^2 + \tfrac34\beta A^2 (a standard estimate, from memory; block I evaluates it), so the peak bends right, and where it overhangs there are two stable amplitudes at one drive frequency. Block I uses ζ=0.02\zeta = 0.02, β=10\beta = 10:

ω\omegastart from reststart from x(0)=1.2x(0) = 1.2linear formula
5.05.00.48920.48920.48920.48921.00001.0000
5.45.40.26360.26360.80260.80260.23270.2327

At the linear peak the true amplitude is under half the prediction, and at 5.45.4 the same force and frequency give 0.26360.2636 or 0.80260.8026 depending on history.

In the sim, sweep mode with ζ=0.02\zeta = 0.02, β=10\beta = 10 and sweep rate 0.0050.005 (readout rate ratio 0.5000.500, not infinitely slow): going up, “largest measured gain” reaches 22.21822.218 (A=0.8887A = 0.8887 m) at ω=5.610\omega = 5.610 and the amplitude then collapses; the estimate above gives ω02+34β(0.8887)2=5.5609\sqrt{\omega_0^2 + \tfrac34\beta(0.8887)^2} = 5.5609. Sweeping down, the largest is 19.13419.134 (A=0.7654A = 0.7654 m) at ω=5.195\omega = 5.195.

Lab: curve, peak, Q, beats, ring-up and power (blocks A to F)

🐍 PythonLab 1: steady-state curve, peak, Q and bandwidth, beats, ring-up, power balance
⌘/Ctrl + Enter

Lab: step size, aliasing, a cubic spring, flutter and Strouhal (blocks G to K)

🐍 PythonLab 2: RK4 steps per drive period, aliasing, a cubic spring, a self-excited toy and f = St U / D
⌘/Ctrl + Enter

Exercise: compute the resonance curve

Implement steady_state (solve the 2×22\times2 system), measured_amplitude (integrate to steady state and measure) and peak_frequency (the argmax of the curve).

🧪 Coding exerciseExercise: the amplitude curve from a 2x2 solve, from an integration, and its peak
⌘/Ctrl + Enter

Problems

✏️ Pen & paperA motor on rubber mounts●○○

A 1212 kg motor on mounts of stiffness 30003000 N/m and damping 6060 N s/m applies a vertical force 40cos⁡ωt40\cos\omega t N. (a) Find ω0\omega_0 and ζ\zeta. (b) At ω=10\omega = 10 rad/s find the steady amplitude and lag. (c) Find ωpeak\omega_{\text{peak}}, the peak amplitude, and the frequency above which the amplitude is below the static deflection.

Show worked solution

(a) ω0=3000/12=15.811388\omega_0 = \sqrt{3000/12} = 15.811388 rad/s, ζ=60/(212⋅3000)=0.158114\zeta = 60/(2\sqrt{12\cdot3000}) = 0.158114.

(b) r2=0.4r^2 = 0.4, ζr=0.1\zeta r = 0.1, so 1−r2=0.61 - r^2 = 0.6, 2ζr=0.22\zeta r = 0.2, gain 1/0.4=1.5811391/\sqrt{0.4} = 1.581139, and with F0/k=0.0133333F_0/k = 0.0133333 m, A=0.0210819A = 0.0210819 m. The lag is atan2⁡(0.2,0.6)=0.321751\operatorname{atan2}(0.2, 0.6) = 0.321751 rad.

(c) ωpeak=15.8113881−0.05=15.41104\omega_{\text{peak}} = 15.811388\sqrt{1 - 0.05} = 15.41104 rad/s, peak gain 1/(2ζ1−ζ2)=3.2025631/(2\zeta\sqrt{1 - \zeta^2}) = 3.202563, i.e. 42.7042.70 mm. The gain is 11 when r4−2r2+4ζ2r2=0r^4 - 2r^2 + 4\zeta^2r^2 = 0, i.e. r2=2−4ζ2=1.9r^2 = 2 - 4\zeta^2 = 1.9, r=1.378405r = 1.378405, ω=21.7945\omega = 21.7945 rad/s.

✏️ Pen & paperReading Q from a shaker test●●○

A model wing shaken with constant force amplitude moves 0.20.2 mm at very low frequency and 5.05.0 mm at resonance, f0=40f_0 = 40 Hz. (a) Estimate QQ and ζ\zeta. (b) Find the exact half-power points of A2A^2 and compare with 2ζf02\zeta f_0. (c) Driven from rest at 4040 Hz, how long until the envelope reaches 90%90\% of steady, and how many cycles is that?

Show worked solution

(a) Q=5.0/0.2=25Q = 5.0/0.2 = 25, so ζ=0.02\zeta = 0.02.

(b) r±2=1−2ζ2±2ζ1+ζ2r_\pm^2 = 1 - 2\zeta^2 \pm 2\zeta\sqrt{1 + \zeta^2} gives r−=0.979383r_- = 0.979383, r+=1.019416r_+ = 1.019416, i.e. 39.17539.175 and 40.77740.777 Hz: a width of 1.60131.6013 Hz against 2ζf0=1.62\zeta f_0 = 1.6 Hz.

(c) ζω0=0.02⋅2π⋅40=5.026548\zeta\omega_0 = 0.02\cdot2\pi\cdot40 = 5.026548 s−1^{-1}, and 1−e−ζω0t=0.91 - e^{-\zeta\omega_0t} = 0.9 gives t=ln⁡10/5.026548=0.458085t = \ln10/5.026548 = 0.458085 s, which is 18.318.3 cycles.

✏️ Pen & paperBeats, then damping●●●

An undamped oscillator x′′+64x=3cos⁡7.8tx'' + 64x = 3\cos7.8t starts at rest. (a) Solve it. (b) Write it as a product of sines and find the largest possible amplitude, the beat period and the carrier cycles per beat. (c) Add damping ζ=0.02\zeta = 0.02 (x′′+0.32x′+64x=3cos⁡7.8tx'' + 0.32x' + 64x = 3\cos7.8t): find the steady amplitude and lag, and the time for the transient to fall to 5%5\% of its start. How much of one beat appears?

Show worked solution

(a) ω02−ω2=64−60.84=3.16\omega_0^2 - \omega^2 = 64 - 60.84 = 3.16 and 3/3.16=0.9493673/3.16 = 0.949367, so x=0.949367(cos⁡7.8t−cos⁡8t)x = 0.949367(\cos7.8t - \cos8t).

(b) x=1.898734 sin⁡(0.1t)sin⁡(7.9t)x = 1.898734\,\sin(0.1t)\sin(7.9t), with largest amplitude at most 1.8987341.898734. The envelope ∣sin⁡0.1t∣|\sin0.1t| repeats every π/0.1=31.4159\pi/0.1 = 31.4159 s, and the carrier cycles per beat are 7.9/0.2=39.57.9/0.2 = 39.5.

(c) P(7.8i)=3.16+2.496iP(7.8i) = 3.16 + 2.496i, ∣P∣=4.026861|P| = 4.026861, so the steady amplitude is 3/4.026861=0.7449973/4.026861 = 0.744997 and the lag is atan2⁡(2.496,3.16)=0.668536\operatorname{atan2}(2.496, 3.16) = 0.668536 rad. The transient falls as e−0.16te^{-0.16t}, to 5%5\% at ln⁡20/0.16=18.7233\ln20/0.16 = 18.7233 s, about 60%60\% of the beat. Beats are a transient here: the long-time response is 0.7450.745, 39%39\% of the undamped maximum.

✅ Concept check

For a damped oscillator with $0 \lt \zeta \lt 1/\sqrt2$, which ordering of the forced-amplitude peak $\omega_{\text{peak}}$, the free ringing frequency $\omega_d$ and $\omega_0$ is correct?

✅ Concept check

A structure's resonant gain (amplitude at resonance divided by the static deflection) is measured to be 40. What is its damping ratio ζ?

✅ Concept check

Why is the Tacoma Narrows failure better described as flutter than as simple resonance?

Key Takeaways

  • The steady state has gain 1/(1−r2)2+(2ζr)21/\sqrt{(1 - r^2)^2 + (2\zeta r)^2} in units of F0/kF_0/k and lag atan2⁡(2ζr,1−r2)\operatorname{atan2}(2\zeta r, 1 - r^2), exactly π/2\pi/2 at r=1r = 1.
  • The amplitude peaks at ω01−2ζ2\omega_0\sqrt{1 - 2\zeta^2} (only for ζ<1/2\zeta \lt 1/\sqrt2), below ωd\omega_d and ω0\omega_0; the gain at ω0\omega_0 is Q=1/(2ζ)Q = 1/(2\zeta).
  • The half-power width has r±2=1−2ζ2±2ζ1+ζ2r_\pm^2 = 1 - 2\zeta^2 \pm 2\zeta\sqrt{1 + \zeta^2}; 2ζω02\zeta\omega_0 is 2%2\% low at ζ=0.1\zeta = 0.1, 21%21\% low at 0.30.3, and exact for absorbed power 12cω2A2\tfrac12c\omega^2A^2.
  • Undamped forcing from rest gives beats (repeat 2π/∣ω0−ω∣2\pi/|\omega_0 - \omega|); ω→ω0\omega \to \omega_0 gives tsin⁡ω0t/(2ω0)t\sin\omega_0t/(2\omega_0), capped by damping on the time scale 1/(ζω0)1/(\zeta\omega_0).
  • Tacoma Narrows was flutter, a self-excited instability from negative effective damping (from memory); vortex shedding at f=St U/Df = \mathrm{St}\,U/D is a genuine forcing that can lock in.
  • Numerically, hω0h\omega_0 and hωh\omega both matter, forcing sampled at hh cannot tell ω\omega from ω+2π/h\omega + 2\pi/h, and a cubic spring bends the peak and makes it history-dependent.
Finished this lesson?

Mark it complete to track your progress (saved in this browser). Reading every section does this for you.

Where this leads

These later lessons build directly on this one: