
Physics I course sequence · Interactive lesson
Driven Oscillations & Resonance
This is where an oscillator meets the world outside it: a motor, a crowd on a footbridge, a vibrating floor, pushing again and again. You will predict how hard a system shakes at any drive frequency, read its phase lag as the sign that energy is flowing in, and use Q to decide whether to exploit a resonance or design it out.
UPI-14.08 · Lesson 8 of 8 in this unit. Position is not a completion score.Build the model
The drive picks the frequency; the system picks the amplitude and the lag.
The key idea
Resonance is the large steady-state response of a damped oscillator to a periodic force whose frequency lies close to the natural frequency ω₀ = √(k/m); the drive sets the frequency of the motion, the system sets its amplitude and lag.
Example. A 0.20 kg mass on an 80 N m⁻¹ spring with b = 0.40 kg s⁻¹ has ω₀ = 20 rad s⁻¹ and Q = 10: a steady 0.40 N pull stretches the spring 5.0 mm, but 0.40 N oscillating at 20 rad s⁻¹ drives the mass 50 mm either way, lagging the force by a quarter cycle.
The complete physical model and assumptions
Push a damped oscillator with a force F₀ cos ωt and Newton's second law gains one term: m d²x/dt² + b dx/dt + kx = F₀ cos ωt. The equation is linear, so its solution is a sum: the free damped motion of the previous lesson, which fades on the time scale 2m/b, plus a steady oscillation at the drive frequency ω whose amplitude and lag are fixed by the system, not by how it started. Both depend on a contest between the net stiffness k − mω², spring against inertia, and the drag term bω. At ω₀ = √(k/m) the net stiffness vanishes, drag alone limits the motion, the velocity falls into step with the force, and the mean power delivered peaks. One number, Q = mω₀/b, then fixes three things at once: the resonant amplitude is Q times the static stretch F₀/k, the strong response fills a band ω₀/Q wide, and settling into or out of the steady state takes 2Q/ω₀. Sharp means slow, and that coupling is why resonance is both an instrument and a hazard. The model ends where linearity does: large swings stiffen many springs and soften every pendulum, so the peak leans over and the amplitude can jump between two values, and a force that depends on the motion itself, as in wind-driven flutter, is not an external drive at all.
1Linearity splits the motion into transient and steady state
Add a periodic push to the damped oscillator and the equation m d²x/dt² + b dx/dt + kx = F₀ cos ωt stays linear, so every solution is the sum of two pieces. One is the free damped motion of the previous lesson: it rings at ω′, its size is set by how the motion started, and its envelope decays as exp(−t/τ) with τ = 2m/b. The other is a particular solution that oscillates at the drive frequency ω, with an amplitude and lag fixed by m, b, k, F₀ and ω alone. For the 0.20 kg mass with b = 0.40 kg s⁻¹, τ = 1.0 s: driven at ω₀ from rest, its amplitude closes the gap to the steady value by a factor e every second, reaching 95% of it after 3 s, and after 5 s the transient is below 1% of its starting size. Until then the two pieces add, and if ω differs from ω′ their sum swells and fades at the difference frequency, beats that die with the transient. Everything that follows is about the piece that stays.
2Match the coefficients and the forces close a triangle
Try x = A cos(ωt − δ) and write θ = ωt − δ. Then kx = kA cos θ, m d²x/dt² = −mω²A cos θ and b dx/dt = −bωA sin θ. Expand the drive about the same angle: F₀ cos ωt = F₀ cos(θ + δ) = F₀ cos δ cos θ − F₀ sin δ sin θ. The equation must hold at every instant, and cos θ and sin θ are independent functions of time, so their coefficients match separately: (k − mω²)A = F₀ cos δ and bωA = F₀ sin δ. Those are the two legs of a right triangle whose hypotenuse is F₀, the picture drawn in the figure. Squaring and adding gives A = F₀/√((k − mω²)² + (bω)²); dividing gives tan δ = bω/(k − mω²). The across leg is set by the net stiffness, the spring constant less the mω² needed to accelerate the mass, and it changes sign at ω₀. The up leg is drag, which is never negative, so δ runs from 0 to π and never beyond.
3Spring, drag and inertia each take a turn
The denominator names which term controls the motion. Take the 0.20 kg mass on the 80 N m⁻¹ spring with b = 0.40 kg s⁻¹, pushed with F₀ = 0.40 N, whose static stretch is F₀/k = 5.0 mm. At ω = 10 rad s⁻¹ the net stiffness is 80 − 0.20 × 100 = 60 N m⁻¹ against a drag term of only 4.0 N m⁻¹, so A = 0.40/√(60² + 4.0²) = 6.7 mm and δ = 3.8°: the spring rules, and the mass moves almost in step with the push. At ω₀ = 20 rad s⁻¹ the net stiffness is zero, spring and inertia cancelling exactly, and drag alone limits the motion: A = 0.40/(0.40 × 20) = 50 mm with δ = 90°. At ω = 40 rad s⁻¹ the net stiffness is 80 − 320 = −240 N m⁻¹, so inertia wins: A = 1.7 mm, a third of the static stretch, and δ = 176°, the mass moving against the push. The lags at ω₀/2 and 2ω₀ sit the same 3.8° from 0° and 180°. That is general: δ at ω₀²/ω equals 180° minus δ at ω, for any Q.
4Energy enters through the velocity, so power peaks at ω₀
The drive delivers power P = Fv. The velocity v = −ωA sin(ωt − δ) leads the displacement by a quarter cycle, so averaging F₀ cos ωt × v over a cycle gives P̄ = ½ F₀ωA sin δ. With bωA = F₀ sin δ this becomes P̄ = ½ bω²A², exactly the mean rate at which drag removes energy: in steady state every joule supplied is dissipated. Written as P̄ = (F₀²/2b) sin²δ, it peaks where δ = 90°, at ω₀ whatever b is. For the running example the peak is 0.16/0.80 = 0.20 W; at 10 rad s⁻¹ the drive delivers only sin²(3.8°) = 0.44% of it, although the amplitude there is 13% of the resonant value. Power halves where δ = 45° or 135°, where |k − mω²| = bω: for the example 19.0 and 21.0 rad s⁻¹, a band b/m = 2.0 rad s⁻¹ wide whose geometric mean is ω₀. The displacement peak sits slightly lower, at ω₀√(1 − 1/(2Q²)) = 19.95 rad s⁻¹; only at small Q do the power peak, the displacement peak and ω′ visibly separate.
5One number, Q, sets height, width and waiting time
With Q = mω₀/b, three results that look separate become one. Height: at ω₀, A = F₀/(bω₀) = Q × F₀/k, so a matched drive multiplies the static stretch by Q; for the example, 10 × 5.0 mm = 50 mm. Width: the half-power band is Δω = b/m = ω₀/Q. Time: the transient, and with it the build-up from rest, decays with τ = 2m/b = 2Q/ω₀, which is Q/π periods. Multiply width by time and Q cancels, Δω × τ = 2: a sharp resonance is necessarily a slow one. A 440 Hz tuning fork with Q ≈ 1000, a representative value, responds strongly only within a band 0.44 Hz wide and takes τ ≈ 0.72 s to ring up or down. Q also counts energy: in steady state at ω₀, Q = 2π × (energy stored) ÷ (energy dissipated per cycle), so the example oscillator, with Q = 10, loses and is handed back 63% of its stored energy every cycle.
6Use the peak, or keep well away from it
Resonance lets a small periodic force produce a large, frequency-selective motion, and that cuts both ways. Tuning forks, quartz clock crystals and ultrasonic cleaners are driven at ω₀ on purpose, so that a weak drive gives a strong response in a narrow band. The same arithmetic is a hazard when the drive is unwanted. A pump running near the ω₀ of its mounting shakes up to Q times harder than its static load suggests, and a washing-machine drum crosses the resonance of its suspension on every spin-up, where a quick crossing helps because the motion needs a time of order τ to build. The formula names the cures: detune by changing k or m so the operating frequency sits far from ω₀; add damping, which lowers the peak in proportion to Q and widens the band as ω₀/Q; or run well above ω₀, where A falls as F₀/(mω²). London's Millennium Bridge, which swayed under crowds when it opened in 2000, was fixed with dampers. Not every famous failure fits the model: the 1940 Tacoma Narrows collapse was aeroelastic flutter, a wind force that depends on the deck's own motion.
Key equations
Read the units, assumptions and limits before substituting.
Driven, damped equation of motion
m d²x/dt² + b dx/dt + kx = F₀ cos ωt
Linear, so its solution is the decaying free motion plus a steady oscillation at ω. After a few τ = 2m/b only the steady part is left.
m in kg, b in kg s⁻¹, k in N m⁻¹, force amplitude F₀ in N, drive angular frequency ω in rad s⁻¹, x in mSteady-state amplitude
x = A cos(ωt − δ), A = F₀ ÷ √((k − mω²)² + (bω)²)
Far below ω₀ it gives the static stretch F₀/k; far above, F₀/(mω²) as inertia takes over; at ω₀ only bω is left to limit the motion.
A in m; k − mω² is the net stiffness and bω the drag term, both in N m⁻¹Phase lag of displacement behind force
tan δ = bω ÷ (k − mω²), with 0 ≤ δ ≤ π
δ is exactly π/2 at ω₀ whatever the damping, so a measured 90° lag locates ω₀ more sharply than the amplitude peak does.
δ in rad; take δ between π/2 and π once k − mω² turns negative, above ω₀Mean power delivered by the drive
P̄ = ½ bω²A² = (F₀² ÷ 2b) sin²δ
The peak F₀²/(2b) sits at ω₀ for any b, and sin²δ says what fraction of it a given drive frequency collects; it halves at δ = 45° and 135°.
P̄ in W, averaged over one cycle; in steady state it equals the mean rate of loss to dragQuality factor: height and width
Q = mω₀/b, A(ω₀) = Q F₀/k, Δω = b/m = ω₀/Q
The factor that multiplies the static stretch at resonance is the same factor that narrows the band in which the drive is effective.
Q dimensionless; Δω in rad s⁻¹ is the full width between the two half-power frequenciesSettling time and bandwidth
τ = 2m/b = 2Q/ω₀, so Δω × τ = 2
Sharpness costs time. Halve the bandwidth and the oscillator takes twice as long to settle into, or out of, its steady motion.
τ in s: the time for the transient, and so the gap to the steady amplitude, to shrink by a factor eInteractive lab
Predict first. Explore the model. Compare the result.
Explore the model
Try this. Set Q = 8 and drag ω through ω₀: the peak reaches 8.0 F₀/k while the across leg of the triangle shrinks to nothing, so δ swings from 45° to 135° inside a band only ω₀/8 wide. Drop Q to 1: the peak sinks to 1.15 F₀/k at 0.71 ω₀, yet δ is still exactly 90° at ω = ω₀.
Live results
- AMPLITUDE A ÷ (F₀/k)
- 2.43
- PHASE LAG δ
- 29.1 °
- POWER ÷ PEAK = sin²δ
- 0.236
- HALF-POWER WIDTH Δω ÷ ω₀
- 0.250
Results follow this lesson’s model and assumptions. Displayed values are rounded.
Predict → change one variable → check
Plan a change before running it. Every other input stays at the saved baseline.
Saved baseline. Drive frequency ω ÷ ω₀: 0.80; Quality factor Q: 4.00
This experiment is kept only while this page is open. It does not change saved course progress.
Catch the common trap
“At resonance the mass moves in step with the push.” It does not. At ω = ω₀ the displacement lags the force by a quarter cycle; it is the velocity that keeps step, and that is exactly why energy pours in. The work the drive does in one cycle is πF₀A sin δ. Far below ω₀, where δ is near 0, the force pushes the mass outward and is then pushed back by it: positive work on the way out, nearly equal negative work on the way back, so almost nothing accumulates even though the motion follows the force faithfully. At δ = 90° the force points along the velocity at every instant and never takes energy back. A related slip is to expect the steady motion at ω₀ or ω′. Those frequencies belong to the transient; the steady state always runs at the drive frequency ω, and resonance is a statement about its amplitude, not about its frequency.
Practice & worked examples
3 authored examples. Make an attempt, then reveal the reasoning one step at a time.
EasyExample 1
A 0.60 kg lab cart is held between springs of combined stiffness 135 N m⁻¹, and rolling losses give b = 0.30 kg s⁻¹. A motor pushes it with a force of amplitude 0.12 N at exactly the cart's natural frequency. Find ω₀, the steady amplitude, the phase lag and the mean power the motor supplies.
Sketch your answer or reasoning, then compare it with the authored solution. Your writing is not automatically graded or saved.
0 of 5 steps revealed.
- Natural frequency: ω₀ = √(k/m) = √(135/0.60) = √225 = 15 rad s⁻¹, about 2.4 Hz.
- At ω = ω₀ the net stiffness k − mω₀² = 135 − 0.60 × 225 = 0, so only the drag term limits the motion: A = F₀/(bω₀) = 0.12/(0.30 × 15) = 0.12/4.5 = 0.0267 m = 26.7 mm.
- Check with Q: the static stretch is F₀/k = 0.12/135 = 0.89 mm, and Q = mω₀/b = 0.60 × 15/0.30 = 30, so Q × F₀/k = 30 × 0.89 mm = 26.7 mm, as it must be at resonance.
- With zero net stiffness, tan δ = bω/(k − mω²) has a zero denominator, so δ = 90°: the displacement lags the push by a quarter period and the velocity is in step with it.
- Mean power: P̄ = F₀²/(2b) = 0.0144/0.60 = 0.024 W. Cross-check against the drag loss: ½bω₀²A² = 0.5 × 0.30 × 225 × 0.0267² = 0.024 W.
Answer. ω₀ = 15 rad s⁻¹; A = 26.7 mm, 30 times the 0.89 mm static stretch; δ = 90°; P̄ = 0.024 W = 24 mW.
MediumExample 2
A 2.0 kg instrument sits on an isolation mount with k = 3200 N m⁻¹ and b = 24 kg s⁻¹. A pump in the same room transmits a force of amplitude 6.0 N through the floor at the pump's running speed. Find the steady amplitude and phase lag of the instrument at (a) 30 rad s⁻¹ and (b) 50 rad s⁻¹, then (c) the worst amplitude the instrument meets while the pump runs up from rest to 50 rad s⁻¹.
Sketch your answer or reasoning, then compare it with the authored solution. Your writing is not automatically graded or saved.
0 of 5 steps revealed.
- Natural frequency and quality: ω₀ = √(3200/2.0) = 40 rad s⁻¹ and Q = mω₀/b = 2.0 × 40/24 = 3.3. The static deflection is F₀/k = 6.0/3200 = 1.9 mm.
- (a) At 30 rad s⁻¹: net stiffness k − mω² = 3200 − 2.0 × 900 = 1400 N m⁻¹, drag term bω = 24 × 30 = 720 N m⁻¹. A = 6.0/√(1400² + 720²) = 6.0/1574 = 3.8 mm, and tan δ = 720/1400 gives δ = 27°.
- (b) At 50 rad s⁻¹: net stiffness 3200 − 2.0 × 2500 = −1800 N m⁻¹, drag term 24 × 50 = 1200 N m⁻¹. A = 6.0/√(1800² + 1200²) = 6.0/2163 = 2.8 mm. The net stiffness is negative, so δ = 180° − arctan(1200/1800) = 180° − 34° = 146°.
- (c) The run-up passes through ω₀ = 40 rad s⁻¹, where A = F₀/(bω₀) = 6.0/(24 × 40) = 6.25 mm = Q × 1.9 mm, with δ = 90°. If the pump lingers there this is the amplitude it settles to; a quick crossing, faster than τ = 2m/b = 0.17 s, gives it less time to build.
- Reading the numbers: the half-power band runs from about 34 to 46 rad s⁻¹, so both running speeds sit outside it, yet the low Q makes the response broad: 2.0 times the static deflection at 30 rad s⁻¹ and 1.5 times at 50 rad s⁻¹. Running faster helps only a little; a softer mount that puts ω₀ well below 30 rad s⁻¹ would help more.
Answer. (a) A = 3.8 mm, δ = 27°; (b) A = 2.8 mm, δ = 146°; (c) up to 6.3 mm as the pump passes 40 rad s⁻¹.
HardExample 3
A 0.200 kg mass hangs from a spring. A steady 0.250 N pull stretches the spring 2.00 mm. Driven by a force of amplitude 0.250 N at ω = 24.0 rad s⁻¹, its steady amplitude is 18.2 mm. From only these data find k, ω₀, b and Q, the phase lag at 24.0 rad s⁻¹, the amplitude the mass will reach if driven at ω₀, and the time after switch-on for that amplitude to reach 90% of its steady value. Then name one measurement that could falsify the linear model.
Sketch your answer or reasoning, then compare it with the authored solution. Your writing is not automatically graded or saved.
0 of 6 steps revealed.
- Stiffness and natural frequency from the static data: k = F/x = 0.250/0.00200 = 125 N m⁻¹, so ω₀ = √(k/m) = √(125/0.200) = √625 = 25.0 rad s⁻¹ (f₀ = 3.98 Hz).
- At 24.0 rad s⁻¹ the net stiffness is k − mω² = 125 − 0.200 × 576 = 9.8 N m⁻¹. Invert the amplitude formula for the drag term: (bω)² = (F₀/A)² − (k − mω²)² = (0.250/0.0182)² − 9.8² = 188.7 − 96.0 = 92.7 N² m⁻², so bω = 9.63 N m⁻¹ and b = 9.63/24.0 = 0.40 kg s⁻¹.
- Quality factor and its companions: Q = mω₀/b = 0.200 × 25.0/0.40 = 12.5, half-power width Δω = b/m = 2.0 rad s⁻¹, settling time τ = 2m/b = 1.0 s.
- Phase lag at 24.0 rad s⁻¹: tan δ = 9.63/9.8 = 0.98, so δ = 44°. The measurement was taken almost exactly at the lower half-power frequency, where the two legs of the force triangle are equal.
- At resonance A = F₀/(bω₀) = 0.250/(0.40 × 25.0) = 0.0250 m = 25.0 mm, which is Q × 2.00 mm. From rest the envelope grows as A(1 − exp(−t/τ)), so 90% is reached when exp(−t/τ) = 0.10, at t = τ ln 10 = 2.3 s, about 9 cycles of T₀ = 0.251 s.
- Falsification: the linear model says doubling F₀ to 0.500 N doubles every amplitude and leaves the peak at 25.0 rad s⁻¹ with δ = 90° there. A peak that shifts or leans as the drive grows, or an amplitude that jumps during a slow sweep, means k or b is not constant and the model fails.
Answer. k = 125 N m⁻¹, ω₀ = 25.0 rad s⁻¹, b = 0.40 kg s⁻¹, Q = 12.5; δ = 44° at 24.0 rad s⁻¹; A = 25.0 mm at resonance, 90% of it about 2.3 s after switch-on. Doubling F₀ must double A without moving the peak.
Quick check
Select an answer and explain it before checking.
Review & continue
Resonance is the large steady-state response of a damped oscillator to a periodic force whose frequency lies close to the natural frequency ω₀ = √(k/m); the drive sets the frequency of the motion, the system sets its amplitude and lag.
- Explain the relationship without looking at the equation.
- Check a sign, a unit and a model assumption in your working.
- Return to the lab or worked example if a step is still unclear.
Optional supporting lessons
These chapter resources can help with earlier ideas; they are not additional required Physics I completion tasks.