Waves on a String · Guide · glidepathscience.org
GuideWaves on a String
Contents
This Guide covers the ideas and equations behind every Walkthrough step, Concept Challenge question and Predict round. Read Part 1 before the Intro track and Part 2 before the Advanced track, about ten minutes each. Come back whenever a question sends you here.
Part 1 · The essentials
A wave moves on; the medium stays home
Flick the end of a rope and a pulse runs along it. Each bit of rope is pulled up by its neighbor on the side the pulse comes from, then pulled back down as that neighbor falls, so it moves and returns to where it started. What travels is the shape, and the energy with it, never the rope.
In a transverse wave the medium moves across the direction of travel: a rope, a guitar string, a stadium wave, seismic S waves. In a longitudinal wave it moves along it: a slinky pushed along its length, sound in air, seismic P waves. Water at the surface does both, moving in loops. Only transverse waves can be polarized.
What sets the speed
A tighter rope snaps back faster, and a heavier one responds more slowly. For a string of tension $F_T$ and mass per meter $\mu$,
$v = \sqrt{F_T/\mu}$ $\mu = m/L$ $t = L/v$
The speed depends only on the medium. How hard or how fast you flick changes a pulse’s height and width, never its speed; a quicker flick makes a narrower pulse, not a faster one. Because of the square root, four times the tension doubles the speed, and so does a quarter of the mass per meter. To reach a chosen speed, the rope needs $F_T = \mu v^2$.
Ends and knots
At a fixed end (a clamp), the arriving pulse pulls up on the clamp, so the clamp pulls down on the rope (Newton’s third law) and sends the pulse back inverted, the same size. At a free end (a light ring sliding on a frictionless rod), nothing can pull the end down: the ring overshoots to twice the pulse’s height for an instant and the pulse comes back upright. Neither end takes any energy. Either way it comes back front first: a steep front still leads.
At a knot between two ropes, part of a pulse goes on and part comes back. Into a heavier rope the knot acts partly like a clamp: the reflection is inverted. Into a lighter rope it acts partly like a free end: the reflection is upright. The part that goes on is always upright, and the frequency never changes at a knot, so in a slower rope the pulse is shorter in proportion to the speed.
Pulses pass through each other
Where two waves overlap, their displacements add:
$y = y_1 + y_2$
and each carries on as if the other were not there. An upright pulse and an inverted one of the same shape cancel for an instant, and the rope is flat; but it is moving fastest then, so all of its energy is kinetic, and both pulses re-form and travel on. Two upright pulses make one taller pulse while they overlap, then separate unchanged.
Steady waves and the wave function
A hand shaking steadily at frequency $f$ sends a train of crests along the rope, a wavelength λ apart. One wavelength passes each period $T = 1/f$, so
$v = f\lambda$
The source sets $f$, the medium sets $v$, and λ follows: shaking faster shortens the wavelength, while the speed stays the same. A sine wave traveling toward $+x$ is
$y(x,t) = A\sin(kx - \omega t + \varphi)$ $k = 2\pi/\lambda$, $\omega = 2\pi f$, $v = \omega/k$
A crest is where $kx - \omega t$ keeps its value, so it moves at $\omega/k$. With $+\omega t$ instead, the wave travels toward $-x$.
Standing waves and harmonics
Two equal waves running opposite ways add to a standing wave. At the nodes the two are always equal and opposite, so the string never moves there; halfway between, at the antinodes, they add and the string swings by 2A. Nodes are λ/2 apart.
On a string held at both ends, waves reflect back and forth, and a standing wave builds only when a whole number of half-wavelengths fits: $\lambda_n = 2L/n$. Then
$f_n = nv/2L$
The $n$-th harmonic has $n$ loops and $n + 1$ nodes, counting the ends. In the classic setup a vibrator shakes one end and the string runs over a pulley to a hanging mass, so $F_T = mg$; four times the mass doubles the speed and every frequency. A slightly mistuned string beside a reference tone beats $f_{beat} = |f - f_{ref}|$ times a second.
Part 2 · Going further
The wave equation and d’Alembert
For small slopes, the net sideways force on a short piece of string is $F_T$ times its change of slope. Newton’s second law then gives the wave equation:
$\partial^2 y/\partial t^2 = v^2\,\partial^2 y/\partial x^2$
Curvature drives acceleration. Every solution is d’Alembert’s $y = f(x - vt) + g(x + vt)$: any shape moving right plus any shape moving left. A bump held still and released must start with $f$ and $g$ equal, so it splits into two half-height bumps that move apart.
Impedance, power and a matched end
What a knot does depends on each rope’s impedance, $Z = \sqrt{F_T\mu} = \mu v$. The amplitude coefficients are
$r = (Z_1 - Z_2)/(Z_1 + Z_2)$ $\tau = 2Z_1/(Z_1 + Z_2)$
and the power splits $r^2$ back and $(Z_2/Z_1)\tau^2$ on, which add to 1, so the transmitted share is $T = 4Z_1Z_2/(Z_1 + Z_2)^2$. From heavy to light, $\tau$ exceeds 1: a taller pulse that still carries less power. A sine wave carries average power
$\bar P = \tfrac12 Z\omega^2A^2$
through every point, with its kinetic and elastic energy equal. A dashpot pushing back with $-b\,\partial y/\partial t$ reflects $r = (Z - b)/(Z + b)$; with $b = Z$ it reflects nothing, and a short string acts like an endless one. A measured $r$ finds a hidden rope: $\mu_2 = \mu_1[(1 - r)/(1 + r)]^2$. A standing wave carries equal power both ways: none on average.
Modes for every pair of ends, and plucks
A fixed end is a node and a free end an antinode. Fixed–fixed and free–free strings fit whole half-wavelengths, $f_n = nv/2L$ (free–free adds a zero-frequency slide of the whole string). A string fixed at one end and free at the other fits odd quarter-wavelengths:
$f_k = (2k-1)v/4L$
odd multiples only, like a tube closed at one end. A string plucked at a fraction $p$ of its length starts as a triangle, and its modes have amplitudes
$b_n \propto \sin(n\pi p)/n^2$
so a harmonic with a node at the pluck point is missing (Young’s rule), and a pluck near the end feeds the high harmonics: it sounds brighter. Friction inside the string damps mode $n$ at $\beta + \nu k_n^2/2$, so high harmonics die first: a plucked note mellows.
Resonance and Q
Each mode of a damped string is one damped, driven oscillator, as in Simple Harmonic Motion, with amplitudes decaying as $e^{-\beta t}$. Driven from an end that moves by $Y_0$, the steady state is $y(x) = Y_0\sin k(L-x)/\sin kL$ with a complex $k$. While the damping is weak ($\beta L/v \ll 1$, so the resonances stay well apart), at each harmonic the antinodes reach about
$A \approx Y_0v/(\beta L)$ $\Delta f \approx \beta/\pi$ $Q_n \approx \pi f_n/\beta$
In that limit the gain is the same for every harmonic, and so is the half-power width, so Q grows with frequency. The sim solves the driven string exactly; with strong damping its peaks widen past β/π, and a low peak can lose its half-power width altogether. The driven end barely moves compared with an antinode: it is nearly a node. Q can also be read from a ring-down: switch the vibrator off and fit the logarithm of the swing against time; its slope is −β.
A slinky: two speeds and a cutoff
A slinky of mass $M$ and coil constant $K$, stretched to length $L$, is a spring of mass per meter $M/L$ and stiffness $KL$, so longitudinal pulses travel at $v = L\sqrt{K/M}$ and cross in $\sqrt{M/K}$, however far it is stretched. Under its tension $F_T = K(L - L_0)$, with closed length $L_0$, transverse pulses take almost the same time, because a closed-wound slinky is nearly a zero-length spring.
A slinky is also a chain of $N$ turns: masses $m$ joined by springs $k$, spaced $a$. Long waves travel at $a\sqrt{k/m}$, but short ones slow down, $\omega = 2\sqrt{k/m}\,\sin(qa/2)$ for wavenumber $q$, and above $f_{max} = (1/\pi)\sqrt{k/m}$ no wave travels at all.
Beyond the uniform, linear string
A hanging rope. At height $y$ above its free bottom end the rope holds up the rope below it, $F_T = \mu g y$, so $v = \sqrt{gy}$. A pulse speeds up and stretches as it climbs, and its height falls as $y^{-1/4}$. Adding up $dy/v$ gives a climb time of $2\sqrt{L/g}$, whatever the rope’s mass per meter. Its modes are Bessel functions, with frequency ratios 2.295 and 3.598: not harmonic.
Tension modulation. A big swing stretches a stretchy string, raising its tension by $EA$ times the strain (Kirchhoff and Carrier), so the fundamental rises by about $(3/32)(EA/F_T)(\pi A/L)^2$. This is the only place where a bigger amplitude goes faster.
Reference
Common mistakes
- Thinking the medium travels with the wave. It moves and returns; the shape and the energy travel.
- Thinking a taller pulse or a quicker flick goes faster. Only the medium sets the speed.
- Scaling the speed in proportion to the tension. It goes as the square root.
- Using the whole rope’s mass for μ. It is the mass per meter, $m/L$.
- Getting the reflections backward. A clamp, or a heavier rope, inverts; a free end, or a lighter rope, does not.
- Thinking canceled pulses are gone. At the flat instant the rope moves fastest; both pulses carry on.
- Thinking a higher frequency travels faster. The frequency changes λ, not $v$.
- Taking one loop as a whole wavelength. A loop is half a wavelength: $\lambda_1 = 2L$.
- Using $nv/2L$ for a fixed–free string. It has only odd multiples of $v/4L$.
Equation sheet
| Part 1 | |
|---|---|
| speed | $v = \sqrt{F_T/\mu}$, $\mu = m/L$, $t = L/v$ |
| tension for a speed | $F_T = \mu v^2$ |
| superposition | $y = y_1 + y_2$ |
| steady wave | $v = f\lambda$, $T = 1/f$ |
| wave function | $y(x,t) = A\sin(kx - \omega t + \varphi)$ |
| $k = 2\pi/\lambda$, $\omega = 2\pi f$, $v = \omega/k$ | |
| fixed at both ends | $\lambda_n = 2L/n$, $f_n = nv/2L$ |
| hanging mass | $F_T = mg$ |
| Part 2 | |
|---|---|
| wave equation | $\partial^2 y/\partial t^2 = v^2\,\partial^2 y/\partial x^2$ |
| impedance | $Z = \sqrt{F_T\mu}$ |
| at a knot | $r = (Z_1 - Z_2)/(Z_1 + Z_2)$, $\tau = 2Z_1/(Z_1 + Z_2)$ |
| power through | $T = 4Z_1Z_2/(Z_1 + Z_2)^2$ |
| average power | $\bar P = \tfrac12 Z\omega^2A^2$ |
| fixed–free | $f_k = (2k-1)v/4L$ |
| a pluck at p | $b_n \propto \sin(n\pi p)/n^2$ |
| resonance (weak damping) | $A \approx Y_0v/(\beta L)$, $\Delta f \approx \beta/\pi$ |
| hanging rope | $t = 2\sqrt{L/g}$ |
| slinky | $v = L\sqrt{K/M}$, $f_{max} = (1/\pi)\sqrt{k/m}$ |
Symbols: $F_T$ tension (N) · μ mass per meter (kg/m) · $v$ wave speed (m/s) · $L$ length (m) · $A$ amplitude (m) · λ wavelength (m) · $f$ frequency (Hz) · $T$ period (s), or the transmitted power fraction · $k$ wavenumber (rad/m) · ω angular frequency (rad/s) · $Z$ impedance (kg/s) · $r$, τ amplitude coefficients · $b$ a dashpot’s constant (kg/s) · β amplitude decay rate (s⁻¹) · $Y_0$ the driven end’s amplitude (m) · $K$ a slinky’s coil constant (N/m) · $k$, $m$ per-turn spring and mass of a chain.
Sources and further reading
Samuel J. Ling, Jeff Sanny and William Moebs, University Physics Volume 1 (OpenStax, 2016), free online under CC BY-NC-SA 4.0: Chapter 16, Waves; the Intro track follows its order of topics and its notation. A. P. French, Vibrations and Waves (1971); F. S. Crawford Jr., Waves (Berkeley Physics Course vol. 3, 1968). And this lab’s Field Manual (H): how the model works, and what it leaves out.
Waves on a String v1.2 · © 2026 Dr. David Meier · CC BY-NC-SA 4.0 · glidepathscience.org