Sound & Resonance Tubes · Guide · glidepathscience.org
GuideSound & Resonance Tubes
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, and come back whenever a question sends you here.
Part 1 · The essentials
Sound is a longitudinal wave
A tuning fork’s prongs push on the air and pull back, hundreds of times a second. Each parcel of air swings back and forth along the direction the wave travels, by a tiny distance, about its own place. Where neighbors crowd together the pressure is a little higher (a compression); where they spread apart it is a little lower (a rarefaction). That pattern travels; the air itself does not flow. Air resists being squeezed but not being sheared sideways, so sound in air is longitudinal and, unlike light, cannot be polarized.
Like every wave, sound travels one wavelength in each period:
$v = f\lambda$
In one gas at one temperature every frequency travels at the same speed; a higher pitch just packs more wavelengths into each meter.
The gas sets the speed
For an ideal gas,
$v = \sqrt{\gamma RT/M}$
with $R$ = 8.314 J/(mol·K), the temperature $T$ in kelvins and the molar mass $M$ in kg/mol. γ is the ratio of heat capacities: 5/3 for a single atom such as helium, about 1.40 for air’s nitrogen and oxygen, 1.29 for carbon dioxide. Light molecules carry sound fast: 1,007 m/s in helium, 343 m/s in air and 267 m/s in CO₂, all at 20 °C. The pressure cancels, since $P/\rho = RT/M$, so a city high in the mountains has the same speed of sound as the coast at the same temperature. For air, the speed is often written
$v = (331\text{ m/s})\sqrt{T/273\text{ K}}$
about 0.6 m/s faster for each degree Celsius near room temperature.
Resonance in a closed tube
Hold a fork over a tube whose bottom is closed by water, and lower the water. At certain lengths the sound swells: the air column resonates. The water is a displacement node, since the air there cannot move; the open mouth is close to an antinode. The column resonates when it holds an odd number of quarter wavelengths:
$L_n + \delta = (2n - 1)\lambda/4$
Here δ is the end correction: the air just outside the mouth moves with the column, so the antinode sits a little beyond the open end, about 0.61 times the radius for a plain (unflanged) pipe. Each resonance is $\lambda/2$ farther down than the one before, and since every resonance shifts by the same δ, the spacing cancels it. That is why the lab measures two resonances: $v = 2f(L_2 - L_1)$. With one resonance, $v = 4fL_1$ comes out a few percent low.
Displacement and pressure
A standing wave has two faces. Where the air moves most (a displacement antinode), neighboring parcels move together and are hardly squeezed: the pressure barely changes, a pressure node. Where the air is held still (a displacement node), the parcels on either side squeeze and stretch it hardest: a pressure antinode. So a closed end is an s node and a p antinode, and an open end is an s antinode and a p node, held near the room’s pressure.
Open and closed pipes
A pipe open at both ends has antinodes at both, so it holds whole numbers of half wavelengths, and every harmonic is present:
$f_n = nv/2L$
Close one end and it holds odd quarter wavelengths instead, so its lowest note drops an octave and only the odd harmonics remain:
$f_n = (2n - 1)v/4L$
That is why a clarinet, whose reed closes one end, plays about an octave below a flute of nearly the same length. (Real instruments differ from ideal pipes because of tone holes, the bore and the mouthpiece.)
Decibels and the ear
The ear spans a factor of a trillion in intensity, so levels are logarithmic:
$\beta = (10\text{ dB})\log(I/I_0)$, $I_0 = 10^{-12}$ W/m²
Ten times the intensity adds 10 dB, twice adds 3 dB, so two equal unrelated sources give 3 dB more than one, not twice the decibels. A small source spreads its power over a sphere, $I \propto 1/r^2$, so $\Delta\beta = -(20\text{ dB})\log(r_2/r_1)$: twice as far is 6 dB quieter.
Hearing runs from about 20 Hz to 20 kHz and is most sensitive near 3 kHz. The ear canal, about 2.5 cm long and closed by the eardrum, is a closed tube whose quarter-wave resonance, $f = v/4L$, falls near 3.5 kHz.
Part 2 · Going further
Why γ, and where it comes from
A compression wave travels at $v^2 = (\partial P/\partial\rho)_S$. Newton treated the compressions as isothermal, $v = \sqrt{RT/M}$, and came out about 15 % low. Laplace saw that heat cannot cross half a wavelength in half a period, so the gas warms as it is squeezed: the compressions are adiabatic, which multiplies by $\sqrt{\gamma}$, giving $v = \sqrt{\gamma P/\rho} = \sqrt{\gamma RT/M}$.
γ = $(C_V + R)/C_V$ counts how a molecule stores energy. A rigid linear molecule has 3 translations and 2 rotations, γ = 7/5. Carbon dioxide is linear, yet its γ is 1.29, because its bending vibration (667 cm⁻¹) is low enough for $kT$ to excite partly; as $T$ rises the bend fills further and γ falls, so CO₂’s speed does not scale as √T. Humid air is a mixture: mix the molar masses and the heat capacities, $C_V/R = 1/(\gamma - 1)$, by mole fraction, with $x_w = p_{sat}/P$. Water vapor is lighter than air, so humid air is faster.
The two fields, exactly
The pressure swing follows from the displacement: $p = -B\,\partial s/\partial x$, with the bulk modulus $B = \gamma P = \rho v^2$. In a standing wave $s \propto \sin kx$, so $p \propto \cos kx$: the two fields are a quarter wavelength apart along the tube but in step in time (it is the air’s velocity, $\partial s/\partial t$, that is a quarter period out of step with the pressure). A tube closed at both ends has displacement nodes at both, so it resonates at $L = n\lambda/2$, like an open tube, with pressure antinodes at both ends.
The end correction and the losses that set Q
Two resonances give the end correction as well as the speed: from $L_1 + \delta = \lambda/4$ and $L_2 + \delta = 3\lambda/4$,
$\delta = (L_2 - 3L_1)/2$
Levine and Schwinger found 0.6133a for an unflanged pipe; a pipe set in a wide flange has 0.8216a. These low-frequency results hold while $ka \le 0.5$.
Each resonance is a driven, damped oscillator, and its sharpness is its quality factor, $Q = f_0/\Delta f$, with Δf the full width at half power. Two losses set it: the viscous and thermal boundary layer at the walls, which grows as the bore narrows, and radiation from the open mouth, about $(ka)^2$ of the energy per reflection, which grows as it widens; $1/Q = 1/Q_{wall} + 1/Q_{rad}$.
Two unknowns: speed and density
A measured speed fixes only the ratio γ/M, so two gases can share a speed: oxygen and argon differ by 2.3 %, neon and methane by 0.7 %. Weighing a bulb of the gas gives its density, and with the pressure that separates them:
$\gamma = \rho v^2/P$ $M = \rho RT/P$
Sound pressure, displacement and A-weighting
Microphones measure pressure, so meters report the sound pressure level,
$L_p = (20\text{ dB})\log(p_{rms}/20\ \mu\text{Pa})$ $I = p_{rms}^2/\rho v$
The two references, 20 µPa and 10⁻¹² W/m², agree only because ρv ≈ 400 Pa·s/m in air (413 at 20 °C). For a traveling wave, $p_{max} = \rho v\omega\,s_{max}$, so
$s_{max} = \sqrt{2}\,p_{rms}/\rho v\omega$
At the threshold of hearing the air swings by about 10⁻¹¹ m, a tenth of an atom. The ear is not flat: A-weighting (IEC 61672-1) discounts low and very high frequencies roughly as the ear does at moderate levels, $L_A = L_p + A(f)$. Occupational limits use it: NIOSH’s full daily dose is 8 hours at 85 dB(A), half the time for every 3 dB more.
Formants and the helium voice
The vocal folds buzz at the pitch; the vocal tract, roughly a 17 cm tube closed at the folds and open at the lips, shapes that buzz with its resonances, the formants: about 520 Hz, 1,560 Hz and 2,600 Hz in air. In a lighter gas the formants rise with $v$, boosting higher harmonics, while the pitch, set by the folds’ mass and tension, stays put.
Reference
Common mistakes
- Sound as a sideways wave. In air the parcels swing along the direction of travel.
- “Denser gases carry sound faster.” At one temperature, heavier molecules carry it slower.
- Grams and Celsius. $M$ in kg/mol and $T$ in kelvins in $v = \sqrt{\gamma RT/M}$.
- Using one resonance. The end correction makes $4fL_1$ low; use the spacing.
- A whole wavelength between resonances. They are $\lambda/2$ apart.
- Pressure peaks where the air moves most. It peaks where the air is held still.
- Even harmonics in a closed pipe. Only odd ones.
- Adding decibels. Add intensities: two equal, independent sources are +3 dB.
- 10 log for pressures. Pressure levels use 20 log.
- Helium raises the pitch. It raises the formants.
Equation sheet
| Part 1 | |
|---|---|
| wave speed | $v = f\lambda$ |
| ideal gas | $v = \sqrt{\gamma RT/M}$ |
| air | $v = (331\text{ m/s})\sqrt{T/273\text{ K}}$ |
| closed tube | $L_n + \delta = (2n - 1)\lambda/4$ |
| ideal ends | $f_n = (2n - 1)v/4L$ |
| open pipe | $f_n = nv/2L$ (ideal ends) |
| spacing | $\lambda/2$; $v = 2f(L_2 - L_1)$ |
| intensity level | $\beta = (10\text{ dB})\log(I/I_0)$, $I_0 = 10^{-12}$ W/m² |
| point source | $I \propto 1/r^2$, $\Delta\beta = -(20\text{ dB})\log(r_2/r_1)$ |
| ear canal | $f = v/4L$ |
| Part 2 | |
|---|---|
| isothermal | $v = \sqrt{RT/M}$ |
| adiabatic | $v = \sqrt{\gamma P/\rho} = \sqrt{\gamma RT/M}$ |
| humid air | mix $M$ and $C_V/R = 1/(\gamma - 1)$; $x_w = p_{sat}/P$ |
| fields | $p = -B\,\partial s/\partial x$, $B = \gamma P = \rho v^2$ |
| end correction | $\delta = (L_2 - 3L_1)/2$ |
| quality factor | $Q = f_0/\Delta f$ |
| from a density | $\gamma = \rho v^2/P$, $M = \rho RT/P$ |
| pressure level | $L_p = (20\text{ dB})\log(p_{rms}/20\ \mu\text{Pa})$, $I = p_{rms}^2/\rho v$ |
| displacement | $s_{max} = \sqrt{2}\,p_{rms}/\rho v\omega$ |
| A-weighting | $L_A = L_p + A(f)$ |
Symbols: v speed of sound (m/s) · f frequency (Hz) · λ wavelength (m) · γ ratio of heat capacities · R gas constant · T temperature (K) · M molar mass (kg/mol) · L length (m) · δ end correction (m) · a radius (m) · n mode number · β intensity level (dB) · I intensity (W/m²) · r distance (m) · P pressure (Pa) · ρ density (kg/m³) · B bulk modulus (Pa) · s displacement (m) · p pressure swing (Pa) · Q quality factor · $L_p$ sound pressure level (dB SPL) · ω angular frequency (rad/s).
Further reading
OpenStax, College Physics 2e (2022), Chapter 17, Physics of Hearing, and University Physics Volume 1 (2016), Chapter 17, Sound, both free online under CC BY-NC-SA 4.0. L. E. Kinsler, A. R. Frey, A. B. Coppens and J. V. Sanders, Fundamentals of Acoustics, 4th ed. (Wiley, 2000), for the wall losses. H. Levine and J. Schwinger, Phys. Rev. 73, 383 (1948), for the end correction. E. Terhardt, Hearing Research 1, 155 (1979), for the threshold. And this lab’s Field Manual (H): how the model works and what it leaves out.
Sound & Resonance Tubes v1.3 · © 2026 Dr. David Meier · CC BY-NC-SA 4.0 · glidepathscience.org