ARCHETYPE: Liquid in a U-Tube

Blow gently across one arm of a U-shaped tube of water, then step back. The liquid seesaws: one column falls as the other rises, overshoots, and swings back — a liquid pendulum with no string, no bob and no visible spring. Isaac Newton himself put this system in the Principia (Book II, Proposition 44) and proved its punchline: water sloshing in a U-tube keeps time exactly like a pendulum whose length is half the liquid column — and, remarkably, the tempo does not care whether the tube holds water or mercury, or whether the bore is a capillary or a drainpipe. The U-tube is the purest "find the effective spring" exercise in the whole oscillations chapter. Displace the liquid a distance $x$ along the tube and the two surfaces move oppositely — one down $x$ , one up $x$ — so the level difference is $2x$ . The unbalanced weight is that of a liquid column of height $2x$ : there is the factor of 2 that decides every mark in this archetype. Meanwhile the thing that must be accelerated is the entire liquid column, bend and all. Stiffness from a $2x$ slab, inertia from the whole snake of liquid: the ratio hands you $T = 2\pi\sqrt{L/2g}$ in three lines. JEE poses this archetype straight (find $T$ ), inverted (find the column length from a stopwatch), and tilted (V-shaped tubes with inclined arms — the JEE-Advanced dressing). Beyond the exam, the same physics is poured into skyscrapers: tuned liquid column dampers are building-sized U-tubes whose slosh period is deliberately matched to the tower's sway. One factor of 2, many disguises.

Key takeaways

Liquid of total column length $L$ (both arms plus the bend) in a uniform vertical U-tube, displaced by $x$ along the tube: the level difference becomes $2x$ , and the restoring force is the weight of a liquid column of height $2x$ : $F = -\,\rho S g\,(2x)$ .
The whole column (mass $\rho S L$ ) accelerates together, giving $a = -\dfrac{2g}{L}x$ : SHM with $\omega = \sqrt{2g/L}$ and $T = 2\pi\sqrt{\dfrac{L}{2g}}$ .
The period is independent of the liquid's density and of the tube's bore — mercury and water in the same tube keep identical time; only $L$ and $g$ enter.
Equivalent pendulum length $= L/2$ (Newton's own result): a 50-cm liquid column swings like a 25-cm pendulum, $T \approx 1.0$ s.
Inclined arms at angles $\theta_1, \theta_2$ to the horizontal: $T = 2\pi\sqrt{\dfrac{L}{g(\sin\theta_1 + \sin\theta_2)}}$ — the vertical tube is the special case $\sin\theta = 1$ twice.
Key idea: displace the liquid by $x$ and gravity levies the weight of a $2x$ column on the entire $L$ -long snake of liquid — SHM at $T = 2\pi\sqrt{L/2g}$ , a pendulum of length $L/2$ in liquid costume.

The core insight

In a uniform U-tube the two surfaces move oppositely, so a displacement $x$ creates a $2x$ level difference whose weight $\rho S g(2x)$ restores the whole column of mass $\rho S L$ : $a = -(2g/L)x$ , giving $T = 2\pi\sqrt{L/2g}$ — density and bore cancel, and the equivalent pendulum has length $L/2$ .

Where this lives today

The U-tube oscillator has been scaled from laboratory glassware to civil-engineering size. Tuned liquid column dampers — enormous U-shaped water tanks — sit near the tops of towers such as Philadelphia's Comcast Center (whose 300,000-gallon TLCD is the world's largest): the water thread's length is chosen so $2\pi\sqrt{L/2g}$ matches the building's sway period, and orifice plates in the bend add deliberate damping, so wind energy pours into the slosh and drains into turbulence — the damped-resonance module's energy-theft trick, played with this module's clock. Ships got there first: Frahm's 1911 anti-roll tanks used tuned water columns to fight rolling, and modern versions still do. Nature runs the archetype at geographic scale as seiches — whole harbours and lakes sloshing between their shores with periods of minutes (Lake Geneva's is about 73 minutes), the standing oscillation Forel first quantified. And in every chemistry lab, the settle-wobble of a manometer after a pressure change is this module's clock plus viscous damping: read the period and you have measured the thread; watch the decay and you have met the next chapter.

The setup

A U-tube of uniform cross-section $S$ holds a liquid of density $\rho$ ; the total length of the liquid thread — up one arm, around the bend, up the other — is $L$ . The liquid is displaced a small distance $x$ along the tube (one meniscus down $x$ , the other up $x$ ) and released. Find the period; invert it for $L$ ; generalise to arms inclined at angles $\theta_1$ , $\theta_2$ . Idealisations, stated up front because the exam states them too: inviscid liquid (no damping), uniform bore, both menisci staying within straight arm sections.

Figure

The U-tube displaced — one meniscus down x, the other up x, and the 2x slab that restores

The master solution pattern

Derivation AThe liquid pendulum — stiffness of a 2x slab, inertia of the whole thread

We derive $T = 2\pi\sqrt{L/2g}$ , watching the two places students bleed marks: the factor 2 in the restoring force and the whole-column mass in the inertia. Strategy: use the incompressibility of the liquid to describe everything by one coordinate, weigh the unbalanced column, and feed the master pattern $T = 2\pi\sqrt{\text{inertia}/\text{stiffness}}$ . Then generalise to inclined arms with one substitution.

Step 1. One coordinate rules the thread. The liquid is incompressible and the bore uniform, so if the left meniscus drops by $x$ , liquid at every point of the tube has moved distance $x$ along the tube, and the right meniscus has risen by the same $x$ . Velocity and acceleration are likewise shared: the thread moves as one body of mass $m = \rho S L$ .

Step 2. Weigh the imbalance. With levels at $-x$ and $+x$ , the right column stands $2x$ higher than the left. Pressure at the bend from the right exceeds that from the left by $\rho g(2x)$ , so the net force pushing the thread back toward balance is that pressure difference times the bore: $F = -\rho S g\,(2x)$ . Equivalently: the system is out of equilibrium by exactly one slab of liquid, height $2x$ , whose weight is unbalanced. Effective stiffness: $k_{\text{eff}} = 2\rho S g$ .

Step 3. Newton's second law for the thread: $\rho S L\,a = -2\rho S g\,x$ , so $a = -\dfrac{2g}{L}\,x$ . Density and bore have cancelled — they multiplied both the stiffness and the inertia, and the ratio forgot them. This is the same democratic cancellation that made the pendulum mass-blind.

Step 4. Read the clock: $\omega = \sqrt{2g/L}$ , $T = 2\pi\sqrt{\dfrac{L}{2g}}$ . Compare with the pendulum $2\pi\sqrt{L_{\text{pend}}/g}$ : they match when $L_{\text{pend}} = L/2$ — Newton's Proposition 44, the liquid column as a pendulum of half its length.

Step 5. Inclined arms. If the arms make angles $\theta_1, \theta_2$ with the horizontal, a displacement $x$ along the tube raises one surface by $x\sin\theta_1$ and lowers the other by $x\sin\theta_2$ — vertical height is what gravity taxes. The unbalanced weight is $\rho S g\,x(\sin\theta_1 + \sin\theta_2)$ , and the same bookkeeping gives $T = 2\pi\sqrt{\dfrac{L}{g(\sin\theta_1 + \sin\theta_2)}}$ . Vertical arms ( $\theta_1 = \theta_2 = 90^\circ$ ) recover Step 4; tilting the arms softens gravity's grip and slows the slosh. Verified: dimensions ✓; $L = 50$ cm gives $T = 2\pi\sqrt{0.50/19.6} = 2\pi(0.1597) = 1.00$ s — the half-metre column as a seconds-beating slosh ✓; inclined formula at $\theta_1 = \theta_2 = 90^\circ$ returns $2\pi\sqrt{L/2g}$ ✓; as $\theta \to 0$ (horizontal arms) $T \to \infty$ — a level tube has no gravity to restore it ✓.

The independence results deserve one more beat, because they are where the distractor options live. Density cancels: mercury is 13.6 times heavier than water, but the restoring slab is 13.6 times heavier too — the clock ties exactly (what differs is damping: mercury's low relative viscosity lets it ring longer, which is why mercury barometers wobble so persistently). Bore cancels: a fat tube moves more liquid and restores with a fatter slab, in exact proportion. Only the length of the liquid thread and gravity survive into the period — a two-variable formula, which is why examiners must dress it up (inclined arms, mass-to-length conversions) to make it earn three minutes.

Figure

Inclined arms — gravity taxes only the vertical rise, sin θ per arm

Limiting cases

case	result	meaning
$L \to 4L$	$T \to 2T$	the standard square root: a longer thread is a slower pendulum
mercury ↔ water, same tube	same $T$	density cancels; only damping differs
fat bore ↔ capillary	same $T$	cross-section cancels (uniform bore assumed)
arms tilted to $\theta$ each	$T = 2\pi\sqrt{L/(2g\sin\theta)}$	tilting dilutes gravity: slower slosh
one arm horizontal ( $\theta_2 = 0$ )	$T = 2\pi\sqrt{L/(g\sin\theta_1)}$	only the tilted arm's surface pays gravity
both arms horizontal	$T \to \infty$	no height change anywhere: no restoring force
on the Moon ( $g/6$ )	$T \times \sqrt6$	a gravity clock, like the pendulum and the float — unlike the spring

Common trap“The restoring force is the weight of the displaced column x”

Off by the factor that decides the mark. Push one meniscus down by $x$ and the other rises by $x$ — the level difference is $2x$ , and the unbalanced weight is that of a $2x$ column: $F = -2\rho S g x$ , stiffness $2\rho Sg$ . Forget the 2 and your period is too long by $\sqrt2$ — and $2\pi\sqrt{L/g}$ is always sitting among the options waiting for exactly this slip. The mirror-image trap lives in the inertia: the mass being accelerated is the entire liquid thread $\rho S L$ (every particle of it moves), not the little $2x$ slab and not just the liquid in the arms. Stiffness from the slab, inertia from the snake: cross the two up in either direction and the formula walks away from you.

Worked examples

1 · Board level — the straight computation (JEE-Main early slot)

A vertical U-tube of uniform bore holds a liquid column of total length $20\ \mathrm{cm}$ . The liquid is slightly depressed in one arm and released. Find the period of oscillation and the length of the equivalent simple pendulum. ( $g = 9.8\ \mathrm{m/s^2}$ .)

Step 1. Direct: $T = 2\pi\sqrt{\dfrac{L}{2g}} = 2\pi\sqrt{\dfrac{0.20}{19.6}} = 2\pi \times 0.101 = \boxed{0.63\ \mathrm{s}}$ .

Step 2. Equivalent pendulum: $L_{\text{pend}} = L/2 = \boxed{10\ \mathrm{cm}}$ . Audit: $2\pi\sqrt{0.10/9.8} = 0.63$ s ✓ — Newton's half-length rule in numbers.

Step 3. Note what the problem never told you: the liquid's identity and the tube's bore. It didn't need to — they cancel, and recognising that is the tested skill.

2 · Board level — mercury versus water (conceptual + number)

Two identical U-tubes hold columns of length $50\ \mathrm{cm}$ — one of water, one of mercury (13.6 times denser). Compare their oscillation periods, and compute them. ( $g = 9.8\ \mathrm{m/s^2}$ .)

Step 1. The comparison first: density multiplies both the restoring slab's weight and the thread's inertia, so it cancels — the periods are equal, despite the 13.6× density gap.

Step 2. The number, once for both: $T = 2\pi\sqrt{\dfrac{0.50}{19.6}} = 2\pi \times 0.160 = \boxed{1.00\ \mathrm{s}}$ — the half-metre column is a one-second slosh, a worthwhile anchor.

Step 3. What does differ: the decay. Mercury (lower kinematic viscosity) rings for many cycles; water's oscillation dies faster. Period identical, damping different — the pairing examiners use to check you separate the clock from the leak.

3 · Competitive level — column length from the mercury's mass (JEE-Main pattern, ~2.5 minutes)

A U-tube of uniform bore $0.50\ \mathrm{cm^2}$ contains $200\ \mathrm{g}$ of mercury (density $13{,}600\ \mathrm{kg/m^3}$ ). Find the period of oscillation of the mercury column. ( $g = 10\ \mathrm{m/s^2}$ .)

Step 1. The formula wants $L$ , the problem gives mass — convert through the thread's volume: $L = \dfrac{m}{\rho S} = \dfrac{0.200}{13600 \times 0.50\times10^{-4}} = \dfrac{0.200}{0.68} = 0.294\ \mathrm{m}$ .

Step 2. Clock: $T = 2\pi\sqrt{\dfrac{L}{2g}} = 2\pi\sqrt{\dfrac{0.294}{20}} = 2\pi \times 0.121 = \boxed{0.76\ \mathrm{s}}$ .

Step 3. Audit: density and bore did appear — but only inside the geometry conversion to $L$ , never in the period law itself. That division of labour (data → $L$ → clock) is the entire architecture of the harder variants. Timing note: ~2–2.5 min. One conversion, one formula. Distractor watch: feeding $m = 200$ g into some invented $2\pi\sqrt{m/\text{something}}$ — the period law has no mass slot; mass only buys you the length.

4 · Competitive level — JEE-Advanced stretch: the V-tube with unequal arms (~3.5 minutes)

A tube of uniform bore is bent so that one straight arm makes $30^\circ$ and the other $90^\circ$ with the horizontal. It holds a liquid thread of total length $60\ \mathrm{cm}$ . Find the period of small oscillations along the tube, and compare with the same thread in a standard vertical U-tube. ( $g = 9.8\ \mathrm{m/s^2}$ .)

Step 1. Displace the thread $x$ along the tube: the $30^\circ$ surface changes height by $x\sin 30^\circ = 0.5x$ , the vertical one by $x\sin 90^\circ = x$ . Level difference: $x(\sin\theta_1 + \sin\theta_2) = 1.5x$ — gravity taxes vertical rise only.

Step 2. Inclined-arms formula: $T = 2\pi\sqrt{\dfrac{L}{g(\sin\theta_1 + \sin\theta_2)}} = 2\pi\sqrt{\dfrac{0.60}{9.8 \times 1.5}} = 2\pi\sqrt{0.0408} = 2\pi \times 0.202 = \boxed{1.27\ \mathrm{s}}$ .

Step 3. Vertical U-tube benchmark: $T = 2\pi\sqrt{\dfrac{0.60}{19.6}} = 2\pi \times 0.175 = 1.10\ \mathrm{s}$ . Tilting one arm from $90^\circ$ to $30^\circ$ weakened the restoring ( $1.5g$ effective instead of $2g$ ) and slowed the slosh by 15%.

Step 4. Limit audits, the JEE-Advanced hygiene: both arms at $90^\circ$ recovers $2\pi\sqrt{L/2g}$ ✓; flatten one arm to $0^\circ$ and only the other surface pays, $T = 2\pi\sqrt{L/g\sin\theta_1}$ ; flatten both and $T \to \infty$ ✓. Timing note: ~3–3.5 min. The single idea: vertical height, not tube distance, is what gravity charges — everything else is the standard pattern. Distractor watch: using $\cos\theta$ (angle measured from the wrong reference) and halving $L$ "because two arms" — $L$ is the whole thread, already.

Exam traps

The factor 2. Level difference $= 2x$ , restoring $= \rho Sg(2x)$ , effective gravity-stiffness $2\rho Sg$ . Its absence turns $T$ into $2\pi\sqrt{L/g}$ — the most-planted wrong option in this archetype.

Whole-thread inertia. The mass in Newton's law is $\rho SL$ — arms and bend. Using the displaced slab's mass (or only the vertical arms' liquid) breaks the derivation at Step 3.

$L$ is the liquid's length, not the tube's height. A tube 30 cm tall can hold a 70 cm thread. Convert given data (mass, volume, fill height per arm) to total thread length first, then clock it.

Density and bore are bait. They cancel in the period. Options featuring $\rho$ or $S$ in the final $T$ are wrong by construction; data mentioning them exists to feed the $L$ conversion (Example 3) or to distract.

Equivalent pendulum is $L/2$ . Not $L$ . "A liquid column of length 1 m oscillates like a pendulum of length…" — 50 cm, Newton's own theorem.

Inclined arms use $\sin$ of the angle from the horizontal. Vertical rise per unit tube-displacement is $\sin\theta$ ; the vertical tube's $\sin 90^\circ = 1$ is the sanity anchor. A $\cos$ in your formula means the angle was measured from the vertical — re-read the figure.

Which $g$ ? JEE convention 10 (clean: $L = 50$ cm gives $T = 2\pi\sqrt{0.025} = 0.99$ s ≈ 1 s); with 9.8 the same column is 1.00 s. The Moon question tags this clock with the pendulum and float, not the spring: $T \times \sqrt6$ .

Origins · 1 minIsaac Newton and how this was discovered

Isaac Newton

1642–1727 · Cambridge & London

Newton put the U-tube in print before "simple harmonic motion" had a name. Proposition 44 of Principia Book II (1687) analyses water reciprocating in a bent channel and proves it oscillates isochronously — amplitude-independent — with the same period as a pendulum of half the column's length: precisely this module's $L/2$ theorem, derived geometrically two decades before calculus notation could tidy it. The result was no curiosity to him; Book II is Newton's assault on fluid resistance and wave motion, and the sloshing channel was his stepping stone to estimating the speed of water waves and, ultimately, sound. It is a characteristic Newton move: take the era's hardest continuous-media problem, find inside it a clean oscillator obeying the pendulum law, and nail the equivalence exactly. Engineers tuning liquid column dampers in skyscrapers today are running Proposition 44 with better plumbing.

History

1687

— Newton's Principia, Book II Prop. 44: liquid oscillating in a U-shaped channel keeps time with a pendulum of half its length.

1738

— Daniel Bernoulli's Hydrodynamica systematises liquid oscillation problems alongside the energy methods that solve them fastest.

1890

s — Forel's studies of Lake Geneva's seiches show whole lakes sloshing as giant open-channel oscillators.

1911

— Frahm fits ships with anti-roll tanks: U-tubes of water tuned so their slosh opposes the ship's rolling.

1980

s — Tuned liquid column dampers enter skyscraper design; buildings begin carrying Newton's Prop. 44 at their crowns.

Going further

Texts. HC Verma, Concepts of Physics Vol. 1, Ch. 12 (U-tube oscillation) · DC Pandey SHM chapter, liquid-column drill set · Newton, Principia Book II, Prop. 44 (in translation) for the original half-length theorem · NCERT Ch. 14 for the SHM base.

Provenance. All results are standard JEE canon. A verifier should recompute: Example 1's $T = 0.63$ s and 10 cm equivalent pendulum; Example 2's shared $T = 1.00$ s at $L = 50$ cm; Example 3's $L = 0.294$ m and $T = 0.76$ s; Example 4's $1.27$ s (inclined, $1.5g$ ) vs $1.10$ s (vertical) with the limit audits.

Prerequisites. Periodic Motion & the Definition of SHM · The Bobbing Float (archetype). Leads to. Time-Period Pattern Recognition (technique).

Why is the level difference 2x and not x? Only one side was pushed.

Because the liquid is incompressible and the tube is closed at the bottom: whatever volume leaves one arm must arrive in the other. Depress the left meniscus by $x$ and that missing liquid stands in the right arm — the right surface rises by exactly $x$ (equal bore). One surface at $-x$ , the other at $+x$ : the difference — which is what pressure at the bend cares about — is $2x$ . The push created the imbalance; the plumbing doubled it. Watch for the unequal-bore generalisation, where the rises differ and the bookkeeping follows the volumes — a JEE-Advanced dressing of the same idea.

Why must I use the whole column's mass when only the little 2x slab is "out of place"?

Because force and moving mass answer different questions. The force is set by the imbalance — the $2x$ slab's weight is what pressure fails to cancel. But that force must accelerate everything that moves, and in an incompressible thread every particle moves together: liquid deep in the bend accelerates exactly as hard as the menisci. So Newton's law reads (weight of $2x$ slab) $= (\rho SL) \times a$ . Using the slab's mass on both sides would predict $a = g$ always — a freely falling slab, which the trapped liquid plainly is not.

How can mercury and water have the same period when mercury is so much heavier?

The same way heavy and light pendulum bobs tie: gravity is democratic. Density multiplies the restoring slab's weight (more force) and the thread's inertia (more reluctance) by the same 13.6, and the ratio — which is all $\omega^2$ is — forgets it. Any oscillator whose stiffness and inertia come from the same material shows this cancellation; it is the liquid version of the hammer-and-feather. What density does buy mercury is persistence: relative to its huge inertia, viscous drag is small, so mercury's oscillation is lightly damped and famously long-ringing, while water's dies in a few swings.

Where does Newton's "pendulum of half the length" come from physically?

Match the clocks. Pendulum: $T = 2\pi\sqrt{L_{\text{pend}}/g}$ — gravity acts on the bob at full strength. U-tube: displace by $x$ and gravity's net levy is the $2x$ slab — per unit of moving mass, twice the pull a pendulum of length $L$ would feel at the same $x/L$ . Twice the stiffness at the same inertia means the column ticks like a pendulum shorter by that factor 2: $L_{\text{pend}} = L/2$ . The doubled level difference is the entire content of the theorem — Newton's geometry and your three lines of algebra agree across 340 years.

Is a harbour seiche really this same oscillator?

The same family, one storey up in sophistication. A seiche is water sloshing between two boundaries — a lake, bay or swimming pool as an open U-tube whose "arms" are the basin's ends. The restoring physics is identical (raised water at one end, gravity levying the difference); what changes is that the water column isn't a thin uniform thread, so the period follows the basin's length and depth (Merian's formula) rather than $L/2g$ . Lake Geneva's fundamental seiche runs about 73 minutes; storm surges and even distant earthquakes excite it. Same physics on your bench and across twenty kilometres of lake — with tuned liquid column dampers sitting exactly between the scales.

What would change if one arm of the U-tube were sealed at the top?

You would add a second spring. As the liquid rises into the sealed arm it compresses the trapped air, which pushes back — a gas spring in parallel with gravity's $2\rho Sg$ . The oscillation stiffens and quickens, by an amount set by the trapped column's length and pressure (Boyle's law linearised for small $x$ ). That hybrid — gravity plus gas stiffness — is genuinely a thermodynamics-meets-SHM problem and appears in JEE-Advanced as such; at this module's level you need only the classification reflex: sealed arm = extra restoring = shorter period, and the open U-tube formula as the limiting case of a very long air column.