ARCHETYPE: Drops Merging & Terminal Velocity

A cloud is a suspension of a billion droplets per cubic metre, each a few micrometres across, each falling — as the viscosity article computed — at a centimetre per second or less. Any updraught keeps them airborne indefinitely. So why does it ever rain? Because droplets are not loners: they collide and merge. Every merger conserves volume but re-scales everything else — a drop of double radius falls four times faster (Stokes' $r^2$ ), sweeps up more stragglers, grows further, falls faster still. Coalescence is a runaway, and the runaway is the difference between fog and a downpour. Examiners love this setup because one little scenario — $n$ droplets, each of radius $r$ , falling at terminal speed $v$ , merge into a single drop — audits three chapters at once. Volume conservation fixes the new radius ( $n^{1/3}r$ : cube root, the first trap). Stokes' law fixes the new terminal speed ( $n^{2/3}v$ : the second trap). And surface tension pays out an energy dividend — the merged drop has less total surface than the swarm, so surface energy is released, warming the water by a computable (and delightfully tiny) amount. Run in reverse — one drop sprayed into $n$ — every sign flips: slower fall, energy absorbed, slight cooling. This is a JEE-Main and NEET perennial (the $n^{2/3}$ ratio is close to an annual event) with a JEE-Advanced tail (the temperature-rise formula). It is also the lane where two concept modules — terminal velocity and surface tension — are made to work one problem together.

Key takeaways

Volume conservation: $n$ droplets of radius $r$ merge into one of radius $R = n^{1/3}\,r$ — cube root, always.
Stokes terminal velocity scales as radius squared, so the merged drop falls at $v' = n^{2/3}\,v$ .
Surface area shrinks on merging: from $n\,4\pi r^2$ to $4\pi R^2 = n^{2/3}\cdot 4\pi r^2$ — energy $E = T\,\Delta A$ is released.
Released energy per merged drop: $E = 4\pi T R^3\left(\dfrac1r - \dfrac1R\right)$ ; if it all heats the water, $\Delta\theta = \dfrac{3T}{\rho c}\left(\dfrac1r - \dfrac1R\right)$ .
Splitting reverses everything: $v' = v/n^{2/3}$ , energy $T\Delta A$ absorbed (work must be supplied; slight cooling).
The exponent ladder to memorize: radius $n^{1/3}$ · area $n^{2/3}$ · terminal speed $n^{2/3}$ · drag force at terminal $n$ (= weight ratio).
Key idea: merging conserves volume ( $r \to n^{1/3}r$ ) but not area or speed — $v_t$ jumps by $n^{2/3}$ and the lost surface pays an energy dividend.

The core insight

When $n$ identical droplets falling at Stokes terminal velocity $v$ coalesce, volume conservation gives radius $n^{1/3}r$ , the $r^2$ law lifts the terminal velocity to $n^{2/3}v$ , and the destroyed surface area releases energy $T\Delta A$ — the complete exam scenario in three scalings.

Where this lives today

Warm-cloud rain is this archetype run as a chain reaction: micrometre cloud droplets collide, the slightly-larger winners fall faster ( $n^{2/3}$ compounding), sweep a widening column, and cross from cloud droplet (10 μm, 1 cm/s) to drizzle (100 μm, ~1 m/s... and beyond Stokes) to raindrop (1 mm, metres per second) in twenty minutes of coalescence — climate models carry an explicit "autoconversion" term for it, and cloud-seeding attempts to jump-start the cascade. Engineering runs the reverse gear on purpose: fuel injectors atomize diesel into micrometre mist (huge total area — fast evaporation and clean burn, paid for in $T\Delta A$ of pump work), spray dryers turn milk and coffee into powder through the same trade, and inkjet printheads fight satellite droplets — unwanted mini-drops pinched off by surface tension during ejection — using exactly the merge/split energy ledger. Even your shower demonstrates the archetype: droplets merging on the glass suddenly streak downward the moment their radius promotes their terminal physics past the sticking forces.

The setup

$n$ identical spherical droplets (radius $r$ , density $\rho$ , surface tension $T$ ) drift down at their common Stokes terminal velocity $v = \dfrac{2r^2(\rho - \sigma)g}{9\eta}$ . They coalesce into a single spherical drop of radius $R$ , which settles to its own terminal velocity $v'$ . Asked, in every permutation: $R$ , $v'$ , the energy released, and occasionally the temperature rise. The only physics inputs are the two concept modules; the archetype is their choreography.

Figure

The merger — volume conserved, area destroyed, speed promoted

The whole archetype compresses into an exponent ladder. With volume fixed, every quantity is some power of $n$ :

quantity	scaling	for $n = 8$
radius $R$	$n^{1/3}\,r$	$2r$
total surface area	$n^{2/3} \times 4\pi r^2$ vs $n\times4\pi r^2$ before	halved
terminal velocity $v'$	$n^{2/3}\,v$	$4v$
weight (= drag at terminal)	$n\times$ one droplet's	$8\times$
momentum at terminal	$n\cdot n^{2/3} = n^{5/3}$	$32\times$

The master solution pattern

Derivation AMerged radius and the n-to-the-two-thirds law

We derive the two headline results — $R = n^{1/3}r$ and $v' = n^{2/3}v$ — from volume conservation and the Stokes terminal-velocity formula. The strategy is pure scaling: never compute either terminal velocity, only their ratio, so $\eta$ , $g$ , and the densities all cancel. Watch the exponents: the cube root from volume, squared by the $r^2$ law — fractional exponents are exactly where hurried candidates bleed marks.

Step 1. Volume conservation (water is neither created nor compressed in the merger):

n\cdot\tfrac43\pi r^3 = \tfrac43\pi R^3 \quad\Rightarrow\quad R = n^{1/3}\,r

Step 2. Stokes terminal velocity from the viscosity module: $v_t = \dfrac{2 a^2(\rho-\sigma)g}{9\eta}$ for a sphere of radius $a$ — everything except the radius is shared between droplet and merged drop.

Step 3. Ratio:

\frac{v'}{v} = \left(\frac{R}{r}\right)^2 = \left(n^{1/3}\right)^2 = n^{2/3}

Why the merged drop must fall faster at zero net force: at terminal velocity, drag equals effective weight. Weight grew by $n$ (all the water is still there) but Stokes drag at a given speed grew only by $n^{1/3}$ (drag ∝ radius) — so the drop must speed up until drag catches up: $n^{1/3}\times n^{2/3} = n$ ✓, the books balance exactly. *Verified: units — pure ratios ✓. Limits — $n = 1$ : nothing changes ✓; $n = 8$ : radius 2×, speed 4× (the classic numbers) ✓. Consistency: drag at the new terminal $= 6\pi\eta R v' \propto n^{1/3}\cdot n^{2/3} = n$ = weight ratio ✓. Standard: NEET/JEE-Main archives; HC Verma Ch. 13-14 exercises.*

Go deeper · Derivation B · 1 minThe energy dividend and the temperature rise

Derivation BThe energy dividend and the temperature rise

We derive the energy released on coalescence and the resulting temperature rise — the JEE-Advanced tail of the archetype. Physics input from the surface-tension module: surface energy $= T\times$ area, so destroying area pays out $E = T\,\Delta A$ . Strategy: write both areas, factor cleverly using $nr^3 = R^3$ , then divide by the drop's heat capacity. The elegant final form contains only $(1/r - 1/R)$ — small drops in, big drop out, and the smaller the ingredients the bigger the payout.

Step 1. Area before: $n\cdot4\pi r^2$ . Area after: $4\pi R^2$ . Energy released:

E = T\,\Delta A = 4\pi T\,(n r^2 - R^2)

Step 2. Eliminate $n$ with volume conservation, $n = R^3/r^3$ , so $n r^2 = R^3/r$ :

E = 4\pi T\left(\frac{R^3}{r} - R^2\right) = 4\pi T R^3\left(\frac{1}{r} - \frac{1}{R}\right)

Step 3. If the energy heats the water (mass $m = \rho\cdot\tfrac43\pi R^3$ , specific heat capacity $c$ ):

\Delta\theta = \frac{E}{mc} = \frac{4\pi T R^3\left(\tfrac1r - \tfrac1R\right)}{\tfrac43\pi R^3 \rho\, c} = \frac{3T}{\rho c}\left(\frac{1}{r} - \frac{1}{R}\right)

Why $(1/r - 1/R)$ : surface-to-volume ratio of a sphere is $3/a$ — merging is a trade between the ingredients' surface-per-volume ( $3/r$ , large) and the product's ( $3/R$ , small); the released energy per unit volume is $T$ times that difference. *Why the effect is real but tiny: for water, $3T/\rho c = 3(0.072)/(1000\times4200) = 5.1\times10^{-8}\ \mathrm{K\cdot m}$ — micro-kelvins unless $1/r$ is enormous, i.e. unless the ingredient droplets are micrometre-scale. Worked example 3 runs the honest number. Verified: units — $\mathrm{(N/m)\,m^{-1}/(kg/m^3 \cdot J/(kg\,K))} = \mathrm{K}$ ✓. Limits — $R = r$ ( $n=1$ ): zero ✓; splitting ( $R < r$ roles swapped): negative, i.e. energy absorbed, cooling ✓. Standard: JEE-Advanced archive; HC Verma surface-tension exercises.*

The reverse gear. Spraying one drop into $n$ droplets flips every sign: radius down by $n^{1/3}$ , terminal speed down by $n^{2/3}$ (why mist hangs — atomization is anti-rain), area up by $n^{1/3}$ , and the energy $T\Delta A$ must be supplied — by the sprayer's work — with a slight adiabatic cooling. The surface-tension article's atomizer example was this row of the ledger; the archetype's symmetry is worth saying in the exam: merging releases, splitting absorbs.

Figure

The coalescence cascade — why merging makes rain

Limiting cases

case	result	meaning
$n = 1$	everything unchanged, $E = 0$	the null test every formula must pass
$n = 8$	$R = 2r$ , $v' = 4v$ , area halved	the canonical exam numbers
$n \to$ large	$v' \sim n^{2/3}v$ grows without bound — until Stokes fails	big raindrops exit the laminar regime; real rain caps near 9 m/s
$r \to$ μm-scale ingredients	$\Delta\theta$ approaches measurable	fine fog pays the best energy dividend per volume
splitting ( $1 \to n$ )	$v/n^{2/3}$ , energy absorbed	atomizers, spray cooling, why mist floats
two unequal drops merge	use volumes, not the $n$ -formulas	$R^3 = r_1^3 + r_2^3$ — the general-case fallback

Common trap“Eight droplets merged, so the drop is 8× heavier and needs 8× the speed for drag to balance”

The instinct is half right — drag must indeed grow to $8\times$ — but Stokes drag is $6\pi\eta R v$ : it grew by $n^{1/3} = 2$ through the radius already, so the speed needs to supply only the remaining $n^{2/3} = 4$ . Speed jumps $4\times$ , not $8\times$ . The mirrored slip assumes "speed scales like radius" and answers $2v$ . Both wrong exponents — $n$ and $n^{1/3}$ — sit among the options on every version of this MCQ; the ladder (radius $\tfrac13$ · area/speed $\tfrac23$ · weight 1) is your armour. And a second ambush rides with it: all of this is Stokes-regime physics — micrometre-to-sub-millimetre drops. A "merged drop" question with millimetre ingredients and $n$ in the thousands would exit the laminar warranty; exams stay inside it, but the honesty clause ("assuming Stokes' law remains valid") is a free rigour mark on long-form papers.

Worked examples

1 · Board level — the canonical eight (NEET/JEE-Main pattern, ~90 seconds)

Eight identical water droplets, each of radius $1.0\ \mathrm{mm}$ , fall through air at a terminal velocity of $5.0\ \mathrm{cm/s}$ each. They coalesce into a single drop. Find (a) the drop's radius and (b) its terminal velocity.

Step 1. (a) Cube root of the count: $R = n^{1/3} r = 8^{1/3}\times1.0 = \boxed{2.0\ \mathrm{mm}}$ .

Step 2. (b) The $r^2$ law as a ratio: $v' = n^{2/3}v = 8^{2/3}\times5.0 = 4\times5.0 = \boxed{20\ \mathrm{cm/s}}$ .

Step 3. Balance check: weight grew $8\times$ ; drag $\propto Rv'$ grew $2\times4 = 8\times$ ✓ — terminal condition holds without computing a single force.

2 · Board level — energy released by a merger (JEE-Main pattern, ~2 minutes)

One thousand water droplets of radius $0.10\ \mathrm{mm}$ merge into a single drop. ( $T = 0.072\ \mathrm{N/m}$ .) Find (a) the final radius and (b) the surface energy released.

Step 1. (a) $R = 1000^{1/3}\times0.10 = 10\times0.10 = \boxed{1.0\ \mathrm{mm}}$ .

Step 2. (b) Areas: before $= n4\pi r^2 = 1000\times4\pi(10^{-4})^2 = 4\pi\times10^{-5}\ \mathrm{m^2}$ ; after $= 4\pi(10^{-3})^2 = 4\pi\times10^{-6}\ \mathrm{m^2}$ . Lost area $\Delta A = 4\pi\times(10^{-5} - 10^{-6}) = 4\pi\times9.0\times10^{-6} = 1.13\times10^{-4}\ \mathrm{m^2}$ — 90% of the surface vanished.

Step 3. $E = T\Delta A = 0.072\times1.13\times10^{-4} = \boxed{8.1\times10^{-6}\ \mathrm{J}}$ .

Step 4. Structure note: with $n = 1000$ , area shrinks by the factor $n^{1/3} = 10$ — the "before" term dominates, so $E \approx T\cdot n4\pi r^2$ to 10%: a fast estimate for MCQ cross-checking.

3 · Competitive level — the temperature rise, honestly (JEE-Adv pattern, ~3 minutes)

Fog droplets of radius $1.0\ \mathrm{\mu m}$ coalesce into a raindrop of radius $1.0\ \mathrm{mm}$ . Assuming all released surface energy heats the water ( $T = 0.072\ \mathrm{N/m}$ , $\rho = 1000\ \mathrm{kg/m^3}$ , $c = 4200\ \mathrm{J\,kg^{-1}K^{-1}}$ ), find the temperature rise. Comment on the size of the answer.

Step 1. The closed form from the derivation: $\Delta\theta = \dfrac{3T}{\rho c}\left(\dfrac1r - \dfrac1R\right)$ .

Step 2. Prefactor: $\dfrac{3\times0.072}{1000\times4200} = 5.14\times10^{-8}\ \mathrm{K\cdot m}$ .

Step 3. Bracket: $\dfrac{1}{10^{-6}} - \dfrac{1}{10^{-3}} = 10^{6} - 10^{3} \approx 9.99\times10^{5}\ \mathrm{m^{-1}}$ .

Step 4. $\Delta\theta = 5.14\times10^{-8}\times9.99\times10^{5} = \boxed{0.051\ \mathrm{K}}$ — a twentieth of a degree, from a million-fold radius consolidation. Comment: surface energy is real but minuscule against water's enormous heat capacity; only micrometre ingredients make even this much. (An option offering tens of kelvin fails the sanity gate instantly.) Timing note: the closed form makes this 90 seconds; re-deriving via areas costs three minutes. Note the bracket is utterly dominated by the SMALL radius — for quick work, $\Delta\theta \approx 3T/\rho c r$ , and the big drop's term is a 0.1% correction.

4 · Competitive level — the reverse gear (JEE-Main pattern, ~2 minutes)

A spherical drop falling at its Stokes terminal velocity of $32\ \mathrm{cm/s}$ breaks into 64 identical droplets. Find (a) each droplet's terminal velocity and (b) whether energy was released or absorbed in the break-up, with the factor by which total surface area changed.

Step 1. (a) Splitting: $v' = \dfrac{v}{n^{2/3}} = \dfrac{32}{64^{2/3}} = \dfrac{32}{16} = \boxed{2.0\ \mathrm{cm/s}}$ .

Step 2. (b) Area after/before $= n^{1/3} = 64^{1/3} = 4$ : total surface quadrupled, so energy $T\Delta A$ was absorbed — supplied by whatever tore the drop (aerodynamic forces, the sprayer), with a slight cooling if adiabatic.

Step 3. Sanity: the fragments fall 16× slower — precisely why crop-sprayers and perfume atomizers produce hanging mists, and why break-up of large raindrops (which really happens above ~4 mm) recycles rain into slower drizzle. Timing note: reverse-gear items are the same ladder read right-to-left — divide by $n^{2/3}$ , area × $n^{1/3}$ , energy absorbed. If you catch yourself re-deriving, you've spent a minute buying nothing.

Exam traps

Cube root, then square. $R = n^{1/3}r$ (volume), $v' = n^{2/3}v$ (Stokes $r^2$ ). The planted wrong answers are $nv$ (weight-only reasoning), $n^{1/3}v$ (linear-radius reasoning), and $n^{1/2}v$ (panic). The ladder kills all three.

Merging releases; splitting absorbs. Direction of the energy flow = direction of the area change. Signs are a stated part of many answers; "energy is released when a drop is atomized" is auto-wrong.

Total area, not per-drop area. Splitting shrinks each droplet's area but grows the total by $n^{1/3}$ . The energy ledger runs on the total.

The $(1/r - 1/R)$ bracket. Small radius first; for merging the bracket is positive (heating), for splitting negative (cooling). Dominated overwhelmingly by the smaller radius — a useful estimation collapse and a units check ( $\mathrm{m^{-1}}$ ).

Unequal drops: back to volumes. $R^3 = r_1^3 + r_2^3$ ; the $n$ -power ladder assumes identical ingredients. A "2 mm and 1 mm drop merge" item wants $R = (8+1)^{1/3} = 9^{1/3}$ mm ≈ 2.08 mm, not any $n^{1/3}$ shortcut.

Stokes-regime honesty. $v' = n^{2/3}v$ inherits Stokes' warranty (small, slow, laminar). For millimetre-plus products, flag the assumption; real raindrops saturate near 9 m/s in the $v^2$ regime.

Momentum is not conserved into the answer. Droplets falling at terminal speed that merge do NOT keep $n\times$ momentum — the new drop accelerates to a faster terminal state; forces (gravity, drag) act throughout. Momentum-conservation reasoning belongs to collisions, not steady-state fall — a cross-chapter confusion the capstone module drills.

Which constants matter. Ratios need none of $\eta, g, \sigma$ ; energy needs $T$ ; temperature needs $\rho c$ . Reading what's given tells you which sub-question you're in.

Origins · 1 minRobert A. Millikan and how this was discovered

Robert A. Millikan

1868–1953 · University of Chicago & Caltech

Robert Millikan turned this archetype's physics into the most famous precision measurement of the early twentieth century. His oil-drop experiment (1909–1913) watched single micrometre droplets fall at Stokes terminal velocity between capacitor plates: the fall speed gave each drop's radius (via $v_t \propto r^2$ ), switching on the field gave the electric force, and the charges always came out as integer multiples of $1.6\times10^{-19}$ C — the electron, weighed one droplet at a time. He chose oil over water for a reason students of this article will appreciate: water droplets evaporate mid-measurement, changing $r$ and wrecking the $r^2$ bookkeeping. Millikan won the 1923 Nobel Prize (also citing his photoelectric work that confirmed Einstein's equation — reluctantly, as he'd set out to disprove it), built Caltech into a physics power, and left a methods lesson every examiner echoes: the drop's radius is measured by its fall.

History

1851

— Stokes' drag law: terminal velocity becomes computable, with its decisive

r^2

1904

— Philipp Lenard studies drop break-up and coalescence in waterfalls and sprays — surface energy in the field.

1909

— Millikan begins the oil-drop measurements: terminal-velocity radii as a precision tool.

1948

— Bowen and Ludlam formalize collision–coalescence as the engine of warm rain (no ice needed) — the merging-drops archetype as meteorology.

1980

s — Cloud radars watch drizzle "autoconversion": droplet populations crossing the merge-and-fall runaway in real time.

Going further

Texts. HC Verma Vol. 1, Ch. 13–14 exercises · NCERT Ch. 9 (surface energy) · JEE/NEET archives (the $n^{2/3}$ family). The regime-crossing (Stokes → quadratic) lives in the UG drag deep dive.

Provenance. All results standard. A verifier should recompute: the ladder's exponents and the drag-balance audit ( $n^{1/3}\cdot n^{2/3} = n$ ); example 1's 2.0 mm / 20 cm/s; example 2's $\Delta A = 1.13\times10^{-4}$ m² and $8.1\times10^{-6}$ J; example 3's $5.14\times10^{-8}\times9.99\times10^5 = 0.051$ K; example 4's 2.0 cm/s and area ×4.

Prerequisites. Viscosity, Stokes' Law & Terminal Velocity · Surface Tension & Capillarity. Leads to. Bubbles, Drops & Excess Pressure (archetype) · Option Elimination & Answer Checking (track capstone).

Why do the droplets merge at all — don't they bounce off each other?

Often they do! Colliding water droplets can bounce (an air film cushions them), coalesce, or even shatter, depending on sizes and impact speed — the "collision efficiency" of cloud physics. What tips the balance toward merging is drainage of that air film until the surfaces touch, whereupon surface tension does the rest: one sphere has less surface energy than two, so coalescence is energetically downhill and proceeds in microseconds. Exams grant the merger; meteorology spends careers on the efficiency factor.

Where exactly does the released surface energy go?

Watch a merger in slow motion: the combined blob oscillates wildly — prolate, oblate, ringing through shapes — before settling into a sphere. The surface energy first becomes this oscillation's kinetic energy, then viscosity damps the ringing into heat (with a tiny acoustic squeak radiated away — literally the sound of raindrops merging). The exam's "assume it all becomes heat" is the end state of that cascade; the temperature-rise formula prices it.

Does the merged drop reach its new terminal velocity instantly?

No — it emerges at roughly the old speed $v$ (momentum's brief say in the matter), finds itself far below its new terminal $n^{2/3}v$ , and accelerates toward it along the asymptotic approach curve from the viscosity article — never overshooting. The relaxation is quick for small drops (time constant $m/6\pi\eta R$ , milliseconds at these scales), which is why exam questions safely compare "terminal velocity before" with "terminal velocity after" and skip the transient. If a question asks about the moment just after merging: speed ≈ $v$ , acceleration directed downward, drag momentarily less than weight.

Why does rain hit at ~9 m/s if the n^(2/3) law says big drops should go arbitrarily fast?

Because the law inherits Stokes' warranty, and millimetre drops at metres per second are far outside it. In the real turbulent-wake regime drag grows as $v^2$ and area $r^2$ , giving $v_t \propto \sqrt{r}$ — a much flatter reward for growing. Compounded with drops flattening (more drag) and finally breaking up above ~4 mm, real raindrop speeds saturate near 9 m/s. The $n^{2/3}$ answer is correct for the exam's Stokes-regime premise and wrong for a thunderstorm — knowing which regime you're in is the transferable skill.

Millikan used charged droplets — does charge change the merging story?

Beautifully, yes. Charged droplets of the same sign repel, suppressing coalescence; opposite signs attract and merge enthusiastically. Thunderclouds are charge-separation machines, so electric forces genuinely modulate rain formation — a live research area (and one motivation behind ionization-based rain-enhancement proposals). Also, charge itself resists confinement: a highly charged drop becomes unstable and splits when electrostatic repulsion beats surface tension (the Rayleigh limit, 1882) — the operating principle of electrospray, which the EM track will price properly.

Is there a per-volume way to see why small droplets store so much surface energy?

Yes — the sphere's surface-to-volume ratio is $3/a$ : energy per unit volume of "being droplets of radius $a$ " is $3T/a$ . A cubic metre of water as 1 μm droplets carries $3(0.072)/10^{-6} \approx 2\times10^{5}$ J of surface energy; as one blob, essentially none. That $1/a$ divergence is why atomization is costly, why fine powders and emulsions are energetic (and sometimes explosive — flour dust), and why the temperature-rise bracket $(1/r - 1/R)$ cares almost only about the small radius.