Elastic Collisions in 1D

Lift the end ball of a Newton's cradle and let it fall. Click — the four balls in the middle do not visibly move, and the far ball leaps out with the arriving ball's speed, as if the momentum had teleported through the line. Play the same trick on a billiard table: a straight, clean strike on a stationary ball stops the cue ball dead and sends the target away at the full incoming speed. Both are showing you the cleanest collision physics allows: the elastic collision, in which the kinetic-energy ledger balances to the last joule. Elasticity is a powerful straitjacket. A collision that only conserves momentum has a whole range of possible outcomes — sticky, bouncy, anything between. Demand that kinetic energy also survives and the outcome is completely determined: two equations, two unknowns, no freedom left. The results are worth owning outright: the velocity-exchange formulas, the equal-mass swap, the light ball rebounding off the heavy one, the heavy ball ploughing through — and the strange, exam-loved fact that a moving heavy object can fling a light one forward at up to twice its own speed, which is exactly how spacecraft steal speed from planets. This article derives the 1D machinery the fast way (a relative-velocity trick that kills the quadratic), stocks the three mass-ratio outcomes you should recognize on sight, and works the energy-transfer question — what mass ratio steals the most energy? — whose answer decided how humanity builds nuclear reactors.

Key takeaways

Elastic = both ledgers balance: momentum (every collision) and kinetic energy (the elastic promise). Two equations ⇒ the outcome is fully determined.
The working tool: in 1D elastic collisions the relative velocity reverses — separation speed = approach speed. (In restitution language: $e = 1$ .)
Closed forms: $v_1' = \dfrac{(m_1-m_2)v_1 + 2m_2v_2}{m_1+m_2}$ , $\;v_2' = \dfrac{(m_2-m_1)v_2 + 2m_1v_1}{m_1+m_2}$ — signed velocities, always.
The three sight-readable outcomes (target at rest): equal masses swap ( $v_1' = 0$ , $v_2' = v_1$ ) · heavy hits light: barrels on at $\approx v_1$ , light flies at $\approx 2v_1$ · light hits heavy: bounces back at $\approx v_1$ , heavy barely moves.
Energy transferred to a stationary target: fraction $\dfrac{4m_1m_2}{(m_1+m_2)^2}$ — maximum (100%) at equal masses, tiny for very unequal ones.
Truly elastic collisions live at the atomic scale (molecules, neutrons) and in ideal problems; steel balls and billiards come close; the word in the problem is your licence to use the KE equation.
Key idea: momentum conservation plus the reversal of relative velocity solves every 1D elastic collision — and the mass ratio alone dictates who leaves with the speed and who leaves with the energy.

The core insight

An elastic collision conserves kinetic energy as well as momentum, which in one dimension forces the relative velocity to exactly reverse (separation speed = approach speed); combining that linear rule with momentum conservation yields the closed-form outcomes — equal masses swap velocities, and the energy transferred to a stationary target peaks at 100% for equal masses, falling as $4m_1m_2/(m_1+m_2)^2$ for unequal ones.

Where this lives today

The equal-mass ideal runs the world's nuclear reactors: water's hydrogen nuclei match the neutron's mass, so each collision can strip up to all of a fission neutron's energy, thermalizing it in a couple of dozen bounces — worked example 3 is literally a moderator-design calculation. The moving-wall result ( $+2V$ per bounce) is NASA's free fuel: gravitational slingshots let Voyager, Cassini and every outer-planet mission "bounce" elastically off moving planets, gaining twice the planet's orbital speed per encounter — no propellant, pure conservation laws. The same $+2V$ physics, run trillions of times per second, is why compressing a gas heats it: molecules rebound off the advancing piston slightly faster than they arrived. And at the LHC and in every particle detector, elastic-scattering events — where the books balance perfectly — serve as the calibration standard against which the interesting, inelastic physics is measured, three and a half centuries after two carts on a Royal Society table started the subject.

The physical picture

Slow the collision down mentally. As the two bodies meet, they compress — microscopically for steel, visibly for a tennis ball — and the kinetic energy of approach pours into elastic strain energy, like loading an invisible spring. Then the spring pays back. The question that classifies every collision is: how much of the deposit is refunded? An elastic collision is the perfect bank: every joule stored during compression returns as kinetic energy during rebound. Nothing leaks into heat, sound, or permanent dents. (A silent thought experiment: a truly elastic collision would be inaudible — the click of the cradle is, strictly, a small elasticity tax.) Perfect refunds are an idealization, but a superbly useful one. Collisions between gas molecules and between neutrons and nuclei are genuinely elastic (there is no "heat" for a single molecule to hide energy in — at low energy the books must balance); hardened steel spheres and billiard balls refund 90-plus percent; the exam's phrase "perfectly elastic" is your contract to use both conservation laws.

Figure

The three outcomes to know on sight — mass ratio decides everything (target at rest)

The one-line summary of elasticity — and the tool you will actually compute with — is about the gap between the bodies: whatever speed the gap was closing at before contact, it opens at exactly that speed after. Approach speed equals separation speed. A superball meets the floor at 8 m/s, leaves at 8 m/s. Two carts closing at 5 m/s will be separating at 5 m/s, whatever their masses. This "reversal rule" is where the next article's coefficient of restitution comes from: real collisions separate slower than they approached, and the ratio (separation ÷ approach) — equal to 1 here — becomes the grading scale $e$ for everything less than perfect.

The math: two ledgers, one straitjacket

Set up the standard 1D encounter: masses $m_1$ and $m_2$ with signed velocities $u_1, u_2$ before and $v_1, v_2$ after, all along one line. Every collision — elastic or not — conserves momentum (internal forces cancel in third-law pairs, last article's law):

m_1u_1 + m_2u_2 = m_1v_1 + m_2v_2 \qquad (1)

where: velocities are signed; a leftward mover carries a minus sign into the ledger, and most wrong answers in this topic are dropped signs. One equation, two unknowns ( $v_1, v_2$ ): momentum alone cannot decide the outcome — a sticky crash and a perfect bounce can share the same total $p$ . Elasticity is the second promise — the kinetic-energy ledger also balances:

\tfrac12m_1u_1^2 + \tfrac12m_2u_2^2 = \tfrac12m_1v_1^2 + \tfrac12m_2v_2^2 \qquad (2)

where: only kinetic energy appears — no heat, no sound, no dents allowed. Two equations, two unknowns: the outcome is now unique. But (2) is quadratic, and brute-force simultaneous solving wastes exam minutes. The classic manoeuvre (derivation A) collapses the pair into a linear rule:

u_1 - u_2 = -(v_1 - v_2) \qquad \text{(approach speed = separation speed)} \qquad (3)

where: $u_1 - u_2$ — the rate the gap closes before; $v_2 - v_1$ — the rate it opens after. Equal, always, for 1D elastic collisions, independent of the masses. In restitution language this is $e = 1$ ; the next article grades every lesser collision on the same scale. In practice you solve collisions with (1) + (3) — two linear equations — and reserve (2) as a free final check. Solving (1) and (3) simultaneously (derivation B) gives the closed forms:

v_1 = \frac{(m_1 - m_2)\,u_1 + 2m_2 u_2}{m_1 + m_2}, \qquad v_2 = \frac{(m_2 - m_1)\,u_2 + 2m_1 u_1}{m_1 + m_2} \qquad (4)

where: each final velocity mixes a "keep some of your own" term (weighted by the mass difference) with a "receive from the partner" term (weighted by twice the partner's mass). Note the symmetry — swap labels 1↔2 and the formulas swap. Sight-read the special cases before ever plugging numbers:

mass ratio (target $u_2 = 0$ )	$v_1$	$v_2$	picture
$m_1 = m_2$	$0$	$u_1$	the swap: cue ball stops, target leaves with everything
$m_1 \gg m_2$	$\approx u_1$	$\approx 2u_1$	freight train: barrels on, flings the light target at double speed
$m_1 \ll m_2$	$\approx -u_1$	$\approx \frac{2m_1}{m_2}u_1$	the wall: bounces back, wall barely moves
general	eq. (4)	eq. (4)	energy handed over: $\dfrac{4m_1m_2}{(m_1+m_2)^2}$ of the incoming KE

Derivation AThe relative-velocity reversal rule

We derive equation (3) — approach speed equals separation speed — from the two conservation laws, by the factor-and-divide trick that kills the quadratic. The strategy: rearrange both ledgers as transfer statements (what body 1 lost, body 2 gained), factor the energy statement as a difference of squares, and divide. Watch the division's fine print: it assumes $u_1 \neq v_1$ — i.e. that a collision actually happened — discarding the ghost solution where the bodies pass through each other unchanged.

Step 1. Momentum (1), rearranged as a transfer:

m_1(u_1 - v_1) = m_2(v_2 - u_2)

Step 2. Energy (2), rearranged the same way and factored:

m_1(u_1^2 - v_1^2) = m_2(v_2^2 - u_2^2) \;\;\Rightarrow\;\; m_1(u_1-v_1)(u_1+v_1) = m_2(v_2-u_2)(v_2+u_2)

Why factoring works: each side is a difference of squares — the algebra is begging you to reuse Step 1.

Step 3. Divide Step 2 by Step 1 (legal since $u_1 \ne v_1$ — the collision happened):

u_1 + v_1 = v_2 + u_2 \;\;\Rightarrow\;\; u_1 - u_2 = v_2 - v_1

Why this is the reversal rule: the left side is the closing speed of the gap before contact; the right side is its opening speed after. Masses cancelled wholesale — the rule is universal for elastic collisions in 1D.

Step 4. The discarded root: $v_1 = u_1$ (with $v_2 = u_2$ ) also solves both ledgers — the "no collision" solution. Algebra cannot tell contact from a miss; physics can. When your quadratic route hands you two answer pairs, this ghost is the one to reject. Verified: superball on floor (Earth as body 2, $u_2 = v_2 \approx 0$ ): rule gives $v_1 = -u_1$ — rebound at incoming speed ✓. Equal-mass swap: $u_1 - 0 = v_2 - v_1$ with momentum $u_1 = v_1 + v_2$ forces $v_1 = 0$ , $v_2 = u_1$ ✓.

Go deeper · Derivation B · 1 minThe closed forms — and the maximum-energy-transfer theorem

Derivation BThe closed forms — and the maximum-energy-transfer theorem

We derive the final velocities (4) from the two linear equations (momentum + reversal rule), then extract the energy fraction handed to a stationary target and find the mass ratio that maximizes it — the folded-in worked result that runs reactor design. The strategy: substitution between two linear equations (no quadratic anywhere), then evaluate $KE_2/KE_{\rm in}$ and read off its peak by symmetry. Watch the transfer fraction turn out symmetric in $m_1 \leftrightarrow m_2$ : a heavy ball hitting a light one wastes energy in exactly the same proportion as a light hitting a heavy.

Step 1. The linear system: $m_1u_1 + m_2u_2 = m_1v_1 + m_2v_2$ and $v_2 = v_1 + (u_1 - u_2)$ (reversal rule rearranged).

Step 2. Substitute $v_2$ into momentum and collect $v_1$ :

m_1u_1 + m_2u_2 = m_1v_1 + m_2v_1 + m_2(u_1 - u_2) \;\Rightarrow\; v_1 = \frac{(m_1-m_2)u_1 + 2m_2u_2}{m_1+m_2}

Back-substitute for $v_2 = v_1 + u_1 - u_2$ :

v_2 = \frac{(m_2-m_1)u_2 + 2m_1u_1}{m_1+m_2}

Step 3. Stationary target ( $u_2 = 0$ ): $v_2 = \dfrac{2m_1u_1}{m_1+m_2}$ . Energy handed over, as a fraction of the incoming $\tfrac12m_1u_1^2$ :

f = \frac{\tfrac12 m_2 v_2^2}{\tfrac12 m_1u_1^2} = \frac{m_2}{m_1}\cdot\frac{4m_1^2}{(m_1+m_2)^2} = \frac{4m_1m_2}{(m_1+m_2)^2}

Step 4. Maximize. Write $f = \dfrac{4x}{(1+x)^2}$ with $x = m_2/m_1$ : numerator and denominator balance best at $x = 1$ — check by symmetry ( $f$ is unchanged under $x \to 1/x$ , so its single peak must sit at the self-symmetric point $x=1$ ), giving $f_{\max} = \dfrac{4}{4} = 1$ . Equal masses transfer ALL the energy; mismatch in either direction strands energy with the projectile. Why exams love this: it is a formula, a maximization, and a real technology in one line — neutrons (mass 1) are slowed best by hydrogen (mass 1), which is why reactors moderate with water and why Chadwick could weigh the neutron from recoil speeds. *Verified: $m_1 = m_2$ : $f = 1$ and (4) gives the swap ✓. Neutron on carbon-12: $f = 48/169 = 28.4\%$ ✓ (worked example 3). Energy check of (4) for 2 kg at 6 m/s on 4 kg at rest: $v_1 = -2, v_2 = 4$ ; KE $36 = 4 + 32$ ✓.*

Figure

The energy-transfer curve — matched masses steal everything, mismatch strands it

Limiting cases

case	result	meaning
$m_1 = m_2$ , any $u_2$	velocities swap	equal masses trade motion completely (cradle, billiards)
$m_2 \to \infty$ , target at rest	$v_1 \to -u_1$ , $v_2 \to 0$	the wall: perfect rebound, no transfer
$m_1 \to \infty$	$v_1 \to u_1$ , $v_2 \to 2u_1$	the freight train: light target flung at double speed
moving wall ( $m_2 \gg m_1$ , $u_2 \ne 0$ )	$v_1 \to 2u_2 - u_1$	bat, racket, planet slingshot: rebound + twice the wall's speed
$u_2 = u_1$	nothing happens	no gap closing, no collision
transfer fraction	$4m_1m_2/(m_1+m_2)^2$ , peak 1 at $m_1 = m_2$	mismatch strands energy with the projectile

Common trapThey bounced apart, so the collision was elastic

"Elastic" is an energy audit, not a description of bouncing. Almost every real bounce is partly inelastic — a tennis ball leaves the floor slower than it arrived, the cradle clicks (sound is escaped energy), even billiard balls refund only ~95%. The exam consequences run both ways. You may only invoke the KE equation (2) — or its offspring, the reversal rule and closed forms (4) — when the problem says "elastic" (or "perfectly elastic"); using the swap or the $2u$ ceiling on a merely "bouncy" collision imports a false ledger. Conversely, bodies that don't stick are not automatically elastic: between the perfect refund ( $e=1$ ) and the perfect stick ( $e=0$ ) lies the whole graded spectrum of real collisions, which is the next article's territory. The word "elastic" in a problem is a contract; no word, no contract — then momentum (always true) plus the given data is all you may write.

Worked examples

1 · Board level — the standard two-cart collision (CBSE/ISC pattern, 3 marks)

A $2.0\ \mathrm{kg}$ cart moving at $6.0\ \mathrm{m/s}$ collides elastically with a stationary $4.0\ \mathrm{kg}$ cart. Find both final velocities and verify the kinetic-energy ledger.

Step 1. Closed forms with $u_2 = 0$ : $v_1 = \dfrac{(2-4)(6.0)}{6} = \boxed{-2.0\ \mathrm{m/s}}$ (it bounces back — lighter than the target); $v_2 = \dfrac{2(2)(6.0)}{6} = \boxed{+4.0\ \mathrm{m/s}}$ .

Step 2. Momentum audit: before $2(6) = 12$ ; after $2(-2) + 4(4) = -4 + 16 = 12$ ✓.

Step 3. Energy audit: before $\tfrac12(2)(36) = 36$ J; after $\tfrac12(2)(4) + \tfrac12(4)(16) = 4 + 32 = 36$ J ✓.

Step 4. Reversal-rule audit (free): approach $6.0$ ; separation $4.0 - (-2.0) = 6.0$ ✓ — three independent checks, ten seconds.

2 · Board level — the equal-mass swap and the cradle (ICSE/A-Level explain-plus-compute pattern)

Ball A ( $0.50\ \mathrm{kg}$ , $3.0\ \mathrm{m/s}$ ) elastically strikes identical ball B moving the same way at $1.0\ \mathrm{m/s}$ . Find the final velocities, and use the result to explain the Newton's cradle.

Step 1. Equal masses in 1D swap velocities — with a moving target too (put $m_1 = m_2$ in (4): $v_1 = u_2$ , $v_2 = u_1$ ). So A leaves at $\boxed{1.0\ \mathrm{m/s}}$ , B at $\boxed{3.0\ \mathrm{m/s}}$ .

Step 2. Audit: momentum $0.5(3+1) = 2.0$ before, $0.5(1+3) = 2.0$ after ✓; the KE ledger balances identically (same speeds, relabelled).

Step 3. The cradle: the arriving ball swaps velocities with ball 2 (stops); ball 2 swaps with ball 3… a bucket brigade of swaps down the line, until the last ball — with no partner — leaves at the original speed. Momentum alone would allow two balls at half speed, but that would break the energy ledger ( $2 \times \tfrac12 m(v/2)^2 = \tfrac14 mv^2 \ne \tfrac12mv^2$ ): the cradle is nightly proof that both laws operate.

3 · Competitive level — slowing a neutron (JEE-Main/NEET pattern, ~3 minutes)

A neutron (mass $1$ u) makes a head-on elastic collision with a stationary carbon nucleus ( $12$ u). What fraction of the neutron's kinetic energy is transferred? Compare hydrogen ( $1$ u) and lead ( $207$ u) as targets, and hence explain why reactors use water or graphite — not lead — as moderators.

Step 1. Transfer fraction: $f = \dfrac{4m_1m_2}{(m_1+m_2)^2}$ . Carbon: $f = \dfrac{4(1)(12)}{13^2} = \dfrac{48}{169} = \boxed{28\%}$ per collision.

Step 2. Hydrogen: $f = \dfrac{4}{4} = \boxed{100\%}$ (equal masses — the swap). Lead: $f = \dfrac{4(207)}{208^2} \approx \boxed{1.9\%}$ .

Step 3. The engineering read: a fission neutron must shed ~99.99999% of its energy to reach "thermal" speeds. Against hydrogen (water) that takes ~20 collisions; against carbon (graphite) ~110; against lead, thousands — the neutron just bounces off, keeping its energy, like the light ball off the wall. Moderators are chosen by this formula.

Step 4. MCQ audit: $f$ is symmetric in the masses and peaks at 1 for equal — any option exceeding 1 or asymmetric is dead on arrival. Timing note: ~3 min including the explanation sentence. Memorize $f = 4m_1m_2/(m_1+m_2)^2$ outright — deriving it mid-exam costs four minutes; quoting it costs ten seconds, and it appears in JEE/NEET in nuclear clothing more often than in cart clothing.

4 · Competitive level — the moving wall's gift of 2V (JEE-Main pattern, ~3 minutes)

A ball moving at $5.0\ \mathrm{m/s}$ meets a massive truck head-on; the truck approaches at $10\ \mathrm{m/s}$ and its velocity is unchanged by the collision. If the collision is elastic, find the ball's rebound speed.

Step 1. Reversal rule — no closed forms needed. Approach speed: gap closes at $5 + 10 = 15\ \mathrm{m/s}$ .

Step 2. Elastic ⇒ separation speed is also $15\ \mathrm{m/s}$ . The truck still moves at $10\ \mathrm{m/s}$ (massive), so the ball must be moving at $10 + 15 = \boxed{25\ \mathrm{m/s}}$ in the truck's direction.

Step 3. Pattern to keep: rebound speed $= u + 2V$ — your own speed back, plus twice the wall's. Check with the closed form's heavy limit $v_1 \to 2u_2 - u_1 = 2(10) - (-5) = 25$ ✓ (signs: truck's direction positive).

Step 4. This " $+2V$ " is a bat lifting a cricket ball, a racket's smash — and a gravitational slingshot: a spacecraft "bounces" elastically off a planet's moving gravity well and leaves with two planetary speeds added. Same formula, no contact required. Timing note: ~3 min via closed forms, ~60 s via the reversal rule — which is the point. For any "massive partner" problem (walls, bats, planets, pistons), go straight to approach = separation; the heavy body's velocity is unchanged and the light body's answer falls out in one line. Compressing gas heats up for exactly this reason: molecules rebound off the incoming piston at $+2V$ each hit.

Exam traps

Signs before formulas. The closed forms (4) use signed velocities. A head-on approach means $u_2$ is negative if $u_1$ is positive. Feeding speeds (all positive) into (4) is the single biggest error source in this topic.

Using elastic machinery without the contract. No "elastic" in the problem ⇒ no equation (2), no reversal rule, no closed forms, no swap. Momentum alone, plus whatever else is given ( $e$ , sticking, energy loss).

Misremembered closed forms. The checks are free: (4) must give the swap at $m_1 = m_2$ , $v_1 \to -u_1$ for a wall, and $v_2 \to 2u_1$ for a train. A formula that fails any corner is misremembered — the limiting-cases technique, applied to your own memory.

The ghost root. Solving (1)+(2) as simultaneous equations yields TWO solutions; one is "nothing happened" ( $v_1 = u_1$ ). If your algebra returns the initial velocities, you found the ghost — take the other root, or better, use (1)+(3) and never meet it.

"Equal masses swap" quoted without elasticity. The swap needs BOTH ledgers. Equal clay balls do not swap; they stick. The cradle argument (example 2) shows why: momentum alone permits many outcomes.

Transfer fraction confused with retained fraction. $f = 4m_1m_2/(m_1+m_2)^2$ is what the target receives; the projectile keeps $1 - f = \left(\frac{m_1-m_2}{m_1+m_2}\right)^2$ . MCQs offer both; read which body the question asks about.

Approach speed mis-assembled. Closing speed for head-on movers is $u_1 + u_2$ (speeds toward each other add); for a chase it is $u_1 - u_2$ . Draw the two arrows before writing (3).

Assuming 1D machinery in 2D. A glancing billiard shot is a two-dimensional collision — the swap and the closed forms don't apply; that geometry (and the 90° theorem) is two articles ahead. The 1D forms govern head-on collisions only.

Origins · 1 minJames Chadwick and how this was discovered

James Chadwick

1891–1974 · Cavendish Laboratory, Cambridge

In 1932 Chadwick had an invisible problem: a mysterious neutral radiation from beryllium that no detector could see directly. His solution was pure elastic-collision arithmetic. He let the radiation strike targets of hydrogen and nitrogen and measured the recoil speeds of the struck nuclei — and since an elastic collision's outcome depends only on the mass ratio, two recoil measurements pinned the projectile's mass: about 1 u, neutral — the neutron. It was collision kinematics as a weighing scale, and it won him the 1935 Nobel Prize. The same formulas immediately turned practical: fission neutrons run too fast to sustain a chain reaction, and the energy-transfer fraction $4m_1m_2/(m_1+m_2)^2$ — maximal for matched masses — told engineers to slow them with the lightest nuclei available. Fermi's 1942 Chicago pile used graphite; most power reactors since use water, whose hydrogen is the neutron's equal-mass ideal partner. Every worked example 3 on every exam is Chadwick's calculation, re-run.

History

1668

— the Royal Society's collision challenge: Wren and Huygens deliver the elastic rules, Wallis the inelastic.

1687

— Newton's Principia absorbs the collision laws into the scholium of the laws of motion, restitution experiments included.

1738

— Bernoulli's Hydrodynamica models gases as elastically colliding molecules — kinetic theory's opening bet.

1932

— Chadwick weighs the neutron using elastic recoil kinematics: collision algebra as an instrument.

1942

— Fermi's CP-1 goes critical on graphite-moderated neutrons — the transfer-fraction formula, built at reactor scale.

Going further

Texts. NCERT Physics Class 11, Ch. 6 (collisions section) · HC Verma, Concepts of Physics Vol. 1, Ch. 9 · Morin, Introduction to Classical Mechanics, Ch. 5 — and Kleppner & Kolenkow Ch. 4 for the centre-of-mass frame view.

Provenance. All results are standard. A verifier should recompute: the reversal rule via the factor-and-divide route; the closed forms from the linear pair; the transfer fraction $4m_1m_2/(m_1+m_2)^2$ with peak 1 at equal masses; Example 1 ( $-2.0$ and $+4.0$ m/s; ledgers $12 = 12$ , $36 = 36$ ); Example 3 (carbon $48/169 = 28.4\%$ , hydrogen 100%, lead $\approx 1.9\%$ ); Example 4 ( $25$ m/s via approach = separation $= 15$ ).

Prerequisites. Conservation of Linear Momentum · Kinetic Energy & the Work–Energy Theorem. Leads to. Inelastic Collisions & Coefficient of Restitution · Collisions in Two Dimensions · Two Blocks & a Spring (archetype) · Elastic Collisions (UG deep dive).

Do perfectly elastic collisions actually exist, or is this all an idealization?

At the atomic scale they genuinely exist: two low-energy gas molecules, or a slow neutron and a nucleus, have no accessible place to hide energy — no permanent dents, no heat reservoir inside a single particle — so the kinetic-energy books must balance exactly. At everyday scale, elasticity is a spectrum: hardened steel balls and billiard balls refund 90–95%, a superball ~85%, clay ~0%. The idealization earns its keep because the perfect case brackets reality from one side (the perfect stick brackets it from the other) and because many collisions are close enough that the elastic answer is right to a few percent.

Why does the relative velocity reverse — is there an intuition beyond the algebra?

Ride the collision from the centre-of-mass frame — the viewpoint gliding along so the total momentum is zero. There, the two bodies approach with equal-and-opposite momenta, compress the invisible spring, and — if every joule is refunded — must leave with the same speeds they brought, just reversed. A perfect mirror. Viewed from any other frame (yours), that mirror-image bounce is the statement "the gap opens as fast as it closed". Huygens invented exactly this trick in the 1660s, imagining collisions on a smoothly drifting boat; it remains the professional's way to see collision results before computing them.

Where does the "up to twice the speed" for a light target come from?

Stand on the heavy projectile. In its frame, the light ball approaches at $u_1$ and — the heavy body being effectively a wall — rebounds at $u_1$ . Now step back to the ground: the ball carries that rebound plus the ride, $u_1 + u_1 = 2u_1$ . The moving-bat version stacks the same way: approach speed $u + V$ , rebound $u + V$ off the bat, plus the bat's $V$ gives $u + 2V$ . It is always "your speed back, plus two of theirs" — and it caps at exactly 2 (or $+2V$ ): no elastic collision can fling a stationary target faster than twice the projectile's speed.

If equal masses transfer ALL the energy, why does a moderator need ~20 collisions, not 1?

The 100% figure is for a perfectly head-on collision, and head-on is rare. Real neutron–proton encounters happen at random impact parameters — mostly glancing — so each collision transfers a random fraction between 0 and 100%, averaging about half (in log terms, a fixed average "lethargy" gain). Twenty-ish collisions in water gets a 2-MeV fission neutron down to thermal energies. The formula still rules the choice of moderator: against lead even the perfect head-on hit moves under 2%, so no number of glancing blows is competitive.

Newton's cradle: why don't two balls fly out at half speed? Momentum would balance.

Momentum would — $mv = 2m(v/2)$ ✓ — but energy would not: $\tfrac12mv^2$ in, $2 \times \tfrac12m(v/2)^2 = \tfrac14mv^2$ out. Half the energy would have to vanish in a nearly-elastic steel-ball collision, which nature refuses. The observed outcome (one ball, full speed) is the only one satisfying both ledgers simultaneously. This makes the cradle the cleanest classroom demonstration that two independent conservation laws operate at once — and a standing MCQ: any proposed cradle outcome can be executed by checking the two books.

How is e = 1 related to all this — and what happens to these formulas when e < 1?

The reversal rule is the statement $e = 1$ : the coefficient of restitution is defined as separation speed ÷ approach speed, and elastic collisions score a perfect 1. Real collisions score less, and remarkably, the entire machinery survives the downgrade: momentum conservation stays exact, and the reversal rule generalizes to "separation $= e \times$ approach". Solving that pair gives closed forms that reduce to (4) at $e = 1$ and to perfect sticking at $e = 0$ . That graded world — including the KE lost as a function of $e$ , and $e$ as its own exam industry — is precisely the next article.