ARCHETYPE: Rotational Collisions

Every time a driver releases the clutch, two spinning discs grab each other and settle to a shared speed in a fraction of a second — a collision, except nothing translates: the crash happens entirely in rotation. A lump of putty smacking into a hinged rod, a child hopping onto a moving merry-go-round, a figure skater catching a thrown ball mid-spin: these are the rotational cousins of the head-on collisions you have already mastered, and they run on a conservation law of their own. Here is the twist that catches almost everyone. In these collisions, linear momentum is often not conserved — the hinge or axle fires off an impulsive force that wrecks the linear books in an instant. What survives is angular momentum about the pivot, because the hinge force, acting at the pivot, has zero lever arm there and can exert no torque no matter how violently it spikes. Choosing the right conserved quantity — and the right point to measure it about — is the entire game, and it is exactly what JEE-Advanced, NEET and IB HL examiners test. This article treats the rotational collision as one master setup with three faces: the particle that sticks to a pivoted rod, the disc dropped onto a spinning disc, and the person walking on a turntable. One law, one discipline about pivots, and a family of fearsome-looking problems becomes three lines of algebra.

Key takeaways

In a rotational collision, conserve angular momentum about the pivot/axle: the impulsive hinge force acts at that point, has zero lever arm, and therefore zero torque.
Linear momentum is generally not conserved when a pivot exists — the hinge delivers its own impulse. Never write $mv = (M+m)V$ for a hinged target.
A particle moving in a straight line has angular momentum about a point: $L = mv\,d$ , where $d$ is the perpendicular distance from the point to the velocity's line.
Putty $m$ at speed $v$ sticking to the end of a rod ( $M$ , $L$ , pivoted at the far end): $\omega = \dfrac{mvL}{\frac{1}{3}ML^2 + mL^2} = \dfrac{3mv}{(M + 3m)L}$ .
Disc on spinning disc (the clutch): $I_1\omega_1 = (I_1 + I_2)\,\omega$ — and kinetic energy is always lost: $\Delta KE = \dfrac{1}{2}\dfrac{I_1 I_2}{I_1+I_2}\,\omega_1^2$ .
Sticking collisions are perfectly inelastic: KE is never conserved, in rotation exactly as in translation.
Key idea: pick the pivot as your reference point — the violent unknown force then has zero torque, and angular momentum sails through the collision untouched.

The core insight

During a rotational collision, angular momentum about the axis or pivot is conserved because the impulsive reaction at that point has no lever arm there — while linear momentum and kinetic energy generally are not; equate $L$ just before to $L$ just after, with the struck object's moment of inertia updated for whatever now rides on it.

Where this lives today

Every manual gearshift replays worked example 1: clutch engagement is a disc-on-disc angular-momentum merger, and the compulsory $\Delta KE = \tfrac12 \tfrac{I_1I_2}{I_1+I_2}\omega_1^2$ is the heat your clutch plates are engineered to survive, ten thousand engagements a lifetime. Spacecraft run the turntable face in reverse: reaction wheels spin up so the satellite body counter-rotates to aim a telescope — total $L$ frozen, orientation traded internally, no propellant spent. The putty-on-rod ledger governs crash-test engineering (a car door struck off-centre acquires exactly $mvd$ of angular momentum about its hinges) and cricket bats and tennis rackets, whose "sweet spot" is the strike point that zeroes the impulsive reaction at your hands — the same hinge impulse computed in worked example 2, engineered away. And at the cosmic end, accreting neutron stars are slow-motion rotational collisions: infalling gas carries $mvd$ per kilogram, and millisecond pulsars are the spun-up result of a hundred million years of such deposits.

The setup

Three faces of one archetype, and the bookkeeping that unites them.

Face 1 — the particle and the pivoted rod. A rod of mass $M$ , length $L$ , hangs from a frictionless hinge. A lump of putty (or bullet, or ball) of mass $m$ arrives horizontally at speed $v$ and sticks at distance $d$ from the hinge (the classic: $d = L$ , the far end). During the impact the hinge yanks impulsively on the rod — an unknown, violent, brief force. About the hinge, that force has zero lever arm: angular momentum about the hinge is untouchable.

Face 2 — disc on disc. A disc spinning at $\omega_1$ on a frictionless axle; a second disc, at rest, is dropped coaxially onto it. Friction between the discs grinds them to a common $\omega$ — but that friction is internal to the two-disc system, and the axle force acts on the axis itself. Angular momentum about the axle survives; kinetic energy does not.

Face 3 — the walker on the turntable. A person strolls along the rim of a free turntable. Nothing collides, but the same law runs the show: the feet's friction is internal, the axle exerts no torque, so the total $L$ of person-plus-table is frozen — walk one way, the table counter-rotates.

Figure

The master collision — putty meets pivoted rod, and only L about the hinge survives

Before any algebra, one essential idea about what has angular momentum. A particle flying in a straight line, nowhere near any circle, still carries angular momentum about your chosen point:

L = m v\, d \qquad (1)

where: $d$ — the perpendicular distance from the point to the straight line of the velocity. This is the piece students refuse to believe: the putty in the figure, moving dead straight, has $L = mvL$ about the hinge, because its line of motion misses the hinge by $L$ . If it aimed straight at the hinge ( $d = 0$ ), it would carry no angular momentum, and the rod would not rotate at all after capture — it would just hang there while the hinge absorbed the blow.

The master solution pattern

Every face runs the same three lines. Line one: declare the reference point — the hinge, the axle — and justify it: the impulsive reaction there has zero torque about it. Line two: angular momentum just before equals angular momentum just after:

L_{\text{before}} = L_{\text{after}} \qquad (2)

where: "just before/just after" flags the collision's brevity — gravity and other finite forces have no time to build any angular impulse during the crash (their torque × the microsecond duration is negligible), so even non-conserved-looking situations conserve $L$ across the instant. Line three: update the moment of inertia — whatever stuck now rides at its distance $r$ , contributing $mr^2$ :

\omega = \frac{L_{\text{before}}}{I_{\text{after}}} = \frac{mvd}{I_{\text{target}} + md^2} \qquad (3)

where: $I_{\text{target}}$ — the struck body's moment of inertia about the pivot; $md^2$ — the passenger's new contribution. For the end-struck rod ( $I = \tfrac13 ML^2$ , $d = L$ ) this collapses to the exam-standard $\omega = \dfrac{3mv}{(M+3m)L}$ .

face	conserve	result
putty sticks to pivoted rod at $d$	$L$ about hinge	$\omega = \dfrac{mvd}{\frac13 ML^2 + md^2}$
disc dropped on spinning disc	$L$ about axle	$\omega = \dfrac{I_1\omega_1}{I_1 + I_2}$
person walks on free turntable	total $L$ (often zero)	$I_p\omega_p + I_t\omega_t = L_{\text{initial}}$
two-stage: collide, then swing up	$L$ across impact, then energy	$\tfrac12 I\omega^2 = (\text{total } m) g\, h_{\text{cm}}$

Derivation AWhy L survives the crash when p does not

We justify the central discipline of this archetype: during an impulsive collision with a pivoted body, angular momentum about the pivot is conserved while linear momentum is not. The strategy is to compare the impulse ledger with the angular-impulse ledger over the tiny collision time $\Delta t$ , and watch which external influences survive the limit $\Delta t \to 0$ . This derivation is short, but it is the answer to the most common wrong equation in the whole chapter — $mv = (M+m)V$ written for a hinged target.

Step 1. Linear ledger. Over the collision, the system (putty + rod) feels two external forces: gravity (finite) and the hinge reaction $F_h$ (impulsive — it spikes enormously). Impulse–momentum:

\Delta p = \int F_h\,dt + \int (M{+}m)g\,dt

Gravity's term $\sim g\Delta t \to 0$ , but the hinge's integral does not vanish — an impulsive force is precisely one whose time-integral stays finite as $\Delta t \to 0$ . Linear momentum changes by exactly the hinge's impulse: not conserved.

Step 2. Angular ledger, torques about the hinge. The hinge reaction acts at the hinge: lever arm zero, torque zero — no matter how large the force spikes. Gravity's torque is finite, so its angular impulse $\sim \tau_g\,\Delta t \to 0$ across the instant:

\Delta L_{\text{hinge}} = \int \tau_{\text{ext}}\,dt \to 0

Step 3. Conclusion. About the hinge — and in general about any point where the unknown impulsive force acts — angular momentum is conserved across the collision instant: $L_{\text{before}} = L_{\text{after}}$ . That single sentence, written on the exam script, is worth a mark by itself. Why the choice of point is strategy, not ritual: about any other point, the hinge impulse has a nonzero arm and $L$ is NOT conserved. The law is "angular momentum about the pivot," not "angular momentum, generally." For a free body (no pivot), there is no external impulse at all, and then both $p$ and $L$ about any fixed point are conserved. *Verified: the free-body limit — remove the hinge and Step 1's impulsive term disappears, recovering the familiar full conservation of collisions on ice. Consistency: the same "finite forces × vanishing time" argument is what legalizes momentum conservation in ordinary collisions despite gravity acting throughout.*

Go deeper · Derivation B · 1 minPutty on a pivoted rod — the end-strike standard

Derivation BPutty on a pivoted rod — the end-strike standard

We derive $\omega = 3mv/((M+3m)L)$ for putty of mass $m$ , speed $v$ , striking and sticking to the free end of a uniform rod ( $M$ , $L$ ) hinged at its top, hanging at rest. The strategy is the three-line pattern: angular momentum about the hinge just before (the flying particle's $mvd$ ) equals just after (combined body spinning), with the moment of inertia updated for the putty now riding at the end. Watch that the rod's own $I$ about the hinge is the end-axis value $\tfrac13 ML^2$ , not the centre value $\tfrac{1}{12}ML^2$ .

Step 1. Before. The rod is at rest ( $L_{\text{rod}} = 0$ ); the putty moves horizontally with its line of motion passing a perpendicular distance $L$ below the hinge:

L_{\text{before}} = mvL

Step 2. After. One rigid object — rod plus embedded putty — rotates about the hinge with angular speed $\omega$ :

I_{\text{after}} = \frac{1}{3}ML^2 + mL^2, \qquad L_{\text{after}} = \left(\frac{1}{3}ML^2 + mL^2\right)\omega

Why the putty contributes $mL^2$ : once embedded, it is a point mass at distance $L$ from the axis — its moment of inertia is $md^2$ with $d = L$ , the same rule that builds every composite-body $I$ .

Step 3. Conserve and solve:

mvL = \left(\frac{1}{3}ML^2 + mL^2\right)\omega \quad\Longrightarrow\quad \omega = \frac{mv}{\left(\frac{M}{3} + m\right)L} = \frac{3mv}{(M + 3m)L}

Step 4. Audit the energy books — the collision's cost:

KE_{\text{after}} = \frac{1}{2}I_{\text{after}}\omega^2 = \frac{(mvL)^2}{2 I_{\text{after}}} = \frac{m^2v^2}{2(\tfrac{M}{3}+m)} < \frac{1}{2}mv^2

Sticking always destroys kinetic energy — into deformation, sound and heat — exactly as in linear perfectly-inelastic collisions. Verified: limits — $M \to 0$ : $\omega \to v/L$ , the putty just swings on a massless stick at its own speed ✓; $M \gg m$ : $\omega \to 3mv/ML$ , tiny ✓. Numbers for worked example 2: $m = 0.10$ , $v = 6.0$ , $M = 1.2$ , $L = 0.90$ : $L_{before} = 0.54\ \mathrm{kg\,m^2/s}$ , $I_{after} = 0.324 + 0.081 = 0.405$ , $\omega = 1.33$ rad/s — matching the boxed answer there.

Go deeper · Derivation C · 1 minThe clutch — disc on spinning disc, and the unavoidable energy bill

Derivation CThe clutch — disc on spinning disc, and the unavoidable energy bill

We derive the common angular speed and the kinetic-energy loss when a stationary disc ( $I_2$ ) is dropped coaxially onto a disc spinning at $\omega_1$ ( $I_1$ ) — the rotational twin of the perfectly inelastic collision, and the physics of every clutch engagement. Strategy: friction between the discs is internal (equal and opposite torques on the pair), the axle exerts no torque about its own axis, so total $L$ is conserved; then we audit the energy books and prove the loss is structural, not accidental.

Step 1. Conservation about the axle:

I_1\omega_1 = (I_1 + I_2)\,\omega \quad\Longrightarrow\quad \omega = \frac{I_1\omega_1}{I_1 + I_2}

Why friction can't touch the total: the top disc speeds up under the friction torque, the bottom disc slows under the exactly opposite torque — Newton's third law in rotational dress. The pair's total $L$ just shuffles internally.

Step 2. Energy before and after:

KE_i = \frac{1}{2}I_1\omega_1^2, \qquad KE_f = \frac{1}{2}(I_1+I_2)\omega^2 = \frac{1}{2}\frac{I_1^2\omega_1^2}{I_1+I_2}

Step 3. The loss:

\Delta KE = KE_i - KE_f = \frac{1}{2}\,\frac{I_1 I_2}{I_1 + I_2}\,\omega_1^2

Why the loss is compulsory: while the discs grind toward a common speed, they slip against each other, and slipping friction converts work into heat at a rate it cannot refuse. The formula is the perfect analogue of the linear $\frac12 \mu_{red} u^2$ loss with reduced mass replaced by "reduced moment of inertia" $I_1I_2/(I_1+I_2)$ — the deep structural rhyme between the two chapters. *Verified: limits — $I_2 \to 0$ : $\omega \to \omega_1$ , no loss ✓; $I_2 \to \infty$ : $\omega \to 0$ , all KE lost ✓. Fraction check: $KE_f/KE_i = I_1/(I_1+I_2)$ , so equal discs keep exactly half — matching worked example 1's numbers ( $40$ rpm from $120$ rpm, one-third kept, for $I_2 = 2I_1$ ).*

Figure

The clutch — a collision that happens entirely in rotation

Limiting cases

case	result	meaning
putty aimed at the hinge ( $d = 0$ )	$\omega = 0$	no lever arm, no angular momentum delivered — the hinge eats everything
$M \to 0$ (massless rod)	$\omega = v/d$	the particle keeps its speed, now on a leash
$M \gg m$	$\omega \approx \dfrac{3mv}{ML}$	massive doors barely notice bird strikes
$I_2 = I_1$ (identical discs)	$\omega = \omega_1/2$ , half the KE gone	the 50% rule — memorize it
free rod on ice (no hinge)	$p$ AND $L$ both conserved	strike causes translation plus rotation
walker on frictionless-axle table, starts at rest	total $L = 0$ forever	walk clockwise, table turns counterclockwise

Worked examples

1 · Board level — the clutch plates (NEET/AP-1 pattern)

A disc with moment of inertia $I_1 = 0.10\ \mathrm{kg\,m^2}$ spins freely at $120\ \mathrm{rpm}$ . A second, stationary disc with $I_2 = 0.20\ \mathrm{kg\,m^2}$ is dropped coaxially onto it, and friction brings them to a common speed. Find the final angular speed and the fraction of kinetic energy lost.

Step 1. Justify the law for the mark: axle torque-free, disc-disc friction internal ⇒ conserve $L$ about the axle.

Step 2. $I_1\omega_1 = (I_1+I_2)\omega \Rightarrow \omega = \dfrac{0.10 \times 120}{0.30} = \boxed{40\ \mathrm{rpm}}$ . (Ratios of $\omega$ permit rpm directly — no conversion needed until energy enters.)

Step 3. Energy fraction kept: $\dfrac{KE_f}{KE_i} = \dfrac{I_1}{I_1+I_2} = \dfrac{0.10}{0.30} = \dfrac{1}{3}$ .

Step 4. Fraction lost $= 1 - \tfrac13 = \boxed{\tfrac{2}{3}}$ — turned to heat in the slipping interface. Note the loss needed no collision details: it is fixed by the two inertias alone.

2 · Board level — putty strikes a hanging rod (IB-HL/CBSE-culture pattern)

A uniform rod of mass $1.2\ \mathrm{kg}$ and length $0.90\ \mathrm{m}$ hangs at rest from a frictionless hinge at its upper end. A $0.10\ \mathrm{kg}$ lump of putty flying horizontally at $6.0\ \mathrm{m/s}$ strikes the lower end and sticks. Find the angular speed just after impact, and show that linear momentum was not conserved.

Step 1. Conserve $L$ about the hinge: $L_i = mvL = 0.10 \times 6.0 \times 0.90 = 0.54\ \mathrm{kg\,m^2/s}$ .

Step 2. New inertia: $I = \tfrac13 ML^2 + mL^2 = \tfrac13(1.2)(0.81) + (0.10)(0.81) = 0.324 + 0.081 = 0.405\ \mathrm{kg\,m^2}$ .

Step 3. $\omega = \dfrac{0.54}{0.405} = \boxed{1.33\ \mathrm{rad/s}}$ .

Step 4. Momentum audit. Before: $p_i = 0.60\ \mathrm{kg\,m/s}$ . After: the system's centre of mass — rod's at $0.45$ m, putty at $0.90$ m — gives $p_f = M(\omega \times 0.45) + m(\omega \times 0.90) = 1.2(0.60) + 0.1(1.20) = 0.84\ \mathrm{kg\,m/s}$ . $p_f \ne p_i$ : the hinge impulsively pushed forward on the rod (impulse $+0.24\ \mathrm{kg\,m/s}$ ). Angular momentum about the hinge, by contrast, balances exactly — the whole lesson in one audit.

3 · Competitive level — the child walks inward (JEE-Main/NEET pattern, ~2.5 minutes)

A merry-go-round of moment of inertia $360\ \mathrm{kg\,m^2}$ carries a $40\ \mathrm{kg}$ child standing at its rim, $R = 2.0\ \mathrm{m}$ ; the system spins freely at $0.50\ \mathrm{rad/s}$ . The child walks to the centre. Find the new angular speed, and explain where the extra kinetic energy came from.

Step 1. Conserve $L$ about the axle (axle torque-free; the child's footsteps are internal): $L = (I_{\text{table}} + mR^2)\,\omega_i = (360 + 160)(0.50) = 260\ \mathrm{kg\,m^2/s}$ .

Step 2. At the centre the child's $mr^2 \to 0$ : $\omega_f = \dfrac{260}{360} = \boxed{0.72\ \mathrm{rad/s}}$ .

Step 3. Energy check: $KE_i = \tfrac12(520)(0.25) = 65$ J; $KE_f = \tfrac12(360)(0.522) = 94$ J. The books rose by 29 J.

Step 4. Source: the child's muscles. Walking inward, she does work against the centrifugal tendency — biochemical energy in, rotational KE out. Conservation of $L$ never promises conservation of KE; the skater's spin-up is the same transaction. Timing note: the entire problem is one product $I\omega$ held constant — 2.5 minutes including the energy sentence. Distractor to dodge: adding the child's $mR^2$ to the final inertia too.

4 · Competitive level — JEE-Advanced stretch: strike, stick, swing full circle (~4 minutes)

A uniform rod ( $M = 0.60\ \mathrm{kg}$ , $L = 0.50\ \mathrm{m}$ ) hangs from a frictionless hinge. A ball of putty ( $m = 0.20\ \mathrm{kg}$ ) moving horizontally at speed $v$ strikes the lower end and sticks. Find the minimum $v$ for the combination to swing up and complete a full vertical circle. ( $g = 10\ \mathrm{m/s^2}$ .)

Step 1. Two-stage discipline. Stage one (collision — conserve $L$ , energy dies): $I = \tfrac13 ML^2 + mL^2 = \tfrac13(0.6)(0.25) + (0.2)(0.25) = 0.050 + 0.050 = 0.100\ \mathrm{kg\,m^2}$ ; $\omega = \dfrac{mvL}{I} = \dfrac{0.2 \times 0.5\,v}{0.100} = v$ (numerically, in rad/s).

Step 2. Stage two (swing — conserve energy, $L$ now changes as gravity torques). Rigid-body "full circle" on a hinge needs only to crawl past the top. Height gains from bottom to top: rod's centre rises $2 \times \tfrac{L}{2} = L$ ; putty rises $2L$ :

\Delta PE = g(M L + 2mL) = 10\,(0.6 \times 0.5 + 2 \times 0.2 \times 0.5) = 10(0.30 + 0.20) = 5.0\ \mathrm{J}

Step 3. Minimum condition — arrive at the top with $\omega \approx 0$ : $\tfrac12 I\omega^2 = \Delta PE \Rightarrow \tfrac12(0.100)\,v^2 = 5.0 \Rightarrow v^2 = 100$ .

Step 4. $v_{\min} = \boxed{10\ \mathrm{m/s}}$ . Timing note: the stages must not blur — L-conservation ACROSS the impact (energy forbidden), energy conservation DURING the swing (L forbidden, gravity torques freely). Rigid rod ⇒ "reach the top with zero speed" suffices; the string-versus-rod distinction from the vertical-circle archetype carries over unchanged. Blend the stages into one energy equation from $v$ and the answer is wrong with certainty.

Exam traps

Conserving the wrong quantity. The three-way sort, worth rehearsing until reflexive: pivoted target → conserve $L$ about the pivot only; free target (ice, space) → conserve $p$ and $L$ ; any sticking collision → KE is lost, never conserved. The favourite JEE trap is offering $mv = (M+m)V$ as a tempting first line for a hinged rod.

Angular momentum about which point? Choose it once — the pivot — and stay there for both sides of the equation. $L$ before about the hinge, $L$ after about the centre of mass is a hybrid that equals nothing.

Forgetting the passenger's inertia. After sticking, the projectile rides at radius $d$ : the final inertia is $I + md^2$ . Omitting $md^2$ inflates $\omega$ — and examiners provide exactly that wrong answer among the options.

"The particle moves straight, so it has no angular momentum." It has $L = mvd$ about any point a perpendicular distance $d$ from its line of motion — the entire before-side of the master equation. Aimed at the pivot, $d = 0$ : capture without rotation.

rpm arithmetic. Ratios of angular speeds ( $\omega_f/\omega_i = I_i/I_f$ ) work in any unit; energies need rad/s. Convert with $\omega = 2\pi n/60$ before squaring anything.

Two-stage blurring. Collide-then-swing problems: $L$ -conservation across the impact, then energy up the swing. One combined "energy conservation from bullet speed to final height" silently assumes the collision was elastic — it never is when things stick.

Which $g$ ? JEE/NEET declare $g = 10$ almost always; IB uses $9.8$ ; AP permits $10$ for estimates. Use the stated value.

Origins · 1 minChristiaan Huygens and how this was discovered

Christiaan Huygens

1629–1695 · The Hague & Paris

When the Royal Society announced a competition in 1668 to settle the laws of collision, three men answered — Wallis, Wren, and a Dutch polymath who had already solved it years earlier: Christiaan Huygens. His rules for elastic impact, built on the relativity of motion (he imagined collisions viewed from a drifting canal boat), gave physics its first conservation bookkeeping beyond bare momentum. Huygens seeded this article twice more: his Horologium Oscillatorium introduced the "centre of oscillation," effectively discovering the moment of inertia a generation before Euler named it, and his analysis of the compound pendulum is the direct ancestor of every hinged-rod problem. The clutch-plate law and the putty-on-rod equation are Huygens' two lifelong obsessions — collisions and rotating rigid bodies — finally meeting in a single setup.

History

1668

— the Royal Society's collision competition: Wallis handles the inelastic case, Wren and Huygens the elastic — conservation bookkeeping enters physics.

1673

— Huygens' Horologium Oscillatorium analyses the compound pendulum, smuggling in the moment of inertia before it had a name.

1687

— Newton's Principia derives Kepler's equal-areas law from what we now read as angular-momentum conservation under a central force.

1765

— Euler's Theoria motus corporum solidorum gives rigid-body rotation its full mathematical machinery:

L = I\omega

and its tensor descendants.

1918

— Noether's theorem reveals the deep source: angular momentum is conserved because physics does not care which way space points.

Going further

Texts. NCERT Physics Class 11, Ch. 7 (angular momentum sections) · HC Verma, Concepts of Physics Vol. 1, Ch. 10 · Halliday–Resnick–Walker Ch. 11 · Morin Ch. 8 for free-body strikes and the sweet-spot calculation.

Provenance. All results are standard rotation canon. A verifier should recompute: $\omega = 3mv/((M+3m)L)$ from $mvL = (\tfrac13 ML^2 + mL^2)\omega$ ; example 1's $40$ rpm and $\tfrac23$ loss; example 2's $\omega = 1.33$ rad/s and hinge impulse $+0.24\ \mathrm{kg\,m/s}$ ; example 3's $0.72$ rad/s and $+29$ J; example 4's $I = 0.100$ , $\Delta PE = 5.0$ J, $v_{\min} = 10$ m/s.

Prerequisites. Angular Momentum & Its Conservation · Moments of Inertia of Standard Bodies · Conservation of Linear Momentum. Leads to. The Falling Hinged Rod (archetype) · Angular Momentum of a Rigid Body (UG deep dive).

Why exactly is linear momentum not conserved when the rod is hinged?

Because the hinge fights back during the collision. As the putty slams the rod, the rod's top end tries to kick backward (or forward — worked example 2 shows the direction can surprise you), and the hinge supplies whatever impulsive force is needed to keep that end pinned. That force is external to the putty-rod system and has a finite impulse even though it acts for microseconds — so $\Delta p \ne 0$ . The same force has zero lever arm about the hinge, which is why the angular books survive untouched.

How can a particle moving in a straight line have angular momentum?

Angular momentum is defined about a point: $L = mvd$ with $d$ the perpendicular miss-distance. A truck driving straight past your gatepost genuinely carries angular momentum about the post — if it snagged a rope tied there, it would start circling. What feels wrong is conflating "angular momentum" with "already rotating"; the quantity actually measures capacity to cause rotation about that point upon connection. Kepler's planets sweep equal areas for exactly this reason: even on the straightest stretch of orbit, $mvd$ about the Sun is constant.

Where does the lost kinetic energy actually go?

Into the collision's microscopic chaos: plastic deformation of the putty, sound, and — in the disc case — frictional heat generated while the surfaces slip against each other before locking. The loss is not negligence you can engineer away; it is structural. Given "sticks" or "reaches a common speed," the final state is fully fixed by $L$ -conservation, and the energy difference between that state and the initial one must be dissipated. Only an elastic rebound (rare in rotation problems) preserves KE.

For a free rod floating on ice, what do I conserve?

Everything at once: linear momentum ( $mv =$ total mass × centre-of-mass velocity), and angular momentum about any fixed point — the centre of mass of the final system is the standard choice. The struck free rod both translates and rotates: $v_{cm} = mv/(M+m)$ , plus a spin from $L$ about the combined CM. That double bookkeeping is a genuinely harder problem than the hinged case, which is why exams flag the hinge so prominently — it removes one law and one unknown.

In the turntable problem, the person and platform push on each other with friction. Why doesn't that ruin conservation?

Those pushes are internal to the person-plus-platform system — action and reaction, equal and opposite torques about the axle, summing to zero. Only torques from outside the system can change its total $L$ , and the frictionless axle declines to provide any (a real axle's bearing friction would, slowly — that's why real merry-go-rounds run down). This is the rotational mirror of "internal forces cannot change total momentum" from the linear chapter.

Why does the two-stage problem forbid using energy across the collision?

Because the collision is perfectly inelastic — the putty deforms and grips, dissipating a fixed, unavoidable fraction of the kinetic energy (Step 4 of the rod derivation computes it). Energy written straight from bullet speed to swing height pretends that loss never happened and overestimates the rise. The correct relay: $L$ -conservation delivers the post-impact $\omega$ ; then energy conservation, now clean because the swing involves no slipping or deformation, carries $\omega$ into height. Two laws, two stages, one boundary — never blended.

Is the skater pulling in her arms a "collision" too?

Same law, opposite energy sign. No impact occurs, but the axle-free, torque-free condition holds, so $I\omega$ stays fixed while she rearranges her own inertia: arms in, $I$ drops, $\omega$ jumps. Kinetic energy increases — her muscles do real work hauling mass inward against the centrifugal tendency (worked example 3's child pays the same bill). Rotational collisions and skater spin-ups are one chapter: $L$ frozen, KE negotiable, with the sign of the energy change telling you who paid.