Center-of-Mass Frame and Collision Analysis

In 1911 Ernest Rutherford announced something absurd: once in a while, an alpha particle fired at gold foil came straight back. "It was almost as incredible as if you fired a fifteen-inch shell at a piece of tissue paper and it came back and hit you," he said. Decoding what those bounces meant — a tiny, massive nucleus — required a habit of mind that has powered collision physics ever since: stop watching from the lab. Jump instead onto the one frame where the total momentum is exactly zero, and the messy asymmetric encounter becomes a perfectly symmetric dance — two bodies approaching with equal and opposite momenta, colliding, and (if the collision is elastic) leaving with the same speeds in merely rotated directions. That frame is the center-of-mass frame, and it is less a trick than a decomposition theorem. Kinetic energy splits cleanly into a bulk part $\tfrac12 MV_{\mathrm{cm}}^2$ that momentum conservation locks away, and an internal part $\tfrac12\mu v_{\mathrm{rel}}^2$ — the only energy a collision can spend, deform, or convert. Elastic, inelastic, explosive: every collision type is one sentence in the CM frame. The payoff is a set of lab-frame results that look like magic until you see their CM shadow: equal-mass billiard balls parting at exactly 90°, a heavy projectile that cannot deflect beyond a maximum angle, a neutron that thermalizes in 18 collisions against hydrogen but needs over a hundred against carbon, and the reason particle physicists build colliders instead of firing beams at stationary targets. One Galilean boost, applied twice, explains them all.

Key takeaways

The center-of-mass frame is the inertial frame in which total momentum vanishes; it drifts through the lab at the constant velocity $\mathbf{V}_{\mathrm{cm}} = \mathbf{P}/M$ .
In the CM frame two colliding bodies always carry equal and opposite momenta of magnitude $\mu v_{\mathrm{rel}}$ — an elastic collision merely rotates the momentum line, leaving both CM speeds unchanged.
König's theorem splits the kinetic energy: $T_{\mathrm{lab}} = \tfrac12 MV_{\mathrm{cm}}^2 + T_{\mathrm{cm}}$ ; the bulk term is frozen by momentum conservation, so only $T_{\mathrm{cm}} = \tfrac12\mu v_{\mathrm{rel}}^2$ is available for transfer or loss.
Lab kinematics follow by adding $\mathbf{V}_{\mathrm{cm}}$ back: $\tan\theta_{\mathrm{lab}} = \sin\theta_{\mathrm{cm}}/(\cos\theta_{\mathrm{cm}} + m_1/m_2)$ for elastic scattering off a stationary target.
Consequences fall out geometrically: equal masses separate at 90° in the lab; a projectile heavier than its target has a maximum deflection $\sin\theta_{\max} = m_2/m_1$ ; the energy fraction transferred head-on is $4m_1m_2/(m_1+m_2)^2$ .
Key idea: boost to the frame where the total momentum is zero — the collision becomes a pure rotation of equal-and-opposite momenta, and every lab result is that rotation plus the constant drift $\mathbf{V}_{\mathrm{cm}}$ .

The core insight

Boost to the frame where total momentum is zero: the collision becomes a pure rotation of equal-and-opposite momenta, and lab results follow by adding back the constant CM velocity.

Where this lives today

Every collider on Earth is a monument to König's theorem. In a fixed-target experiment the bulk term $\tfrac12MV_{\mathrm{cm}}^2$ is pure waste — for ultrarelativistic beams the useful (invariant) energy grows only as the square root of the beam energy — so modern machines collide two beams head-on, making the lab frame be the CM frame and every joule spendable. The LHC's 13.6 TeV of proton–proton CM energy would require a fixed-target beam of roughly $10^{17}$ eV to match: five orders of magnitude beyond feasible. The same bookkeeping runs dark-matter direct detection: a WIMP scattering off a nucleus is elastic two-body kinematics, the maximum recoil energy is set by the $4m_1m_2/(m_1+m_2)^2$ mass-matching factor, and that is why experiments hunting light dark matter have shifted to light targets — helium, hydrogen-rich scintillators, even electrons — while xenon detectors rule at heavier masses. Spacecraft navigators use the elastic-collision picture with a planet as the target: a gravity assist is an elastic scattering event in the planet–spacecraft CM frame (speed preserved, direction rotated), and boosting back to the Sun's frame adds the planet's 10–47 km/s of orbital velocity to the outgoing leg. ESA's JUICE and NASA's Europa Clipper both threaded multiple such CM-frame rotations through 2025–26 en route to Jupiter — each flyby a worked example of equation (6) with the planet's orbital motion as $\mathbf{V}_{\mathrm{cm}}$ .

The physical picture

Watch two ice skaters glide toward each other — a heavy one drifting slowly, a light one rushing in. In the lab the encounter is lopsided and hard to intuit. Now ride a camera dolly that moves so the total momentum reads zero. From the dolly the scene is perfectly balanced: the skaters approach along a line with equal and opposite momenta, meet, and part with equal and opposite momenta again. If no kinetic energy was lost, the outgoing speeds are identical to the incoming ones — the collision did nothing but swing the line of motion through an angle. All the apparent complexity of the lab view is just this symmetric event watched from a moving platform.

Figure

The same collision in two frames — lopsided in the lab, symmetric in the center of mass

Why is this frame so powerful? Because it separates the two conserved-quantity stories. Momentum conservation is the statement that the platform's drift $\mathbf{V}_{\mathrm{cm}}$ never changes — so the bulk kinetic energy $\tfrac12 MV_{\mathrm{cm}}^2$ riding with it is locked. Everything a collision can actually do — transfer energy, dissipate it, release it — happens in the internal account $T_{\mathrm{cm}}$ , and in the CM frame that account is laid bare with no bulk motion obscuring it. Elastic collision: $T_{\mathrm{cm}}$ unchanged, momenta rotate. Perfectly inelastic: $T_{\mathrm{cm}}$ spent to zero, both bodies stop dead in this frame. Explosion: $T_{\mathrm{cm}}$ increases. Every collision type is one sentence here, and the lab-frame answer is always the same sentence plus a constant drift. The geometric bonus: since the outgoing CM momenta have fixed magnitude (elastic case), the possible outcomes trace a circle. Lab velocities are that circle shifted by $\mathbf{V}_{\mathrm{cm}}$ — and nearly every celebrated scattering fact (the 90° billiards rule, the maximum deflection of a heavy projectile) is a one-line statement about whether the shift is smaller or larger than the circle's radius.

Building the frame

Start from a system of particles with no external forces. The total momentum $\mathbf{P} = \sum_i m_i\mathbf{v}_i$ is conserved, so the velocity

\mathbf{V}_{\mathrm{cm}} = \frac{\sum_i m_i \mathbf{v}_i}{\sum_i m_i} = \frac{\mathbf{P}}{M} \qquad (1)

where: $M = \sum_i m_i$ — total mass. Because $\mathbf{P}$ is constant, $\mathbf{V}_{\mathrm{cm}}$ is constant: the CM frame moves uniformly, and is therefore itself inertial — Newton's laws hold there without any fictitious forces. That is not a small point; it is what licenses everything below. Boost every velocity by this one vector and the total momentum vanishes identically:

\mathbf{v}_i' = \mathbf{v}_i - \mathbf{V}_{\mathrm{cm}}, \qquad \sum_i m_i \mathbf{v}_i' = \mathbf{P} - M\mathbf{V}_{\mathrm{cm}} = 0 \qquad (2)

where: primes denote CM-frame quantities throughout. The boost is precisely the one that nulls the momentum — equation (1) is not a definition pulled from a hat but the unique solution of $\sum m_i\mathbf{v}_i' = 0$ . Kinetic energy now splits with surgical cleanliness. Substituting $\mathbf{v}_i = \mathbf{v}_i' + \mathbf{V}_{\mathrm{cm}}$ into $T = \sum \tfrac12 m_i v_i^2$ and expanding, the cross term is $\mathbf{V}_{\mathrm{cm}}\cdot\sum m_i \mathbf{v}_i'$ — zero by (2). What survives is König's theorem:

T_{\mathrm{lab}} = \tfrac12 M V_{\mathrm{cm}}^2 + T_{\mathrm{cm}} \qquad (3)

where: $T_{\mathrm{cm}} = \sum_i \tfrac12 m_i v_i'^2$ — the internal kinetic energy. Read it as a ledger: the first term belongs to the system as a whole and is frozen by momentum conservation; the second is the spendable balance. A collision is elastic if $T_{\mathrm{cm}}$ is unchanged, inelastic if it decreases, superelastic (explosive) if it increases — and the lab energy change equals the internal change exactly, $\Delta T_{\mathrm{lab}} = \Delta T_{\mathrm{cm}}$ . For two bodies the internal account has a famous closed form. With the reduced mass

\mu = \frac{m_1 m_2}{m_1+m_2}, \qquad T_{\mathrm{cm}} = \tfrac12\, \mu\, v_{\mathrm{rel}}^2 \qquad (4)

where: $v_{\mathrm{rel}} = |\mathbf{v}_1 - \mathbf{v}_2|$ — the relative speed, the same in every inertial frame. The two-body internal motion is exactly that of one fictitious particle of mass $\mu$ moving with the relative velocity — the same reduction that tames the Kepler problem. And the individual CM momenta are forced into the symmetric pattern:

\mathbf{p}_1' = -\,\mathbf{p}_2', \qquad |\mathbf{p}_1'| = |\mathbf{p}_2'| = \mu\, v_{\mathrm{rel}} \qquad (5)

where: the equal-and-opposite structure is not an assumption but the content of $\sum m_i\mathbf{v}_i' = 0$ for two bodies. This is the single most useful line in collision practice: in the CM frame there is only one momentum magnitude to track.

Derivation AThe CM velocities of a two-body system

We are deriving the explicit CM-frame velocities of two bodies in terms of their relative velocity — the formulas $\mathbf{v}_1' = (m_2/M)\,\mathbf{v}_{\mathrm{rel}}$ and $\mathbf{v}_2' = -(m_1/M)\,\mathbf{v}_{\mathrm{rel}}$ that make equations (4) and (5) one-line checks. The route is direct substitution of the boost (2) with the definition (1); the algebra is short, but the structure of the answer is what to watch: each body's CM velocity is proportional to the relative velocity, weighted by the other body's mass fraction. That inverse weighting is the seed of every "the light one does the moving" intuition in two-body physics, and it is why the reduced mass emerges when the energies are added.

Step 1. Boost body 1: substitute (1) into $\mathbf{v}_1' = \mathbf{v}_1 - \mathbf{V}_{\mathrm{cm}}$ and put everything over the common denominator $M = m_1 + m_2$ :

\mathbf{v}_1' = \frac{M\mathbf{v}_1 - m_1\mathbf{v}_1 - m_2\mathbf{v}_2}{M} = \frac{m_2\,(\mathbf{v}_1 - \mathbf{v}_2)}{M} = \frac{m_2}{M}\,\mathbf{v}_{\mathrm{rel}}

Why the other body's mass appears: subtracting the mass-weighted average leaves body 1 with exactly the fraction of the relative velocity that body 2's weight contributes to that average.

Step 2. Body 2 by the identical algebra (or the swap $1 \leftrightarrow 2$ , which flips the sign of $\mathbf{v}_{\mathrm{rel}}$ ):

\mathbf{v}_2' = -\,\frac{m_1}{M}\,\mathbf{v}_{\mathrm{rel}}

Step 3. Check the momenta: $m_1\mathbf{v}_1' = \frac{m_1 m_2}{M}\mathbf{v}_{\mathrm{rel}} = \mu\,\mathbf{v}_{\mathrm{rel}}$ and $m_2\mathbf{v}_2' = -\mu\,\mathbf{v}_{\mathrm{rel}}$ — equal and opposite with magnitude $\mu v_{\mathrm{rel}}$ , which is equation (5). Why this is the memorable form: both momenta equal the reduced mass times the relative velocity, so the CM frame carries one momentum scale, not two.

Step 4. Check the energy: $T_{\mathrm{cm}} = \tfrac12 m_1 \big(\tfrac{m_2}{M}\big)^2 v_{\mathrm{rel}}^2 + \tfrac12 m_2 \big(\tfrac{m_1}{M}\big)^2 v_{\mathrm{rel}}^2 = \tfrac12\,\frac{m_1m_2}{M}\,v_{\mathrm{rel}}^2 = \tfrac12\mu v_{\mathrm{rel}}^2$ , which is equation (4). Verified: matches Goldstein §3.11 (eqs. 3.110–3.112) and Landau–Lifshitz §16; the momentum check gives zero total momentum identically.

From the CM rotation back to the lab

Here is the working core of scattering kinematics. Let body 1 (projectile, lab speed $v_1$ ) strike body 2 at rest. The boost velocity is $V_{\mathrm{cm}} = \tfrac{m_1}{M}v_1$ along the beam, and the projectile's CM speed is $u_1 = \tfrac{m_2}{M}v_1$ — both constants of the problem. An elastic collision rotates the CM momentum line by the scattering angle $\theta_{\mathrm{cm}}$ without changing $u_1$ . The lab deflection is then fixed by trigonometry alone:

\tan\theta_{\mathrm{lab}} = \frac{\sin\theta_{\mathrm{cm}}}{\cos\theta_{\mathrm{cm}} + m_1/m_2} \qquad (6)

where: the ratio $m_1/m_2 = V_{\mathrm{cm}}/u_1$ compares the boost to the CM speed — the entire mass dependence of elastic scattering enters through this one number. For $m_1 \ll m_2$ the denominator's offset vanishes and $\theta_{\mathrm{lab}} \approx \theta_{\mathrm{cm}}$ (the heavy target barely recoils); for $m_1 = m_2$ the half-angle identity gives $\theta_{\mathrm{lab}} = \theta_{\mathrm{cm}}/2$ . The energy carried off by the recoiling target follows from the same velocity triangle. The target's CM speed is $V_{\mathrm{cm}}$ itself (it started at rest), so its lab speed after the rotation is the base of an isosceles triangle, $v_2 = 2V_{\mathrm{cm}}\cos(\theta_{\mathrm{cm}}/2)$ — wait for the derivation below to see the geometry — and its kinetic energy, as a fraction of the beam energy, is

\frac{\Delta T}{T_1} = \frac{4\, m_1 m_2}{(m_1+m_2)^2}\,\sin^2\!\frac{\theta_{\mathrm{cm}}}{2} \qquad (7)

where: $T_1 = \tfrac12 m_1 v_1^2$ . The prefactor $4m_1m_2/(m_1+m_2)^2$ is the maximum fractional transfer, reached head-on ( $\theta_{\mathrm{cm}} = 180°$ ). It equals 1 only for equal masses — a neutron can hand its entire energy to a proton in one head-on collision — and collapses like $4m_1/m_2$ for a very heavy target. When the projectile is heavier than its target, something qualitatively new happens: the boost $V_{\mathrm{cm}}$ exceeds the CM speed $u_1$ , so no CM rotation can turn the lab velocity backwards. The lab deflection has a hard ceiling:

\sin\theta_{\max} = \frac{m_2}{m_1} \qquad (m_1 > m_2) \qquad (8)

where: the maximum occurs where the lab velocity is tangent to the circle of possible outcomes. An alpha particle ( $m \approx 4$ u) scattering off an electron-loaded target's nuclei can backscatter; off an electron ( $m_2/m_1 \approx 1/7300$ ) it can deflect by at most $0.008°$ — which is precisely how Rutherford knew the large-angle events in gold foil could not come from electrons. For completeness at this article's depth: intensities transform too. A detector at fixed lab solid angle collects a CM solid angle distorted by the boost, so differential cross-sections convert as

\frac{\mathrm{d}\sigma}{\mathrm{d}\Omega_{\mathrm{lab}}} = \frac{\mathrm{d}\sigma}{\mathrm{d}\Omega_{\mathrm{cm}}}\;\frac{\left(1 + 2x\cos\theta_{\mathrm{cm}} + x^2\right)^{3/2}}{\big|\,1 + x\cos\theta_{\mathrm{cm}}\,\big|}, \qquad x = \frac{m_1}{m_2} \qquad (9)

where: the Jacobian factor is the stretching of solid angle under the angle map (6). Forward directions get compressed in the lab for heavy projectiles — the same physics as the maximum angle, seen in intensity rather than geometry.

Which formula answers which question

you want	use	notes
the frame's drift velocity	$\mathbf{V}_{\mathrm{cm}} = \mathbf{P}/M$ , eq. (1)	constant when no external force acts
spendable collision energy	$T_{\mathrm{cm}} = \tfrac12\mu v_{\mathrm{rel}}^2$ , eq. (4)	all of it is lost in a perfect stick
lab deflection from CM angle	eq. (6)	equal masses: $\theta_{\mathrm{lab}} = \theta_{\mathrm{cm}}/2$
energy given to a stationary target	eq. (7)	max $= 4m_1m_2/M^2$ , head-on
can the projectile bounce back?	eq. (8)	only if $m_1 \le m_2$ ; else ceiling $\arcsin(m_2/m_1)$
detector intensities between frames	eq. (9)	Jacobian of the angle map

Go deeper · Derivation B · 2 minThe lab scattering angle and the maximum-deflection ceiling

Derivation BThe lab scattering angle and the maximum-deflection ceiling

The goal is the angle transformation (6) and the maximum-angle result (8) in one geometric picture: the velocity circle. The strategy is to write the projectile's final lab velocity as the constant boost $\mathbf{V}_{\mathrm{cm}}$ plus a vector of fixed length $u_1$ (its CM speed, unchanged by an elastic collision) pointing in the rotated direction $\theta_{\mathrm{cm}}$ — so the tip of the lab velocity moves on a circle of radius $u_1$ centered at the tip of $\mathbf{V}_{\mathrm{cm}}$ . Everything reduces to asking angles of that figure. The one case demanding care is $m_1 > m_2$ : then the circle's center lies farther from the origin than the circle's radius, the origin sits outside the circle, and the lab angle can no longer reach all values — that geometric fact is the maximum-deflection theorem.

Step 1. Set up along the beam ( $x$ -axis). Before the collision, $V_{\mathrm{cm}} = \tfrac{m_1}{M}v_1$ and the projectile's CM speed is $u_1 = v_1 - V_{\mathrm{cm}} = \tfrac{m_2}{M}v_1$ . Why $u_1$ survives the collision: elastic means $T_{\mathrm{cm}}$ unchanged, and with equal-and-opposite momenta of one magnitude, unchanged energy forces unchanged magnitude — only the direction rotates.

Step 2. After the rotation by $\theta_{\mathrm{cm}}$ , the lab velocity components are

v_{1x} = u_1\cos\theta_{\mathrm{cm}} + V_{\mathrm{cm}}, \qquad v_{1y} = u_1\sin\theta_{\mathrm{cm}}

Why only the $x$ component shifts: the boost back to the lab adds the purely-forward vector $\mathbf{V}_{\mathrm{cm}}$ .

Step 3. Divide to form the lab angle, then divide numerator and denominator by $u_1$ and use $V_{\mathrm{cm}}/u_1 = m_1/m_2$ :

\tan\theta_{\mathrm{lab}} = \frac{u_1\sin\theta_{\mathrm{cm}}}{u_1\cos\theta_{\mathrm{cm}} + V_{\mathrm{cm}}} = \frac{\sin\theta_{\mathrm{cm}}}{\cos\theta_{\mathrm{cm}} + m_1/m_2}

which is equation (6).

Step 4. Read the circle. If $m_1 < m_2$ then $V_{\mathrm{cm}} < u_1$ : the origin lies inside the velocity circle, every lab angle from 0 to 180° is reachable, and backscattering is allowed. If $m_1 > m_2$ the origin lies outside the circle; the lab velocity direction is bounded by the two tangent lines from the origin, giving

\sin\theta_{\max} = \frac{u_1}{V_{\mathrm{cm}}} = \frac{m_2}{m_1}

which is equation (8). Why tangency: the extreme direction from an external point to a circle is along the tangent — at that outcome, $\mathrm{d}\theta_{\mathrm{lab}}/\mathrm{d}\theta_{\mathrm{cm}} = 0$ and the lab intensity (9) momentarily diverges, a caustic real detectors see as a bright rim. *Verified: matches Taylor ch. 14 and Goldstein eq. (3.110); at $m_1 = m_2$ , $\tan\theta_{\mathrm{lab}} = \sin\theta_{\mathrm{cm}}/(1+\cos\theta_{\mathrm{cm}}) = \tan(\theta_{\mathrm{cm}}/2)$ , so $\theta_{\mathrm{lab}} = \theta_{\mathrm{cm}}/2$ exactly.*

Go deeper · Derivation C · 1 minThe right-angle theorem for equal masses

Derivation CThe right-angle theorem for equal masses

We are proving the billiard player's theorem: when two equal masses collide elastically and one starts at rest, the final lab velocities are perpendicular — they part at exactly 90°. The strategy is deliberately frame-free this time: momentum and energy conservation in the lab, three lines of vector algebra. Doing it both ways (here directly, and via the CM angle map in the previous derivation) is the best possible check that the CM machinery is honest. Watch the degenerate cases: the theorem presumes both bodies actually move afterward.

Step 1. Momentum conservation with the masses cancelled:

\mathbf{v}_0 = \mathbf{v}_1 + \mathbf{v}_2

Why the masses drop out: they are equal on every term.

Step 2. Square the vector equation: $v_0^2 = v_1^2 + v_2^2 + 2\,\mathbf{v}_1\cdot\mathbf{v}_2$ .

Step 3. Energy conservation (elastic, equal masses): $v_0^2 = v_1^2 + v_2^2$ . Subtracting from Step 2:

\mathbf{v}_1 \cdot \mathbf{v}_2 = 0

Why this means 90°: a vanishing dot product between two nonzero vectors is orthogonality. The degenerate outcomes are head-on ( $\mathbf{v}_1 = 0$ : full velocity exchange) and a miss ( $\mathbf{v}_2 = 0$ ).

Step 4. Cross-check in the CM picture: the projectile emerges at $\theta_{\mathrm{lab}} = \theta_{\mathrm{cm}}/2$ and the target recoils at $(\pi - \theta_{\mathrm{cm}})/2$ on the other side; the two lab angles sum to $\pi/2$ for every $\theta_{\mathrm{cm}}$ . Same theorem, now visibly a property of the isosceles velocity triangle. Verified: consistent with Morin ch. 5 problems and with equation (6) at $m_1 = m_2$ ; nuclear-emulsion and cloud-chamber tracks of equal-mass collisions (p–p) show the 90° opening angle at non-relativistic energies.

Figure

The velocity circle — why a heavy projectile has a maximum lab deflection

Limiting cases

Case	Result	Meaning
$m_1 = m_2$ , target at rest	$\theta_{\mathrm{lab}} = \theta_{\mathrm{cm}}/2$ ; opening angle 90°; head-on transfer 100%	billiards geometry; neutron–proton moderation in one hit
$m_1 \ll m_2$	$\mathbf{V}_{\mathrm{cm}} \to 0$ , $\theta_{\mathrm{lab}} \to \theta_{\mathrm{cm}}$ , $\mu \to m_1$	lab ≈ CM; the fixed-scattering-center (Rutherford) limit
$m_1 \gg m_2$	$\sin\theta_{\max} = m_2/m_1 \to 0$ ; transfer $\le 4m_2/m_1$	the heavy projectile plows on, barely deflected — alphas vs electrons
$\theta_{\mathrm{cm}} = 180°$	$\Delta T/T_1 = 4m_1m_2/M^2$ ; projectile CM velocity reversed	head-on: the maximum-transfer collision
perfectly inelastic	both bodies at rest in CM; loss $= T_{\mathrm{cm}} = \tfrac12\mu v_{\mathrm{rel}}^2$	the CM frame shows the stick as "all internal motion spent"
relativistic energies	CM frame → center-of-momentum frame; invariant $\sqrt{s}$ replaces $T_{\mathrm{cm}}$	same logic, Lorentz boost instead of Galilean

Common trapThe CM angle and the lab angle are the same thing

They are equal only in the limit $m_1 \ll m_2$ (immovable target) — in general the boost back to the lab compresses angles toward the forward direction, and for equal masses it exactly halves them, equation (6). Applying CM-frame facts directly to lab velocities produces classic wrong answers: "elastic collisions preserve each body's speed" (true only in the CM frame), "the projectile can always bounce straight back" (false for $m_1 > m_2$ : the ceiling is $\arcsin(m_2/m_1)$ , equation (8)), and "the target flies off sideways" (it never does — a target initially at rest always recoils into the forward hemisphere, at $(\pi-\theta_{\mathrm{cm}})/2$ from the beam). Always transform: rotate in the CM frame, then add $\mathbf{V}_{\mathrm{cm}}$ .

Worked examples

1 · A one-dimensional collision, solved by round trip through the CM frame (confidence builder)

A 2.0 kg cart moving at 6.0 m/s strikes a stationary 4.0 kg cart head-on, elastically. Find both final velocities by boosting to the CM frame and back.

Step 1. The boost, equation (1): $V_{\mathrm{cm}} = \dfrac{2.0 \times 6.0 + 4.0 \times 0}{6.0} = 2.0\ \mathrm{m/s}$ .

Step 2. Boost in, equation (2): $v_1' = 6.0 - 2.0 = 4.0\ \mathrm{m/s}$ , $v_2' = 0 - 2.0 = -2.0\ \mathrm{m/s}$ . Momentum check: $2.0(4.0) + 4.0(-2.0) = 0$ . ✓

Step 3. Collide. A head-on elastic collision in the CM frame is the 180° rotation: each velocity simply reverses. $v_1' \to -4.0\ \mathrm{m/s}$ , $v_2' \to +2.0\ \mathrm{m/s}$ . (No algebra — this is the entire point of the frame.)

Step 4. Boost out: $v_1 = -4.0 + 2.0 = \boxed{-2.0\ \mathrm{m/s}}$ (bounces back), $v_2 = 2.0 + 2.0 = \boxed{+4.0\ \mathrm{m/s}}$ .

Step 5. Verify in the lab: momentum $2.0(-2.0) + 4.0(4.0) = 12\ \mathrm{kg\,m/s}$ = initial $2.0 \times 6.0$ . ✓ Energy: before $\tfrac12(2.0)(36) = 36\ \mathrm{J}$ ; after $\tfrac12(2.0)(4.0) + \tfrac12(4.0)(16) = 4.0 + 32 = 36\ \mathrm{J}$ . ✓ The standard elastic formulas give $v_1 = \tfrac{m_1-m_2}{M}v_0 = -\tfrac{2}{6}(6.0) = -2.0$ and $v_2 = \tfrac{2m_1}{M}v_0 = \tfrac{4}{6}(6.0) = 4.0$ — same answers with far more algebra.

2 · Choosing a neutron moderator (exam level)

Fission neutrons are born at about 2 MeV and are useful at thermal energy, 0.025 eV. Using CM-frame kinematics, compare hydrogen ( $A = 1$ ) and carbon ( $A = 12$ ) as moderators: (a) maximum fractional energy transfer per elastic collision; (b) the average number of collisions to thermalize.

Step 1. Maximum transfer is head-on, equation (7) at $\theta_{\mathrm{cm}} = 180°$ : $f_{\max} = \dfrac{4m_1m_2}{(m_1+m_2)^2}$ . Hydrogen ( $m_2 = m_1$ ): $f_{\max} = 4/4 = \boxed{100\%}$ . Carbon ( $m_2 = 12m_1$ ): $f_{\max} = \dfrac{4 \times 12}{13^2} = \dfrac{48}{169} = \boxed{28.4\%}$ .

Step 2. Between collisions the energy falls by a random factor; the right average is logarithmic. For isotropic CM scattering the mean logarithmic energy decrement is $\xi = 1 + \dfrac{\alpha\ln\alpha}{1-\alpha}$ with $\alpha = \left(\dfrac{A-1}{A+1}\right)^2$ .

Step 3. Hydrogen: $A = 1 \Rightarrow \alpha = 0$ , so $\xi = 1$ exactly. Carbon: $\alpha = (11/13)^2 = 0.7160$ , $\ln\alpha = -0.3342$ , so $\xi = 1 + \dfrac{0.7160 \times (-0.3342)}{0.2840} = 1 - 0.8425 = 0.1575$ .

Step 4. Collisions needed: $n = \dfrac{1}{\xi}\ln\dfrac{E_0}{E_f}$ , and $\ln\dfrac{2\times10^6}{0.025} = \ln(8.0\times10^7) = 18.2$ . Hydrogen: $n = 18.2/1 = \boxed{\approx 18\ \text{collisions}}$ . Carbon: $n = 18.2/0.1575 = \boxed{\approx 116\ \text{collisions}}$ .

Step 5. Read the physics: the mass-matching factor in equation (7) is the whole story of moderator choice. Water (hydrogen-rich) thermalizes neutrons in ~18 collisions but also captures some; graphite needs ~116 but barely captures — which is why Fermi's 1942 pile used graphite, and why heavy nuclei like lead (transfer $\le 4/208 \approx 1.9\%$ ) are useless as moderators and excellent as reflectors.

Origins · 1 minErnest Rutherford and how this was discovered

Ernest Rutherford

1871–1937 · Cavendish Laboratory, Cambridge

A New Zealand farm boy who became the greatest experimental physicist of his age, Rutherford had already earned a Nobel Prize (Chemistry, 1908, for radioactive decay) before the work that immortalized him. When Geiger and Marsden reported alpha particles bouncing backward off gold foil, Rutherford spent over a year on the kinematics before publishing the nuclear atom in 1911 — and the argument is pure center-of-mass reasoning: electrons, thousands of times lighter than an alpha, could deflect it by at most a few hundredths of a degree (the $\sin\theta_{\max} = m_2/m_1$ ceiling), so the backscattering demanded a scattering partner heavier than the alpha itself, concentrated in a minute nucleus. His scattering formula, derived for a fixed center, is the $m_1 \ll m_2$ limit of the transformations on this page; the finite-mass corrections are made with the reduced mass. He tipped experiments toward collision thinking so thoroughly that his lab's motto might as well have been: when in doubt, boost to the CM frame and count what is heavy.

History

1668

— Christiaan Huygens derives the elastic collision laws by viewing the experiment from a uniformly drifting boat — the first frame-boost argument in physics.

1751

— Johann Samuel König states the kinetic-energy decomposition

T = \tfrac12MV_{\mathrm{cm}}^2 + T_{\mathrm{cm}}

now taught as König's theorem.

1911

— Rutherford infers the nuclear atom from alpha backscattering, using the mass-ratio ceiling on deflection angles to rule out electrons as the scatterers.

1932

— Chadwick discovers the neutron by applying elastic-collision kinematics to recoiling protons and nitrogen nuclei, extracting the neutron mass from the two recoil speeds.

1942

— Fermi's CP-1 reactor goes critical in Chicago: graphite chosen as moderator by exactly the energy-transfer arithmetic of equation (7).

Going further

Texts. Morin, Introduction to Classical Mechanics, ch. 5 (Conservation of Momentum) & ch. 4 · Goldstein, Poole & Safko, Classical Mechanics, 3rd ed., ch. 3 §3.11 (Scattering in a Central Force Field) · Landau & Lifshitz, Mechanics, 3rd ed., §16–17 (Collisions of Particles) · Taylor, Classical Mechanics, ch. 14 (Collision Theory).

Provenance. Equations derivable from momentum conservation and Galilean transformations. Verifier should recompute: (1) König's theorem cross-term vanishing using $\sum m_i\mathbf{v}_i'=0$ ; (2) reduced-mass internal energy $T_{\mathrm{cm}}=\tfrac12\mu v_{\mathrm{rel}}^2$ ; (3) angle transformation $\tan\theta_{\mathrm{lab}}=\sin\theta_{\mathrm{cm}}/(\cos\theta_{\mathrm{cm}}+m_1/m_2)$ ; (4) max-angle $\sin\theta_{\max}=m_2/m_1$ ; (5) fractional energy transfer $4m_1m_2/(m_1+m_2)^2$ . Cross-check against Goldstein §3.11 (Eqs. 3.110–3.112), Landau-Lifshitz Mechanics §16–17, Taylor Ch. 14, Morin Ch. 5. Reduced mass $\mu=m_1m_2/(m_1+m_2)$ is standard. No numerical physical constants asserted beyond mass-ratio identities. Worked-example numbers on this page (the ±2.0/±4.0 m/s cart round-trip; moderator figures $f_{\max} = 100\%$ vs $28.4\%$ , $\xi = 0.1575$ for carbon, and ≈18 vs ≈116 collisions from $\ln(8\times10^7) = 18.2$ ) are illustrative arithmetic recomputed from the boxed formulas.

Builds on: Relative Motion and Galilean Transformations · Newton's Third Law · Newton's First Law and Inertial Frames · Vectors in Kinematics and Coordinate Systems · Velocity and Acceleration. Leads to: Rutherford Scattering · The Two-Body Problem and Reduced Mass · Relativistic Collisions and Invariant Mass.

Is the CM frame just the rest frame of the heavier body?

No — it is the rest frame of the system as a whole, and it generally coincides with neither body's frame. For a projectile $m_1$ hitting a stationary $m_2$ , the CM frame moves at $\tfrac{m_1}{m_1+m_2}v_1$ : for equal masses that is half the beam velocity, clearly not either body's rest frame. It approaches the heavy body's rest frame only in the limit $m_2 \gg m_1$ , which is why Rutherford could treat the gold nucleus as a fixed scattering center. The defining property is dynamical, not geometrical: it is the unique inertial frame in which $\sum m_i\mathbf{v}_i' = 0$ , equation (2), and that zero is what makes collisions symmetric there.

Why are the CM speeds unchanged by an elastic collision — could the two bodies not share the energy differently?

In the CM frame the momenta must be equal and opposite both before and after (that is what the frame means), so there is only one momentum magnitude $p'$ to solve for. The internal kinetic energy is then fixed by that single number: $T_{\mathrm{cm}} = p'^2/2m_1 + p'^2/2m_2 = p'^2/2\mu$ . Demanding $T_{\mathrm{cm}}$ unchanged (elastic) forces $p'$ unchanged — there is simply no second degree of freedom to reshuffle energy into. The only freedom left is the direction of the momentum line, which is why an elastic collision in the CM frame is a pure rotation, and why the outcome circle in the velocity diagram has fixed radius.

A gravity assist speeds the spacecraft up, but an elastic collision conserves kinetic energy — where does the energy come from?

From the planet — legally. In the planet–spacecraft CM frame (essentially the planet's frame, since $m_2 \sim 10^{26}$ kg) the encounter is elastic: the spacecraft's speed relative to the planet is the same on the way in and the way out; only the direction rotates. But the Sun-frame velocity is that rotated vector plus the planet's orbital velocity $\mathbf{V}_{\mathrm{cm}}$ , and if the rotation aligns the outgoing leg with the planet's motion, the magnitudes add. The spacecraft can gain up to $2V_{\mathrm{planet}}$ in the ideal reversal. Momentum bookkeeping is exact: the planet loses precisely the momentum the spacecraft gains, decelerating by an utterly negligible $\sim m_{\mathrm{sc}}/m_{\mathrm{planet}}$ fraction.

What actually changes when the collision is relativistic?

The logic survives; the arithmetic upgrades. The CM frame becomes the center-of-momentum frame (defined the same way: total three-momentum zero), the boost becomes a Lorentz transformation instead of a Galilean one, and the invariant that measures spendable energy becomes $\sqrt{s} = \sqrt{(\sum E)^2 - |\sum \mathbf{p}c|^2}$ — the invariant mass, the relativistic heir of $T_{\mathrm{cm}}$ . The punchline sharpens dramatically: for a fixed-target machine $\sqrt{s} \approx \sqrt{2E_{\mathrm{beam}}m_2c^2}$ grows only as the square root of beam energy, while a collider gets $\sqrt{s} = 2E_{\mathrm{beam}}$ . König's split — bulk motion is wasted motion — is the reason the LHC is a ring of two counter-rotating beams and not a beam on a block of iron.

Why does the maximum energy transfer formula peak at equal masses?

Look at the factor $f = 4m_1m_2/(m_1+m_2)^2$ as a function of the ratio $r = m_2/m_1$ : it is $4r/(1+r)^2$ , which vanishes at both extremes and peaks at exactly $r = 1$ , where $f = 1$ . Physically, a too-heavy target is a wall — the projectile bounces, keeping its energy (reversed velocity, same speed); a too-light target is a fly — it flees at up to twice the projectile's speed but carries almost no energy because it has almost no mass. Perfect handoff needs impedance matching, and for collisions the impedance is mass. The same $4m_1m_2/M^2$ appears in phonon transmission across interfaces and in maximum power transfer between mechanical oscillators — mass-matching is a universal design rule.

Can I still use the CM frame when external forces act during the collision, like gravity on colliding balls?

Yes, for the collision itself — by the impulse argument. During the brief contact the internal forces are overwhelmingly larger than gravity, so over the collision's millisecond the CM velocity is constant to high accuracy and the CM analysis of "before vs after" holds. Over longer stretches the CM frame is no longer inertial if external forces act: the CM accelerates at $\mathbf{F}_{\mathrm{ext}}/M$ (it falls on a parabola for projectiles), and a frame glued to it acquires fictitious forces. The practical rule: use the CM frame across the impact, then let the CM resume its externally-forced trajectory. That is exactly the two-stage logic of the ballistic pendulum, and it fails only when the "collision" is slow enough for external impulses to compete.

Is the reduced mass a real physical mass? What exactly moves with mass mu?

Nothing you could put on a scale — $\mu$ is the effective inertia of a coordinate, the relative separation $\mathbf{r} = \mathbf{r}_1 - \mathbf{r}_2$ . When you rewrite two-body dynamics in terms of $\mathbf{R}_{\mathrm{cm}}$ and $\mathbf{r}$ , the internal equation of motion becomes $\mu\ddot{\mathbf{r}} = \mathbf{F}_{12}$ : one fictitious particle of mass $\mu$ responding to the interaction force. The payoff in this article is equation (4): the entire spendable energy of a collision is $\tfrac12\mu v_{\mathrm{rel}}^2$ , as if only that fictitious particle existed. The concept recurs everywhere two bodies interact — hydrogen's electron orbits with $\mu = m_e m_p/(m_e + m_p)$ , binary stars orbit their common CM — always the same change of variables.

For equal masses the lab angle is half the CM angle — where does the factor of 2 come from?

From an isosceles triangle. The final lab velocity is $\mathbf{V}_{\mathrm{cm}}$ plus the rotated CM velocity $\mathbf{u}_1$ , and for equal masses these two vectors have equal length ( $V_{\mathrm{cm}} = u_1 = v_1/2$ ). The lab velocity is then the diagonal of a rhombus whose sides make angle $\theta_{\mathrm{cm}}$ with each other, and a rhombus diagonal bisects its vertex angle — hence $\theta_{\mathrm{lab}} = \theta_{\mathrm{cm}}/2$ . The same triangle shows the recoiling target leaving at $(\pi - \theta_{\mathrm{cm}})/2$ on the other side of the beam, and adding the two lab angles gives $\pi/2$ : the right-angle theorem, seen now as pure geometry rather than algebra.