Technique Masterclasses

Mechanics 1

Deep Dive: Momentum Impulse Sign Convention

Provenance: written 2026-09-29. Read in full for this pass: research/veridian/WME01-verified-facts.md (the 17-series verified-facts bank — §1 spec text, §2 formula-booklet section, §3 mark-scheme conventions, §4 topic 4.3, §5 source archive, §6–§8, and the lesson-audit entry for this topic at §9, lines 214–222); content/veridian/curriculum/wme01/momentum-impulse-sign-convention.ts in full (its own teach block, mechanism derivation, diagram, worked chain, method-comparison and all MCQs/prequestions are not reproduced below — only referenced); all three existing WME01 technique-hub documents in full (codex-formula-and-method-bank.md §1/§2.9, forge-exam-technique-guide.md §1–§4, signal-frequency-and-boundaries.md §1); and content/veridian/question-bank/wme01.ts (grepped for "momentum/impulse/collision/4.3" — no hits, confirming no real past-paper momentum question with a recovered numeric stem exists anywhere in this repo to cross-check against). No new primary-source PDF was fetched for this pass; every claim below traces to one of the five documents just named, each of which already traces its own claims to a real spec, formula-booklet, mark-scheme, examiner-report or question-paper PDF. Every quotation from a Pearson document below is kept to a short fragment with attribution; longer passages are paraphrased.

CODEX Deep-Dive — Momentum, Impulse, and the Sign-Convention Trap

Pearson Edexcel IAL Mechanics 1 (WME01/01), spec 4.3

This is not a fourth re-teach of the topic. The lesson file owns the full worked chain, the diagram, the method-comparison block and every MCQ, and none of that is repeated here. codex-formula-and-method-bank.md already gives this spec point a formula-masterlist entry and a short method skeleton with its own numeric check (M/N, 7 kg and 2 kg). forge-exam-technique-guide.md already documents the general M/A/B marking rules and the paper-wide sign-convention finding. What none of the four existing documents does is set out, in one place, the complete structural model this topic rests on — including exactly why real questions on it are always shaped the way they are — name every distinct sub-case spec 4.3's own two sentences actually split into, and assemble the full real exam record for this specific topic in one synthesis. That is this document's job.

1. Spec 4.3 is two sentences, and they name two different pieces of content

The entire spec text for this point reads: momentum and impulse; the impulse–momentum principle; "the principle of conservation of momentum applied to two particles colliding directly." That is not one idea repeated for emphasis — it is two separate, spec-named techniques stacked in one short item:

Both formulae are confirmed absent from the real M1 formula booklet (Issue 2, January 2021) and named directly in the spec's own "must be memorised" list (spec M1.2 §3) — the single sharpest planning fact for this topic, already the lesson's own opening point and not re-argued here. What is worth adding is that this two-part spec structure is also why the topic's real exam record (§7 below) contains two genuinely different question shapes, not one: a pure single-particle impulse calculation, and a two-particle collision. The lesson's own worked chain and its method-comparison block already show sub-case A living inside a sub-case B scenario (finding the impulse on one particle of a two-particle collision) — but sub-case A does not require that context, and a candidate should recognise it as fully self-contained whenever it appears alone.

The spec's own guidance adds one explicit exclusion and one explicit scope limit, both load-bearing for everything that follows: "Knowledge of Newton's law of restitution is not required," and problems are confined to "those of a one-dimensional nature." Restitution is the usual second physical relationship a general mechanics course uses to fix the outcome of a collision (it links the particles' separation speed after to their approach speed before); WME01 deliberately withholds it.

2. Why every real collision question has to hand you one velocity — a structural consequence, not a house style

This is the genuinely new piece of theory this document adds: a plain reading of why sub-case B's real questions are shaped the way they are, rather than simply noting that shape as a pattern.

For two particles colliding directly, with both masses and both initial velocities known, conservation of momentum is exactly one scalar equation. The two velocities after the collision, v1v_1 and v2v_2, are in general two unknowns. One equation cannot fix two unknowns on its own. In a mechanics course that teaches Newton's law of restitution, the missing second equation comes from the restitution law itself, e=v2−v1u1−u2e = \frac{v_2 - v_1}{u_1 - u_2}, and the system becomes solvable for any collision. Spec 4.3 explicitly withholds that law. The consequence follows by pure algebra, not by observation of past papers: a WME01 collision question is only solvable at all if something else closes that second degree of freedom — either one of the two post-collision velocities is given directly as data, or the collision is specified as a coalescence (the two particles moving off together, v1=v2v_1 = v_2, which collapses two unknowns into one by definition rather than by extra information).

This is exactly the shape every real and VERIDIAN-original example already collected across this course's own WME01 material takes, checked directly: the lesson's own worked chain gives D's post-collision velocity directly and solves for C's; codex-formula-and-method-bank.md's own short check gives N's post-collision velocity directly and solves for M's; and the one anchor question whose full mark breakdown this course's research recovered, Jan 2020 Q1, is structured as an impulse part followed by a "find the other particle's velocity" part — consistent with, though not itself proof of, the same one-velocity-given shape (the exact wording of that question's stem was not recovered; only its mark-scheme and examiner-report language was — see §7). No example anywhere in the material reviewed for this course gives both post-collision velocities as independent unknowns to be found from momentum alone, and the algebra above is the reason no such question could exist under this spec's own stated restriction.

One further, purely mathematical consequence worth stating explicitly, since the lesson's own diagram asserts it without proving it: reversing which direction is called positive multiplies every velocity, and every term of the conservation equation, by −1-1. Since −(m1u1)+−(m2u2)=−(m1v1)+−(m2v2)-(m_1u_1) + -(m_2u_2) = -(m_1v_1) + -(m_2v_2) is just the original equation multiplied through by −1-1, it is exactly as true as the original — which is why the convention is a free, arbitrary choice with no physical content of its own, and why it must nonetheless be fixed once, consistently, across every velocity in the working: switching it mid-calculation does not multiply everything by −1-1 consistently, and the two sides stop describing the same collision.

3. Sub-case A in isolation — a single particle, no second body at all

A single-particle impulse question needs no collision, no second mass and no conservation equation — just one substitution into I=m(v−u)I = m(v - u). VERIDIAN-original micro-example, independently checked: a particle of mass 2.5 kg has velocity 4 m/s; a force acts on it briefly, after which its velocity is −2-2 m/s (taking the particle's original direction as positive). The impulse is I=2.5(−2−4)=2.5(−6)=−15I = 2.5(-2 - 4) = 2.5(-6) = -15 N s.

The sign here is not a final-answer technicality of the kind the lesson's trap taxonomy addresses — it is the content of the answer. A negative impulse means the net force that produced it acted, on balance, in the direction opposite to the particle's original motion; here, one strong enough not just to slow the particle but to reverse it. If the question instead asked for the impulse's magnitude, the answer would be 15 N s — the identical sign-strip-when-asked-for-a-magnitude discipline the lesson teaches for speed, applied to impulse instead of velocity. This is the same discipline generalised one level, not a second discipline to learn.

4. Sub-case B's second branch — coalescence, and a check the archive cannot supply

The lesson's own mi-mcq-1 already illustrates a coalescing collision (VERIDIAN-original particles E and F, joining into one body) — that worked example is not reproduced here. What is worth adding, because it is genuine model-completeness content rather than a restatement, is the general result and an independent numeric check of it.

When two particles coalesce, v1=v2=vv_1 = v_2 = v by definition, and conservation of momentum reduces immediately to v=m1u1+m2u2m1+m2v = \dfrac{m_1u_1 + m_2u_2}{m_1 + m_2} — the two unknowns collapsing to one is what makes this variant solvable without any given post-collision velocity at all, exactly the second route out of the underdetermination described in §2. A coalescing collision also necessarily loses kinetic energy (unless the two particles already share a common velocity before impact, in which case there is no collision to speak of): ΔKE=12(m1m2m1+m2)(u1−u2)2\Delta KE = \tfrac{1}{2}\left(\dfrac{m_1 m_2}{m_1+m_2}\right)(u_1 - u_2)^2, a standard identity, independently verified here with a fresh numeric check rather than asserted: m1=3m_1 = 3 kg, u1=10u_1 = 10 m/s, m2=2m_2 = 2 kg, u2=0u_2 = 0. Combined velocity v=3(10)+2(0)5=6v = \frac{3(10)+2(0)}{5} = 6 m/s. KEbefore=12(3)(10)2=150KE_{\text{before}} = \frac12(3)(10)^2 = 150 J; KEafter=12(5)(6)2=90KE_{\text{after}} = \frac12(5)(6)^2 = 90 J; loss =60= 60 J. The identity above gives 12⋅(3)(2)5⋅(10−0)2=12(1.2)(100)=60\frac12 \cdot \frac{(3)(2)}{5} \cdot (10-0)^2 = \frac12(1.2)(100) = 60 J — matching exactly, confirming both the combined-velocity formula and the energy-loss identity on an independent set of numbers from the lesson's own E/F pair.

This paragraph is flagged honestly as model-completeness content, not exam-frequency content: nothing in the 17-series archive this course has reviewed confirms a real WME01 question built around a coalescing collision specifically (a grep of the facts bank for "coalesc" returns nothing). Spec 4.3's own wording — "two particles colliding directly" — does not itself rule the case out, and it is a real, distinct mathematical sub-case of the same general principle, which is why it belongs in a document about the topic's complete theory; it should not be read as evidence that a coalescing question has actually been set.

5. The sign convention as a constraint on an unknown mass — completing an algebraic step the lesson leaves symbolic

The lesson's own trap-taxonomy already names this variant and quotes its real source directly: Jan 2020 Q1(c)'s mark scheme requires "correct inequality for their v, containing u," scoring the move from a signed-velocity requirement to a mass inequality — verbatim, the real answer is m2>34m1m_2 > \tfrac34 m_1 — and the matching examiner report confirms candidates had to track "whether their expression for velocity... was > 0 or < 0" against their own chosen direction. The lesson applies this to its own C/D pair only in the abstract, leaving both masses symbolic and stopping before an explicit inequality is derived. Completing that step with concrete numbers is genuinely additive.

VERIDIAN-original micro-example, independently checked: particle G, mass 6 kg, moving at 8 m/s, collides directly with a stationary particle of unknown mass mm kg. After the collision, the previously-stationary particle moves off at 5 m/s in the direction G was originally travelling (given). Taking G's original direction as positive: 6(8)+m(0)=6vG+5m6(8) + m(0) = 6v_G + 5m, so vG=48−5m6v_G = \dfrac{48 - 5m}{6}. Requiring G to rebound — vG<0v_G < 0 — gives 48−5m<048 - 5m < 0, i.e. m>9.6m > 9.6 kg: exactly the mark scheme's own technique (form the inequality from conservation of momentum, then impose the required sign), reproduced end-to-end on fresh numbers rather than left symbolic. The algebra itself is what the real M1 mark rewards; it does not, by itself, guarantee that every value of mm in that range corresponds to an equally realistic collision from this same starting setup. One representative point inside the range was checked directly for that reason: at m=12m = 12 kg, vG=48−606=−2v_G = \frac{48-60}{6} = -2 m/s, momentum after =6(−2)+12(5)=48= 6(-2) + 12(5) = 48 ✓ (matching momentum before), and kinetic energy 192192 J before, 162162 J after — a loss, not a gain, confirming this specific point describes a physically coherent collision rather than an equation that merely balances on paper.

6. How strict is "report the exact word asked for"? A boundary case from a different spec point, worth knowing before over-generalising

forge-exam-technique-guide.md records a real finding that complicates a too-simple version of this topic's own rule ("always strip the sign if the wording resembles speed"). On a different spec point entirely — Jan 2025 Q5(a), vectors (spec 2.1–2.2), not momentum — a real WME01 question asked for a velocity (a signed quantity), and a significant number of candidates instead reported it as an unsigned speed; the mark scheme applied no penalty for that specific substitution. This is the mirror image of the momentum trap (there, a signed answer is given where an unsigned one was required; here, an unsigned answer is given where, strictly, a signed one was asked for) — and the two real findings sit side by side without contradicting each other, because the actual rule is narrower than either error suggests on its own: the exact wording of the specific question part decides which format earns the mark, and different series and different question parts have been shown to enforce that requirement with different strictness. The momentum topic's own three-series trap (Jan 2020, Oct 2022, Jun 2024) is the single most consistently and strictly enforced instance of this wording-sensitivity anywhere in the WME01 record reviewed — that much is not weakened by this boundary case — but a learner should not conclude from it that every sign-format mismatch on this paper costs a mark regardless of direction. This document did not itself re-fetch the Jan 2025 primary PDF; the claim is sourced to FORGE's own prior deep-read of it, not independently re-verified this pass.

7. The real exam record for this topic, assembled in one place

No likelihood percentage and no grade boundary is stated anywhere above for this specific spec point. Every frequency claim in this section is a named count of series this course has actually read, not an estimate.

8. What remains unconfirmed

Stated plainly rather than left implicit: Oct 2022 Q1's and Jun 2024 Q1's own total marks and part structure were never recovered in this course's research — only the examiner-report sentence describing each series' sign-convention error, and, for Jun 2024, the particle-choice sentence. Whether Newton's third law is ever tested as its own separate step (rather than as a route to the same answer an alternative method reaches without it) is not confirmed either way. No real WME01 question anywhere in the archive reviewed frames impulse as force × time (I=FtI = Ft) rather than via a before/after velocity pair — this document does not assert that variant is examined here, since spec 4.3's own text names only the impulse–momentum principle. No coalescing collision, and no question involving more than two particles or more than one collision in sequence, was found tested anywhere in the material reviewed (§4). Whether a bare, no-second-clue single-particle impulse question (sub-case A entirely on its own, with no colliding second body anywhere in the stem) has ever anchored its own standalone question, as opposed to appearing as one part of a two-particle scenario the way every real example recovered for this course does, is genuinely unknown from the material reviewed.

Sources