ARCHETYPE: Turning on a Flat Road

A car takes a bend on a perfectly level road. No banking helps it, no rail guides it — so what force actually turns two tonnes of metal? Not the engine: the engine pushes the car forward. Not the steering wheel: it only re-aims the tyres. The entire inward pull, every newton of the $mv^2/r$ the curve demands, is supplied by friction between four palm-sized patches of rubber and the tarmac. When that friction can no longer pay, the car does not "fly outward" — it simply stops turning and continues straight, along the tangent, into the scenery. That single budget — friction's ceiling $\mu_s mg$ against the curve's invoice $mv^2/r$ — generates one of the most-asked archetypes in circular dynamics. It sets the maximum speed on a flat curve, $v_{\max} = \sqrt{\mu_s g r}$ , with the famous surprise that the car's mass cancels: a loaded truck and a scooter share the same speed limit on the same bend. The same inequality, rearranged, governs a coin riding a turntable, a cyclist leaning into a corner at exactly $\tan\theta = v^2/gr$ , and the racing driver's dilemma that braking and turning draw from one grip budget. The concept module audited who pays for circular motion; this article drills the flat-road setup to exam mastery — the skid condition, the cyclist's lean, the turntable variant, the grip circle, and the traps (kilometres per hour, kinetic-versus-static, phantom centrifugal forces) that convert right physics into wrong marks.

Key takeaways

On a level curve, static friction is the only inward force available: $\dfrac{mv^2}{r} \le \mu_s N = \mu_s mg$ .
Maximum safe speed: $v_{\max} = \sqrt{\mu_s g r}$ — the mass cancels; grip and geometry alone set the limit.
Equivalently, at speed $v$ the tightest possible turn is $r_{\min} = \dfrac{v^2}{\mu_s g}$ ; doubling speed quadruples the required radius.
Past the limit the car skids along the tangent — friction fails and the path straightens; nothing pushes it radially outward.
A cyclist must lean so the ground's total push lines up through the centre of mass: $\tan\theta = \dfrac{v^2}{gr}$ .
Braking while turning shares one budget: $\sqrt{f_{\rm turn}^2 + f_{\rm brake}^2} \le \mu_s mg$ — the grip circle.
Key idea: a flat turn is an audit — the invoice $mv^2/r$ against friction's ceiling $\mu_s mg$ ; every variant asks where that inequality bites.

The core insight

On a level road the entire centripetal requirement $mv^2/r$ must be paid by static friction, whose ceiling $\mu_s mg$ is fixed by grip and weight — so safe cornering is the inequality $v^2 \le \mu_s g r$ , mass cancels out of it, and every flat-curve problem is a question about where the budget runs out.

Where this lives today

Every modern car runs this article's audit in silicon. Electronic Stability Control compares the yaw rate the driver requests (steering angle, speed) with the yaw rate physics delivers, and when the flat-curve inequality is about to fail it brakes individual wheels to buy back the difference — regulators credit it with cutting single-vehicle skid fatalities by roughly a third. Highway engineers still set advisory speed plates on unbanked bends from $v = \sqrt{\mu g r}$ with a conservative wet-road $\mu$ ; the winter-road community publishes live "friction maps" from instrumented vehicles, since ice's $\mu \approx 0.1$ slashes every limit to a third of dry. MotoGP riders live inside the lean-angle law — modern machines reach $\theta > 60°$ , demanding $\tan 60° \approx 1.7$ of effective grip from race slicks — and the traction-circle display on a racing telemetry screen is equation (5) drawn live, lap after lap.

The setup

Geometry first. View the turning car head-on: three forces only. Weight $mg$ down, normal force $N$ up from the level road, and friction $f$ pointing horizontally toward the centre of the curve — sideways across the tyres. Vertically nothing accelerates, so $N = mg$ . Horizontally, the car is accelerating at $v^2/r$ toward the centre (that is what moving on a circle means), and friction is the only force with a horizontal component to pay for it.

Figure

The flat turn, head-on — friction alone pays the centripetal invoice

Two readings of the picture protect marks before any algebra:

The friction is static, and it is sideways. The contact patch of a rolling tyre is momentarily at rest on the road — the tyre rolls, it does not skate — so the relevant coefficient is $\mu_s$ , not $\mu_k$ . And the friction we need points across the direction of travel: the road pushes the tyres toward the centre because the steered tyres are trying to slide outward. The moment the demanded sideways force exceeds $\mu_s N$ , the patch lets go, the friction collapses to the (smaller) kinetic value, and the turn is lost — which is why the skid, once begun, is so hard to recover.

No new force appears on the FBD. There is no "centripetal force" arrow — centripetal is the job description, and friction is the employee doing it. There is no "centrifugal force" arrow either: in the ground frame nothing pushes the car outward. What a skidding car does is revert to Newton's first law — no adequate inward force, so it travels in a straighter line than the road. Crash investigators read this signature at every curved-road accident: the tyre marks leave the bend along the tangent.

The master solution pattern

Every flat-curve problem is the same audit. Write the invoice, write the payer's ceiling, and impose payment:

\frac{mv^2}{r} \le \mu_s N = \mu_s m g \qquad (1)

where: $N = mg$ came from the vertical balance on a level road (banking changes this — that is the next archetype). The mass appears on both sides and cancels — the single most examined feature of the result. Grip quality $\mu_s$ and geometry $r$ decide everything; loading the car changes both invoice and ceiling in the same proportion. At the ceiling, the maximum speed:

v_{\max} = \sqrt{\mu_s\, g\, r} \qquad (2)

where: the square root enforces gentle scaling — four times the radius buys only twice the speed. Anchor numbers: dry road ( $\mu_s \approx 0.8$ ), $r = 50$ m: $v_{\max} = \sqrt{0.8 \times 10 \times 50} = 20$ m/s $= 72$ km/h. Rain ( $\mu_s \approx 0.5$ ): $15.8$ m/s. Ice ( $\mu_s \approx 0.1$ ): $7.1$ m/s — a quarter of the dry limit, because $\sqrt{1/8} \approx 0.35$ . The same inequality solved for the geometry gives the tightest turn at a given speed:

r_{\min} = \frac{v^2}{\mu_s\, g} \qquad (3)

where: the $v^2$ is the danger — at double the speed you need four times the room to turn. A motorway exit designed for 60 km/h needs sixteen times the radius a car-park loop at 15 km/h needs, not four. And the coin-on-turntable variant runs the audit at fixed angular speed $\omega$ (so the invoice is $m\omega^2 r$ , growing with radius):

m\omega^2 r \le \mu_s m g \qquad\Longrightarrow\qquad r_{\max} = \frac{\mu_s g}{\omega^2} \qquad (4)

where: on a turntable the outer seats are the dangerous ones — invoice grows with $r$ at fixed $\omega$ — exactly opposite to the car, where (at fixed $v$ ) a larger radius makes life easier. Reading which variable is held fixed is the entire question.

you want	use	notes
max speed on a flat bend	$v_{\max} = \sqrt{\mu_s g r}$	mass cancels; static $\mu$
tightest turn at speed $v$	$r_{\min} = v^2/\mu_s g$	doubles speed → 4× radius
coin/passenger on turntable	$r_{\max} = \mu_s g/\omega^2$	fixed $\omega$ : outer = worse
cyclist's lean angle	$\tan\theta = v^2/gr$	same triangle as banking
braking while turning	$\sqrt{(v^2/r)^2 + a_t^2} \le \mu_s g$	one grip budget for both jobs

Derivation AThe flat-curve speed limit — and why the mass cancels

We derive $v_{\max} = \sqrt{\mu_s g r}$ from the two axis equations, establish the mass cancellation honestly, and read the skid condition. The strategy is the standard circular-dynamics routine: axes horizontal (toward the centre — the direction of the acceleration) and vertical; friction is capped, not fixed, so the result is an inequality. Watch the logic of static friction: below the limit it supplies exactly what the invoice demands, no more.

Step 1. Vertical equation (no vertical acceleration): $N - mg = 0$ , so $N = mg$ . On a level road, the full weight sets the friction ceiling.

Step 2. Horizontal (centripetal) equation: the net inward force must equal $mv^2/r$ , and friction is the only candidate: $f = \dfrac{mv^2}{r}$ . Why $f$ is not automatically $\mu_s N$ : static friction is self-adjusting. At 30 km/h on a wide bend it supplies a small fraction of its ceiling; the ceiling matters only at the limit. Writing $f = \mu_s N$ for a sub-limit speed is the classic error the "does it skid?" questions are built to catch.

Step 3. Impose the ceiling: $\dfrac{mv^2}{r} \le \mu_s N = \mu_s m g$ . The mass divides out on both sides:

v^2 \le \mu_s\, g\, r \qquad\Rightarrow\qquad v_{\max} = \sqrt{\mu_s g r}

Why the cancellation is physical, not algebraic luck: a heavier car needs more inward force (invoice ∝ m) but also presses harder on the road, raising the friction ceiling (∝ m). The two effects scale identically. What loading actually changes is tyre behaviour — real μ falls slightly with load — but that is engineering, not this model.

Step 4. Read the failure mode. At $v > v_{\max}$ the required force exceeds the ceiling; the tyres slip, friction drops to kinetic, and the inward supply collapses below even the old ceiling. The car's path straightens toward the tangent — the radius it can negotiate at this speed, $v^2/\mu_k g$ , is larger than the road's. It exits the bend on the outside, moving initially along the tangent line, not radially outward. Verified: numbers — $\mu_s = 0.8$ , $r = 50$ m, $g = 10$ : $v_{\max} = 20$ m/s ✓ (worked example 3); NCERT cyclist check ( $\mu_s = 0.1$ , $r = 3$ m, $g = 9.8$ ): $v_{\max}^2 = 2.94\ \mathrm{m^2/s^2}$ , and $18$ km/h $= 5$ m/s gives $v^2 = 25 \gg 2.94$ — he skids ✓. Treatment as in Halliday Ch. 6, NCERT Ch. 5.

Go deeper · Derivation B · 2 minThe cyclist's lean — why two wheels must tilt at tan θ = v²/gr

Derivation BThe cyclist's lean — why two wheels must tilt at tan θ = v²/gr

We derive the lean angle of a cyclist (or motorcyclist) rounding a flat curve. Forces alone cannot distinguish a leaning rider from an upright one — the new ingredient is torque: on two wheels, the ground's push must point through the centre of mass, or the rider topples. The strategy: compute the direction of the ground's total push (normal + friction), then demand the body line up with it. Watch the punchline — the lean angle law is the banking law with the roles reversed: here the vehicle tilts, not the road.

Step 1. Forces at the contact: normal $N = mg$ (vertical balance) and friction $f = mv^2/r$ (paying the invoice). The ground's total push is their vector sum: magnitude $\sqrt{N^2 + f^2}$ , tilted inward from the vertical by angle $\phi$ with

\tan\phi = \frac{f}{N} = \frac{v^2}{gr}

Step 2. Torque about the centre of mass. Gravity acts at the centre of mass (no torque). The only other force is the ground's push at the contact patch. If that push's line of action misses the centre of mass, it exerts an unbalanced torque and the bike rotates — falls inward or outward. Steady cornering therefore requires the push to aim straight at the centre of mass.

Step 3. The bike's frame must lie along that line: the lean angle from the vertical equals $\phi$ ,

\tan\theta = \frac{v^2}{g\,r}

Why cars do not need this: a car's four contact patches form a rectangle — the ground can redistribute normal forces between inner and outer wheels to cancel the torque. A single-track vehicle has no width to redistribute across; it must borrow width by leaning.

Step 4. Numbers: $v = 10$ m/s, $r = 20$ m, $g = 10$ : $\tan\theta = 100/200 = 0.5$ , $\theta = 26.6°$ — a real, visible motorcycle lean. And the friction demanded is $f = mg\tan\theta = 0.5\,mg$ , so the stunt requires $\mu_s \ge 0.5$ : the lean does not reduce the friction needed — it only prevents toppling. Grip and balance are separate audits, both of which must pass. Verified: limits — $v \to 0$ : $\theta \to 0$ , upright at rest ✓; $\tan\theta$ formula identical to the banked-road design equation with the tilt moved from road to rider (same force triangle) ✓. Standard treatment: NCERT Ch. 5, Kleppner Ch. 2.

Braking while turning — the grip circle

Friction has one ceiling, $\mu_s N$ , but two jobs may bid for it: turning (sideways component $mv^2/r$ ) and braking or accelerating (fore-aft component $ma_t$ ). The two components add as vectors, so the audit becomes

\sqrt{\left(\frac{v^2}{r}\right)^2 + a_t^2} \;\le\; \mu_s\, g \qquad (5)

where: the allowed combinations of cornering and braking fill a circle of radius $\mu_s g$ in acceleration space — the racing driver's "traction circle". Corner at the limit and you may not brake at all; brake hard and you must unwind the steering. Every "a car decelerates while rounding a bend" JEE problem is equation (5), and every racing line ever driven is an attempt to spend the whole circle without leaving it.

Figure

The grip circle — cornering and braking spend one friction budget

Limiting cases

case	result	meaning
$\mu_s \to 0$ (ice)	$v_{\max} \to 0$	no grip, no turning: any speed skids
$r \to \infty$	$v_{\max} \to \infty$	a straight road demands nothing
$v \to 2v$	$r_{\min} \to 4r_{\min}$	the $v^2$ law: speed is paid for in quadratic room
load doubled	$v_{\max}$ unchanged	invoice and ceiling both scale with $m$
turntable, $r \to 0$	always safe	invoice $m\omega^2 r$ vanishes at the axis
lean angle, $v \to 0$	$\theta \to 0$	a stationary cyclist stands upright
braking at $\mu_s g$	zero turning capacity	the grip circle: one budget, two bidders

Common trapCentrifugal force threw the car off the road

No force threw it anywhere — the road simply asked for more inward force than friction could supply, and the car obeyed Newton's first law. In the ground frame (the frame of every free-body diagram here) the forces on the car are weight, normal force, and friction; there is nothing pointing outward. A car that exceeds $v_{\max}$ leaves the bend along the tangent — the near-straight path of an under-forced body — not along the radius, and skid marks at real accidents show exactly that. The companion error is drawing a "centripetal force" arrow in addition to friction: centripetal is the total that real forces must add up to, not an extra member of the cast. Double-booking it produces $2mv^2/r$ nonsense downstream. Keep the audit clean: invoice on the right of Newton's law, real payers on the left, and the word "centrifugal" reserved for rotating frames (the pseudo-forces module), where it is legitimate bookkeeping — never a real push on a real car in the ground frame.

Worked examples

1 · Board level — will the cyclist skid? (NCERT/NEET pattern, 3 marks)

A cyclist speeding at $18$ km/h on a level road takes a sharp circular turn of radius $3.0$ m without slowing. The coefficient of static friction between tyres and road is $0.1$ . Will the cyclist slip? ( $g = 9.8\ \mathrm{m/s^2}$ .)

Step 1. Convert units first: $v = 18$ km/h $= 18 \times \dfrac{1000}{3600} = 5.0$ m/s. (This conversion is a designed trap — see Exam traps.)

Step 2. The audit. Required: $\dfrac{v^2}{r} = \dfrac{25}{3} = 8.3\ \mathrm{m/s^2}$ of centripetal acceleration. Available: $\mu_s g = 0.1 \times 9.8 = 0.98\ \mathrm{m/s^2}$ .

Step 3. $8.3 \gg 0.98$ : the demand is eight times the ceiling. The cyclist slips. Equivalently $v_{\max} = \sqrt{\mu_s g r} = \sqrt{2.94} = 1.7$ m/s — he is at nearly three times the safe speed.

Step 4. State the failure honestly for the last mark: the tyres cannot supply $mv^2/r$ , so the bicycle does not follow the $3$ m circle; it slides along a flatter arc toward the outside of the turn (initially near the tangent).

2 · Board level — the coin on the turntable (CBSE/ICSE pattern)

A coin sits $10$ cm from the axis of a turntable rotating at $\omega = 5.0$ rad/s. The coefficient of static friction is $0.4$ . (a) Does the coin ride or slide? (b) What is the largest radius at which a coin can ride? ( $g = 10\ \mathrm{m/s^2}$ .)

Step 1. Invoice at fixed $\omega$ : $a_{\rm req} = \omega^2 r = 25 \times 0.10 = 2.5\ \mathrm{m/s^2}$ . Ceiling: $\mu_s g = 4.0\ \mathrm{m/s^2}$ .

Step 2. (a) $2.5 < 4.0$ : friction can pay. The coin rides, with friction supplying exactly $2.5m$ newtons — not its full ceiling.

Step 3. (b) The ceiling binds where $\omega^2 r_{\max} = \mu_s g$ : $r_{\max} = \dfrac{4.0}{25} = 0.16$ m $= \boxed{16\ \mathrm{cm}}$ .

Step 4. Read the turntable's logic: at fixed $\omega$ , danger grows outward ( $a \propto r$ ). A coin nudged past 16 cm slides off along — say it precisely — the tangent to its circle at the point of release.

3 · Competitive level — dry versus wet speed limits (JEE-Main pattern, ~2.5 minutes)

A level highway bend has radius $50$ m. Find the maximum negotiable speed when the road is (a) dry, $\mu_s = 0.8$ , (b) wet, $\mu_s = 0.5$ . (c) By what percentage must a driver at the dry limit slow down when rain starts? ( $g = 10\ \mathrm{m/s^2}$ .)

Step 1. (a) $v_{\max} = \sqrt{0.8 \times 10 \times 50} = \sqrt{400} = \boxed{20\ \mathrm{m/s}}$ ( $72$ km/h).

Step 2. (b) $v_{\max} = \sqrt{0.5 \times 10 \times 50} = \sqrt{250} = \boxed{15.8\ \mathrm{m/s}}$ ( $57$ km/h).

Step 3. (c) The ratio is $\sqrt{\mu_{\rm wet}/\mu_{\rm dry}} = \sqrt{0.5/0.8} = 0.79$ : the driver must shed about $\boxed{21\%}$ of speed.

Step 4. Structural read: speed limits scale as $\sqrt{\mu}$ , so even a big drop in grip demands a milder-looking drop in speed — which is exactly why drivers underestimate rain. Halving $\mu$ cuts the limit by only 29%, but exceeding the limit fails catastrophically, not gradually. Timing note: ~2.5 min. Do not recompute part (b) from scratch — reuse $v \propto \sqrt{\mu}$ . MCQ setters build the options from the wrong scalings ( $\propto \mu$ , $\propto \mu^2$ ) hoping you will compute one case and guess the pattern.

4 · Competitive level — braking on the bend (JEE-Advanced flavour, ~4 minutes)

A car rounds a level bend of radius $80$ m at a steady $16$ m/s, with $\mu_s = 0.5$ . (a) Check that the steady turn is possible. (b) The driver now brakes while holding the same curve. What is the maximum deceleration available? (c) What total deceleration would just cause a skid at this instant? ( $g = 10\ \mathrm{m/s^2}$ .)

Step 1. (a) Cornering demand: $\dfrac{v^2}{r} = \dfrac{256}{80} = 3.2\ \mathrm{m/s^2}$ . Ceiling: $\mu_s g = 5.0\ \mathrm{m/s^2}$ . Possible, with room to spare.

Step 2. (b) The grip circle (5): sideways and fore-aft demands add vectorially inside the ceiling: $\sqrt{3.2^2 + a_t^2} \le 5.0$ .

Step 3. $a_t^2 \le 25 - 10.24 = 14.76$ , so $a_t \le \boxed{3.84\ \mathrm{m/s^2}}$ — barely three-quarters of the straight-line braking limit of $5\ \mathrm{m/s^2}$ , because cornering has already spent $3.2$ of the budget sideways.

Step 4. (c) The skid begins when the vector total hits the ceiling: total acceleration magnitude $= \mu_s g = \boxed{5.0\ \mathrm{m/s^2}}$ , in whatever mix. Note the trap: the answer to (c) is not $3.2 + 3.84 = 7.0$ — components at right angles add as squares, not as numbers. Timing note: ~4 min. The whole problem is recognizing ONE friction budget spent in two perpendicular directions. Write the circle inequality first; the rest is Pythagoras. JEE-Advanced variants ask for the safe-speed profile while braking — same circle, solved for $v(t)$ conceptually.

Exam traps

Kilometres per hour. The single most reliable mark-killer in this archetype: $18$ km/h is $5$ m/s, not $18$ . Convert in the first line, every time — the NCERT cyclist problem is designed around this slip.

Kinetic instead of static. A rolling tyre's contact patch is momentarily at rest: use $\mu_s$ . The moment it slides you are in kinetic territory with a smaller coefficient — which is why exceeding the limit is unrecoverable, and why "the car skids; find its new path radius" questions switch you to $\mu_k$ .

Assuming $f = \mu N$ below the limit. Static friction supplies what the invoice demands, up to the ceiling. In "does it skid?" problems, compare demand with ceiling; never set them equal unless you are at the maximum.

Mass-dependence distractors. "A heavier car can take the bend faster / slower" — neither: $m$ cancels. Options quoting $v_{\max}$ with an $m$ in it are automatically wrong on a flat curve.

Linear-scaling guesses. $v_{\max} \propto \sqrt{\mu r}$ : double the radius gives $\times1.41$ the speed, not $\times2$ ; and $r_{\min} \propto v^2$ : double the speed needs four times the room.

Skid direction. Along the tangent (then curving as kinetic friction acts) — never radially outward. Any option with "flung outward along the radius" is describing a force that does not exist.

Adding perpendicular demands linearly. Braking-while-turning uses $\sqrt{a_c^2 + a_t^2} \le \mu g$ . The arithmetic sum overestimates the demand and fails part (c) of every grip-circle problem.

Which $g$ . NCERT numericals here habitually use $9.8$ ; JEE/NEET papers usually declare $10$ . The cyclist example above is a $9.8$ classic — use what is stated.

Origins · 1 minCharles-Augustin de Coulomb and how this was discovered

Charles-Augustin de Coulomb

1736–1806 · Paris Academy of Sciences

Before electrostatics made his name a unit, Coulomb was a military engineer who wrote the definitive eighteenth-century treatise on friction. His 1781 Théorie des machines simples — winner of the Academy's prize — established from meticulous experiments that sliding friction is proportional to the normal load, largely independent of contact area, and roughly independent of speed: the "Coulomb friction" model that still rules every $\mu N$ in this article. He distinguished the grip before sliding from the resistance during it — our $\mu_s$ versus $\mu_k$ — a distinction on which the entire flat-curve archetype balances: the safe turn lives on static friction's ceiling, and the skid falls into kinetic friction's weaker hands. That two palm-sized rubber patches can swing a two-tonne car around a bend is Coulomb's proportionality doing quiet, daily work.

History

1781

— Coulomb's prize memoir establishes the laws of dry friction: proportional to load, split into static and kinetic regimes.

1900

s — Early motoring exposes the flat-curve limit; the first posted advisory speeds on bends encode √(μgr) empirically.

1930

s — Tyre-testing rigs measure μ systematically; road engineers begin designing curve radii from target speeds.

1959

— The traction-circle concept enters racing theory: cornering and braking share one friction budget.

1995

— Electronic Stability Control (ESC) ships, sensing incipient skids and braking individual wheels — the grip audit run 25 times a second.

Going further

Texts. NCERT Physics Class 11, Ch. 5 (motion in a circle, the cyclist example) · HC Verma, Concepts of Physics Vol. 1, Ch. 7 · Halliday–Resnick–Walker, Ch. 6 (the flat-curve and traction problems).

Provenance. All results are standard. A verifier should recompute: the NCERT cyclist audit ( $8.3$ vs $0.98\ \mathrm{m/s^2}$ , slips); the turntable pair ( $2.5 < 4.0$ , $r_{\max} = 16$ cm); the dry/wet limits ( $20$ and $15.8$ m/s, $21\%$ ); the grip-circle numbers ( $a_t \le 3.84\ \mathrm{m/s^2}$ from $\sqrt{25 - 10.24}$ ); the lean angle ( $26.6°$ at $v = 10$ , $r = 20$ ).

Prerequisites. Centripetal Force: Dynamics of Circular Motion · Static Friction · Circular Motion: Centripetal & Tangential Acceleration. Leads to. Banked Curves · Motion in a Vertical Circle · Rotor & Well of Death · Centripetal Force Dynamics (UG deep dive).

If friction from the road turns the car, what is the steering wheel actually doing?

Re-aiming the tyres. A tyre grips hardest along its rolling plane; turn the front wheels and the road's grip acquires a sideways component relative to the car's motion — that component is the centripetal supply. The engine drives the car forward, the steering chooses the direction of the grip, and friction does the turning. No steering input, no sideways friction, no turn: on ice all three statements collapse together.

Why does the mass really cancel? It feels wrong that a truck corners as fast as a car.

Both sides of the audit scale with $m$ : the invoice $mv^2/r$ doubles with a doubled load, but so does the weight pressing tyre to road, and hence the ceiling $\mu_s mg$ . In the idealized Coulomb model the limit speed is mass-free. Reality adds second-order effects — real tyre $\mu$ decreases slightly under heavier load, and a tall truck may topple before it skids (a torque problem, met in rotation) — which is why actual trucks take bends slower. For every exam at this level: mass cancels.

Is it the same friction that stops the car when braking? How do the two jobs share?

Yes — one contact patch, one ceiling, two demands. Cornering wants sideways force $mv^2/r$ ; braking wants backward force $ma_t$ . They add as perpendicular vectors, so the constraint is $\sqrt{(v^2/r)^2 + a_t^2} \le \mu_s g$ — the traction circle. Spend the whole budget cornering and there is nothing left for braking: touch the brakes at the cornering limit and the total demand leaves the circle — the car skids. Racing drivers "trail-brake" precisely along the circle's rim.

Why doesn't leaning the motorcycle reduce the friction it needs?

Because the lean solves a different problem. The friction needed is set by the turn — $f = mv^2/r$ , lean or no lean. What leaning fixes is torque: on two wheels, the ground's total push must aim through the centre of mass or the bike rotates and falls. The lean angle $\tan\theta = v^2/gr$ is the unique tilt at which the push (normal + friction combined) lines up with the body. Grip is an audit of forces; balance is an audit of torques; a safe corner passes both.

What happens physically at the moment a skid begins?

The contact patch, which had been momentarily at rest (static regime), starts sliding. Friction instantly switches personnel: from "up to $\mu_s N$ , self-adjusting" to "exactly $\mu_k N$ , opposing the slide" — and $\mu_k < \mu_s$ , so the inward supply drops just when it was already insufficient. The path straightens; steering input mostly stops mattering because a sliding patch responds weakly to wheel angle. That discontinuous loss is why the limit must be respected rather than probed, and why ESC intervenes before the ceiling.

On the turntable, why does the coin near the edge fly off first when the whole table turns together?

Everything on the table shares the same $\omega$ , but the required centripetal acceleration is $\omega^2 r$ — proportional to radius. The outer coin needs more force from the same friction ceiling $\mu_s g$ ; past $r_{\max} = \mu_s g/\omega^2$ the ceiling cannot pay. Contrast the car at fixed speed $v$ , where larger radius means a gentler demand $v^2/r$ . Fixed- $\omega$ systems punish the outside; fixed- $v$ systems punish the inside. Identify which is fixed before anything else.

Do banked roads make this whole analysis obsolete?

They upgrade it. Banking tilts the normal force so that part of $N$ points inward — a second payer joins friction, and at the "design speed" the geometry pays the entire invoice with friction idle. The flat road is the $\theta = 0$ limit of the banked-curve formulas, and everything here (the audit logic, the grip circle, the skid signature) carries over. That richer machine — design speed, $v_{\max}$ and $v_{\min}$ bands — is the next archetype, Banked Curves.