Kinetic & Rolling Friction

On a dry road, a car braking hard with its wheels locked — tyres screeching, rubber smearing — takes longer to stop than the same car braking smoothly with its wheels rolling. That sounds backwards: surely the dramatic skid is the harder stop? The skid marks say otherwise, and the reason is one inequality from the Static Friction module: $\mu_k < \mu_s$ . The moment a surface starts to slide, friction abandons its self-adjusting behaviour, drops to a fixed lower value, and stays there. Every anti-lock braking system on Earth is engineered around that drop. This module is friction's two sliding personalities. Kinetic friction is the honest one: a constant force $\mu_k N$ , aimed against the relative slipping of the surfaces — a direction rule subtle enough to push a parcel forward on a conveyor belt. Rolling friction is the quiet one: the tiny resistance a wheel feels even when nothing slides, born from the squash-and-recover of the contact patch, ten to a hundred times weaker than sliding — why the wheel, not fire, may be humanity's best invention. CBSE and ISC want the kinetic-friction laws, the friction-versus-push graph, and rolling friction in words. JEE turns the direction rule and the $\mu_k < \mu_s$ drop into traps; NEET loves the mass-free stopping-distance formula; AP, IB and A-Level test the static-versus-kinetic decision inside dynamics problems. The "will it move?" question belongs to the Static Friction module — here, everything is already moving.

Key takeaways

Once surfaces slide, friction is an equality: $f_k = \mu_k N$ — fixed magnitude, no self-adjustment, nearly independent of speed and apparent area.
Direction: kinetic friction opposes the relative sliding ( $-\hat{v}_{\text{rel}}$ ), not necessarily the body's motion — a parcel dropped on a moving belt is pushed forward.
$\mu_k < \mu_s$ always: keeping a thing moving beats starting it. That drop is the breakaway lurch and the reason ABS out-brakes a skid.
A skidding block decelerates at $a = \mu_k g$ (mass cancels) and stops in $d = \dfrac{v^2}{2\mu_k g}$ — double the speed, quadruple the skid.
Rolling resistance $f_r = \mu_r N$ comes from contact-patch deformation: $\mu_r \sim 0.01$ – $0.02$ for tyres, $\sim 0.001$ for train wheels — far below sliding values.
Key idea: sliding friction is a fixed bill, $f_k = \mu_k N$ , charged against the relative slip — smaller than the static ceiling, and tiny again once the object rolls instead of slides.

The core insight

During sliding, friction has the definite magnitude $f_k = \mu_k N$ directed opposite the relative velocity of the surfaces, with $\mu_k < \mu_s$ ; a rolling contact replaces this with a much smaller rolling resistance $f_r = \mu_r N$ caused by deformation at the contact patch.

Where this lives today

Rolling resistance is now a headline number in vehicle engineering. In an electric car, tyres take roughly a fifth of the battery's energy, so manufacturers certify low-rolling-resistance tyres (EU labels grade them A to E) and every notch shows up as range. The extreme case is rail: steel-on-steel's $\mu_r \sim 0.001$ lets one locomotive haul what hundreds of trucks would — continental freight economics rest on equation (5). Meanwhile the $\mu_k < \mu_s$ drop is patrolled by software: anti-lock braking senses incipient wheel lock and pulses the brakes tens of times a second, holding each tyre at the top of the static budget instead of the kinetic plateau — Example 1's skid, prevented in real time. At the frontier, tribologists chase the estimated ~20% of world energy output that friction consumes: atomic-force microscopes watch junctions form and shear one atom at a time, and "superlubric" contacts with nearly zero kinetic friction are moving from curiosity toward bearing engineering. The flat line on this module's first graph hides, at the nanoscale, some of the liveliest physics in materials science.

The physical picture

Drag a heavy crate across a floor with a spring balance and watch the needle. While the crate is stuck, the reading climbs — static friction matching you newton for newton, as the Static Friction module showed. At the breakaway the crate lurches forward and the needle falls back, settling at a steady value below the peak. From here on, push harder and the crate accelerates; push gentler and it decelerates — but the friction force itself stays parked at one fixed value, $\mu_k N$ , however fast the crate slides. The negotiation is over: friction has switched from a self-adjusting constraint to a flat fee.

Figure

The full friction graph — the static ramp, the breakaway drop, and the flat kinetic regime

Why the drop? Zoom into the interface. When surfaces sit still, their microscopic peaks — asperities — have time to settle, flatten, and weld into tiny junctions; the longer they sit, the stronger the joint. Sliding never grants that time: each junction is torn apart within microseconds of forming, always weaker than a matured one, so the average resistance during sliding sits below the static ceiling — $\mu_k < \mu_s$ , universally. The same picture explains sliding friction's two strangest habits: it barely cares how fast you slide, and it ignores the apparent contact area (the true welded area tracks the load $N$ , not the footprint — the Static Friction module's asperity bookkeeping, still in force). And rolling? A wheel that rolls without slipping never shears junctions — its contact point touches down and lifts off like a footstep, with zero relative sliding. What remains is gentler: wheel and road squash slightly at the contact patch, the front of the patch compressing while the rear springs back, and compression costs a little more than recovery repays. The wheel continuously leaks energy — but at a rate ten to a hundred times below sliding friction. That factor is the entire engineering case for the wheel.

Figure

Sliding shears junctions — rolling only squashes and releases the contact patch

The kinetic friction law

The experimental facts compress into one equation plus one direction rule. Magnitude first: during sliding, friction is proportional to how hard the surfaces are pressed together, and to nothing else that matters in an exam:

f_k = \mu_k N \qquad (1)

where: $f_k$ — kinetic (sliding) friction force; $\mu_k$ — the coefficient of kinetic friction, a dimensionless number for the material pair; $N$ — the normal force. Note what is absent: no speed, no area, no mass on its own. And this is an equality — contrast the sibling module's inequality $f_s \le \mu_s N$ . You never solve for kinetic friction from equilibrium; you compute it directly, always with $\mu_k$ , whose value sits below $\mu_s$ for every dry pair (rubber on dry concrete: $\mu_s \approx 0.9$ , $\mu_k \approx 0.7$ ; waxed ski on snow: $\mu_s \approx 0.1$ , $\mu_k \approx 0.05$ ). Direction is the half of the law that separates top scorers. Kinetic friction acts on each surface opposite to that surface's velocity relative to the other:

\vec{f}_k = -\mu_k N \,\hat{v}_{\text{rel}} \qquad (2)

where: $\hat{v}_{\text{rel}}$ — the unit vector of the body's velocity relative to the surface it slides on. Usually the floor is at rest and "opposite relative sliding" collapses to "opposite the motion" — which is why the slogan survives. But drop a parcel on a conveyor belt moving forward: relative to the belt the parcel slides backward, so friction on the parcel points forward, accelerating it up to belt speed — friction as engine, not brake. The Conveyor Belts archetype drills this setup; here, own the rule. While the parcel is dragged up to speed, mechanical energy is being lost: sliding friction is the canonical energy sink, and the sheared junctions show up as heat — why brake discs glow and a hard skid smells. On level ground the bookkeeping is simple, and it produces this module's most-tested formula.

Derivation ASkidding to a stop — deceleration and stopping distance

A block (or a car with locked wheels) slides on level ground at speed $v_0$ , and kinetic friction alone brings it to rest; we want the deceleration and the stopping distance — the skid-mark formula accident investigators run backwards to estimate a driver's speed. The route has three moves: the vertical balance fixes $N$ , equation (1) converts $N$ into a force, and the time-free kinematic equation $v^2 = v_0^2 + 2as$ turns constant deceleration into distance. Watch for two cancellations: the mass drops out, and the speed enters squared.

Step 1. Vertical balance. No vertical acceleration, so $N = mg$ . The friction magnitude is now fixed: $f_k = \mu_k mg$ . Why: on level ground with no other vertical pushes, the floor must carry exactly the weight — this is the step that changes on an incline or when someone presses down on the block, so make it consciously every time.

Step 2. Horizontal second law. Friction is the only horizontal force and it opposes the sliding:

ma = -\mu_k m g \quad\Rightarrow\quad a = -\mu_k g \qquad (3)

Why: the mass cancels because friction itself is proportional to weight — heavier objects get proportionally more grip and need proportionally more stopping force. Every skidding object on a given surface decelerates identically.

Step 3. Time-free kinematics. With $v = 0$ at the end, $0 = v_0^2 - 2\mu_k g\, d$ , so

d = \frac{v_0^2}{2\mu_k g} \qquad (4)

Why: $v^2 = v_0^2 + 2as$ because the question asks for distance, not time; the time version $t = v_0/\mu_k g$ is a one-line bonus. *Verified: energy check — heat $f_k d = \mu_k mg \cdot \dfrac{v_0^2}{2\mu_k g} = \tfrac{1}{2}mv_0^2$ , exactly the kinetic energy destroyed. Limits: $\mu_k \to 0$ gives $d \to \infty$ ; doubling $v_0$ quadruples $d$ .*

The $v_0^2$ in equation (4) is the whole science of following distances. At $50\ \mathrm{km/h}$ on a dry road ( $\mu_k = 0.7$ , $g = 10$ ), a locked-wheel skid runs $(13.9)^2/14 \approx 14\ \mathrm{m}$ ; at $100\ \mathrm{km/h}$ — twice the speed — $55\ \mathrm{m}$ , four times as far. Exams convert this into ratio questions because the calculation takes five seconds if you see the scaling. Measuring $\mu_k$ uses the equality itself: drag a block at constant velocity with a horizontal spring balance — zero acceleration means the pull exactly matches friction, so $\mu_k = F_{\text{pull}}/mg$ . (For $\mu_s$ you tilt instead — the protractor method of the Angle of Friction & Angle of Repose module.)

situation	law	direction
body sliding on a fixed surface	$f_k = \mu_k N$	opposite the body's velocity
body sliding on a moving surface (belt, truck bed)	$f_k = \mu_k N$	opposite the relative velocity — may point forward
body at rest or moving with the surface	static law — see Static Friction	whatever the second law demands
wheel rolling without slipping	$f_r = \mu_r N$ (+ static friction as needed)	resists the rolling

Rolling friction: the quiet resistance

A wheel rolling without slipping has no relative sliding at its contact point — so where does its resistance come from? Not from shearing junctions but from deformation. The tyre flattens against the road; as the wheel advances, fresh material at the front of the contact patch is squeezed while the rear recovers, and materials return slightly less energy on recovery than they absorb in compression (the loss is called hysteresis). The effect behaves, to a good approximation, like a small backward force proportional to the load:

f_r = \mu_r N \qquad (5)

where: $f_r$ — rolling friction (rolling resistance); $\mu_r$ — the coefficient of rolling friction: typically $0.01$ – $0.02$ for car tyres on tarmac, $0.002$ – $0.005$ for a firm bicycle tyre, $\sim 0.001$ for steel train wheels on steel rail. Against sliding coefficients of $0.3$ – $0.9$ , rolling undercuts sliding by a factor of ten to several hundred — why a locomotive hauls thousands of tonnes, and why you pump your tyres: a hard tyre deforms less, wastes less, rolls further.

Go deeper · Derivation B · 1 minWhere the rolling resistance formula comes from

Derivation BWhere the rolling resistance formula comes from

We want to see why rolling resistance takes the same $\mu N$ form as sliding friction despite a completely different mechanism. The strategy: locate the shifted normal force on a deformed rolling wheel, compute the torque with which it fights the rotation, and convert that torque into an equivalent backward force. Boards want the idea and the formula, not elasticity theory — and the payoff is seeing why $\mu_r$ is sometimes quoted as a length.

Step 1. Picture the deformed contact. A loaded wheel of radius $r$ flattens over a small patch. The front of the patch is being compressed while the rear recovers (and recovery repays less), so the ground's push is stronger toward the front: the resultant normal force $N$ acts a small distance $e$ ahead of the axle line.

Step 2. That offset creates a torque: about the axle, the shifted normal force exerts $\tau = N e$ opposing the wheel's rotation. Why: a force through the axle line produces no torque; the entire resistance to rolling lives in that few-millimetre forward shift.

Step 3. Convert to an equivalent force. Keeping the wheel rolling at steady speed requires a drive torque $Ne$ , which at the rim is equivalent to overcoming a backward force $f_r = \dfrac{Ne}{r}$ . Defining $\mu_r = e/r$ recovers equation (5): $f_r = \mu_r N$ . Why: some tables quote rolling resistance as the length e itself, others as the dimensionless e/r — same physics, two conventions. *Verified: a stiffer tyre deforms less, shrinking $e$ and hence $\mu_r$ (pumping tyres works); a larger radius lowers $\mu_r = e/r$ (why carts used big wheels); $e \to 0$ (rigid bodies) gives zero rolling resistance.*

One distinction protects you from a subtle exam wound. Rolling resistance is not the friction that makes rolling possible. When a ball rolls down a ramp without slipping, static friction enforces the no-slip condition — it can be large, does no dissipative work, and stars in the Rolling With & Without Slipping module in Rotation. Rolling resistance is the separate, small, energy-eating deformation effect of equation (5). An ideal rigid wheel on a rigid road has static friction but zero rolling resistance. Separate mental boxes.

The third family: resistance that grows with speed

Kinetic and rolling friction share one defining trait: their magnitudes ignore how fast you move. Nature runs a third family that does not — the fluid resistances. Air drag on a cyclist, water drag on a swimmer, the syrup inside a shock absorber: all grow with speed, roughly as $v$ at low speeds in thick fluids and as $v^2$ through air. That is why drag, unlike dry friction, enforces a terminal velocity — resistance grows until it balances the weight, something a constant $\mu_k N$ could never do. Own the contrast: a skid decelerates at a constant $\mu_k g$ whatever the speed; a drag-dominated fall decelerates ever less as it slows. The formulas live in the Viscosity, Stokes' Law & Terminal Velocity module.

Limiting cases

case	result	meaning
$\mu_k \to 0$	$a \to 0$ , $d \to \infty$	ice-rink physics: once sliding, nearly nothing stops you
$v_{\text{rel}} \to 0$ during sliding	law switches back to static	the parcel reaching belt speed stops feeling kinetic friction
rigid wheel, rigid road ( $e \to 0$ )	$f_r \to 0$	rolling resistance is purely a deformation effect
speed doubled in a skid	$d$ quadruples, stopping time doubles	$d \propto v_0^2$ from (4), $t \propto v_0$
load doubled	$f_k$ , $f_r$ double; $a = \mu_k g$ unchanged	friction scales with $N$ ; deceleration stays mass-free

Worked examples

1 · Board level — skid marks as a speedometer (CBSE/ISC pattern, 3 marks)

A car's wheels lock and it skids $25\ \mathrm{m}$ on a level road before stopping ( $\mu_k = 0.45$ , $g = 10\ \mathrm{m/s^2}$ ). How fast was it going when the wheels locked?

Step 1. Deceleration from equation (3): $a = \mu_k g = 0.45 \times 10 = 4.5\ \mathrm{m/s^2}$ — the car's mass is irrelevant.

Step 2. Invert formula (4): $v_0 = \sqrt{2\mu_k g\, d} = \sqrt{2 \times 4.5 \times 25} = \sqrt{225} = \boxed{15\ \mathrm{m/s}}$ — that is $54\ \mathrm{km/h}$ .

Step 3. Check forwards: $d = 225/(2 \times 4.5) = 25\ \mathrm{m}$ ✓. This inversion — marks to speed — is precisely what crash investigators compute.

2 · Board level — measuring μₖ with a spring balance (ISC/A-Level practical pattern)

A $20\ \mathrm{kg}$ sled is dragged across level snow at constant velocity by a horizontal $70\ \mathrm{N}$ force ( $g = 9.8\ \mathrm{m/s^2}$ ). Find $\mu_k$ . The pull then rises to $90\ \mathrm{N}$ — find the acceleration.

Step 1. Constant velocity means zero net force, so friction exactly balances the pull: $f_k = 70\ \mathrm{N}$ .

Step 2. Normal force: $N = mg = 20 \times 9.8 = 196\ \mathrm{N}$ . From equation (1): $\mu_k = \dfrac{f_k}{N} = \dfrac{70}{196} = \boxed{0.36}$ .

Step 3. Harder pull: friction stays at $\mu_k N = 70\ \mathrm{N}$ — the flat plateau of the graph; it does not rise to match. So $a = \dfrac{90 - 70}{20} = \boxed{1.0\ \mathrm{m/s^2}}$ .

Step 4. Contrast with the static regime: below breakaway, friction would have matched the extra pull and $a$ would stay zero. Once sliding, every extra newton of pull is pure acceleration.

3 · Competitive level — JEE-Main pattern, two surfaces (~2.5 minutes)

A block slides at $8.0\ \mathrm{m/s}$ onto $10\ \mathrm{m}$ of ice ( $\mu_k = 0.05$ ), then onto rough concrete ( $\mu_k = 0.50$ ); $g = 10\ \mathrm{m/s^2}$ . How far into the concrete does it travel?

Step 1. On the ice, $a_1 = \mu_{k1} g = 0.5\ \mathrm{m/s^2}$ , so at the end of the ice $v^2 = 64 - 2 \times 0.5 \times 10 = 54\ \mathrm{m^2/s^2}$ .

Step 2. Do not square-root — the next step wants $v^2$ anyway. On concrete, $a_2 = \mu_{k2} g = 5.0\ \mathrm{m/s^2}$ , and the block stops when $0 = 54 - 2 \times 5.0 \times d$ .

Step 3. Solve: $d = \dfrac{54}{10} = \boxed{5.4\ \mathrm{m}}$ .

Step 4. Read it: ten metres of ice barely taxed the block ( $\sqrt{54} \approx 7.35\ \mathrm{m/s}$ of its $8$ survived); the concrete devoured the rest in half that distance. Timing note: budget ~2.5 min. The efficiency move is carrying $v^2$ between surfaces without square-rooting — segment problems chain cleanly through $v^2$ . The planted distractor uses one $\mu$ for the whole journey.

4 · Competitive level — NEET pattern, the v² scaling (~1.5 minutes)

A car skids to rest in $20\ \mathrm{m}$ from $40\ \mathrm{km/h}$ on a certain road. On the same road, its skid from $120\ \mathrm{km/h}$ is…

Step 1. From equation (4), $d \propto v_0^2$ at fixed $\mu_k$ : no need for $\mu_k$ , masses, or unit conversions — only the ratio.

Step 2. Speed ratio $= 120/40 = 3$ , so distance ratio $= 3^2 = 9$ : $d = 9 \times 20 = \boxed{180\ \mathrm{m}}$ .

Step 3. The planted wrong answers: $60\ \mathrm{m}$ (linear thinking) and $360\ \mathrm{m}$ . Nearly two hundred metres from highway speed is the truth that following-distance rules encode. Timing note: ~1.5 min, under 30 seconds if the $v^2$ scaling is reflexive. Any stopping-distance MCQ that changes only the speed is a ratio question in disguise — never recompute from scratch.

Exam traps

Using μₛ in a sliding problem (or μₖ at rest). The single biggest friction error. Sliding ⇒ $\mu_k$ , always; at rest or rolling without slipping ⇒ the static law. A problem quoting both coefficients is testing whether you can pick — the wrong-coefficient answer will be among the options.

Letting kinetic friction "adjust". Only static friction self-adjusts. Once sliding, $f_k = \mu_k N$ regardless of the push (Example 2: pull rose from $70$ to $90\ \mathrm{N}$ , friction stayed at $70$ ).

Friction arrows auto-drawn "backwards". Direction follows the relative sliding (equation 2). Belt, truck-bed and stacked-block problems put friction along the body's motion. Ask "sliding relative to what?" before drawing.

N is not always mg. Press down on a sliding block and $N$ (hence $f_k$ ) grows; pull the rope at an upward angle and $N$ shrinks. Recompute $N$ from the vertical balance every time — equation (1) is only as good as the $N$ you feed it.

Rolling resistance vs the static friction of rolling. Rolling resistance ( $\mu_r N$ , tiny, dissipative) slows a freewheeling wheel; static friction at a rolling contact (can be large, zero dissipative work) accelerates cars and enforces no-slip. Different forces, different modules.

Skid heat "disappearing". Energy questions want the equality $f_k d = \tfrac{1}{2}mv_0^2$ — kinetic energy becomes heat at the interface. Writing "energy is lost" without it drops marks.

Which g? CBSE/ISC use $g = 9.8\ \mathrm{m/s^2}$ ; JEE/NEET usually state $g = 10$ (use what the paper says); A-Level uses $9.81$ . In ratio questions (Example 5), $g$ cancels — do not substitute it.

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

Charles-Augustin de Coulomb

1736–1806 · Paris

Coulomb came to friction as a military engineer with a practical brief: the Académie des Sciences had offered a prize for the laws governing ropes, pulleys and the launching of ships. Working with full-scale naval hardware — greased oak ways, hemp cables, loads of tonnes rather than grams — he won the 1781 prize with experiments of unprecedented care, published as Théorie des machines simples (1785). Coulomb cleanly separated the static ceiling from the sliding value, established $\mu_k < \mu_s$ , and showed kinetic friction nearly independent of speed. The dry-friction law is still called Coulomb friction — and the same man, a few years later with a torsion balance, gave his name to the unit of electric charge. Two chapters of physics, one signature.

History

1493

— Leonardo da Vinci's notebooks record friction proportional to load and independent of contact area — unpublished for four centuries.

1699

— Guillaume Amontons publishes the dry-friction laws; the Académie doubts the strange area-independence.

1785

— Coulomb's prize memoir separates static from kinetic friction, establishing μₖ < μₛ and speed-independence.

1888

— John Boyd Dunlop patents the pneumatic tyre, slashing rolling resistance on rough roads and rewriting transport.

1950

— Bowden and Tabor explain the laws microscopically: true contact at asperities grows with load; sliding junctions never mature to static strength.

Going further

Texts. NCERT Physics Class 11, Ch. 5 (Laws of Motion — friction) · HC Verma, Concepts of Physics Vol. 1, Ch. 6 · Halliday, Resnick & Walker, Ch. 6 · Bowden & Tabor, Friction: An Introduction to Tribology.

Provenance. Coulomb friction laws presented at HS depth; asperity explanation follows Bowden–Tabor. A verifier should recompute: Example 1 ( $v_0 = \sqrt{2 \times 0.45 \times 10 \times 25} = 15\ \mathrm{m/s}$ ); Example 2 ( $N = 196\ \mathrm{N}$ , $\mu_k = 70/196 = 0.357 \approx 0.36$ , $a = 20/20 = 1.0\ \mathrm{m/s^2}$ ); Example 3 ( $v^2 = 64 - 10 = 54$ , $d = 54/10 = 5.4\ \mathrm{m}$ , $\sqrt{54} \approx 7.35$ ); Example 4 ( $3^2 \times 20 = 180\ \mathrm{m}$ ); the skid comparison ( $13.9^2/14 \approx 13.8\ \mathrm{m}$ ; $27.8^2/14 \approx 55\ \mathrm{m}$ ); the cyclist anchor ( $f_r = 0.005 \times 90 \times 10 = 4.5\ \mathrm{N}$ , $P = 4.5 \times 6 = 27\ \mathrm{W}$ ); energy identity $f_k d = \tfrac{1}{2}mv_0^2$ . Coefficient values ( $\mu_r$ : tyres $0.01$ – $0.02$ , bicycle $0.002$ – $0.005$ , rail $\sim 0.001$ ; rubber/concrete $\mu_s \approx 0.9$ , $\mu_k \approx 0.7$ ) are standard order-of-magnitude figures, not precise constants.

Prerequisites. Static Friction. Leads to. Angle of Friction & Angle of Repose · Conveyor Belts · Stacked Blocks & Friction Between — and, for the deep dive, the university-tier treatment in Friction Forces (Advanced Mechanics).

Why is μₖ always less than μₛ — where does the drop actually come from?

Time. While surfaces rest against each other, their microscopic contact junctions creep, flatten and bond — the interface literally strengthens as it sits. Sliding never allows that: each junction exists for microseconds and is sheared while still immature, so the average resistance is lower — typically by 10–30% for dry pairs. The drop is also why a squeaking door and a violin string work at all: the stick-slip cycle needs sticking to be stronger than slipping.

Does kinetic friction really not depend on speed at all?

To exam accuracy, yes — Coulomb tested this and modern measurements agree that dry sliding friction changes only weakly across everyday speeds. That is what makes $a = \mu_k g$ and $d = v_0^2/2\mu_k g$ legal. The independence fails at extremes: interface heating can melt a lubricating layer (why skis glide), and lubricated or fluid-mediated contacts belong to the drag family, where force genuinely grows with $v$ . If an exam wants speed-dependence, it will say "viscous" or "drag" — dry friction is constant.

How can friction push something forward? Doesn't that contradict its whole job?

Friction's job is to oppose relative sliding between the touching surfaces — not motion through space. The parcel on the belt slides backward relative to the belt, so friction on it acts forward, exactly as equation (2) commands. The same holds when you walk (shoe pushes back on ground, ground pushes you forward) and when a car accelerates. Check the relative velocity at the contact and aim friction against it — the rule never fails, the slogan often does.

Is rolling friction just a smaller version of kinetic friction?

No — different mechanism entirely. Kinetic friction is asperity junctions being sheared because the surfaces slide; a wheel rolling without slipping shears nothing, since its contact point is momentarily at rest on the road. Rolling resistance comes from deformation: tyre and road squash at the front of the patch and rebound at the rear, and the rebound repays less than the squash cost. That is why $\mu_r$ tracks stiffness (pump the tyres, drop the bill) while $\mu_k$ tracks surface chemistry, and why the two differ by up to a factor of a hundred.

Where does the kinetic energy go when something skids to a stop?

Into heat at the interface, with scraps to sound and wear. The bookkeeping is exact: friction force times sliding distance, $f_k d$ , equals the kinetic energy destroyed, $\tfrac{1}{2}mv_0^2$ — the Verified check of the stopping-distance derivation. Brake discs glowing and tyres smoking are that arithmetic made visible. When a question says "find the heat generated", compute $f_k \times$ (relative sliding distance) — on a belt that is the relative displacement, not the ground displacement.

When a ball rolls down a hill without slipping, which friction acts on it?

Static friction — plus a little rolling resistance. "Without slipping" means zero relative sliding at the contact, so the kinetic law never engages; static friction takes whatever value enforces rolling (that story belongs to the Rolling With & Without Slipping module). Rolling resistance nibbles at the energy, but boards treat it as negligible unless stated. The discipline: "rolls without slipping" ⇒ static friction, unknown value, solve from the equations of motion; "slides" or "skids" ⇒ kinetic friction, known value $\mu_k N$ .