Work-Energy Theorem

Every driving handbook carries the same stark table: double your speed and your braking distance quadruples. That is not a rule of thumb about tires or reaction times — it is the work-energy theorem doing arithmetic in public. A car at speed $v$ carries kinetic energy $\tfrac12 mv^2$ , the brakes can withdraw that energy only at a fixed rate per metre of road, and $v^2$ means four times the energy at twice the speed. One scalar equation, $W_{\mathrm{net}} = \Delta K$ , sets the spacing of motorway chevrons and the physics of every crash barrier. The theorem is the scalar shadow of Newton's second law. Where $\mathbf{F} = m\mathbf{a}$ is a vector statement — three components, full trajectory geometry, the whole time history — the work-energy theorem collapses the dynamics along the path into a single line of bookkeeping: the running total of work delivered to a particle equals the change in its kinetic energy. Nothing is assumed about the force. It may vary with position like a bowstring, depend on velocity like drag, or be as crooked as a mountain road; conservative or dissipative, the ledger balances exactly, in every inertial frame. That generality is the theorem's power and its trap. It answers "how fast at the end?" questions with almost insulting ease — no trajectory solving, no time variable at all — precisely when constant-acceleration kinematics is helpless. But it is a statement about a particle, it counts only the net work, and applying it carelessly to cars, people, and other deformable bodies is the seat of some of the most instructive paradoxes in mechanics. Mastering both the shortcut and its fine print is what this article is for.

Key takeaways

The work-energy theorem: the net work done on a particle equals the change in its kinetic energy, $W_{\mathrm{net}} = \Delta K = \tfrac12 mv_f^2 - \tfrac12 mv_i^2$ .
It is Newton's second law dotted with velocity and integrated along the path — no new physics, but a vector equation traded for one scalar that never asks about time.
Work is the line integral $W = \int_C \mathbf{F}\cdot\mathrm{d}\mathbf{r}$ : only the force component along the motion counts. Perpendicular forces — normal forces, string tensions in circular motion, magnetic forces — do exactly zero work.
The theorem accepts any force whatsoever: position-dependent, velocity-dependent, dissipative. That makes it more general than conservation of mechanical energy, which needs conservative forces.
Kinetic energy grows as $v^2$ , so stopping distance does too: at fixed braking force, $d = v^2/2\mu_k g$ — the energetic origin of speed limits.
Fine print: the theorem governs a point particle (or, with care, the center of mass). For deformable bodies — a sprinter, a braking car — the energy source can be internal even when an external force sets the momentum budget.
Key idea: dotting Newton's second law with velocity turns the vector equation of motion into a scalar identity — the net work accumulated along the actual path equals the change in $\tfrac12 mv^2$ , for any force, in any inertial frame.

The core insight

Dotting Newton's second law with velocity converts the vector equation of motion into a scalar identity: the net work integrated along the actual path equals the change in $\tfrac12 mv^2$ — for any force whatsoever, in any inertial frame.

Where this lives today

Every electric vehicle on the road runs the work-energy theorem in reverse, on purpose, at every red light. Regenerative braking makes the motor apply a retarding force through a displacement — negative work on the car — and routes the withdrawn kinetic energy into the battery instead of into hot brake discs. The stakes are concrete: an 1800 kg EV slowing from 100 km/h carries $\tfrac12 mv^2 \approx 690\ \mathrm{kJ}$ , about 0.19 kWh per stop; over a stop-and-go urban commute, recovering most of that is worth 20–30 percent of range. The engineering ceiling is set by eq. (4): at low speed $P = Fv$ collapses, so regeneration fades and friction brakes must finish the job — which is why your EV blends the two below walking pace. The theorem also runs the largest machines ever built. An accelerator is a device for doing work on charged particles: an electric field $\mathbf{E}$ exerts $q\mathbf{E}$ through kilometres of beamline, and the particle's kinetic energy rises by exactly the work done — which is why particle physicists measure energy in work units: the electron-volt, the work done on one elementary charge through one volt. The LHC's 6.8 TeV protons carry about $1.1\times10^{-6}$ J apiece — a mosquito's kinetic energy compressed into a single proton, deposited there one dotted line integral at a time. (At those speeds the formula for $K$ is relativistic, but the theorem's logic — net work in, kinetic energy out — survives intact.)

The physical picture

Think of kinetic energy as a bank balance of motion, denominated in joules. Every push along the direction you are moving is a deposit; every push against the motion is a withdrawal; a push exactly sideways neither deposits nor withdraws — it only steers, redirecting the velocity without changing its magnitude. The work-energy theorem is the bank statement: the net change in the balance equals the sum of all deposits and withdrawals, no matter who made them or why.

Figure

Deposits, withdrawals, steering — how a force changes or preserves kinetic energy

The factor $\tfrac12$ and the $v^2$ are not decoration. Acceleration accumulates speed with time, but the distance over which a force acts also grows with speed — so the energy deposited grows as speed squared. That is why kinetic energy is a quadratic account: a pitched baseball (0.145 kg at 40 m/s) carries about 116 J; a 1500 kg car at highway speed (30 m/s) carries 675 kJ, roughly six thousand baseballs' worth; a 8 g bullet at 400 m/s carries 640 J — less than one per cent of the car's, which is why the car is the more dangerous projectile. Because only the force component along the path matters, geometry does the bookkeeping for you. A bead sliding down a frictionless wire of any shape — straight ramp, ski-jump curve, corkscrew — reaches the bottom at the same speed, $v = \sqrt{2g\,\Delta h}$ , set purely by the height dropped. The wire's normal force, always perpendicular to the motion, never touches the account. This indifference to path shape is the theorem's most elegant gift, and it is the seed from which the concept of potential energy will grow in the articles that follow.

Figure

Two wires, one answer — perpendicular forces never touch the energy account

Finally, the theorem is path-aware but mechanism-blind. It counts the work whether the force came from a spring, gravity, friction, or a hand — which is why it keeps working for dissipative systems, where naive "energy conservation" fails. Friction simply enters the ledger as a withdrawal. What the theorem never does is ask when: time appears nowhere in $W_{\mathrm{net}} = \Delta K$ . If the question is "how fast after 8 metres?", energy answers instantly; if it is "how fast after 3 seconds?", you must go back to Newton.

The bookkeeping equations

Start with the account itself. For a particle of mass $m$ moving at speed $v$ , the kinetic energy is defined as

K = \tfrac12 mv^2 = \tfrac12 m\,(\mathbf{v}\cdot\mathbf{v}) \qquad (1)

where: $\mathbf{v}$ — the velocity vector; $v = |\mathbf{v}|$ — the speed. $K$ is a scalar, measured in joules ( $1\ \mathrm{J} = 1\ \mathrm{kg\,m^2/s^2}$ ), and it never cares which way the particle is going — a car doing 30 m/s east and one doing 30 m/s north hold identical balances. It is also frame-dependent: a passenger asleep on a train has $K = 0$ for the conductor and hundreds of kilojoules for someone on the platform. Both are right; the theorem will hold for each of them separately. Next, the transactions. When a force $\mathbf{F}$ acts on a particle that undergoes an infinitesimal displacement $\mathrm{d}\mathbf{r}$ along its path, the work done is $\mathrm{d}W = \mathbf{F}\cdot\mathrm{d}\mathbf{r}$ — force component along the motion, times distance moved. Summing the contributions along the actual trajectory $C$ gives the work as a line integral:

W = \int_C \mathbf{F}\cdot\mathrm{d}\mathbf{r} = \int_C F\cos\theta \,\,\mathrm{d}s \qquad (2)

where: $\theta$ — the local angle between the force and the direction of motion; $\mathrm{d}s$ — arc length along the path. The dot product is the whole story: $\theta < 90°$ deposits, $\theta > 90°$ withdraws, $\theta = 90°$ does nothing. For the special case of a constant force along a straight displacement $\mathbf{d}$ , the integral collapses to the school formula $W = \mathbf{F}\cdot\mathbf{d} = Fd\cos\theta$ — carry a 20 kg suitcase horizontally across an airport and gravity, forever perpendicular to your stroll, does zero work no matter how tired your arm insists it is. With both pieces defined, the theorem itself reads

W_{\mathrm{net}} = \Delta K = \tfrac12 mv_f^2 - \tfrac12 mv_i^2 \qquad (3)

where: $W_{\mathrm{net}}$ — the work done by the net force, equal (by linearity of the dot product) to the sum of the works done by each individual force along the actual path; $v_i, v_f$ — initial and final speeds. Read it in both directions: net positive work means the particle leaves faster than it arrived; a particle that returns to its initial speed — whatever looping detour it took — has had exactly zero net work done on it along the way. The theorem also comes in an instantaneous form. Differentiating $K$ in time and using Newton's second law gives the power delivered to the particle:

\frac{\mathrm{d}K}{\mathrm{d}t} = \mathbf{F}_{\mathrm{net}}\cdot\mathbf{v} = P_{\mathrm{net}} \qquad (4)

where: $P$ — power, in watts ( $1\ \mathrm{W} = 1\ \mathrm{J/s}$ ). This form explains a fact every driver knows: at high speed the same engine accelerates the car more feebly. The engine delivers roughly constant power, so the available forward force is $F = P/v$ — it shrinks exactly as fast as the speed grows, and the last 20 km/h before top speed cost more than the first hundred.

Which equation answers which question

you want	use	notes
final speed after a given push or drop	$W_{\mathrm{net}} = \Delta K$ , eq. (3)	no time variable needed — the go-to for "how fast at the end"
work by a varying force along a path	line integral, eq. (2)	reduces to $Fd\cos\theta$ only for constant force, straight path
rate of energy delivery right now	$P = \mathbf{F}\cdot\mathbf{v}$ , eq. (4)	engine ratings, athletic output, terminal-speed arguments
stopping distance at speed $v$	$d = v^2/2\mu_k g$ , from eq. (3)	mass cancels; distance scales as $v^2$
constant-force sanity check	$v_f^2 = v_i^2 + 2a\,\Delta x$	eq. (3) divided by $m/2$ — the familiar kinematics formula is the theorem in disguise

Derivation AThe work-energy theorem in one dimension by the chain rule

We first derive $W = \Delta K$ in one dimension, where the whole argument fits in three lines and the origin of the $\tfrac12 v^2$ is impossible to miss. The strategy is the chain-rule trick that converts a time derivative into a position derivative: acceleration, which knows about time, is rewritten as $v\,\mathrm{d}v/\mathrm{d}x$ , which only knows about the track. This is the standard manoeuvre for any force given as a function of position, and it is worth owning independently of the theorem itself. Watch how the factor $\tfrac12$ appears — not assumed, but generated by the integration.

Step 1. Rewrite the acceleration with the chain rule:

a = \frac{\mathrm{d}v}{\mathrm{d}t} = \frac{\mathrm{d}v}{\mathrm{d}x}\,\frac{\mathrm{d}x}{\mathrm{d}t} = v\,\frac{\mathrm{d}v}{\mathrm{d}x}

Why this move: the force may be known only as a function of position — a bowstring, a spring, gravity with height — so time must be eliminated in favor of position before anything can be integrated.

Step 2. Put this into Newton's second law and separate the variables:

F(x) = m\,v\,\frac{\mathrm{d}v}{\mathrm{d}x} \quad\Longrightarrow\quad F(x)\,\mathrm{d}x = m\,v\,\mathrm{d}v

Why it works: each side now involves only one variable — position on the left, speed on the right — so both sides can be integrated between the endpoints of the motion.

Step 3. Integrate from the start $(x_i, v_i)$ to the end $(x_f, v_f)$ :

\int_{x_i}^{x_f} F(x)\,\mathrm{d}x = m\int_{v_i}^{v_f} v\,\mathrm{d}v = \tfrac12 mv_f^2 - \tfrac12 mv_i^2

where: the left side is the 1D work integral and the right side is $\Delta K$ . Why the one-half: it is the antiderivative of $v$ — the factor is manufactured by the integration, which is precisely why kinetic energy had to be defined with it for the ledger to balance. *As a check, take $F$ constant: the result reads $Fd = \tfrac12 mv_f^2 - \tfrac12 mv_i^2$ , and dividing by $m/2$ gives $v_f^2 = v_i^2 + 2a\,d$ — the familiar constant-acceleration kinematics formula, which the theorem contains as a special case.*

Go deeper · Derivation B · 2 minThe general theorem in three dimensions

Derivation BThe general theorem in three dimensions

Now the full result, with no restriction to one dimension, straight paths, or any particular kind of force. The strategy is the single most fruitful move in mechanics: dot Newton's second law with the velocity. The dot product automatically discards the perpendicular (steering) part of the force and keeps the part that changes the speed, and what remains on the acceleration side turns out to be a perfect time derivative — of exactly $\tfrac12 mv^2$ . Integrating then telescopes. Nothing is assumed about $\mathbf{F}$ : it may depend on position, velocity, or time, and it may be dissipative; the result is an exact identity.

Step 1. Take the scalar product of $\mathbf{F}_{\mathrm{net}} = m\,\mathrm{d}\mathbf{v}/\mathrm{d}t$ with the velocity $\mathbf{v}$ :

\mathbf{F}_{\mathrm{net}}\cdot\mathbf{v} = m\,\frac{\mathrm{d}\mathbf{v}}{\mathrm{d}t}\cdot\mathbf{v}

Why dot with velocity: the projection onto $\mathbf{v}$ is exactly the part of the force that changes speed; whatever is perpendicular drops out of the equation here, once and for all — this is the algebraic reason steering forces never appear in the energy account.

Step 2. Recognize the right side as a perfect derivative:

m\,\frac{\mathrm{d}\mathbf{v}}{\mathrm{d}t}\cdot\mathbf{v} = \frac{\mathrm{d}}{\mathrm{d}t}\left(\tfrac12 m\,\mathbf{v}\cdot\mathbf{v}\right) = \frac{\mathrm{d}K}{\mathrm{d}t}

Why: the product rule gives $\mathrm{d}(\mathbf{v}\cdot\mathbf{v})/\mathrm{d}t = 2\,\mathbf{v}\cdot\dot{\mathbf{v}}$ — the crux step of the whole subject, and the reason the quantity $\tfrac12 mv^2$ , and no other power of $v$ , earns a name.

Step 3. Integrate over time from $t_i$ to $t_f$ ; the right side telescopes by the fundamental theorem of calculus:

\int_{t_i}^{t_f} \mathbf{F}_{\mathrm{net}}\cdot\mathbf{v}\,\,\mathrm{d}t = K_f - K_i

Step 4. Convert the left side into a line integral over the trajectory using $\mathbf{v}\,\mathrm{d}t = \mathrm{d}\mathbf{r}$ :

\int_C \mathbf{F}_{\mathrm{net}}\cdot\mathrm{d}\mathbf{r} = \tfrac12 mv_f^2 - \tfrac12 mv_i^2

where: $C$ is the actual path the particle follows — the theorem evaluates work along the real trajectory, not along some hypothetical straight line between the endpoints. Why this matters: for a non-conservative force like friction the value of the integral genuinely depends on the route taken, and the theorem is still exact — it simply charges the account for the route actually driven. *Notice that every step here is an algebraic identity or the fundamental theorem of calculus: no property of $\mathbf{F}$ was used, which is exactly why the theorem holds for any net force whatsoever. Restricting the motion to one dimension recovers the chain-rule derivation above.*

There is a geometric way to see equation (3) that is worth a picture: in one dimension, the work is the area under the force-position graph, and the theorem says that area equals the change in kinetic energy — however wildly the force varies. This is how energy methods dispose of variable-force problems that constant-acceleration kinematics cannot touch.

Figure

Work is the area under the force-position curve — and area is all the theorem needs

One more reading of equation (3) deserves its own paragraph before the worked problems: the theorem is a statement about the net force. Individual forces may shuffle enormous works among themselves — a crane lowering a container has gravity depositing megajoules while the cable withdraws almost exactly as many — and the account registers only the difference. When a problem hands you several forces, you may either add the force vectors first and compute one work, or compute each force's work along the path and add the numbers; linearity of the dot product guarantees the same answer.

Go deeper · Derivation C · 2 minRound trip on a rough incline — where energy methods beat kinematics

Derivation CRound trip on a rough incline — where energy methods beat kinematics

An exam-grade application, done symbolically: a block is launched up a rough incline (angle $\theta$ , kinetic friction coefficient $\mu_k$ ) with speed $v_0$ ; find how far up the slope it travels and how fast it returns to the launch point. The strategy is to apply the theorem separately to the uphill and downhill legs, because friction reverses direction with the motion — its work is negative on both legs, which is the signature of a dissipative force and the detail that makes a naive "energy is conserved" approach fail. The result exposes an asymmetry that is easy to check against limits.

Step 1. Uphill leg, distance $d$ along the slope: gravity's component along the path is $mg\sin\theta$ backwards, and friction $\mu_k N = \mu_k mg\cos\theta$ also acts backwards (motion is uphill, friction points downhill). The theorem, from speed $v_0$ to the momentary stop, reads

-\,mg(\sin\theta + \mu_k\cos\theta)\,d = 0 - \tfrac12 mv_0^2

Step 2. Solve for the travel distance:

d = \frac{v_0^2}{2g\,(\sin\theta + \mu_k\cos\theta)}

Why the two terms add: on the way up, gravity and friction are allies — both withdraw from the account.

Step 3. Downhill leg over the same distance $d$ : gravity now deposits $+mg\sin\theta\,d$ while friction, reversed with the motion, still withdraws $\mu_k mg\cos\theta\,d$ . From rest to return speed $v_r$ :

\tfrac12 mv_r^2 = mg(\sin\theta - \mu_k\cos\theta)\,d

Why the sign flipped only for gravity: friction always opposes the current motion, so its work is negative on both legs — the hallmark of dissipation and the reason the trip is not reversible.

Step 4. Insert $d$ from Step 2 and cancel:

v_r = v_0\,\sqrt{\frac{\sin\theta - \mu_k\cos\theta}{\sin\theta + \mu_k\cos\theta}}

where: the ratio under the root is less than one whenever $\mu_k > 0$ — the block always comes back slower. Why the block might not come back at all: if $\tan\theta \le \mu_s$ the block simply stays parked at the top of its run; the formula presumes the slope is steep enough to overcome static friction.

Step 5. Check the limits. Frictionless ( $\mu_k = 0$ ): $v_r = v_0$ , the round trip is free, as energy conservation demands. Numbers: $\theta = 37°$ , $\mu_k = 0.25$ , $v_0 = 12\ \mathrm{m/s}$ give $d = \dfrac{144}{2 \times 9.81 \times (0.602 + 0.25\times 0.799)} = 9.16\ \mathrm{m}$ and $v_r = 12\sqrt{0.402/0.801} = 8.50\ \mathrm{m/s}$ . Two sanity checks: the mass cancels throughout, as it must when every force scales with $m$ ; and in the frictionless limit the distance reduces to $d = v_0^2/(2g\sin\theta)$ , so the round trip comes for free.

Limiting cases

Case	Result	Meaning
force perpendicular to velocity	$\mathbf{F}\cdot\mathbf{v} = 0 \Rightarrow \mathrm{d}K/\mathrm{d}t = 0$	normal forces, magnetic forces $q\,\mathbf{v}\times\mathbf{B}$ , and the tension in circular whirling steer without ever changing speed
constant force, straight path	$W = Fd\cos\theta$	the line integral collapses to the school formula; dividing eq. (3) by $m/2$ recovers $v_f^2 = v_i^2 + 2a\,\Delta x$
conservative force	$W = -\Delta U \Rightarrow \Delta K + \Delta U = 0$	when every working force has a potential, the theorem upgrades to conservation of mechanical energy
kinetic friction, sliding over path length $s$ against a surface at rest in the chosen frame	$W_{\mathrm{fric}} = -\mu_k N s < 0$	dissipation enters as a withdrawal on every leg of such a trip, path-dependent by design; against a moving surface (a conveyor belt, a block on a block) friction can do positive work on the body, and only the sum over both surfaces is necessarily negative
zero net work over a closed loop	$v_{\mathrm{final}} = v_{\mathrm{initial}}$	speed is restored though velocity direction may not be; true for any conservative force system
instantaneous (power) form	$P = \mathbf{F}\cdot\mathbf{v} = \mathrm{d}K/\mathrm{d}t$	at fixed engine power the available force is $P/v$ — acceleration necessarily starves at high speed

Common trap“The wall did work on you — that's why you speed up when you push off it”

The wall exerts a very real force on your hands, and you very really gain kinetic energy — yet the wall does zero work, because its force acts on a contact point that never moves. No displacement, no work. The energy came from your muscles: an internal, chemical source. The resolution is that $W_{\mathrm{net}} = \Delta K$ as derived here governs a point particle; a person, a car, a sprinter are deformable systems, and for them the external force still sets the momentum budget while the energy can be supplied internally. The same trap in reverse: a braking car's kinetic energy is destroyed mostly inside the brake discs, not by the road's friction force, whose contact point on a rolling tire is momentarily at rest. Related bookkeeping errors to avoid: using the work of one force where the theorem demands the net work, and forgetting that friction's work is negative on every leg of a round trip — it never refunds on the way back.

Worked examples

1 · Braking distance and the tyranny of v² (confidence builder)

A 1400 kg car travels at 108 km/h when the driver brakes hard, locking into a skid with $\mu_k = 0.70$ on dry asphalt. Find the stopping distance, check the energy ledger explicitly, and compare with braking from 54 km/h.

Step 1. Convert and set up: $v = 108/3.6 = 30.0\ \mathrm{m/s}$ . The only horizontal force is kinetic friction $f = \mu_k mg$ , acting opposite the motion over the whole skid.

Step 2. Apply the theorem from $v$ to rest: $-\mu_k m g\,d = 0 - \tfrac12 mv^2$ . The mass cancels, leaving $d = \dfrac{v^2}{2\mu_k g} = \dfrac{30^2}{2 \times 0.70 \times 9.81} = \boxed{65.5\ \mathrm{m}}$ .

Step 3. Audit the ledger with the actual numbers: friction force $f = 0.70 \times 1400 \times 9.81 = 9614\ \mathrm{N}$ ; work $W = -9614 \times 65.5 = -6.30\times10^5\ \mathrm{J}$ ; kinetic energy surrendered $\Delta K = -\tfrac12 \times 1400 \times 30^2 = -6.30\times10^5\ \mathrm{J}$ . The books balance to the joule — 630 kJ, all of it delivered to hot rubber and asphalt.

Step 4. At half the speed, $v = 15\ \mathrm{m/s}$ : $d = \dfrac{15^2}{2 \times 0.70 \times 9.81} = \boxed{16.4\ \mathrm{m}}$ — exactly one quarter. Halving your speed doesn't halve your stopping distance; it quarters it. Note also what the formula doesn't contain: the mass. A loaded truck and an empty one skid the same distance (at equal $\mu_k$ ), because both the energy to be destroyed and the destroying force scale with $m$ .

2 · A winched crate on a rough ramp (exam level)

A 25 kg crate starts from rest and is dragged 8.0 m up a 30° ramp by a rope parallel to the slope with constant tension $T = 220\ \mathrm{N}$ . The kinetic friction coefficient is $\mu_k = 0.30$ . Find the crate's speed at the top, and show which force won and lost what along the way.

Step 1. List the four forces and their angles to the motion: tension (along, deposits), gravity (component $mg\sin\theta$ against, withdraws), friction (against, withdraws), normal force (perpendicular, does nothing). This is a net work problem — compute each force's work and add.

Step 2. Tension's deposit: $W_T = T d = 220 \times 8.0 = +1760\ \mathrm{J}$ .

Step 3. Gravity's withdrawal: $W_g = -mg\sin\theta\, d = -25 \times 9.81 \times 0.500 \times 8.0 = -981\ \mathrm{J}$ — equivalently $-mgh$ with $h = d\sin\theta = 4.0\ \mathrm{m}$ of height gained.

Step 4. Friction's withdrawal: the normal force on the ramp is $N = mg\cos\theta = 25 \times 9.81 \times 0.866 = 212\ \mathrm{N}$ , so $W_f = -\mu_k N d = -0.30 \times 212 \times 8.0 = -510\ \mathrm{J}$ .

Step 5. Balance the account and read off the speed: $W_{\mathrm{net}} = 1760 - 981 - 510 = +269\ \mathrm{J} = \tfrac12 mv_f^2$ , so $v_f = \sqrt{\dfrac{2 \times 269}{25}} = \boxed{4.6\ \mathrm{m/s}}$ .

Step 6. Sanity checks: the answer needed no time variable and no acceleration — kinematics would have demanded both. And the ledger tells the physical story at a glance: of the 1760 J the winch paid in, 981 J became height (recoverable later — the seed of potential energy) and 510 J became heat (gone for good); only 269 J, about 15 percent, ended up as motion.

Origins · 1 minJames Prescott Joule and how this was discovered

James Prescott Joule

1818–1889 · Manchester, England

Joule was a Manchester brewer who did physics before breakfast, with the thermometers and the patience of a man who measured fermentation vats for a living. Through the 1840s he chased one number — how much mechanical work equals how much heat — with churning paddle wheels in insulated barrels of water, measuring temperature rises of thousandths of a degree. His 1843–1850 experiments pinned the mechanical equivalent of heat at 4.15–4.21 J per calorie (the modern value is 4.186), closing the account that the work-energy theorem leaves open: when friction withdraws kinetic energy, the joules do not vanish — they reappear, to the last decimal, as heat. Physics repaid the brewer by stamping his name on the unit every energy transaction is denominated in. His careful numbers, presented to a mostly indifferent British Association meeting in 1847, caught the attention of a young William Thomson — and their collaboration seeded thermodynamics.

History

1686

— Leibniz argues that the true measure of motion is vis viva,

mv^2

, against the Cartesians' momentum

mv

— the quantity conserved in elastic collisions, he shows, is the square.

1740

— Émilie du Châtelet, synthesizing Leibniz's argument with Willem 's Gravesande's brass-ball-into-clay experiments (double the speed, four times the dent), makes the definitive case that "living force" goes as

v^2

1829

— Gaspard-Gustave de Coriolis defines work as force times displacement and attaches the factor one-half to

mv^2

, turning vis viva into the modern kinetic energy and making the work-energy relation an engineer's tool.

1843

— James Prescott Joule's paddle-wheel experiments fix the mechanical equivalent of heat, showing that work withdrawn from the kinetic-energy account by friction reappears as heat, joule for joule.

1847

— Hermann von Helmholtz generalizes the ledger to all of physics in "On the Conservation of Force," folding mechanics, heat, electricity, and chemistry into a single conservation law.

Going further

Texts. Kleppner & Kolenkow, An Introduction to Mechanics, 2nd ed., ch. 5 (Energy) · Morin, Introduction to Classical Mechanics, ch. 5 (Conservation of energy and momentum) · Taylor, Classical Mechanics, ch. 4 (Energy) · Goldstein, Poole & Safko, Classical Mechanics, 3rd ed., §1.1 · Landau & Lifshitz, Mechanics, §6 (Energy).

Builds on: Newton's Second Law · Velocity and Acceleration · Vectors, Coordinate Systems and Kinematics · Newton's First Law and Inertial Frames. Leads to: Gravitational Force Near Earth's Surface · Normal Force · Circular Motion and Centripetal Acceleration · Projectile Motion.

Is the work-energy theorem the same thing as conservation of energy? When do I use which?

They are relatives, not twins. The work-energy theorem is always true for a particle — it is Newton's second law in disguise, valid for any force including friction. Conservation of mechanical energy, $\Delta K + \Delta U = 0$ , is the special case where every force doing work is conservative, so its work can be pre-paid into a potential energy function. Practical rule: if friction, drag, or an engine is present, write the full theorem, $W_{\mathrm{net}} = \Delta K$ , and charge each force honestly; if only gravity and springs act, use energy conservation and skip the integrals — the potential has already done them for you.

A braking car's wheels roll without skidding, so the road's friction acts on a contact point that is momentarily at rest and does no work. What actually removes the car's kinetic energy?

The brake pads do, internally. For rolling-without-slipping, static friction from the road exerts the backward force that decelerates the car's center of mass — it controls the momentum ledger — but since its application point has zero velocity, it transfers no energy. The kinetic energy is destroyed inside the brake discs, where pad rubs on rotor through a real sliding displacement, converting $\tfrac12 mv^2$ into heat (glowing-orange rotors on a mountain descent are this ledger entry made visible). This is the classic example of why the single-particle theorem needs care for extended bodies: external force and internal energy conversion are separate books.

Kinetic energy depends on the reference frame — a ball at rest on a train has zero KE for a passenger and plenty for the platform. Doesn't that break the theorem?

No — it makes the theorem frame-dependent but valid in every inertial frame. The platform observer sees the same force act through a longer displacement (the train moves), so they book more work; they also book a bigger $\Delta K$ . The two excesses match exactly — the guarantee comes from the impulse-momentum theorem, which ties the mismatch in work to the mismatch in kinetic energy through the boost velocity. Each observer's ledger balances internally. What you must never do is mix frames: force from one frame, displacement from another. Choose an inertial frame, compute everything there, and the theorem holds.

I carry a heavy box across a room at constant height and constant speed. I am exhausted, yet the theorem says I did zero work on the box. Seriously?

Seriously — on the box. Your upward force is perpendicular to the horizontal displacement, so $\mathbf{F}\cdot\mathrm{d}\mathbf{r} = 0$ , and the box's kinetic energy is indeed unchanged. Your exhaustion is real but happens inside your arm: human muscle burns chemical energy just to sustain a force, because muscle fibers continually contract and slip even during a static hold (isometric effort). A table holding the same box burns nothing. Physics work is a precise transaction — force through displacement of the point of application — not a measure of physiological effort, and this question is exactly where the two vocabularies part ways.

Why is kinetic energy ½mv² and not mv² or mv? Where does the one-half actually come from?

From an integral. Push a mass from rest with force $F$ : the work accumulates as $\int F\,\mathrm{d}x = m\int v\,\mathrm{d}v$ , and the antiderivative of $v$ is $\tfrac12 v^2$ — the one-half is manufactured by the integration, and kinetic energy is defined with it so that the ledger balances exactly. The $v^2$ (rather than $v$ ) traces to the fact that a faster particle covers more distance per unit time, so a force deposits energy at a rate proportional to speed. Historically this was a century-long fight: Leibniz's vis viva was $mv^2$ , and Coriolis attached the $\tfrac12$ in 1829 precisely to make work equal the change in the account.

Does a magnetic force really never do work? Electric motors seem like a loud counterexample

On a free charged particle, never: $\mathbf{F} = q\,\mathbf{v}\times\mathbf{B}$ is perpendicular to $\mathbf{v}$ by construction, so $\mathbf{F}\cdot\mathbf{v} = 0$ identically — a magnetic field can bend a proton's path into a circle forever without changing its speed, which is exactly how storage rings work. Motors are subtler: the magnetic force redirects and constrains, while the energy is delivered by the electric fields driving the current (and by induced EMFs as the machine turns). The bookkeeping is delicate enough that "who does the work in a motor" is a beloved graduate-exam question — but the theorem's verdict on the pure magnetic force itself is airtight: zero, always.

On a round trip, friction's work doesn't cancel — but gravity's does. What makes the difference?

Direction policy. Gravity is a fixed vector: on the way up it opposes you, on the way down it helps, and over a closed loop its works cancel exactly — that is what makes it conservative and lets us define potential energy. Kinetic friction instead reverses itself to oppose whatever the current motion is: uphill it points downhill, downhill it points uphill, so its work is negative on both legs and accumulates as $-\mu_k N \times (\text{total path length})$ . That is why the incline round-trip derivation returns the block slower than it left, and why friction's work depends on the route taken, not just the endpoints — the defining mark of a non-conservative force.

If the net work over some stretch is zero, does that mean nothing happened to the motion?

It means precisely one thing: the speed at the end equals the speed at the start. Plenty may still have happened — the velocity's direction can be completely different (a ball whirled half a circle on a string has reversed its velocity at zero net work), the particle may have sped up and slowed down in between, and individual forces may have traded huge equal-and-opposite works (crane lowering at constant speed: gravity +, cable −). Kinetic energy is a scalar account; it is blind to direction and to history. If you need direction or timing, the theorem is the wrong tool — go back to Newton's second law or to momentum.

Work-Energy Theorem

Where this lives today

The physical picture

The bookkeeping equations

Which equation answers which question

Limiting cases

Worked examples

1 · Braking distance and the tyranny of v² (confidence builder)

2 · A winched crate on a rough ramp (exam level)

Going further

Before this

Where this leads