Case Study 2 — The Fastest Slide: Engineering the Brachistochrone

Field: Mechanical and design engineering Calculus used: The cycloid parametrization (§25.7), arc length (§25.4), the chain-rule slope and cusp condition (§25.3), surface area of revolution (§25.5)

A design brief that sounds easy

A children's-museum design team has a deceptively simple commission: build a slide. Not just any slide — a "race" exhibit with two parallel chutes side by side, both dropping a ball bearing from the same height $A$ at the top to the same point $B$ at the bottom, where $B$ is off to the side, not directly below $A$. One chute will be a straight ramp. The other will be a curve of the engineers' choosing. Visitors release both balls at once and watch. The exhibit only delights if the curved chute consistently wins — if the ball on the curve beats the ball on the straight ramp every single time, by a visible margin.

The lead engineer's first instinct is the universal one: the straight ramp is the shortest path from $A$ to $B$, so surely it is the fastest. This instinct is wrong, and the reason it is wrong is one of the most beautiful results in the history of calculus. The curve that wins is the cycloid — the very path traced by a point on a rolling wheel that we derived in §25.7 — turned upside down so its arch hangs below like a valley. Johann Bernoulli posed exactly this puzzle as a public challenge in 1696 and named the winning curve the brachistochrone, from the Greek for "shortest time."

Why shortest distance loses to shortest time

The intuition the engineer needs is a trade between distance and speed. Time is distance divided by speed, and on a gravity-driven slide the ball's speed depends on how far it has already dropped: by energy conservation, after falling a vertical distance $d$ the ball moves at $v = \sqrt{2gd}$. So a path that drops steeply at the start banks speed early, when speed is cheap to acquire, and then spends that speed cruising across the gentler lower stretch. The straight ramp, by contrast, accelerates gently the whole way and never builds the early speed. The cycloid sacrifices a little extra path length to buy a lot of early velocity, and the trade pays. A steeper initial plunge, counterintuitively, wins the race.

This is why the answer could never have been guessed by staring at lengths alone. It is an optimization not over a single number but over an entire curve — the first problem of what became the calculus of variations. We can only point at the full machinery here; the §25.7 sidebar sketches the Euler–Lagrange equation that pins the winner down. What matters for the museum is that the winner is a curve the team can actually cut and build, and its equations are the ones from this chapter.

Laying out the curve

The team parametrizes the chute as an inverted cycloid generated by a wheel of radius $r$:

$$x(t) = r(t - \sin t), \qquad y(t) = -\,r(1 - \cos t),$$

where the minus sign on $y$ flips the standard §25.7 arch downward into a valley. The release point $A$ sits at $t = 0$, where $(x, y) = (0, 0)$. As $t$ runs forward the ball slides down and across. The single design parameter is $r$, the "wheel radius" of the generating cycloid; the team will tune it so that the curve passes through the required endpoint $B$.

Suppose $B$ lies $4$ m to the side. The horizontal reach of a cycloid from the cusp to the bottom of its arch ($t = \pi$) is $x(\pi) = r(\pi - 0) = \pi r$. Setting $\pi r = 4$ gives $r = 4/\pi \approx 1.273$ m, and at that point the drop is $|y(\pi)| = r(1 - \cos\pi) = 2r \approx 2.546$ m. So the bottom of the valley sits about $2.55$ m below the release and $4$ m to the side — a chute roughly two and a half meters deep. (For a $B$ that is not at the lowest point of the arch, the team solves $x(t) = 4$, $y(t) = -h$ simultaneously for $r$ and the terminal $t$; the algebra is messier but the curve is the same family.)

How much track must we cut?

A fabricator does not care about elegance; she cares about how many meters of steel rail to bend. That is an arc-length question, and §25.7 already did the hard part. For the cycloid, $\dot x = r(1 - \cos t)$ and $\dot y = -r\sin t$ (the sign flip does not change the magnitude), so

$$\dot x^2 + \dot y^2 = r^2(1 - \cos t)^2 + r^2\sin^2 t = r^2(2 - 2\cos t) = 4r^2\sin^2(t/2),$$

using the half-angle identity $1 - \cos t = 2\sin^2(t/2)$. Therefore $\sqrt{\dot x^2 + \dot y^2} = 2r\,|\sin(t/2)|$, which is just $2r\sin(t/2)$ on $[0, \pi]$. The length of the chute from $A$ ($t = 0$) to the bottom ($t = \pi$) is

$$L = \int_0^{\pi} 2r\sin(t/2)\,dt = 2r\big[-2\cos(t/2)\big]_0^{\pi} = -4r(\cos(\pi/2) - \cos 0) = -4r(0 - 1) = 4r.$$

Half an arch has length exactly $4r$ — half of the famous $8r$ for a full arch. With $r = 4/\pi$, that is $L = 16/\pi \approx 5.09$ m of rail. The straight ramp from $A = (0,0)$ to $B = (4, -2.546)$ has length $\sqrt{4^2 + 2.546^2} \approx 4.74$ m. So the winning curve is genuinely longer — about $0.35$ m longer — and still faster. The fabricator now has her number; the exhibit's whole premise rests on that counterintuitive length comparison.

The cusp is a warning, not a feature

There is a subtlety the §25.3 cusp condition flags before anyone gets hurt. At $t = 0$, both $\dot x = r(1 - \cos 0) = 0$ and $\dot y = -r\sin 0 = 0$. The point momentarily stops, and the curve has a cusp — a sharp vertical corner at the very top of the chute. A real ball cannot start at a true cusp without an instantaneous infinite curvature, and a real rail cannot be bent into a perfect point. The engineers therefore start the physical chute a small distance past the cusp, at some $t_0 > 0$, where the tangent already has a finite, steep slope. Reading §25.3, the slope there is

$$\frac{dy}{dx} = \frac{\dot y}{\dot x} = \frac{-r\sin t}{r(1 - \cos t)} = -\frac{\sin t}{1 - \cos t} = -\cot(t/2),$$

which blows up to $-\infty$ as $t \to 0^+$ — the chute starts nearly vertical, exactly the steep early plunge the physics wanted, and exactly the place the rail must be eased. The cusp condition from this chapter is doing real safety-engineering work: it locates the one point where the idealized curve and a buildable rail must part company.

A second chute, a second formula

The museum also wants a polished sheet-metal trough around the rail — a surface, not just a wire. If the trough is formed by revolving the chute profile about its long axis (the line of travel, here taken as the $x$-axis), its area comes straight from the §25.5 surface-of-revolution formula,

$$S = \int_{t_0}^{\pi} 2\pi\,|y|\,\sqrt{\dot x^2 + \dot y^2}\,dt = \int_{t_0}^{\pi} 2\pi\,r(1 - \cos t)\cdot 2r\sin(t/2)\,dt,$$

with $|y| = r(1 - \cos t)$ as the radius from the axis. The fabricator feeds this to the CAD kernel, which integrates it numerically (no need for a closed form) to report the square meters of sheet to cut and the mass of zinc plating to budget — the very same integral that priced Archimedes' sphere now prices a museum trough. The cycloid that Huygens chased for a perfect pendulum clock, and that Bernoulli dressed up as a race, is here cut from steel by an integral every engineer in the shop can set up.

The payoff

Opening day: two balls released together, and the cycloid ball reaches the bottom first, every time, by a margin the eye can catch. Children ask why the curvy one wins when it is the longer track, and the docent gives the one-sentence answer that took the Bernoullis, Newton, Leibniz, and L'Hôpital a public feud to settle: because it drops steepest where dropping is cheapest. The exhibit is, quietly, a working monument to the parametric curve of §25.7 — its arc length, its slope, its cusp, and its surface, all earned by the calculus of this chapter.

Discussion Questions

1. Distance versus time. The straight ramp here is about $0.35$ m shorter than the cycloid yet loses the race. Explain, using $v = \sqrt{2gd}$, why early steepness beats overall shortness. At what part of the chute does the cycloid ball build most of its lead?

2. Tuning the radius. The team set $\pi r = 4$ to place $B$ at the bottom of the arch. If instead $B$ were only $2$ m down and $4$ m over (not at the lowest point), describe the two equations you would solve for $r$ and the terminal parameter $t$.

3. The cusp. Why can the physical chute not begin exactly at $t = 0$? Use the result $dy/dx = -\cot(t/2)$ to describe how the tangent behaves as $t \to 0^+$, and explain what the fabricator does about it.

4. Length check. Verify $L = 4r$ for half an arch by re-deriving $\sqrt{\dot x^2 + \dot y^2} = 2r\sin(t/2)$ and integrating over $[0, \pi]$. How does this relate to the full-arch result $8r$ of §25.7?

5. The tautochrone twist. The same inverted cycloid has a second magic property (§25.7): a ball released from rest anywhere on it reaches the bottom in the same time. How might a museum build a second exhibit dramatizing the tautochrone rather than the brachistochrone?

A Short Annotated Reading

• Stillwell, J., Mathematics and Its History (Springer, 3rd ed.). Tells the full brachistochrone story — Bernoulli's 1696 challenge, Newton's overnight anonymous solution, and the birth of the calculus of variations — at exactly the level a reader of this chapter can follow.

• Lawrence, J. D., A Catalog of Special Plane Curves (Dover). A compact reference giving the cycloid, trochoid, astroid, and cardioid each in parametric form with their arc lengths and areas — the practical companion to §25.7–25.8 when you need a classical curve's formula in a hurry.

• Simmons, G. F., Calculus Gems (MAA). Its chapters on the cycloid and the brachistochrone connect the geometry to the physics of least time, and motivate the Euler–Lagrange machinery the §25.7 Math Major Sidebar only gestures at.