12
Part Zero · Foundations

Lagrangian mechanics

A ray of light passing from air into water bends at exactly the angle that gets it from one point to another in the least time, as if it had compared every possible route. A thrown ball follows the one path, out of all imaginable paths, that makes a certain quantity stationary. How can a ray or a ball “know” the best path before it has travelled any of them?

This chapter rewrites mechanics around that idea, the principle of stationary action. It looks like a philosophical curiosity. It is in fact the most powerful way to do physics, and in Part One it will turn out that, in a sense, nature really does try every path.

The lifeguard’s problem

A lifeguard on the beach sees a swimmer in trouble. Running is faster than swimming, so the straight line is not the quickest route; neither is running to the point of shore nearest the swimmer. Choose where to enter the water: 0.50 of the way along. The plate shows the total time.

The quickest route bends at the shore so that sin⁡θ1/v1=sin⁡θ2/v2\sin\theta_1/v_1 = \sin\theta_2/v_2. Light does the same at the surface of water, and with n=c/vn = c/v that is Snell’s law, n1sin⁡θ1=n2sin⁡θ2n_1\sin\theta_1 = n_2\sin\theta_2. Pierre de Fermat proposed in 1662 that light takes the path of least time. Established

The action

For moving bodies the quantity is not time but the action: kinetic energy minus potential energy, added up along the whole path.

S=∫t1t2(T−V) dtS = \int_{t_1}^{t_2} \big(T - V\big)\,dt

Throw a ball straight up so that it lands one second later. The plate shows its true path, a parabola in time, and a distorted path bent by 0.30 m. Every distorted path has more action, and near the true path the action changes only with the square of the distortion. The true path sits at a flat point: the action is stationary there.

From one principle to the equations of motion

Write the integrand as a Lagrangian, L(q,q˙)=T−VL(q, \dot q) = T - V, using whatever coordinate qq is convenient. Requiring the action to be stationary under every small change of path gives one equation for each coordinate, the Euler–Lagrange equation:

ddt∂L∂q˙=∂L∂q\frac{d}{dt}\frac{\partial L}{\partial \dot q} = \frac{\partial L}{\partial q}

For a ball, L=12mx˙2−mgxL = \tfrac12 m\dot x^2 - mgx, and the equation gives mx¨=−mgm\ddot x = -mg: Newton’s law, recovered from a single principle.

Deeper

Change the path by δq(t)\delta q(t), zero at both ends. To first order, δS=∫(∂L∂qδq+∂L∂q˙δq˙)dt\delta S = \int \left(\frac{\partial L}{\partial q}\delta q + \frac{\partial L}{\partial \dot q}\delta\dot q\right)dt. Integrate the second term by parts: ∫∂L∂q˙ddtδq dt=[∂L∂q˙δq]−∫ddt∂L∂q˙ δq dt\int \frac{\partial L}{\partial\dot q}\frac{d}{dt}\delta q\,dt = \left[\frac{\partial L}{\partial\dot q}\delta q\right] - \int \frac{d}{dt}\frac{\partial L}{\partial\dot q}\,\delta q\,dt. The boundary term vanishes, leaving δS=∫(∂L∂q−ddt∂L∂q˙)δq dt\delta S = \int\left(\frac{\partial L}{\partial q} - \frac{d}{dt}\frac{\partial L}{\partial\dot q}\right)\delta q\,dt. For this to vanish for every δq\delta q, the bracket must be zero.

Any coordinates you like

The great practical gift is freedom of coordinates. For a pendulum, use the angle θ\theta directly: L=12mℓ2θ˙2+mgℓcos⁡θL = \tfrac12 m\ell^2\dot\theta^2 + mg\ell\cos\theta. The Euler–Lagrange equation hands you θ¨=−(g/ℓ)sin⁡θ\ddot\theta = -(g/\ell)\sin\theta without ever mentioning the tension in the string. Constraints that would need awkward forces in Newton’s method simply disappear when you choose coordinates that respect them.

Where conservation laws come from

If the Lagrangian does not depend on some coordinate at all, the Euler–Lagrange equation says ddt∂L/∂q˙=0\frac{d}{dt}\partial L/\partial\dot q = 0: the matching momentum never changes. Emmy Noether turned this into a theorem in 1918: every continuous symmetry gives a conserved quantity. Established

The laws of physics are the same today as tomorrow: energy is conserved. The same here as over there: momentum is conserved. The same in every direction: angular momentum is conserved. The conservation laws of Chapter 11 are not separate facts. They are symmetries in disguise.

The fastest slide

A bead slides without friction from a high point to a lower one some distance away. Which shape of wire gets it there first? In 1696 Johann Bernoulli posed this brachistochrone problem, and the answer is not the straight line but a cycloid, the curve traced by a point on a rolling wheel. It dives steeply to build speed early. .

For a drop of 2 metres over a run of π\pi metres, the straight wire takes 1.19 s and the cycloid 1.00 s. The calculus of variations, the same mathematics as the action principle, finds the winner.

Determined but unpredictable

The Euler–Lagrange equations fix the future exactly, given the present. Yet some systems are unpredictable in practice. The plate shows two double pendulums released from angles that differ by one thousandth of a radian. . For a few swings they move as one, then they part company completely.

That is chaos: tiny differences in starting conditions grow exponentially. It is not randomness; the motion is fully determined. It means that predicting it requires knowing the start with impossible precision.

A preview: trying every path

Why should the true path be the stationary one? Richard Feynman’s answer, which you will meet in Part One, is that a quantum particle takes every path, each contributing a little arrow that turns by an angle S/ℏS/\hbar. Far from the stationary path, neighbouring paths have wildly different actions and their arrows cancel. Near it, the action is flat, the arrows line up, and they add. The classical path is simply where the quantum arrows agree.

What you now have

Fermat’s principle, the action, the Euler–Lagrange equations, free choice of coordinates, Noether’s link between symmetry and conservation, and a first sight of chaos.

· · ·

Three questions to carry with you.

What does “stationary action” mean?tap
The true path is one where small changes to the path change the action only at second order: the action is flat there, usually a minimum.
What does Noether’s theorem connect?tap
Symmetries and conservation laws: time symmetry gives energy, translation gives momentum, rotation gives angular momentum.
Is chaotic motion random?tap
No. It is fully determined, but small differences in the start grow so fast that prediction becomes impossible in practice.

And one to work out. Light enters water (n=1.33n = 1.33) from air at 30° to the vertical. At what angle does it travel in the water, in degrees?

θ₂ =°

Sources: L. D. Landau and E. M. Lifshitz, Mechanics. J. R. Taylor, Classical Mechanics. E. Noether, Nachr. Ges. Wiss. Göttingen, 235 (1918). R. P. Feynman, QED (1985). Checked by code: cycloid 1.003 s vs straight line 1.189 s; action of a thrown ball stationary with δS ∝ a².

Next · Chapter 13
Hamiltonian mechanics and Poisson brackets

Quantum mechanics was built by changing one bracket.

Turn the page