This action might not be possible to undo. Are you sure you want to continue?

Welcome to Scribd! Start your free trial and access books, documents and more.Find out more

by John C. Baez

notes by Derek K. Wise Department of Mathematics University of California, Riverside

A L TEX’ed by Blair M. Smith

Department of Physics and astronomy Louisiana State University 2005

i

c 2005 John C. Baez & Derek K. Wise

ii

iii

Preface

These are notes for a mathematics graduate course on classical mechanics. I’ve taught this course twice recently. The ﬁrst time I focused on the Hamiltonian approach. This time I started with the Lagrangian approach, with a heavy emphasis on action principles, and derived the Hamiltonian approach from that. Derek Wise took notes. A The chapters in this L TEX version are in the same order as the weekly lectures, but I’ve merged weeks together, and sometimes split them over chapter, to obtain a more textbook feel to these notes. In some chapters a familiarity with modern diﬀerential geometry (diﬀerential forms, bundles, homology) is helpful but the eﬀort to learn the relevant mathematics is not overly demanding. For reference, the weekly lectures are outlined here. Week 1: (Mar. 28, 30, Apr. 1)—The Lagrangian approach to classical mechanics: deriving F = ma from the requirement that the particle’s path be a critical point of the action. The prehistory of the Lagrangian approach: D’Alembert’s ‘principle of least energy’ in statics, Fermat’s ‘principle of least time’ in optics, and how D’Alembert generalized his principle from statics to dynamics using the concept of ‘inertia force’. Week 2: (Apr. 4, 6, 8)—Deriving the Euler-Lagrange equations for a particle on an arbitrary manifold. Generalized momentum and force. Noether’s theorem on conserved quantities coming from symmetries. Examples of conserved quantities: energy, momentum and angular momentum. Week 3 (Apr. 11, 13, 15)—Example problems: (1) The Atwood machine. (2) A frictionless mass on a table attached to a string threaded through a hole in the table, with a mass hanging on the string. (3) A special-relativistic free particle: two Lagrangians, one with reparametrization invariance as a gauge symmetry. (4) A special-relativistic charged particle in an electromagnetic ﬁeld. Week 4 (Apr. 18, 20, 22)—More example problems: (5) A special-relativistic charged particle in an electromagnetic ﬁeld in special relativity, continued. (6) A general-relativistic free particle. Week 5 (Apr. 25, 27, 29)—How Jacobi uniﬁed Fermat’s principle of least time and Lagrange’s principle of least action by seeing the classical mechanics of a particle in a potential as a special case of optics with a position-dependent index of refraction. The ubiquity of geodesic motion. Kaluza-Klein theory. From Lagrangians to Hamiltonians. Week 6 (May 2, 4, 6)—From Lagrangians to Hamiltonians, continued. Regular and strongly regular Lagrangians. The cotangent bundle as phase space. Hamilton’s equations. Getting Hamilton’s equations directly from a least action principle. Week 7 (May 9, 11, 13)—Waves versus particles: the Hamilton-Jacobi equation. Hamilton’s principal function and extended phase space. How the Hamilton-Jacobi equation foreshadows quantum mechanics

iv

Week 8 (May 16, 18, 20)—Towards symplectic geometry. The canonical 1-form and the symplectic 2-form on the cotangent bundle. Hamilton’s equations on a symplectic manifold. Darboux’s theorem. Week 9 (May 23, 25, 27)—Poisson brackets. The Schrdinger picture versus the Heisenberg picture in classical mechanics. The Hamiltonian version of Noether’s theorem. Poisson algebras and Poisson manifolds. A Poisson manifold that is not symplectic. Liouville’s theorem. Weil’s formula. Week 10 (June 1, 3, 5)—A taste of geometric quantization. K¨hler manifolds. a If you ﬁnd errors in these notes, please email me!

Contents

1 From Newton’s Laws to Lagrange’s Equations 1.1 Lagrangian and Newtonian Approaches . . . . . . . . . . 1.1.1 Lagrangian versus Hamiltonian Approaches . . . 1.2 Prehistory of the Lagrangian Approach . . . . . . . . . . 1.2.1 The Principle of Minimum Energy . . . . . . . . 1.2.2 D’Alembert’s Principle and Lagrange’s Equations 1.2.3 The Principle of Least Time . . . . . . . . . . . . 1.2.4 How D’Alembert and Others Got to the Truth . . 2 Equations of Motion 2.1 The Euler-Lagrange Equations 2.1.1 Comments . . . . . . . 2.1.2 Lagrangian Dynamics . 2.2 Interpretation of Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 2 6 6 7 8 10 12 14 14 16 16 18 20 20 21 22 23 24 25 25 26 28 28 29 31 35 35 36

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .

3 Lagrangians and Noether’s Theorem 3.1 Time Translation . . . . . . . . . . . . . . . . . . . . . 3.1.1 Canonical and Generalized Coordinates . . . . . 3.2 Symmetry and Noether’s Theorem . . . . . . . . . . . 3.2.1 Noether’s Theorem . . . . . . . . . . . . . . . . 3.3 Conserved Quantities from Symmetries . . . . . . . . . 3.3.1 Time Translation Symmetry . . . . . . . . . . . 3.3.2 Space Translation Symmetry . . . . . . . . . . . 3.3.3 Rotational Symmetry . . . . . . . . . . . . . . . 3.4 Example Problems . . . . . . . . . . . . . . . . . . . . 3.4.1 The Atwood Machine . . . . . . . . . . . . . . . 3.4.2 Disk Pulled by Falling Mass . . . . . . . . . . . 3.4.3 Free Particle in Special Relativity . . . . . . . . 3.5 Electrodynamics and Relativistic Lagrangians . . . . . 3.5.1 Gauge Symmetry and Relativistic Hamiltonian . 3.5.2 Relativistic Hamiltonian . . . . . . . . . . . . . v

vi

CONTENTS

3.6 3.7

3.8

3.9

Relativistic Particle in an Electromagnetic Field . . . . . . . . Alternative Lagrangians . . . . . . . . . . . . . . . . . . . . . 3.7.1 Lagrangian for a String . . . . . . . . . . . . . . . . . . 3.7.2 Alternate Lagrangian for Relativistic Electrodynamics . The General Relativistic Particle . . . . . . . . . . . . . . . . 3.8.1 Free Particle Lagrangian in GR . . . . . . . . . . . . . 3.8.2 Charged particle in EM Field in GR . . . . . . . . . . The Principle of Least Action and Geodesics . . . . . . . . . . 3.9.1 Jacobi and Least Time vs Least Action . . . . . . . . . 3.9.2 The Ubiquity of Geodesic Motion . . . . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

. . . . . . . . . .

37 39 39 41 43 44 45 46 46 48 51 51 54 54 55 55 55 56 57 59 61 61 64 70 70 72 72 74 75 77 80 80 82 85 86 86 87

4 From Lagrangians to Hamiltonians 4.1 The Hamiltonian Approach . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Regular and Strongly Regular Lagrangians . . . . . . . . . . . . . . . . . . 4.2.1 Example: A Particle in a Riemannian Manifold with Potential V (q) 4.2.2 Example: General Relativistic Particle in an E-M Potential . . . . . 4.2.3 Example: Free General Relativistic Particle with Reparameterization Invariance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.4 Example: A Regular but not Strongly Regular Lagrangian . . . . . 4.3 Hamilton’s Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3.1 Hamilton and Euler-Lagrange . . . . . . . . . . . . . . . . . . . . . 4.3.2 Hamilton’s Equations from the Principle of Least Action . . . . . . 4.4 Waves versus Particles—The Hamilton-Jacobi Equations . . . . . . . . . . 4.4.1 Wave Equations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4.2 The Hamilton-Jacobi Equations . . . . . . . . . . . . . . . . . . . . 5 Symplectic Geometry 5.1 Towards Symplectic Geometry . . . . . . . . . . . . . . . 5.2 The Canonical Forms on the Cotangent Bundle . . . . . 5.2.1 The Canonical 1-form on the Cotangent Bundle . 5.2.2 The Symplectic 2-form on the Cotangent Bundle 5.3 Hamilton’s Equations on a Symplectic Manifold . . . . . 5.4 Darboux’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

6 Poisson Brackets 6.1 The Schr¨dinger Picture and the Heisenberg Picture in Classical o 6.2 The Hamiltonian Version of Noether’s Theorem . . . . . . . . . 6.3 Poisson Algebras and Poisson Manifolds . . . . . . . . . . . . . 6.3.1 A Poisson Manifold that is Not Symplectic . . . . . . . . 6.4 Liouville’s Theorem . . . . . . . . . . . . . . . . . . . . . . . . . 6.4.1 Phase Space Volumes . . . . . . . . . . . . . . . . . . . .

Mechanics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

CONTENTS

1

6.4.2

Weil’s formula . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

7 Introduction to Geometric Quantization 91 7.1 A Taste of Geometric Quantization . . . . . . . . . . . . . . . . . . . . . . 91 7.2 K¨hler Manifolds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95 a 7.3 General Geometric Quantization . . . . . . . . . . . . . . . . . . . . . . . . 96

**Chapter 1 From Newton’s Laws to Lagrange’s Equations
**

(Week 1, March 28, 30, April 1.) Classical mechanics is a very peculiar branch of physics. It used to be considered the sum total of our theoretical knowledge of the physical universe (Laplace’s daemon, the Newtonian clockwork), but now it is known as an idealization, a toy model if you will. The astounding thing is that probably all professional applied physicists still use classical mechanics. So it is still an indispensable part of any physicist’s or engineer’s education. It is so useful because the more accurate theories that we know of (general relativity and quantum mechanics) make corrections to classical mechanics generally only in extreme situations (black holes, neutron stars, atomic structure, superconductivity, and so forth). Given that GR and QM are much harder theories to use and apply it is no wonder that scientists will revert to classical mechanics whenever possible. So, what is classical mechanics?

1.1

Lagrangian and Newtonian Approaches

We begin by comparing the Newtonian approach to mechanics to the subtler approach of Lagrangian mechanics. Recall Newton’s law: F = ma (1.1) wherein we consider a particle moving in Ên . Its position, say q, depends on time t ∈ Ê, so it deﬁnes a function, q : Ê −→ Ên . From this function we can deﬁne velocity, v = q : Ê −→ Ên ˙ 2

1.1 Lagrangian and Newtonian Approaches

3

where q = ˙

dq , dt

and also acceleration, a = q : Ê −→ Ên . ¨

Now let m > 0 be the mass of the particle, and let F be a vector ﬁeld on Ên called the force. Newton claimed that the particle satisﬁes F = ma. That is: m a(t) = F (q(t)) . (1.2)

This is a 2nd-order diﬀerential equation for q : Ê → Ên which will have a unique solution given some q(t0 ) and q(t0 ), provided the vector ﬁeld F is ‘nice’ — by which we technically ˙ mean smooth and bounded (i.e., |F (x)| < B for some B > 0, for all x ∈ Ên ). We can then deﬁne a quantity called kinetic energy: K(t) := This quantity is interesting because d K(t) = m v(t) · a(t) dt = F (q(t)) · v(t) So, kinetic energy goes up when you push an object in the direction of its velocity, and goes down when you push it in the opposite direction. Moreover,

t1

1 m v(t) · v(t) 2

(1.3)

K(t1 ) − K(t0 ) =

t0 t1

F (q(t)) · v(t) dt F (q(t)) · q(t) dt ˙

t0

=

So, the change of kinetic energy is equal to the work done by the force, that is, the integral of F along the curve q : [t0 , t1 ] → Ên . This implies (by Stokes’s theorem relating line integrals to surface integrals of the curl) that the change in kinetic energy K(t1 ) − K(t0 ) is independent of the curve going from q(t0 ) = a to q(t0 ) = b iﬀ ∇×F = 0. This in turn is true iﬀ F = −∇V (1.4) for some function V : Ên → Ê. This function is then unique up to an additive constant; we call it the potential. A force with this property is called conservative. Why? Because in this case we can deﬁne the total energy of the particle by E(t) := K(t) + V (q(t)) (1.5)

4

From Newton’s Laws to Lagrange’s Equations

where V (t) := V (q(t)) is called the potential energy of the particle, and then we can show that E is conserved: that is, constant as a function of time. To see this, note that F = ma implies d [K(t) + V (q(t))] = F (q(t)) · v(t) + ∇V (q(t)) · (t) dt = 0, (because F = −∇V ). Conservative forces let us apply a whole bunch of cool techniques. In the Lagrangian approach we deﬁne a quantity L := K(t) − V (q(t)) (1.6)

**called the Lagrangian, and for any curve q : [t0 , t1 ] → Ên with q(t0 ) = a, q(t1 ) = b, we deﬁne the action to be t
**

1

S(q) :=

t0

L(t) dt

(1.7)

From here one can go in two directions. One is to claim that nature causes particles to follow paths of least action, and derive Newton’s equations from that principle. The other is to start with Newton’s principles and ﬁnd out what conditions, if any, on S(q) follow from this. We will use the shortcut of hindsight, bypass the philosophy, and simply use the mathematics of variational calculus to show that particles follow paths that are ‘critical points’ of the action S(q) if and only if Newton’s law F = ma holds. To do this,

Ên

qs + sδq

q

Ê

t0 t1

Figure 1.1: A particle can sniﬀ out the path of least action. let us look for curves (like the solid line in Fig. 1.1) that are critical points of S, namely: d S(qs )|s=0 = 0 ds (1.8)

1.1 Lagrangian and Newtonian Approaches

5

where qs = q + sδq for all δq : [t0 , t1 ] → Ên with, δq(t0 ) = δq(t1 ) = 0. To show that, F = ma ⇔ d S(qs )|t=0 = 0, ds ∀δq : [] → Ên , δq(t0 ) = δq(t1 ) = 0 (1.9)

we start by using integration by parts on the deﬁnition of the action, and ﬁrst noting that dqs /ds = δq(t) is the variation in the path, d d S(qs )|s=0 = ds ds

t1 t1 t0

1 mqs (t) · qs (t) − V (qs (t)) dt ˙ ˙ 2

s=0

s=0

=

t0 t1

d 1 2 mq (t) − V (qs (t)) dt ˙ ds 2 s mqs ˙

=

t0 t1

d d qs (t) − V (qs (t)) qs (t) dt ˙ ds ds

s=0

=

t0

d d qs (t) − V (qs (t))δq(t) dt mqs ˙ dt ds

s=0

**so after integrating by parts and noting the boundary terms vanish (because δq = 0 at t1 and t0 ), d S(qs )|s=0 = ds =
**

t0 t1

[−m¨s (t) − ∇V (qs (t))] q

t0 t1

d qs (t) dt ds

s=0

[−m¨s (t) − ∇V (qs (t))] δq(t)dt q

s=0

The variation in the action is then clearly zero for all variations δq iﬀ the term in brackets [. . .] is identically zero, that is, −m¨(t) − ∇V (q(t)) = 0 q So, the path q is a critical point of the action S iﬀ F = ma. The above result applies only for conservative forces, i.e., forces that can be written as the minus the gradient of some potential. However, this seems to be true of the most fundamental forces that we know of in our universe. It is a doctrine of classical mechanics that has withstood the test of time and experiment.

6

From Newton’s Laws to Lagrange’s Equations

1.1.1

Lagrangian versus Hamiltonian Approaches

I am not sure where to mention this, but before launching into the history of the Lagrangian approach may be as good a time as any. In later chapters we will describe another approach to classical mechanics: the Hamiltonian approach. Why do we need two approaches, Lagrangian and Hamiltonian? They both have their own advantages. In the simplest terms, the Hamiltonian approach focuses on position and momentum, while the Lagrangian approach focuses on position and velocity. The Hamiltonian approach focuses on energy, which is a function of position and momentum — indeed, ‘Hamiltonian’ is just a fancy word for energy. The Lagrangian approach focuses on the Lagrangian, which is a function of position and velocity. Our ﬁrst task in understanding Lagrangian mechanics is to get a gut feeling for what the Lagrangian means. The key is to understand the integral of the Lagrangian over time—the ‘action’, S. We shall see that this describes the ‘total amount that happened’ from one moment to another as a particle traces out a path. And, peeking ahead to quantum mechanics, the quantity exp(iS/ ), where is Planck’s constant, will describe the ‘change in phase’ of a quantum system as it traces out this path. In short, while the Lagrangian approach takes a while to get used to, it provides invaluable insights into classical mechanics and its relation to quantum mechanics. We shall see this in more detail soon.

1.2

Prehistory of the Lagrangian Approach

We’ve seen that a particle going from point a at time t0 to a point b at time t1 follows a path that is a critical point of the action,

t1

S=

t0

K − V dt

so that slight changes in its path do not change the action (to ﬁrst order). Often, though not always, the action is minimized, so this is called the Principle of Least Action. Suppose we did not have the hindsight aﬀorded by the Newtonian picture. Then we might ask, “Why does nature like to minimize the action? And why this action K−V dt? Why not some other action?” ‘Why’ questions are always tough. Indeed, some people say that scientists should never ask ‘why’. This seems too extreme: a more reasonable attitude is that we should only ask a ‘why’ question if we expect to learn something scientiﬁcally interesting in our attempt to answer it. There are certainly some interesting things to learn from the question “why is action minimized?” First, note that total energy is conserved, so energy can slosh back and forth between kinetic and potential forms. The Lagrangian L = K − V is big when most of the energy is in kinetic form, and small when most of the energy is in potential

1.2 Prehistory of the Lagrangian Approach

7

form. Kinetic energy measures how much is ‘happening’—how much our system is moving around. Potential energy measures how much could happen, but isn’t yet— that’s what the word ‘potential’ means. (Imagine a big rock sitting on top of a cliﬀ, with the potential to fall down.) So, the Lagrangian measures something we could vaguely refer to as the ‘activity’ or ‘liveliness’ of a system: the higher the kinetic energy the more lively the system, the higher the potential energy the less lively. So, we’re being told that nature likes to minimize the total of ‘liveliness’ over time: that is, the total action. In other words, nature is as lazy as possible! For example, consider the path of a thrown rock in the Earth’s gravitational ﬁeld, as in Fig. 1.2. The rock traces out a parabola, and we can think of it as doing this in order K − V small here. . . good! Spend as much time as possible here

K − V big. . . bad! Get this over with quick! Figure 1.2: A particle’s “lazy” motion, minimizes the action. to minimize its action. On the one hand, it wants to spend a lot much time near the top of its trajectory, since this is where the kinetic energy is least and the potential energy is greatest. On the other hand, if it spends too much time near the top of its trajectory, it will need to really rush to get up there and get back down, and this will take a lot of action. The perfect compromise is a parabolic path! Here we are anthropomorphizing the rock by saying that it ‘wants’ to minimize its action. This is okay if we don’t take it too seriously. Indeed, one of the virtues of the Principle of Least Action is that it lets us put ourselves in the position of some physical system and imagine what we would do to minimize the action. There is another way to make progress on understanding ‘why’ action is minimized: history. Historically there were two principles that were fairly easy to deduce from observations of nature: (i) the principle of minimum energy used in statics, and (ii) the principle of least time, used in optics. By putting these together, we can guess the principle of least action. So, let us recall these earlier minimum principles.

1.2.1

The Principle of Minimum Energy

Before physicists really got going in studying dynamical systems they used to study statics. Statics is the study of objects at rest, or in equilibrium. Archimedes studied the laws of

8

From Newton’s Laws to Lagrange’s Equations

L1 m1

L2 m2

Figure 1.3: A principle of energy minimization determines a lever’s balance. a see-saw or lever (Fig. 1.3), and he found that this would be in equilibrium if m1 L1 = m2 L2 . Later D’Alembert understood this using his “principle of virtual work”. He considered moving the lever slightly, i.e., inﬁnitesimally, He claimed that in equilibrium the inﬁnitesdq1 dθ dq2

Figure 1.4: A principle of energy minimization determines a lever’s balance. imal work done by this motion is zero! He also claimed that the work done on the ith body is, dWi = Fi dθi and gravity pulls down with a force mi g so, dWi = (0, 0, −mg) · (0, 0, −L1 dθ) = m1 gL1 dθ and similarly, dW2 = −m2 gL2 dθ Now D’Alembert’s principle says that equilibrium occurs when the “virtual work” dW = dW1 +dW2 vanishes for all dθ (that is, for all possible inﬁnitesimal motions). This happens when m1 L1 − m2 L2 = 0 which is just as Archimedes wrote.

1.2.2

D’Alembert’s Principle and Lagrange’s Equations

Let’s go over the above analysis in more detail. I’ll try to make it clear what we mean by virtual work.

1.2 Prehistory of the Lagrangian Approach

9

The forces and constraints on a system may be time dependent. So equal small inﬁnitesimal displacements of the system might result in the forces Fi acting on the system doing diﬀerent amounts of work at diﬀerent times. To displace a system by δri for each position coordinate, and yet remain consistent with all the constraints and forces at a given instant of time t, without any time interval passing is called a virtual displacement. It’s called ‘virtual’ because it cannot be realized: any actual displacement would occur over a ﬁnite time interval and possibly during which the forces and constraints might change. Now call the work done by this hypothetical virtual displacement, Fi · δri , the virtual work. Consider a system in the special state of being in equilibrium, i.e., when Fi = 0. Then because by deﬁnition the virtual displacements do not change the forces, we must deduce that the virtual work vanishes for a system in equilibrium, Fi · δri = 0,

i

(when in equilibrium)

(1.10)

Note that in the above example we have two particles in Ê3 subject to a constraint (they are pinned to the lever arm). However, a number n of particles in Ê3 can be treated as a single quasi-particle in Ê3n , and if there are constraints it can move in some submanifold of Ê3n . So ultimately we need to study a particle on an arbitrary manifold. But, we’ll postpone such sophistication for a while. For a particle in Ên , D’Alembert’s principle simply says, q(t) = q0 satisﬁes F = ma, vanishes for all dq ∈ Ên , (it’s in equilibrium) (virtual work is zero for δq → 0) (no force on it!)

dW = F · dq F = 0,

If the force is conservative (F = −∇V ) then this is also equivalent to, ∇V (q0 ) = 0 that is, we have equilibrium at a critical point of the potential. The equilibrium will be stable if q0 is a local minimum of the potential V . We can summarize all the above by proclaiming that we have a “principle of least energy” governing stable equilibria. We also have an analogy between statics and dynamics, Statics equilibrium, a = 0 potential, V critical points of V Dynamics F = ma

t1

action, S =

t0

K − V dt

critical points of S

10

From Newton’s Laws to Lagrange’s Equations

unstable equilibrium unstable equilibrium

V

stable equilibrium

Ên

Figure 1.5: A principle of energy minimization determines a lever’s balance.

d dt d dt

∂T ∂ qi ˙ ∂L ∂ qi ˙

−

∂T = Qi . ∂qi ∂L = 0. ∂qi

(1.11)

−

(1.12)

D’Alembert’s principle is an expression for Newton’s second law under conditions where the virtual work done by the forces of constraint is zero.

1.2.3

The Principle of Least Time

Time now to look at the second piece of history surrounding the principles of Lagrangian mechanics. As well as hints from statics, there were also hints from the behavior of light, hints again that nature likes to minimize eﬀort. In a vacuum light moves in straight lines, which in Euclidean space is the minimum distance. But more interesting than straight lines are piecewise straight paths and curves. Consider reﬂection of light from a mirror, What path does the light take? The empirical answer was known since antiquity, it chooses B such that θ1 = θ2 , so the angle of incidence equals the angle of reﬂection. But this is precisely the path that minimizes the distance of the trajectory subject to the condition that it must hit the mirror (at least at one point). In fact light traveling from A to B takes both the straight paths ABC and AC. Why is ABC the shortest path hitting

1.2 Prehistory of the Lagrangian Approach

11

A

C

θ1 B

θ2 θ2

C′ the mirror? This follows from some basic Euclidean geometry: B minimizes AB + BC ⇔ B minimizes AB + BC ′ ⇔ A, B, C ′ lie on a line ⇔ θ1 = θ2 Note the introduction of the ﬁctitious image C′ “behind” the mirror, this is a trick often used in solving electrostatic problems (a conducting surface can be replaced by ﬁctitious mirror image charges to satisfy the boundary conditions), it is also used in geophysics when one has a geological fault, and in hydrodynamics when there is a boundary between two media (using mirror image sources and sinks). The big clue leading to D’Alembert’s principle however came from refraction of light. Snell (and predecessors) noted that each medium has some number n associated with it, θ1 medium 1

medium 2 θ2

called the index of refraction, such that, n1 sin θ1 = n2 sin θ2

12

From Newton’s Laws to Lagrange’s Equations

(having normalized n so that for a vacuum n = 1). Someone guessed the explanation, realizing that if the speed of light in a medium is proportional to 1/n, then light will satisfy Snell’s law if the light minimizes the time it takes to get from A to C. In the case of refraction it is the time that is important, not just the path distance. But the same is true for the law of reﬂection, since in that case the path of minimum length gives the same results as the path of minimum time. So, not only is light the fastest thing around, it’s also always taking the quickest path from here to there!

1.2.4

How D’Alembert and Others Got to the Truth

Sometimes laws of physics are just guessed using a bit of intuition and a gut feeling that nature must be beautiful or elegantly simple (though occasionally awesomely complex in beauty). One way to make good guesses is to generalize. D’Alembert’s principle of virtual work for statics says that equilibrium occurs when F (q0 ) · δq = 0, ∀δq ∈ Ên D’Alembert generalized this to dynamics by inventing what he called the “inertia force”=−m a, and he postulated that in dynamics equilibrium occurs when the total force = F +inertia force, vanishes. Or symbolically when, F (q(t)) − ma(t) · δq(t) = 0 We then take a variational path parameterized by s, qs (t) = q(t) + s δq(t) where δq(t0 ) = δq(t1 ) = 0 and with these paths, for any function f on the space of paths we can deﬁne the variational derivative, d δf := f (qs ) (1.14) ds s=0 Then D’Alembert’s principle of virtual work implies

t1

(1.13)

F (q) − m¨ · δq dt = 0 q

t0

**for all δq, so if F = −∇V , we get
**

t1

0=

t0 t1

**−∇V (q) − m¨ · δq dt q −∇V (q) · δq + mqδ q dt ˙ ˙
**

t0

=

1.2 Prehistory of the Lagrangian Approach

13

**using d V (qs (t)) ds and δ(q 2 ) = ˙ then we have 0= ˙ dV dqs m dqs (t) dt + dq dt 2 ds t0 d t1 d m = − V qs (t) + qs (t) dt ˙ ds t0 dt 2 s=0 t1 m d ˙ =δ − V qs (t) + qs (t) dt dt 2 t0 −
**

t1 t1

=

s=0

dV dqs (t) dq ds

s=0

dqs (t) ˙2 ds

therefore δ

−V (q) + K dt = 0

t0

so the path taken by the particle is a critical point of the action, S(q) = (K − V ) dt (1.15)

We’ve described how D’Alembert might have arrived at the principle of least action by generalizing previously known energy minimization and least time principles. Still, there’s something unsatisfying about the treatment so far. We do not really understand why one must introduce the ‘inertia force’. We only see that it’s necessary to obtain agreement with Newtonian mechanics (which is manifest in Eq.(1.13)). We conclude with a few more words about this mystery. Recall from undergraduate physics that in an accelerating coordinate system there is a ﬁctional force = ma, which is called the centrifugal force. We use it, for example, to analyze simple physics in a rotating reference frame. If you are inside the rotating system and you throw a ball straight ahead it will appear to curve away from your target, and if you did not know that you were rotating relative to the rest of the universe then you’d think there was a force on the ball equal to the centrifugal force. If you are inside a big rapidly rotating drum then you’ll also feel pinned to the walls. This is an example of an inertia force which comes from using a funny coordinate system. In general relativity, one sees that — in a certain sense — gravity is an inertia force!

**Chapter 2 Equations of Motion
**

(Week 2, April 4, 6, 8.) In this chapter we’ll start to look at the Lagrangian equations of motion in more depth. We’ll look at some speciﬁc examples of problem solving using the Euler-Lagrange equations. First we’ll show how the equations are derived.

2.1

The Euler-Lagrange Equations

We are going to start thinking of a general classical system as a set of points in an abstract conﬁguration space or phase space1 . So consider an arbitrary classical system as living in a space of points in some manifold Q. For example, the space for a spherical double pendulum would look like Fig. 2.1, where Q = S 2 × S 2 . So our system is “a particle in

Q = S2 × S2

Figure 2.1: Double pendulum conﬁguration space. Q”, which means you have to disabuse yourself of the notion that we’re dealing with real

The tangent bundle T Q will be referred to as conﬁguration space, later on when we get to the chapter on Hamiltonian mechanics we’ll ﬁnd a use for the cotangent bundle T ∗ Q, and normally we call this the phase space.

1

14

2.1 The Euler-Lagrange Equations

15

particles, we are not, we are dealing with a single quasi-particle in an abstract higher dimensional space. The single quasi-particle represents two real particles if we are talking about the classical system in Fig. 2.1. Sometimes to make this clear we’ll talk about “the system taking a path”, instead of “the particle taking a path”. It is then clear that when we say, “the system follows a path q(t)” that we’re referring to the point q in conﬁguration space Q that represents all of the particles in the real system. So as time passes “the system” traces out a path q : [t0 , t1 ] −→ Q and we deﬁne it’s velocity, q(t) ∈ Tq(t) Q ˙ to be the tangent vector at q(t) given by the equivalence class [σ] of curves through q(t) with derivatives σ(t) = dq(s)/ds|s=t. We’ll just write is as q(t). ˙ ˙ Let Γ be the space of smooth paths from a ∈ Q to b ∈ Q, Γ = {q : [t0 , t1 ] → Q|q(t0 ) = a, q(t1 ) = b} (Γ is an inﬁnite dimensional manifold, but we won’t go into that for now.) Let the Lagrangian=L for the system be any smooth function of position and velocity (not explicitly of time, for simplicity), L : T Q −→ Ê and deﬁne the action, S: S : Γ −→ Ê by

t1

S(q) :=

t0

L(q, q) dt ˙

(2.1)

The path that the quasi particle will actually take is a critical point of S, in accord with D’Alembert’s principle of least action. In other words, a path q ∈ Γ such that for any smooth 1-parameter family of paths qs ∈ Γ with q0 = q1 , we have d S(qs ) ds We write, d ds so Eq.(2.2) can be rewritten δS = 0 (2.3)

s=0

s=0

=0

(2.2)

as “δ ′′

16

Equations of Motion

2.1.1

Comments

What is a “1-parameter family of paths”? Well, a path is a curve, or a 1D manifold. So the 1-parameter family is nothing more nor less than a set of well-deﬁned paths {qs }, each one labeled by a parameter s. So a smooth 1-parameter family of paths will have q(s) everywhere inﬁnitesimally close to q(s + ǫ) for an inﬁnitesimal hyperreal ǫ. So in Fig. 2.2

qs q0

Figure 2.2: Schematic of a 1-parameter family of curves. we can go from q0 to qs by smoothly varying s from s = 0 to s = s What does the condition δS = 0 imply? Patience, we are just getting to that. We will now start to explore what δS = 0 means for our Lagrangian.

2.1.2

Lagrangian Dynamics

We were given that Q is a manifold, so it admits a covering of coordinate charts. For now, let’s pick coordinates in a neighborhood U of some point q(t) ∈ Q. Next, consider only variations qs such that qs = q outside U. A cartoon of this looks like Fig. 2.3 Then

Q

a U b

Figure 2.3: Local path variation.

2.1 The Euler-Lagrange Equations

17

we restrict attention to a subinterval [t′0 , t′1 ] ⊆ [t0 , t1 ] such that qs (t) ∈ U for t′0 ≤ t ≤ t′1 . Let’s just go ahead and rename t′0 and t′1 as “ t0 and t1 ” to drop the primes. We can use the coordinate charts on U, ϕ : U −→ Ên x −→ ϕ(x) = (x1 , x2 , . . . , xn ) and we also have coordinates for the 1-forms, dϕ : T U −→ T Ên ∼ Ên × Ên = (x, y) −→ dϕ(x, y) = (x1 , . . . , xn , y 1 , . . . , y n ) where y ∈ Tx Q. We restrict L : T M → Ê to T U ⊆ T M, in our case the manifold is M = Q, and then we can describe it, L, using the coordinates xi , y i on T U. The i are generalized position coordinates, the y i are the associated generalized velocity coordinates. (Velocity and position are in the abstract conﬁguration space Q). Using these coordinates we get,

t1

δS = δ

t0 t1

L q(t), q(t) dt ˙ δL(q, q) dt ˙

t0 t1

= =

t0

∂ ∂ δq i (t) + i δ q i dt ˙ i ∂x ∂y

where we’ve used the given smoothness of L and the Einstein summation convention for repeated indices i. Note that we can write δL as above using a local coordinate patch because the path variations δq are entirely trivial outside the patch for U. Continuing, using the Leibniz rule d ∂L ∂L d ∂L δq = δq + q ˙ dt ∂y dt ∂y ∂y we have,

t1

δS =

t0

∂L d ∂L − δq i (t) dt i i ∂x dt ∂y

= 0. If this integral is to vanish as demanded by δS = 0, then it must vanish for all path variations δq, further, the boundary terms vanish because we deliberately chose δq that

18

Equations of Motion

vanish at the endpoints t0 and t1 inside U. That means the term in brackets must be identically zero, or d ∂L ∂L − i =0 (2.4) i dt ∂y ∂x This is necessary to get δS = 0, for all δq, but in fact it’s also suﬃcient. Physicists always give the coordinates xi , y i on T U the symbols “q i ” and “q i ”, despite the fact that these ˙ i i also have another meaning, namely the x and y coordinates of the quantity, q(t), q(t) ∈ T U. ˙ So in any case, physicists write, d ∂L ∂L = i dt ∂ q i ˙ ∂q and they call these the Euler-Lagrange equations.

2.2

Interpretation of Terms

The derivation of the Euler-Lagrange equations above was fairly abstract, the terms “position” and “velocity” were used but were not assumed to be the usual kinematic notions that we are used to in physics, indeed the only reason we used those terms was for their analogical appeal. Now we’ll try to illuminate the E-L equations a bit by casting them into the usual position and velocity terms. So consider, Q=Ê 1 ˙ ˙ L(q, q) = mq · q − V (q) ˙ 2 1 ˙ ˙ = mq i qi − V (q) 2 TERMS ∂L ∂qi ˙ ∂L ∂q i MEANING (in this example) mq ˙ −(∇V )i MEANING (in general) the momentum pi the force Fi

When we write ∂V /∂q i = ∇V we’re assuming the q i are Cartesian coordinates on Q.

2.2 Interpretation of Terms

19

So translating our example into general terms, if we conjure up some abstract Lagrangian then we can think of the independent variables as generalized positions and velocities, and then the Euler-Lagrange equations can be interpreted as equations relating generalized concepts of momentum and force, and they say that p=F ˙ (2.5)

So there’s no surprise that in the mundane case of a single particle moving in Ê3 under time t this just recovers Newton II. Of course we can do all of our classical mechanics with Newton’s laws, it’s just a pain in the neck to deal with the redundancies in F = ma when we could use symmetry principles to vastly simplify many examples. It turns out that the Euler-Lagrange equations are one of the reformulations of Newtonian physics that make it highly convenient for introducing symmetries and consequent simpliﬁcations. Simpliﬁcations generally mean quicker, shorter solutions and more transparent analysis or at least more chance at insight into the characteristics of the system. The main thing is that when we use symmetry to simplify the equations we are reducing the number of independent variables, so it gets closer to the fundamental degrees of freedom of the system and so we cut out a lot of the wheat and chaﬀ (so to speak) with the full redundant Newton equations. One can of course introduce simpliﬁcations when solving Newton’s equations, it’s just that it’s easier to do this when working with the Euler-Lagrange equations. Another good reason to learn Lagrangian mechanics is that it translates better into quantum mechanics.

**Chapter 3 Lagrangians and Noether’s Theorem
**

If the form of a system of dynamical equations does not change under spatial translations then the momentum is a conserved quantity. When the form of the equations is similarly invariant under time translations then the total energy is a conserved quantity (a constant of the equations of motion). Time and space translations are examples of 1-parameter groups of transformations. Invariance under a group of transformations is precisely what we mean by a symmetry in group theory. So symmetries of a dynamical system give conserved quantities or conservation laws. The rigorous statement of all this is the content of Noether’s theorem.

3.1

Time Translation

To handle time translations we need to replace our paths q : [t0 , t1 ] → Q by paths q : Ê → Q, and then deﬁne a new space of paths, Γ = {q : Ê → Q}. The bad news is that the action

∞

S(q) =

−∞

L q(t), q(t) dt ˙

typically will not converge, so S is then no longer a function of the space of paths. Nevertheless, if δq = 0 outside of some ﬁnite interval, then the functional variation,

∞

δS :=

−∞

d L qs (t), qs (t) ˙ ds

dt

s=0

will converge, since the integral is smooth and vanishes outside this interval. Moreover, demanding that this δS vanishes for all such variations δq is enough to imply the Euler20

3.1 Time Translation

21

Lagrange equations:

∞

δS =

−∞ ∞

=

−∞ ∞

=

−∞

d L qs (t), qs (t) ˙ dt ds s=0 ∂L ∂L δqi + δ qi dt ˙ ∂qi ∂ qi ˙ d ∂L ∂L δqi dt − ∂qi dt ∂ qi ˙

where again the boundary terms have vanished since δq = 0 near t = ±∞. To be explicit, the ﬁrst term in ∂L i d ∂L d ∂L i δq = ˙ δq − δq i i ∂q ˙ dt ∂ q ˙ dt ∂ q i ˙ vanishes when we integrate. Then the whole thing vanishes for all compactly supported smooth δq iﬀ ∂L d ∂L = . dt ∂ qi ˙ ∂qi Recall that, ∂L = pi , is the generalized momentum, by defn. ∂ qi ˙ ∂L = pi , is the force, by the E-L eqns. ˙ ∂qi Note the similarity to Hamilton’s equations—if you change L to H you need to stick in a minus sign, and change variables from q to pi and eliminate pi . ˙ ˙

3.1.1

Canonical and Generalized Coordinates

In light of this noted similarity with the Hamilton equations of motion, let’s spend a few moments clearing up some terminology (I hate using jargon, but sometimes it’s unavoidable, and sometimes it can be eﬃcient—provided everyone is clued in). Generalized Coordinates For Lagrangian mechanics we have been using generalized coordinates, these are the {qi , qi }. The qi are generalized positions, and the qi are generalized velocities. The full ˙ ˙ set of independent generalized coordinates represent the degrees of freedom of a particle, or system of particles. So if we have N particles then we’d typically have 6N generalized coordinates (the “6” is for 3 space dimensions, and at each point a position and a momentum). These can be in any reference frame or system of axes, so for example,

22

Lagrangians and Noether’s Theorem

in a Cartesian frame, with two particles, in 3D space we’d have the 2 × 3 = 6 position coordinates, and 2 × 3 = 6 velocities, {x1 , y1 , z1 , x2 , y2 , z2 }, {u1, v1 , w1, u2 , v2 , w2 } where say u = vx , v = vy , w = vz are the Cartesian velocity components. This makes 12 = 6 × 2 = 6N coordinates, matching the total degrees of freedom as claimed. If we constrain the particles to move in a plane (say place them on a table in a gravitational ﬁeld) then we get 2N fewer degrees of freedom, and so 4N d.o.f. overall. By judicious choice of coordinate frame we can eliminate one velocity component and one position component for each particle. It is also handy to respect other symmetries of a system, maybe the particles move on a sphere for example, one can then deﬁne new positions and momenta with a consequent reduction in the number of these generalized coordinates needed to describe the system. Canonical Coordinates In Hamiltonian mechanics (which we have not yet fully introduced) we will ﬁnd it more useful to transform from generalized coordinates to canonical coordinates. The canonical coordinates are a special set of coordinates on the cotangent bundle of the conﬁguration space manifold Q. They are usually written as a set of (q i , pj ) or (xi , pj ) with the x’s or q’s denoting the coordinates on the underlying manifold and the p’s denoting the conjugate momentum, which are 1-forms in the cotangent bundle at the point q in the manifold. It turns out that the q i together with the pj , form a coordinate system on the cotangent bundle T ∗ Q of the conﬁguration space Q, hence these coordinates are called the canonical coordinates. We will not discuss this here, but if you care to know, later on we’ll see that the relation between the generalized coordinates and the canonical coordinates is given by the Hamilton-Jacobi equations for a system.

3.2

Symmetry and Noether’s Theorem

First, let’s give a useful deﬁnition that will make it easy to refer to a type of dynamical system symmetry. We want to refer to symmetry transformations (of the Lagrangian) governed by a single parameter. Deﬁnition 3.1 (one-parameter family of symmetries). A 1-parameter family of symmetries of a Lagrangian system L : T Q → Ê is a smooth map, F : Ê × Γ −→ Γ (s, q) −→ qs ,

with q0 = q

3.2 Symmetry and Noether’s Theorem

23

such that there exists a function ℓ(q, q) for which ˙ δL = for some ℓ : T Q → Ê, that is, d L qs (t), qs (t) ˙ ds for all paths q. Remark: The simplest case is δL = 0, in which case we really have a way of moving paths around (q → qs ) that doesn’t change the Lagrangian—i.e., a symmetry of L in the d most obvious way. But δL = dt ℓ is a sneaky generalization whose usefulness will become clear. =

s=0

dℓ dt d ℓ qs (t), qs (t) ˙ dt

3.2.1

Noether’s Theorem

Here’s a statement of the theorem. Note that ℓ in this theorem is the function associated with F in deﬁnition 3.1. Theorem 3.1 (Noether’s Theorem). Suppose F is a one-parameter family of symmetries of the Lagrangian system, L : T Q → Ê. Then, pi δqi − ℓ is conserved, that is, it’s time derivative is zero for any path q ∈ Γ satisfying the EulerLagrange equations. In other words, in boring detail,Euler-Lagrange!equations!symmetry d ∂L d i q(s)q(s) ˙ q (t) dt ∂y i ds s Proof. d d pi δq i − ℓ = pi δq i + pi δ q i − ℓ ˙ ˙ dt dt ∂L ∂L ˙ = i δq i + i δ q − δL ∂q ∂q ˙ = δL − δL = 0. ˙ − ℓ q(t), q(t)

s=0

=0

OK, big deal you might say. Before this can be of any use we’d need to ﬁnd a symmetry F . Then we’d need to ﬁnd out what this pi δqi − ℓ business is that is conserved. So let’s look at some examples.

24

Lagrangians and Noether’s Theorem

Example 1. Conservation of Energy. (A most important example!) All of our Lagrangian systems will have time translation invariance (because the laws of physics do not change with time, at least not to any extent that we can tell). So we have a one-parameter family of symmetries qs (t) = q(t + s) This indeed gives, ˙ δL = L for d L(qs ) ds =

s=0

d ˙ L=L dt

so here we take ℓ = L simply! We then get the conserved quantity pi δq i − ℓ = pi q i − L ˙ which we normally call the energy. For example, if Q = Ên , and if 1 L = mq 2 − V (q) ˙ 2 then this quantity is mq · q − ˙ ˙ 1 mq · q − V ˙ ˙ 2 1 = mq 2 + V (q) ˙ 2

The term in parentheses is K − V , and the left-hand side is K + V . Let’s repeat this example, this time with a speciﬁc Lagrangian. It doesn’t matter what the Lagrangian is, if it has 1-parameter families of symmetries then it’ll have conserved quantities, guaranteed. The trick in physics is to write down a correct Lagrangian in the ﬁrst place! (Something that will accurately describe the system of interest.)

3.3

**Conserved Quantities from Symmetries
**

Fs : Γ −→ Γ q −→ qs

We’ve seen that any 1-parameter family

3.3 Conserved Quantities from Symmetries

25

which satisﬁes ˙ δL = ℓ for some function ℓ = ℓ(q, q) gives a conserved quantity ˙ pi δq i − ℓ As usual we’ve deﬁned δL := d L qs (t), qs (t) ˙ ds

s=0

Let’s see how we arrive at a conserved quantity from a symmetry.

3.3.1

Time Translation Symmetry

For any Lagrangian system, L : T Q → Ê, we have a 1-parameter family of symmetries qs (t) = q(t + s) because ˙ δL = L so we get a conserved quantity called the total energy or Hamiltonian, H = pi q i − L ˙ (3.1)

(You might prefer “Hamiltonian” to “total energy” because in general we are not in the same conﬁguration space as Newtonian mechanics, if you are doing Newtonian mechanics then “total energy” is appropriate.) 1 For example: a particle on Ên in a potential V has Q = Ên , L(q, q) = 2 mq 2 − V (q). ˙ ˙ This system has ∂L pi q i = i q i = mq 2 = 2K ˙ ˙ ˙ ∂q ˙ so H = pi q i − L = 2K − (K − V ) = K + V ˙ as you’d have hoped.

3.3.2

Space Translation Symmetry

For a free particle in Ên , we have Q = Ên and L = K = 1 mq 2 . This has spatial translation ˙ 2 n symmetries, so that for any v ∈ Ê we have the symmetry qa (t) = q(t) + s v

26

Lagrangians and Noether’s Theorem

with δL = 0 because δ q = 0 and L depends only on q not on q in this particular case. (Since L ˙ ˙ i i does not depend upon q we’ll call q and ignorable coordinate; as above, these ignorables always give symmetries, hence conserved quantities. It is often useful therefore, to change coordinates so as to make some of them ignorable if possible!) In this example we get a conserved quantity called momentum in the v direction: pi δq i = mqi v i = mq · v ˙ ˙ Aside: Note the subtle diﬀerence between two uses of the term “momentum”; here it is a conserved quantity derived from space translation invariance, but earlier it was a diﬀerent thing, namely the momentum ∂L/∂ q i = pi conjugate to q i . These two diﬀerent ˙ “momentum’s” happen to be the same in this example! Since this is conserved for all v we say that mq ∈ Ên is conserved. (In fact that ˙ n whole Lie group G = Ê is acting as a translation symmetry group, and we’re getting a q(= Ên )-valued conserved quantity!)

3.3.3

Rotational Symmetry

The free particle in Ên also has rotation symmetry. Consider any X ∈ so(n) (that is a skew-symmetric n × n matrix), then for all s ∈ Ê the matrix esX is in SO(n), that is, it describes a rotation. This gives a 1-parameter family of symmetries qs (t) = esX q(t) which has δL = ∂L i ∂L i δq + i δ q = mqi δ q i ˙ ˙ ˙ ∂q i ∂q ˙

now qi is ignorable and so ∂L/∂q i = 0, and ∂L/∂ q i = pi , and ˙ δqi = ˙ d i q ˙ ds s s=0 d d sX = e q ds dt d = Xq dt =Xq ˙

s=0

3.3 Conserved Quantities from Symmetries

27

So, δL = mqi Xj q j ˙ i˙ = mq · (X q) ˙ ˙ =0 since X is skew symmetric as stated previously (X ∈ so(n)). So we get a conserved quantity, the angular momentum in the X direction. (Note: this whole bunch of maths above for δL just says that the kinetic energy doesn’t change when the velocity is rotated, without changing the magnitude in other words.) We write, pi δq i = mqi · (X q)i ˙ (δq i = Xq just as δ q i = X q in our previous calculation), or if X has zero entries except ˙ ˙ in ij and ji positions, where it’s ±1, then we get m(qi q j − qj q i ) ˙ ˙ the “ij component of angular momentum”. If n = 3 we write these as, ˙ mq×q Note that above we have assumed one can construct a basis for so(n) using matrices of the form assumed for X, i.e., skew symmetric with ±1 in the respectively ij and ji elements, otherwise zero. I mentioned earlier that we can do mechanics with any Lagrangian, but if we want to be useful we’d better pick a Lagrangian that actually describes a real system. But how do we do that? All this theory is ﬁne but is useless unless you know how to apply it. The above examples were for a particularly simple system, a free particle, for which the Lagrangian is just the kinetic energy, since there is no potential energy variation for a free particle. We’d like to know how to solve more complicated dynamics. The general idea is to guess the kinetic energy and potential energy of the particle (as functions of your generalized positions and velocities) and then let, L=K−V So we are not using Lagrangians directly to tell us what the fundamental physical laws should be, instead we plug in some assumed physics and use the Lagrangian approch to solve the system of equations. If we like, we can then compare our answers with experiments, which indirectly tells us something about the physical laws—but only provided the Lagrangian formulation of mechanics is itself a valid procedure in the ﬁrst place.

28

Lagrangians and Noether’s Theorem

3.4

Example Problems

(Week 3, Apr. 11, 13, 15.) To see how the formalisms in this chapter function in practise, let’s do some problems. It’s vastly superior to the simplistic F = ma formulation of mechanics. The Lagrangian formulation allows the conﬁguration space to be any manifold, and allows us to easily use any coordinates we wish.

3.4.1

The Atwood Machine

A frictionless pulley with two masses, m1 and m2 , hanging from it. We have

x m1 m2 ℓ−x

1 ˙ 2 1 ˙ K = (m1 + m2 )(ℓ − x) = (m1 + m2 )x2 2 2 V = −m1 qx − m2 g(ℓ − x) so 1 L = K − V = (m1 + m2 )x2 + m1 gx + m2 g(ℓ − x) ˙ 2 The conﬁguration space is Q = (0, ℓ), and x ∈ (0, ℓ) (we could use the “owns” symbol ∋ here and write Q = (0, ℓ) ∋ x ). Moreover T Q = (0, ℓ) × Ê ∋ (x, x). As usual L : T Q → Ê. ˙ Note that solutions of the Euler-Lagrange equations will only be deﬁned for some time t ∈ Ê, as eventually the solutions reaches the “edge” of Q. The momentum is ∂L p= = (m1 + m2 )x ˙ ∂x ˙ and the force is, ∂L = (m1 − m2 )g F = ∂x

3.4 Example Problems

29

**The Euler-Lagrange equations say p=F ˙ (m1 + m2 )¨ = (m1 − m2 )g x m1 − m2 g x= ¨ m1 + m2
**

m1 −m2 So this is like a falling object in a downwards gravitational acceleration a = m1 +m2 g. It is trivial to integrate the expression for x twice (feeding in some appropriate initial ¨ conditions) to obtain the complete solution to the motion x(t) and x(t). Note that x = 0 ˙ ¨ when m1 = m2 , and x = g if m2 = 0. ¨

3.4.2

Disk Pulled by Falling Mass

Consider next a disk pulled across a table by a falling mass. The disk is free to move on a frictionless surface, and it can thus whirl around the hole to which it is tethered to the mass below. r m1

ℓ−r m2

no swinging allowed!

Here Q = open disk of radius ℓ, minus it’s center = (0, ℓ) × S 1 ∋ (r, θ) T Q = (0, ℓ) × S 1 × Ê × Ê ∋ (r, θ, r, θ) ˙ ˙ 1 d 1 ˙ ˙ K = m1 (r 2 + r 2 θ2 ) + m2 (ℓ − r)2 2 2 dt V = gm2 (r − ℓ) 1 1 ˙ L = m1 (r 2 + r 2 θ2 ) + m2 r 2 + gm2 (ℓ − r) ˙ ˙ 2 2 having noted that ℓ is constant so d/dt(ℓ − r) = −r. For the momenta we get, ˙ ∂L = (m1 + m2 )r ˙ ∂r ˙ ∂L ˙ = m1 r 2 θ. pθ = ˙ ∂θ pr =

30

Lagrangians and Noether’s Theorem

Note that θ is an “ignorable coordinate”—it doesn’t appear in L—so there’s a symmetry, rotational symmetry, and pθ , the conjugate momentum, is conserved. The forces are, ∂L ˙ = m1 r θ2 − gm2 ∂r ∂L = 0, (θ is ignorable) Fθ = ∂θ Fr = ˙ Note: in Fr the term m1 r θ2 is recognizable as a centrifugal force, pushing m1 radially out, while the term −gm2 is gravity pulling m2 down and thus pulling m1 radially in. So, the Euler-Lagrange equations give, pr = Fr , ˙ pθ = 0, ˙ ˙ (m1 + m2 )¨ = m1 r θ2 − m2 g r ˙ pθ = m1 r 2 θ = J = a constant.

˙ Let’s use our conservation law here to eliminate θ from the ﬁrst equation: J ˙ θ= m1 r so (m1 + m2 )¨ = r J2 − m2 g m1 r 3

Thus eﬀectively we have a particle on (0, ℓ) of mass m = m1 + m2 feeling a force Fr = J2 − m2 g m1 r 3

which could come from an “eﬀective potential” V (r) such that dV /dr = −Fr . So integrate −Fr to ﬁnd V (r): J2 V (r) = + m2 gr 2m1 r 2 this is a sum of two terms that look like Fig. 3.1 ˙ If θ(t = 0) = 0 then there is no centrifugal force and the disk will be pulled into the hole until it gets stuck. At that time the disk reaches the hole, which is topologically the center of the disk that has been removed from Q, so then we’ve hit the boundary of Q and our solution is broken. At r = r0 , the minimum of V (r), our disc mass m1 will be in a stable circular orbit of radius r0 (which depends upon J). Otherwise we get orbits like Fig. 3.2.

3.4 Example Problems

31

V (r)

attractive gravitational potential repulsive centrifugal potential

r0

r

Figure 3.1: Potential function for disk pulled by gravitating mass.

Figure 3.2: Orbits for the disc and gravitating mass system.

3.4.3

Free Particle in Special Relativity

In relativistic dynamics the parameter coordinate that parametrizes the particle’s path in Minkowski spacetime need not be the “time coordinate”, indeed in special relativity there are many allowed time coordinates. Minkowski spacetime is, Ên+1 ∋ (x0 , x1 , . . . , xn ) if space is n-dimensional. We normally take x0 as “time”, and (x1 , . . . , xn ) as “space”, but of course this is all relative to one’s reference frame. Someone else travelling at some high velocity relative to us will have to make a Lorentz transformation to translate from our coordinates to theirs. This has a Lorentzian metric g(v, w) = v 0 w 0 − v 1 w 1 − . . . − v n w n = ηµν v µ w ν

32

Lagrangians and Noether’s Theorem

where

ηµν

In special relativity we take spacetime to be the conﬁguration space of a single point particle, so we let Q be Minkowski spacetime, i.e., Ên+1 ∋ (x0 , . . . , xn ) with the metric ηµν deﬁned above. Then the path of the particle is, q : Ê(∋ t) −→ Q where t is a completely arbitrary parameter for the path, not necessarily x0 , and not ˙ necessarily proper time either. We want some Lagrangian L : T Q → Ê, i.e., L(q i , q i ) such that the Euler-Lagrange equations will dictate how our free particle moves at a constant velocity. Many Lagrangians do this, but the “best” ones(s) give an action that is independent of the parameterization of the path—since the parameterization is “unphysical” (it can’t be measured). So the action

t1

=

1 0 0 ... 0 0 −1 0 . . . 0 0 0 −1 0 . . . .. . . . . . . . 0 0 0 . . . −1

S(q) =

t0

L q i (t), q i (t) dt ˙

**for q : [t0 , t1 ] → Q, should be independent of t. The obvious candidate for S is mass times arclength,
**

t1

S=m

t0

ηij q i (t)q j (t) dt ˙ ˙

or rather the Minkowski analogue of arclength, called proper time, at least when q is a ˙ timelike vector, i.e., ηij q i q j > 0, which says q points into the future (or past) lightcone ˙ ˙ ˙ and makes S real, in fact it’s then the time ticked oﬀ by a clock moving along the path q : [t0 , t1 ] → Q. By “obvious candidate” we are appealing somewhat to physical intu-

Timelike Lightlike Spacelike

ition and generalization. In Euclidean space, free particles follow straight paths, so the arclength or pathlength variation is an extremum, and we expect the same behavior in

3.4 Example Problems

33

Minkowski spacetime. Also, the arclength does not depend upon the parameterization, and lastly, the mass m merely provides the correct units for ‘action’. So let’s take L = ηij q i q j ˙ ˙ (3.2) and work out the Euler-Lagrange equations. We have pi = ∂ ∂L = i i ∂q ˙ ∂q ˙ =m ηij q i q j ˙ ˙

2ηij q j ˙ 2 ηij q i q j ˙ ˙ j ηij q ˙ mqi ˙ =m = iqj q ˙ ηij q ˙ ˙ (Note the numerator is “mass times 4-velocity”, at least when n = 3 for a real single particle system, but we’re actually in a more general n + 1-dim spacetime, so it’s more like the “mass times n + 1-velocity”). Now note that this pi doesn’t change when we change the parameter to accomplish q → αq. The Euler-Lagrange equations say, ˙ ˙ pi = Fi = ˙ ∂L =0 ∂q i

The meaning of this becomes clearer if we use “proper time” as our parameter (like parameterizing a curve by it’s arclength) so that

t1

q dt = t1 − t0 , ˙

t0

∀ t0 , t1

which ﬁxes the parametrization up to an additive constant. This implies q = 1, so that ˙ pi = m and the Euler-Lagrange equations say pi = 0 ⇒ m¨i = 0 ˙ q so our (free) particle moves unaccelerated along a straight line, which is as we desired (expected). Comments This Lagrangian from Eq.(3.2) has lots of symmetries coming from reparameterizing the path, so Noether’s theorem yields lots of conserved quantities for the relativistic free qi ˙ = mqi ˙ q ˙

34

Lagrangians and Noether’s Theorem

particle. This is in fact called “the problem of time” in general relativity, here we see it starting to show up in special relativity. These reparameterization symmetries work as follows. Consider any (smooth) 1parameter family of reparameterizations, i.e., diﬀeomorphisms fs : Ê −→ Ê with f0 = ½Ê . These act on the space of paths Γ = {q : Ê → Q} as follows: given any q ∈ Γ we get qs (t) = q fs (t) where we should note that qs is physically indistinguishable from q. Let’s show that ˙ δL = ℓ, (when E-L eqns. hold)

so that Noether’s theorem gives a conserved quantity pi δq i − ℓ Here we go then. δL = ∂L i ∂L i δq + i δ q ˙ ∂qi ˙ ∂q = pi δ q i ˙ mqi d i ˙ q fs (t) ˙ = q ds ˙ s=0 mqi d d i ˙ = q fs (t) q dt ds ˙ s=0 fs (t) mqi d i ˙ q fs (t) ˙ = q dt ˙ ds s=0 mqi d i ˙ = q δfs ˙ q dt ˙ d pi q i δf ˙ = dt

d ˙ where in the last step we used the E-L eqns., i.e. dt pi = 0, so δL = ℓ with ℓ = pi q i δf . ˙ So to recap a little: we saw the free relativistic particle has

L=m q =m ˙

ηij q i q j ˙ ˙

and we’ve considered reparameterization symmetries qs (t) = q fs (t) , fs : Ê → Ê

3.5 Electrodynamics and Relativistic Lagrangians

35

we’ve used the fact that δq i := so (repeating a bit of the above) δL =

d i q fs (t) ds

= q i δf ˙

s=0

∂L i ∂L i δq + i δ q ˙ ∂q i ∂q ˙ = pi δ q i , (since ∂L/∂q i = 0, and ∂L/∂ q i = p) ˙ ˙

= pi δ q i ˙ d = pi δq i dt d ˙ = pi q i δf dt d ˙ = pi q i δf, dt

and set pi q i δf = ℓ ˙

so Noether’s theorem gives a conserved quantity pi δq i − ℓ = pi q i δf − pi q i δf ˙ ˙ =0 So these conserved quantities vanish! In short, we’re seeing an example of what physicists call gauge symmetries. This is a good topic for starting a new section.

3.5

Electrodynamics and Relativistic Lagrangians

We will continue the story of symmetry and Noether’s theorem from the last section with a few more examples. We use principles of least action to conjure up Lagrangians for our systems, realizing that a given system may not have a unique Lagrangian but will often have an obvious natural Lagrangian. Given a Lagrangian we derive equations of motion from the Euler-Lagrange equations. Symmetries of L guide us in ﬁnding conserved quantities, in particular Hamiltonians from time translation invariance, via Noether’s theorem. This section also introduces gauge symmetry, and this is where we begin.

3.5.1

Gauge Symmetry and Relativistic Hamiltonian

What are gauge symmetries? 1. These are symmetries that permute diﬀerent mathematical descriptions of the same physical situation—in this case reparameterizations of a path.

36

Lagrangians and Noether’s Theorem

2. These symmetries make it impossible to compute q(t) given q(0) and q(0): since if ˙ q(t) is a solution so is q(f (t)) for any reparameterization f : Ê → Ê. We have a high degree of non-uniqueness of solutions to the Euler-Lagrange equations. 3. These symmetries give conserved quantities that work out to equal zero! Note that (1) is a subjective criterion, (2) and (3) are objective, and (3) is easy to test, so we often use (3) to distinguish gauge symmetries from physical symmetries.

3.5.2

Relativistic Hamiltonian

What then is the Hamiltonian for special relativity theory? We’re continuing here with the example problem of §3.4.3. Well, the Hamiltonian comes from Noether’s theorem from time translation symmetry, qs (t) = q(t + s) and this is an example of a reparametrization (with δf = 1), so we see from the previous results that the Hamiltonian is zero! H = 0. Explicitly, H = pi δ q i − ℓ where under q(t) → q(t + s) we have δ q i = q i δf , and so ˙ ˙ ˙ i δL = dℓ/dt, which implies ℓ = pi δq . The result H = 0 follows. Now you know why people talk about “the problem of time” in general relativity theory, it’s glimmerings are seen in the ﬂat Minkowski spacetime of special relativity. You may think it’s nice and simple to have H = 0, but in fact it means that there is no temporal evolution possible! So we can’t establish a dynamical theory on this footing! That’s bad news. (Because it means you might have to solve the static equations for the 4D universe as a whole, and that’s impossible!) But there is another conserved quantity deserving the title of “energy” which is not zero, and it comes from the symmetry, qs (t) = q(t) + s w where w ∈ Ên+1 and w points in some timelike direction. In fact any vector w gives a conserved quantity, δL = ∂L i ∂L i δq + i δ q ˙ ∂q i ∂q ˙ = pi δ q i , (since ∂L/∂q i = 0 and ∂L/∂ q i = pi ) ˙ ˙ = pi 0 = 0

3.6 Relativistic Particle in an Electromagnetic Field

37

qs

q w

˙ since δq i = w i , δ q i = w i = 0. This is our ℓ from Noether’s theorem with ℓ = 0, so ˙ ˙ Noether’s theorem says that we get a conserved quantity pi δq i − ℓ = pi w namely, the momentum in the w direction. We know p = 0 from the Euler-Lagrange ˙ equations, for our free particle, but here we see it coming from spacetime translation symmetry; p =(p0 , p1 , . . . , pn ) p0 is energy, (p1 , . . . , pn ) is spatial momentum. We’ve just about exhausted all the basic stuﬀ that we can learn from the free particle. So next we’ll add some external force via an electromagnetic ﬁeld.

3.6

Relativistic Particle in an Electromagnetic Field

The electromagnetic ﬁeld is described by a 1-form A on spacetime, A is the vector potential, such that dA = F (3.3) is a 2-from containing the electric and magnetic ﬁelds, F µν = ∂Ai ∂Aj − ∂xj ∂xi (3.4)

We’d write (for Q having local charts to Ên+1 ), A = A0 dx0 + A1 dx1 + . . . An dxn

38

Lagrangians and Noether’s Theorem

and then because d2 = 0 dA = dA0 dx0 + dA1 dx1 + . . . dAn dxn and since the “Ai ” are just functions, dAi = ∂µ Ai dxµ using the summation convention and ∂µ := ∂/∂xµ . The student can easily check that the components for F = F01 dx0 ∧ dx1 + F02 dx0 ∧ dx2 + . . ., agrees with the matrix expression below (at least for 4D). So, for example, in 4D spacetime 0 E1 E2 E3 −E1 0 B3 −B2 F = −E2 −B3 0 B1 −E3 B2 −B1 0 where E is the electric ﬁeld and B is the magnetic ﬁeld. The action for a particle of charge e is

t1

S=m

t0

q dt + e ˙

q

A

here

t1

q dt = proper time, ˙

t0

**A = integral of A along the path q.
**

q

Note that since A is a 1-form it can be integrated (it is a linear combination of some basis 1-forms like the {dxi }). (Week 4, April 18, 20, 22.) Note that since A is a 2-form we can integrate it over an oriented manifold, but one can also write the path integral using time t as a parameter, with Ai q i dt the diﬀerential, ˙ i i after dq = q dt. ˙ The Lagrangian in the above action, for a charge e with mass m in an electromagnetic potential A is L(q, q) = m q + eAi q i ˙ ˙ ˙ (3.5) so we can work out the Euler-Lagrange equations: pi = qi ˙ ∂L =m + eAi i ∂q ˙ q ˙ = mvi + e Ai

3.7 Alternative Lagrangians

39

where v ∈ Ên+1 is the velocity, normalized so that v = 1. Note that now momentum is no longer mass times velocity! That’s because we’re in n + 1-d spacetime, the momentum is an n + 1-vector. Continuing the analysis, we ﬁnd the force Fi = ∂ ∂L = i e Aj q j ˙ i ∂q ∂q ∂Aj = e j qj ˙ ∂q

So the Euler-Lagrange equations say (noting that Ai = Aj q(t) : p=F ˙ d ∂Aj mvi + eAi = e i q j ˙ dt ∂q ∂Aj dAi dvi = e i qj − e ˙ m dt ∂q dt dvi ∂Aj ∂Ai m = e i qj − e j qj ˙ ˙ dt ∂q ∂q ∂Aj ∂Ai =e − j qj ˙ i ∂q ∂q the term in parentheses is F ij = the electromagnetic ﬁeld, F = dA. So we get the following equations of motion m dvi = eF ij q j , ˙ dt (Lorentz force law) (3.6)

(Usually called the “Lorenz” force law.)

3.7

Alternative Lagrangians

We’ll soon discuss a charged particle Lagrangian that is free of the reparameterization symmetry. First a paragraph on objects other than point particles!

3.7.1

Lagrangian for a String

So we’ve looked at a point particle and tried S = m · (arclength) + A

40

Lagrangians and Noether’s Theorem

or with ‘proper time’ instead of ‘arclength’, where the 1-from A can be integrated over a 1-dimensional path. A generalization (or specialization, depending on how you look at it) would be to consider a Lagrangian for an extended object. In string theory we boost the dimension by +1 and consider a string tracing out a 2D surface as time passes (Fig. 3.3).

becomes

Figure 3.3: Worldtube of a closed string. Can you infer an appropriate action for this system? Remember, the physical or physico-philosophical principle we’ve been using is that the path followed by physical objects minimizes the “activity” or “aliveness” of the system. Given that we presumably cannot tamper with the length of the closed string, then the worldtube quantity analogous to arclength or proper time would be the area of the worldtube (or worldsheet for an open string). If the string is also assumed to be a source of electromagnetic ﬁeld then we need a 2-form to integrate over the 2D worldtube analogous to the 1-form integrated over the pathline of the point particle. In string theory this is usually the “Kalb-Ramond ﬁeld”, call it B. To recover electrodynamic interactions it should be antisymmetric like A, but it’s tensor components will have two indices since it’s a 2-form. The string action can then be written S = α · (area) + e B (3.7)

We’ve also replaced the point particle mass by the string tension α [mass·length−1 ] to obtain the correct units for the action (since replacing arclength by area meant we had to compensate for the extra length dimension in the ﬁrst term of the above string action). This may still seem like we’ve pulled a rabbit out of a hat. But we haven’t checked that this action yields sensible dynamics yet! But supposing it does, then would it justify our guesswork and intuition in arriving at Eq.(3.7)? Well by now you’ve probably realized that one can have more than one form of action or Lagrangian that yields the same dynamics. So provided we supply reasonabe physically realistic heuristics then whatever Lagrangian or action that we come up with will stand a good chance of describing some system with a healthy measure of physical verisimilitude.

3.7 Alternative Lagrangians

41

That’s enough about string for now. The point was to illustrate the type of reasoning that one can use in conjuring up a Lagrangian. It’s particularly useful when Newtonian theory cannot give us a head start, i.e., in relativistic dynamics and in the physics of extended particles.

3.7.2

Alternate Lagrangian for Relativistic Electrodynamics

In § 3.6, Eq.(3.5) we saw an example of a Lagrangian for relativistic electrodynamics that had awkward reparametrization symmetries, meaning that H = 0 and there were non-unique solutions to the Euler-Lagrange equations arising from applying gauge transformations. This freedom to change the gauge can be avoided. Recall Eq.(3.5), which was a Lagrangian for a charged particle with reparametrization symmetry L = m q + eAi q i ˙ ˙ just as for an uncharged relativistic particle. But there’s another Lagrangian we can use that doesn’t have this gauge symmetry: 1 L = mq · q + eAi q i ˙ ˙ ˙ 2 This one even has some nice features.

1 • It looks formally like “ 2 mv 2 ”, familiar from nonrelativistic mechanics.

(3.8)

• There’s no ugly square root, so it’s everywhere diﬀerentiable, and there’s no trouble with paths being timelike or spacelike in direction, they are handled the same. What Euler-Lagrange equations does this Lagrangian yield? ∂L = mqi + eAi ˙ ∂qi ˙ ∂Aj ∂L ˙ Fi = i = e i q j ∂q ∂q pi = Very similar to before! The E-L eqns. then say d ∂Aj mqi + eAi = e i q j ˙ ˙ dt ∂q m¨i = eF ij q j q ˙ almost as before. (I’ve taken to using F here for the electromagnetic ﬁeld tensor to avoid clashing with F for the generalized force.) The only diﬀerence is that we have m¨i instead q of mvi where vi = qi / q . So the old Euler-Lagrange equations of motion reduce to the ˙ ˙ ˙

42

Lagrangians and Noether’s Theorem

new ones if we pick a parametrization with q = 1, which would be a parametrization ˙ by proper time for example. Let’s work out the Hamiltonian for this 1 L = mq · q + eAi q i ˙ ˙ ˙ 2 for the relativistic charged particle in an electromagnetic ﬁeld. Recall that for our reparametrization-invariant Lagrangian L=m qi q i + eAi q i ˙ ˙ we got H = 0, time translation was a gauge symmetry. With the new Lagrangian it’s not! Indeed H = pi q i − L ˙ and now pi = so

1 H = (mqi + eAi )q i − ( 2 mqi q i + eAi q i ) ˙ ˙ ˙ ˙ ˙

∂L = mq i + eAi ˙ ∂qi ˙

˙ ˙ = 1 mqi q i 2 Comments. This is vaguely like how a nonrelativistic particle in a potential V has H = pi q i − L = 2K − (K − V ) = K + V, ˙ but now the “potential’ V = eAi q i in linear in velocity, so now ˙ H = pi q i − L = (2K − V ) − (K − V ) = K. ˙ As claimed H is not zero, and the fact that it’s conserved says q(t) is constant as ˙ a function of t, so the particle’s path is parameterized by proper time up to rescaling of t. That is, we’re getting “conservation of speed” rather than some more familiar “conservation of energy”. The reason is that this Hamiltonian comes from the symmetry qs (t) = q(t + s) instead of spacetime translation symmetry qs (t) = q(t) + s w, w ∈ Ên+1 the diﬀerence is illustrated schematically in Fig. 3.4. Our Lagrangian L(q, q) = 1 m q 2 + Ai (q)q i ˙ ˙ ˙ 2 has time translation symmetry iﬀ A is translation invariant (but it’s highly unlikely a given system of interest will have A(q) = A(q + sw)). In general then there’s no conserved “energy” for our particle corresponding to translations in time.

3.8 The General Relativistic Particle

43

4 5 3 4 2 3 1 2 0 1

w

qs (t) = q(t + s)

qs (t) = q(t) + sw

Figure 3.4: Proper time rescaling vs spacetime translation.

3.8

The General Relativistic Particle

In GR spacetime, Q, is an (n + 1)-dimensional Lorentzian manifold, namely a smooth (n + 1)-dimensional manifold with a Lorentzian metric g. We deﬁne the metric as follows. 1. For each x ∈ Q, we have a bilinear map g(x) : Tx Q × Tx Q −→ Ê (v, w) −→ g(x)(v, w) or we could write g(v, w) for short. 2. With respect to some basis of Tx Q we have g(v, w) = gij v i w j 1 0 ... 0 0 −1 0 gij = . . .. . . . . . 0 0 . . . −1

Of course we can write g(v, w) = gij v i w j in any basis, but for diﬀerent bases gij will have a diﬀerent form. 3. g(x) varies smoothly with x.

44

Lagrangians and Noether’s Theorem

3.8.1

**Free Particle Lagrangian in GR
**

L(q, q) = m ˙ =m g(q)(q, q) ˙ ˙ gij q i q j ˙ ˙

The Lagrangian for a free point particle in the spacetime Q is

just like in special relativity but with ηij replaced by gij . Alternatively we could just as well use L(q, q) = 1 mg(q)(q, q) ˙ ˙ ˙ 2

1 ˙ ˙ = 2 mgij q i q j

The big diﬀerence between these two Lagrangians is that now spacetime translation symmetry (and rotation, and boost symmetry) is gone! So there is no conserved energymomentum (nor angular momentum, nor velocity of center of energy) anymore! Let’s ﬁnd the equations of motion. Suppose then Q is a Lorentzian manifold with metric g and L : T Q → Ê is the Lagrangian of a free particle, L(q, q) = 1 mgij q i q j ˙ ˙ ˙ 2 We ﬁnd equations of motion from the Euler-Lagrange equations, which in this case start from ∂L ˙ pi = i = mgij q j ∂q ˙ The velocity q here is a tangent vector, the momentum p is a cotangent vector, and we ˙ need the metric to relate them, via g : Tq M × Tq M −→ Ê (v, w) −→ g(v, w) which gives Tq M −→ Tq∗ M v −→ g(v, ). In coordinates this would say that the tangent vector v i gets mapped to the cotangent vector gij v j . This is lurking behind the passage from q i to the momentum mgij v j . ˙ Getting back to the E-L equations, ∂L = mgij q j ˙ i ∂q ˙ ∂ 1 ∂L ˙ ˙ Fi = i = i 2 mgjk (q)q j q k ∂q ∂q pi = ˙ ˙ = 1 m∂i gik q j q k , 2 (where ∂i = ∂ ). ∂q i

3.8 The General Relativistic Particle

45

So the Euler-Lagrange equations say d mgij q j = 1 m∂i gik q j q k . ˙ ˙ ˙ 2 dt The mass factors away, so the motion is independent of the mass! Essentially we have a geodesic equation. We can rewrite this geodesic equation as follows d 1 gij q j = 2 ∂i gik q j q k ˙ ˙ ˙ dt ˙ ˙ ∂k gij q k q j + gij q j = 1 ∂i gjk q j q k ˙ ˙ ¨ 2 ∴ gij q j = ¨ =

1 2 1 ∂g 2 i jk

∴

− ∂k gij q j q k ˙ ˙

∂i gjk − ∂k gij − ∂j gki q j q k ˙ ˙

where the last line follows by symmetry of the metric, gik = gki . Now let, Γijk = − ∂i gjk − ∂k gij − ∂j gki the minus sign being just a convention (so that we agree with everyone else). This deﬁnes what we call the Christoﬀel symbols Γi . Then jk qi = gij q j = −Γijk q j q k ¨ ¨ ˙ ˙ ∴ q i = −Γi q j q k . ¨ jk ˙ ˙ So we see that q can be computed in terms of q and the Christoﬀel symbols Γi , which ¨ ˙ jk is really a particular type of connection that a Lorentzian manifold has (the Levi-Civita connection), a connection is just the rule for parallel transporting tangent vectors around the manifold. Parallel transport is just the simplest way to compare vectors at diﬀerent points in the manifold. This allows us to deﬁne, among other things, a covariant derivative.

3.8.2

Charged particle in EM Field in GR

We can now apply what we’ve learned in consideration of a charged particle, of charge e, in an electromagnetic ﬁeld with potential A, in our Lorentzian manifold. The Lagrangian would be 1 ˙ ˙ ˙ L = 2 mgij q j q k + eAi q i which again was conjured up be replacing the ﬂat space metric ηij by the metric for GR gij . Not surprisingly, the Euler-Lagrange equations then yield the following equations of motion, m¨i = −mΓijk q j q k + eF ij q j . q ˙ ˙ ˙ If you want to know more about Lagrangians for general relativity we recommend the paper by Peldan [Pel94], and also the “black book” of Misner, Thorne & Wheeler [WTM71].

46

Lagrangians and Noether’s Theorem

3.9

The Principle of Least Action and Geodesics

(Week 4, April 18, 20, 22.)

3.9.1

Jacobi and Least Time vs Least Action

We’ve mentioned that Fermat’s principle of least time in optics is analogous to the principle of least action in particle mechanics. This analogy is strange, since in the principle of least action we ﬁx the time interval q : [0, 1] → Q. Also, if one imagines a force on a particle resulting from a potential gradient at an interface as analogous to light refraction then you also get a screw-up in the analogy (Fig. 3.5).

n high

light faster light slower

particle slower particle faster

V high

n low

V low

Figure 3.5: Least time versus least action. Nevertheless, Jacobi was able to reinterpret the mechanics of a particle as an optics problem and hence “unify” the two minimization principles. First, let’s consider light in a medium with a varying index of refraction n (recall 1/n ∝speed of light). Suppose it’s in Ên with its usual Euclidean metric. If the light s trying to minimize the time, its trying to minimize the arclength of its path in the metric gij = n2 δij that is, the index of refraction n : Ên → (0, ∞), times the usual Euclidean metric 1 .. .

δij =

0

0

1

3.9 The Principle of Least Action and Geodesics

47

This is just like the free particle in general relativity (minimizing it’s proper time) except that now gij is a Riemannian metric g(v, w) = gij v i w j where g(v, v) ≥ 0 So we’ll use the same Lagrangian, L(q, q) = ˙ and get the same Euler-Lagrange equations, d2 q i + Γi q j q k jk ˙ ˙ dt2 if q is parameterized by arclength or more generally q = ˙ gij (q)q i q j = constant. ˙ ˙ (3.9) gij (q)q iq j ˙ ˙

As before the Christoﬀel symbols Γ are built from the derivatives of the metric g). Now, what Jacobi did is show how the motion of a particle in a potential could be viewed as a special case of this. Consider a particle of mass m in Euclidean Ên with potential V : Ên → Ê. It satisﬁes F = ma, i.e., m d2 q i = −∂i V dt2 (3.10)

How did Jacobi see (3.10) as a special case of (3.9)? He considered a particle of energy E and he chose the index of refraction to be n(q) = 2 E − V (q) m

**which is just the speed of a particle of energy E when the potential energy is V (q), since 2 (E − V ) = m 21 m q ˙ m2
**

2

= q . ˙

Note: this is precisely backwards compared to optics, where n(q) is proportional to the reciprocal of the speed of light!! But let’s see that it works. L= = = gij (q)q i q j ˙ ˙ n2 (q)q i q j ˙ ˙ 2/m(E − V (q))q 2 ˙

48

Lagrangians and Noether’s Theorem

where q 2 = q · q is just the usual Euclidean dot product, v · w = δij v i w j . We get the ˙ ˙ ˙ Euler-Lagrange equations, pi = Fi = ∂L = ∂qi ˙ ∂L = ∂i ∂q i 1 = 2 q ˙ 2 (E − V ) · m q ˙ 2 (E − V (q)) · q ˙ m −2/m∂i V · q ˙ 2/m(E − V q)

Then p = F says, ˙ d dt 2/m(E − V (q)) · qi ˙ 1 = − ∂i V q ˙ m q ˙ 2/m(E − V )

Jacobi noticed that this is just F = ma, or m¨i = −∂i V , that is, provided we reparameq terize q so that, q = 2/m(E − V (q)). ˙ Recall that our Lagrangian gives reparameterization invariant Euler-Lagrange equations! This is the uniﬁcation between least time (from optics) and least action (from mechanics) that we sought.

3.9.2

The Ubiquity of Geodesic Motion

We’ve seen that many classical systems trace out paths that are geodesics, i.e., paths q : [t0 , t1 ] → Q that are critical points of

t1

S(q) =

t0

gij q i q j dt ˙ ˙

which is proper time when (Q, g) is a Lorentzian manifold, or arclength when (Q, g) is a Riemannian manifold. We have 1. The metric at q ∈ Q is, g(q) : Tq Q × Tq Q → Ê (v, w) → g(v, w) and it is bilinear. 2. w.r.t a basis of Tq Q g(v, w) = δij v i w j

3.9 The Principle of Least Action and Geodesics

49

3. g(q) varies smoothly with q ∈ Q. An important distinction to keep in mind is that Lorentzian manifolds represent spacetimes, whereas Riemannian manifolds represent that we’d normally consider as just space. We’ve seen at least three important things. (1) In the geometric optics approximation, light in Q = Ên acts like particles tracing out geodesics in the metric gij = n(q)2 δij where n : Q → (0, ∞) is the index of refraction function. (2) Jacobi saw that a particle in Q = Ên in some potential V : Q → Ê traces out geodesics in the metric 2 gij = (E − V )δij m 1 if the particle has energy E (where V < E). (3) A free particle in general relativity traces out a geodesic on a Lorentzian manifold (Q, q). In fact all three of these results can be generalized to cover every problem that we’ve discussed! (1′ ) Light on any Riemannian manifold (Q, q) with index of refraction n : Q → (0, ∞) traces out geodesics in the metric h = n2 g. (2′ ) A particle on a Riemannian manifold (Q, q) with potential V : Q → Ê traces out geodesics w.r.t the metric 2 h = (E − V )g m if it has energy E. Lots of physical systems can be described this way, e.g., the Atwood machine, a rigid rotating body (Q = SO(3)), spinning tops, and others. All of these systems have a Lagrangian which is a quadratic function of position, so they all ﬁt into this framework. (3′ ) Kaluza-Klein Theory. A particle with charge e on a Lorentzian manifold (Q, q) in an electromagnetic vector potential follows a path with e qi = −Γijk q j q k + Fij q j ¨ ˙ ˙ ˙ m where Fij = ∂i Aj − ∂j Ai but this is actually geodesic motion on the manifold Q×U(1) where U(1) = {eiθ : θ ∈ Ê} is a circle.

1

The case V > E, if they exist, would be classically forbidden regions.

50

Lagrangians and Noether’s Theorem

Let’s examine this last result a bit further. To get the desired equations for motion on Q × U(1) we need to given Q × U(1) a cleverly designed metric built from g and A where the amount of “spiralling”—the velocity in the U(1) direction is e/m. The metric h on

Q × U(1) a geodesic Q the apparent path Q × U(1) is built from g and A in a very simple way. Let’s pick coordinates xi on Q where i ∈ {0, . . . , n} since we’re in n + 1-dimensional spacetime, and θ is our local coordinate on S 1 . The components of h are hij = gij + Ai Aj hθi = hθi = −Ai hθθ = 1 Working out the equations for a geodesic in this metric we get qi = −Γijk q j q k + ¨ ˙ ˙ qθ = 0, ¨ if qθ = e/m ˙ since Fij is part of the Christoﬀel symbols for h. To summarize this section on least time versus least action we can say that every problem that we’ve discussed in classical mechanics can be regarded as geodesic motion! e Fij q j ˙ m

**Chapter 4 From Lagrangians to Hamiltonians
**

In the Lagrangian approach we focus on the position and velocity of a particle, and compute what the particle does starting from the Lagrangian L(q, q), which is a function ˙ L : T Q −→ Ê where the tangent bundle is the space of position-velocity pairs. But we’re led to consider momentum ∂L pi = i ∂q ˙ since the equations of motion tell us how it changes dpi ∂L = i. dt ∂q

4.1

The Hamiltonian Approach

In the Hamiltonian approach we focus on position and momentum, and compute what the particle does starting from the energy H = pi q i − L(q, q) ˙ ˙ reinterpreted as a function of position and momentum, called the Hamiltonian H : T ∗ Q −→ Ê where the cotangent bundle is the space of position-momentum pairs. In this approach, position and momentum will satisfy Hamiltons equations: ∂H dq i = , dt ∂pi 51 dpi ∂H =− dt ∂qi

52

From Lagrangians to Hamiltonians

where the latter is the Euler-Lagrange equation dpi ∂L =− dt ∂qi in disguise (it has a minus sign since H = pq − L). ˙ To obtain this Hamiltonian description of mechanics rigorously we need to study this map λ : T Q −→ T ∗ Q (q, q) −→ (q, p) ˙ where q ∈ Q, and q is any tangent vector in Tq Q (not the time derivative of something), ˙ and p is a cotangent vector in Tq∗ Q := (Tq Q)∗ , given by q −→ pi = ˙

λ

∂L ∂qi ˙

So λ is deﬁned using L : T Q → Ê. Despite appearances, λ can be deﬁned in a coordinate∂L free way, as follows (referring to Fig. 4.1). We want to deﬁne “ ∂ q˙i ” in a coordinate-free

Tq Q TQ T(q,q) Tq Q ˙

Tq Q q Figure 4.1: way; it’s the “diﬀerential of L in the vertical direction”—i.e., the q i directions. We have ˙ π : T Q −→ Q (q, q) −→ q ˙ and dπ : T (T Q) −→ T Q Q

4.1 The Hamiltonian Approach

53

**has kernel1 consisting of vertical vectors: V T Q = ker dπ ⊆ T T Q The diﬀerential of L at some point (q, q) ∈ T Q is a map from T T Q to Ê, so we have ˙
**

∗ (dL)(q,q) ∈ T(q,q) T Q ˙ ˙

that is, dL(q,q) : T(q,q) T Q −→ Ê. ˙ ˙ We can restrict this to V T Q ⊆ T T Q, getting f : V(q,q) T Q −→ Ê. ˙ But note V(q,q) T Q = T (Tq Q) ˙ and since Tq Q is a vector space, T(q,q) Tq Q ∼ Tq Q = ˙ in a canonical way2 . So f gives a linear map p : Tq Q −→ Ê that is, p∈ this is the momentum! (Week 6, May 2, 4, 6.) Given L : T Q → T ∗ Q, we now know a coordinate-free way of describing the map λ : T Q −→ T ∗ Q (q, q) −→ (q, p) ˙ given in local coordinates by pi =

1

Tq Q V(q,q) T Q ˙ T(q,q) Tq Q ˙

Tq∗ Q

q

Q

∂L . ∂qi ˙

The kernel of a map is the set of all elements in the domain that map to the null element of the range, so ker dπ = {v ∈ T T Q : dπ(v) = 0 ∈ T Q}. 2 The ﬁber Tv V at v ∈ V of a vector manifold V has the same dimension as V .

54

From Lagrangians to Hamiltonians

We say L is regular if λ is a diﬀeomorphism from T Q to some open subset X ⊆ T ∗ Q. In this case we can describe what our system is doing equally well by specifying position and velocity, (q, q) ∈ T Q ˙ or position and momentum (q, p) = λ(q, q) ∈ X. ˙ We call X the phase space of the system. In practice often X = T ∗ Q, then L is said to be strongly regular.

4.2

Regular and Strongly Regular Lagrangians

This section discusses some examples of the above theory.

4.2.1

Example: A Particle in a Riemannian Manifold with Potential V (q)

For a particle in a Riemannian manifold (Q, q) in a potential V : Q → Ê has Lagrangian 1 L(q, q) = mgij q i q j − V (Q) ˙ ˙ 2 Here pi = so λ(q, q) = q, mg(q, −) ˙ ˙ so3 L is strongly regular in this case because Tq Q −→ Tq∗ Q v −→ g(v, −) is 1-1 and onto, i.e., the metric is nondegenerate. Thus λ is a diﬀeomorphism, which in this case extends to all of T ∗ Q.

The missing object there “−” is of course any tangent vector, not inserted since λ itself is an operator on tangent vectors, not the result of the operation.

3

∂L = mij q j ˙ ∂qi ˙

4.2 Regular and Strongly Regular Lagrangians

55

4.2.2

Example: General Relativistic Particle in an E-M Potential

For a general relativistic particle with charge e in an electromagnetic vector potential A the Lagrangian is 1 L(q, q) = mgij q i q j − eAi q i ˙ ˙ ˙ 2 and thus ∂L pi = i = mgij q i q j + eAi . ˙ ∂q ˙ This L is still strongly regular, but now each map λ |Tq Q : Tq Q −→ Tq∗ Q q −→ m g(q, −) + eA(q) ˙ ˙ is aﬃne rather than linear4 .

4.2.3

Example: Free General Relativistic Particle with Reparameterization Invariance

The free general relativistic particle with reparameterization invariant Lagrangian has, L(q, q) = m ˙ gij q i q j . ˙ ˙

This is terrible from the perspective of regularity properties—it’s not diﬀerentiable when gij q i q j vanishes, and undeﬁned when the same is negative. Where it is deﬁned ˙ ˙ pi = mgij q j ˙ ∂L = i ∂q ˙ q ˙

(where q is timelike), we can ask about regularity. Alas, the map λ is not 1-1 where deﬁned ˙ since multiplying q by some number has no eﬀect on p! (This is related to the reparameter˙ ization invariance—this always happens with reparameterization-invariant Lagrangians.)

4.2.4

Example: A Regular but not Strongly Regular Lagrangian

**Here’s a Lagrangian that’s regular but not strongly regular. Let Q = Ê and L(q, q) = f (q) ˙ ˙
**

All linear transforms are aﬃne, but aﬃne transformations include translations, which are nonlinear. In aﬃne geometry there is no deﬁned origin. For the example the translation is the “+eA(q)” part.

4

56

From Lagrangians to Hamiltonians

∂L = f ′ (q) ˙ ∂q ˙ This will be regular if f ′ : Ê → Ê is a diﬀeomorphism from Ê to some proper subset U ⊂ Ê, ∼ but not strongly regular. For example, take f (q) = eq˙ so f ′ : Ê → (0, ∞) ⊂ Ê. So ˙ p= positive slope L(q, q) = eq˙ ˙

G

so that

**or slope between −1 and 1 L(q, q) = ˙ and so forth. 1 + q2 ˙
**

G

4.3

Hamilton’s Equations

λ : T Q −→ X ⊆ T ∗ Q (q, q) −→ (q, p) ˙

∼

Now let’s assume L is regular, so

This lets us have the best of both worlds: we can identify T Q with X using λ. This lets us treat q i , pi , L, H, etc., all as functions on X (or T Q), thus writing qi ˙ for the function q i ◦ λ−1 ˙ In particular (function on X) (function on T Q)

∂L (Euler-Lagrange eqn.) ∂q i which is really a function on T Q, will be treated as a function on X. Now let’s calculate: pi := ˙ dL = ∂L i ∂L i dq + i dq ˙ ∂q i ∂q ˙ = pi dq i + pi dq i ˙ ˙

4.3 Hamilton’s Equations

57

while dH = d(pi q i − L) ˙ = q i dpi + pi dq i − L ˙ ˙ = q i dpi + pi dq i − (pi dq i + pi dq i ) ˙ ˙ ˙ ˙ = q i dpi − pi dq i ˙ ˙ ˙ so dH = q i dpi − pi dqi . ˙ ˙ Assume the Lagrangian L : T Q → Ê is regular, so λ : T Q −→ X ⊆ T ∗ Q (q, q) −→ (q, p) ˙ is a diﬀeomorphism. This lets us regard both L and the Hamiltonian H = pi q i − L as ˙ i i functions on the phase space X, and use (q , q ) as local coordinates on X. As we’ve seen, ˙ this gives us dL = pi dq i + pi dq i , ˙ ˙ dH = q i dpi − pi dq i . ˙ ˙ But we can also work out dH directly, this time using local coordinates (q i , pi ), to get dH = ∂H ∂H dpi + i dq i . i ∂p ∂q

∼

Since dpi , dq i form a basis of 1-forms, we conclude: qi = ˙ These are Hamilton’s Equations. ∂H , ∂pi pi = − ˙ ∂H ∂qi

4.3.1

Hamilton and Euler-Lagrange

Though q i and pi are just functions of X, when the Euler-Lagrange equations hold for ˙ ˙ some path q : [t0 , t1 ] → Q, they will be the time derivatives of q i and pi . So when the Euler-Lagrange equations hold, Hamilton’s equations describe the motion of a point x(t) = q(t), p(t) ∈ X. In fact, in this context, Hamilton’s equations are just the EulerLagrange equations in disguise. The equation qi = ˙ ∂H ∂pi

58

From Lagrangians to Hamiltonians

really just lets us recover the velocity q as a function of q and p, inverting the formula ˙ pi = ∂L ∂qi ˙

which gave p as a function of q and q. So we get a formula for the map ˙ λ−1 : X −→ T Q (q, p) −→ (q, q). ˙ Given this, the other Hamilton equation pi = − ˙ is secretly the Euler-Lagrange equation d ∂L ∂L = i, i dt ∂ q ˙ ∂q These are the same because ∂H ∂ ∂L = i pi q i − L = − i . ˙ ∂q i ∂q ∂q Example: Particle in a Potential V (q) For a particle in Q = Ên in a potential V : Ên → Ê the system has Lagrangian m L(q, q) = ˙ q 2 − V (q) ˙ 2 which gives p = mq ˙ p q= , ˙ m and Hamiltonian H(q, p) = pi q i − L = ˙ 1 p 2 p 2− −V m 2m 1 p 2 + V (q). = 2m p m g ij pj ) m or p= ˙ ∂L . ∂q i ∂H ∂q i

(though really that’s q = ˙

So Hamilton’s equations say ∂H ∂pi ∂H pi = − i ˙ ∂q qi = ˙ ⇒ ⇒ q= ˙

p = −∇V ˙

The ﬁrst just recovers q as a function of p; the second is F = ma. ˙

4.3 Hamilton’s Equations

59

Note on Symplectic Structure Hamilton’s equations push us toward the viewpoint where p and q have equal status as coordinates on the phase space X. Soon, we’ll drop the requirement that X ⊆ T ∗ Q where Q is a conﬁguration space. X will just be a manifold equipped with enough structure to write down Hamilton’s equations starting from any H : X → Ê. The coordinate-free description of this structure is the major 20th century contribution to mechanics: a symplectic structure. This is important. You might have some particles moving on a manifold like S 3 , which is not symplectic. So the Hamiltonian mechanics point of view says that the abstract manifold that you are really interesting in is something diﬀerent, it must be a symplectic manifold. That’s the phase space X. We’ll introduce symplectic geometry more completely in later chapters.

4.3.2

Hamilton’s Equations from the Principle of Least Action

Before, we obtained the Euler-Lagrange equations by associating an “action” S with any q : [t0 , t1 ] → Q and setting δS = 0. Now let’s get Hamilton’s equations directly by assigning an action S to any path x : [t0 , t1 ] → X and setting δS = 0. Note: we don’t impose any relation between p and q, q at all! The relation will follow from δS = 0. ˙ Let P be the space of paths in the phase space X and deﬁne the action S : P −→ Ê by

t1

S(x) =

t0

(pi q i − H)dt ˙

**where pi q i − H = L. More precisely, write our path x as x(t) = q(t), p(t) and let ˙
**

t1

S(x) =

t0

pi (t)

d i q (t) − H q(t), p(t) dt

dt

d we write dt q i instead of q i to emphasize that we mean the time derivative rather than a ˙ coordinate in phase space. Let’s show δS = 0 ⇔Hamilton’s equations.

δS = δ =

(pi q i − H)dt ˙ δpi q i + pi δ q i − δH dt ˙ ˙

60

From Lagrangians to Hamiltonians

then integrating by parts, δS = = = δpi q i − pi δq i − δH dt ˙ ˙ δpi q i − pi δq i − ˙ ˙ ∂H i ∂H δq − δpi dt ∂q i ∂pi ∂H ∂H δpi q i − ˙ + δq i −pi − i ˙ dt ∂pi ∂q ∂H , ∂pi ∂H ∂q i

This vanishes ∀δx = (δq, δp) if and only if Hamilton’s equations qi = ˙ pi = −

hold. Just as we hoped. We’ve now seen two principles of “least action”: 1. For paths in conﬁguration space Q, δS = 0 ⇔Euler-Lagrange equations. 2. For paths in phase space X, δS = 0 ⇔Hamilton’s equations. Additionally, since X ⊆ T ∗ Q, we might consider a third version based on paths in position-velocity space T Q. But when our Lagrangian is regular we have a diﬀeomor∼ phism λ : T Q → X, so this third principle of least action is just a reformulation of principle 2. However, the really interesting principle of least action involves paths in the extended phase space, X × Ê, where we have an additional coordinate for time. Recall the action S(x) = = = (pi q i − H) dt ˙ pi dq i dt − H dt dt

pi dq i − H dt

We can interpet the integrand as a 1-form β = pi dq i − H dt on X × Ê, which has coordinates {pi , q i , t}. So any path x : [t0 , t1 ] −→ X gives a path σ : [t0 , t1 ] −→ X × Ê t −→ (x(t), t)

4.4 Waves versus Particles—The Hamilton-Jacobi Equations

61

**and the action becomes the integral of a 1-form over a curve: S(x) = pi dq i − H dt =
**

σ

β.

“Why is this more interesting?” you may ask, the answer is, “just read onwards”.

4.4

Waves versus Particles—The Hamilton-Jacobi Equations

(Week 7, May 9, 11, 13.) In quantum mechanics we discover that every particle—electrons, photons, neutrinos, etc.—is a wave, and vice versa. Interestingly Newton already had a particle theory of light (his “corpuscules”) and various physicists argued against it by pointing out that diﬀraction is best explained by a wave theory. We’ve talked about geometrized optics, an approximation in which light consists of particles moving along geodesics. Here we start with a Riemannian manifold (Q, g) as space, but we use the new metric hij = n2 gij where n : Q → (0, ∞) is the index of refraction throughout space (generally not a constant).

4.4.1

Wave Equations

Huygens considered this same setup (in simpler language) and considered the motion of a wavefront:

G

62

From Lagrangians to Hamiltonians

and saw that the wavefront is the envelope of a bunch of little wavelets centered at points along the big wavefront:

balls of radius ǫ centered at points of the old wavefront

1$%Ð×ÖÑ&(' 0ÕÒ) ÓÔ 1$%Ð×ÖÑ&(' Ô 0ÕÒ) ÓÔ 1$%Ð×ÖÑÕÒ&(' 0) ÓÔ 1$%Ð×ÖÑ&(' 0ÕÒ) ÓÔ 1$%Ð×ÖÑ&(' 0ÕÒ) ÓÔ 1$%Ð×ÖÑÕÒ&(' 0) ÓÔ 1$%Ð×ÖÑÕÒ&(' 0) ÓÔ 1$%Ð×ÖÑÕÒ&(' 0) ÓÔ 1$%Ð×ÖÑ&(' 0ÕÒ) ÓÔ 1$%Ð×ÖÑ&(' 0ÕÒ) ÓÔ 1$%Ð×ÖÑÕÒ&(' 0) ÓÔ 1$%Ð×ÖÑ&(' 0ÕÒ) ÓÔ 1$%Ð×ÖÑ&(' 0ÕÒ) ÓÔ

In short, the wavefront moves at unit speed in the normal direction with respect to the “optical metric” h. We can think about the distance function d : Q × Q −→ [0, ∞) on the Riemannian manifold (Q, h), where d(q0 , q1 ) = inf (arclength)

Υ

where Υ = {paths from q0 to q1 }. (Secretly this d(q0 , q1 ) is the least action—the inﬁmum of action over all paths from q0 to q1 .) Using this we get the wavefronts centered at q0 ∈ Q as the level sets {q : d(q0 , q) = c} or rather this is so for small c > 0, as depicted in Fig. 4.2. For larger c the level sets can

q0

Figure 4.2: cease to be smooth—we say a catastrophe occurs—and then the wavefronts are no longer the level sets. This sort of situation can happen for topological reasons (Fig. 4.3) or it can happen for geometrical reasons (Fig. 4.4). Assuming no such catastrophes occur, we

4.4 Waves versus Particles—The Hamilton-Jacobi Equations

63

level set of d(q0 , −) Figure 4.3:

Figure 4.4: can approximate the waves of light by a wavefunction: ψ(q) = A(q)eik d(q,q0 ) where k is the wavenumber of the light (i.e., its color) and A : Q → Ê describes the amplitude of the wave, which drops oﬀ far from q0 . This becomes the eikonal approximation in optics5 once we ﬁgure out what A should be. Hamilton and Jacobi focused on distance d : Q × Q → [0, ∞) as a function of two variables and called it W =Hamilton’s principle function. They noticed,

•

∂ W (q0 , q1 ) = (p1 )i , i ∂q1

q1 q0

•

p1

G

**where p1 is a cotangent vector pointing normal to the wavefronts.
**

Eikonal comes form the Greek word for ‘image’ or ‘likeness’, in optics the eikonal approximation is the basis for ray tracing methods.

5

64

From Lagrangians to Hamiltonians

4.4.2

The Hamilton-Jacobi Equations

We’ve seen that in optics, particles of light move along geodesics, but wavefronts are level sets of the distance functions:

qqq qqVq qqq

at least while the level sets remain smooth. In the eikonal approximation, light is described by waves ψ : Q −→ ψ(q1 ) = A(q1 )eik W (q0 ,q1 ) where (Q, h) is a Riemannian manifold, h is the optical metric, q0 ∈ Q is the light source, k is the frequency and W : Q × Q −→ [0, ∞) is the distance function on Q, or Hamilton’s principle function: W (q0 , q1 ) = inf S(q)

q∈Υ

where Υ is the space of paths from q0 to q and S(q) is the action of the path q, i.e., its arclength. This is begging to be generalized to other Lagrangian systems! (At least retrospectively with the advantage of our historical perspective.) We also saw that q1 p1 q0

• •

G

∂ W (q0 , q1 ) = (p1 )i , i ∂q1

“points normal to the wavefront”—really the tangent vector pi = hij (p1 )j 1 points in this direction. In fact kp1i is the momentum of the light passing through q1 . This foreshadows quantum mechanics! (Note: in QM, the momentum is a derivative operator—we get p by diﬀerentiating the wavefunction!) Jacobi generalized this to the motion of point particles in a potential V : Q → Ê, using the fact that a particle of energy E traces out geodesics in the metric hij = 2(E − V ) gij . m

4.4 Waves versus Particles—The Hamilton-Jacobi Equations

65

We’ve seen this reduces point particle mechanics to optics—but only for particles of ﬁxed energy E. Hamilton went further, and we now can go further still. Suppose Q is any manifold and L : T Q → Ê is any function (Lagrangian). Deﬁne Hamilton’s principle function W : Q × Ê × Q × Ê −→ Ê by W (q0 , t0 ; q1 , t1 ) = inf S(q)

q∈Υ

where Υ = q : [t0 , t1 ] → Q, q(t0 ) = q0 , & q(t1 ) = q1 and S(q) =

t0 t1

L q(t), q(t) dt ˙

Now W is just the least action for a path from (q0 , t0 ) to (q1 , t1 ); it’ll be smooth if (q0 , t0 ) and (q1 , t1 ) are close enough—so let’s assume that is true. In fact, we have p (q1 , t1 ) 1 (q0 , t0 )

• •

G

∂ W (q0 , q1 ) = (p1 )i , i ∂q1

where p1 is the momentum of the particle going from q0 to q1 , at time t1 , and ∂W = −(p0 )i , (= −momentum at time t0 ) i ∂q0 ∂W = −H1 , (= −energy at time t1 ) ∂t1 ∂W = H0 , (= +momentum at time t0 ) ∂t0 (H1 = H0 as energy is conserved). These last four equations are the Hamilton-Jacobi equations. The mysterious minus sign in front of energy was seen before in the 1-form, β = pi dq i − H dt on the extended phase space X × Ê. Maybe the best way to get the Hamilton-Jacobi equations is from this extended phase space formulation. But for now let’s see how Hamilton’s principle function W and variational principles involving least action also yield the Hamilton-Jacobi equations. Given (q0 , t0 ), (q1 , t1 ), let q : [t0 , t1 ] −→ Q

66

From Lagrangians to Hamiltonians

be the action-minimizing path from q0 to q1 . Then W (q0 , t0 ; q1 , t1 ) = S(q) Now consider varying q0 and q1 a bit

q t0 t1

and thus vary the action-minimizing path, getting a variation δq which does not vanish at t0 and t1 . We get δW = δS

t1

=δ

t0 t1

L(q, q) dt ˙ ∂L i ∂L i δq δ q dt ˙ ∂q i ∂ q i ˙ ∂L i δq − pi δq i dt + pi δq i ˙ ∂q i ∂L − pi δq i dt + pi δq i ˙ ∂q i

t1 t0 t1 t0

=

t0 t1

=

t0 t1

=

t0

the term in parentheses is zero because q minimizes the action and the Euler-Lagrange equations hold. So we have i i δW = p1i δq1 − p0i δq0 and so ∂W ∂W = p1i , and = −p0i i i ∂q1 ∂q0 These are two of the four Hamilton-Jacobi equations! To get the other two, we need to vary t0 and t1 :

Now change in W will involve ∆t0 and ∆t1

•

• • •

t0 t0 + ∆t0

t1 t1 + ∆t1

4.4 Waves versus Particles—The Hamilton-Jacobi Equations

67

(you can imagine ∆t0 < 0 in that ﬁgure if you like). We want to derive the Hamilton-Jacobi equations describing the derivatives of Hamilton’s principle function W (q0 , t0 ; q1 , t1 ) = inf S(q)

q∈Υ

**where Υ is the space of paths q : [t0 , t1 ] → Q with q(t0 ) = q, q(t1 ) = q1 and
**

t1

S(q) =

t0

L(q, q) dt ˙

where the Lagrangian L : T Q → Ê will now be assumed regular, so that λT Q −→ X ⊆ T ∗ Q (q, q) −→ (q, p) ˙ is a diﬀeomorphism. We need to ensure that (q0 , t0 ) is close enough to (qi , t1 ) that there is a unique q ∈ Υ that minimizes the action S, and assume that this q depends smoothly on U = (q0 , t0 ; q1 , t1 ) ∈ (Q × Ê)2 . We’ll think of q as a function of U:

(Q × Ê)2 → Υ u→q deﬁned only when (q0 , t0 ) and (q1 , t1 ) are suﬃciently close.

q

(t0 , q0 ) •

• (t1 , q1 )

**Then Hamilton’s principle function is W (u) := W (q0 , t0 ; q1 , t1 ) = S(q)
**

t1

=

t0 t1

L(q, q) dt ˙ pq − H(q, p) dt ˙

t0 t1

= =

t0

p dq − H dt β

C

=

where β = pdq − H(q, p)dt is a 1-form on the extended phase space X × Ê, and C is a curve in the extended phase space: C(t) = q(t), p(t), t ∈ X Ê.

68

From Lagrangians to Hamiltonians

Note that C depends on the curve q ∈ Υ, which in turn depends upon u = (q0 , t0 ; q1 , t1 ) ∈ (Q × Ê)2 . We are after the derivatives of W that appear in the Hamilton-Jacobi relations, so let’s diﬀerentiate W (u) = β

C

with respect to u and get the Hamilton-Jacobi equations from β. Let us be a 1-parameter family of points in (Q × Ê)2 and work out d d W (us ) = ds ds where Cs depends on us as above β

Cs

... . . . • .. . . .. . ........ .. . . . . . . . . .. ........... Bs . .. . . . . . . ... As .................. . . ........ . . . . • . . ....... . . •

C0

D

CB s

•

X ×Ê

Let’s compare β

C0

and

As +Cs −Bs

=

As

β+

Cs

β−

Bs

β

Since C0 minimizes the action among paths with the given end-points, and the curve As + Cs − Bs has the same end-points, we get d ds β=0

As +Cs −Bs

(although As + Cs − Bs is not smooth, we can approximate it by a path that is smooth). So d d d β= β− β at s = 0. ds Cs ds Bs ds As Note d d β= β(A′r ) dr ds As ds = β(A′0 ) where A′0 = v is the tangent vector of As at s = 0. Similarly, d ds β = β(w)

Bs

4.4 Waves versus Particles—The Hamilton-Jacobi Equations

69

′ where w = B0 . So,

d W (us ) = β(w) − β(v) ds where w keeps track of the change of (q1 , p1 , t1 ) as we move Cs and v keeps track of (q0 , p0 , t0 ). Now since β = pi dqi − Hdt, we get ∂W = pi , 1 i ∂q1 ∂W = −H ∂t1 and similarly ∂W = −pi , 0 i ∂q0 ∂W = H. ∂t0

So, if we deﬁne a wavefunction: ψ(q0 , t0 ; q1 , t1 ) = eiW (q0 ,t0 ;q1 ,t1 )/ then we get ∂ψ i = − H1 ψ, ∂t1 i ∂ψ = p1 ψ i ∂q1 which at the time of Hamilton and Jacobi’s research was interesting enough, but nowadays it is thoroughly familiar from quantum mechanics!

Chapter 5 Symplectic Geometry

(Week 8, May 16, 18, 20.)

5.1

**Towards Symplectic Geometry
**

β = pi dq i − H dt

Last time we saw that the 1-form

on the extended phase space X × Ê ∋ (q, p, t) can be integrated to get the action, with least action giving “Hamilton’s principle function” W (q0 , t0 ; q1 , t1 ). Diﬀerentiating this we get momentum and energy—the HamiltonJacobi equations. These foreshadow quantum mechanics, since we can deﬁne a wavefunction ψ(q0 , t0 ; q1 , t1 ) = ei W (q0 ,t0 ;q1,t1 )/ which (approximately) gives the amplitude for a quantum particle to get from (q0 , t0 ) to (q1 , t1 )

×

**least action path
**

Ò

•

wavefunction

(q0 , t0 ) •

(q1 , t1 )

70

5.1 Towards Symplectic Geometry

71

**In full-ﬂedged quantum mechanics we instead calculate ψ exactly by ψ(q0 , t0 ; q1 , t1 ) =
**

Υ

ei S(q) Dq

where Υ = q : [t0 , t1 ] → Q | q(t0 ) = q0 & q(t1 ) = q1 and Dq makes no sense (not yet at least—except in certain special cases, like a particle in a potential on a Riemannian manifold Q). The case when ψ = eiW/ is a good approximation to the full-ﬂedged “path integral”, is precisely the case when classical mechanics is approximately right, though really ψ = eiW/ works better (the eikonal or WKB approximation). But in classical mechanics, people focus less on the extended phase space X × Ê than on the phase space X. The 1-form β doesn’t live on X, but something else does. If x : [t0 , t1 ] → X is a path, then we get C : [t0 , t1 ] → X × Ê with C(t) = (x(t), t), and then S=

C

β=

C

pi dq i − H dt

**and if x satisﬁes Hamilton’s equations, so H is conserved then, S = −E(t1 − t0 ) +
**

X

pi dq i

(doing the integral over dt leaves us just with a 1-form to integrate on X), and where E is the energy. So people focus on α = pi dq i on X. If the system executes periodic motion, we can take x to be a loop on X, and the

X

x

Σ

action is S = −E · period +

x

α dα

Σ

= −E · period +

72

Symplectic Geometry

if Σ is any surface with ∂Σ = x by Stokes’ theorem (this works provided x is a contractible loop). Note dα = dpi ∧ dq i is a 2-form on X—a lot of theorists focus on this: the symplectic structure. In the BohrSommerfeld “old quantum mechanics”, energy eigenstates correspond to periodic orbits for which dpi ∧ dq i = n2π = nh

Σ

so that qi

R

Σ

dpi ∧dq i

= 1.

In the modern “symplectic geometry” approach, people focus on the 1-form α on T ∗ Q and especially on the 2-form ω := dα. In these lectures we will follow suit and (re)develop classical mechanics starting with these objects. The point of all the extra work is to gain deeper insight through generalizations and geometrization of physics.

5.2

The Canonical Forms on the Cotangent Bundle

These forms will be the natural objects of study in classical mechanics in the Hamiltonian approach. Why? Because one can integrate forms on a manifold and, as we have seen, the Hamiltonian picture involves a symplectic manifold, the cotangent bundle T ∗ Q of the conﬁguration space Q. But we will see that symplectic geometry is a powerful tool because with it we can extend the study of classical mechanics to more general phase spaces X that are not subsets of T ∗ Q. We’ll get to this later.

5.2.1

**The Canonical 1-form on the Cotangent Bundle
**

α = pi dq i

First lets describe the 1-form, in a coordinate free way (which as usual means looking for unique mappings between appropriate spaces). Any tangent vector at (q, p) ∈ T ∗ Q is of the form v = ai ∂ ∂ + bi i ∂q ∂pi

using coordinates (q i , pi ). Note that this means v ∈ T(q,p) T ∗ Q. So α(v) = pi ai .

5.2 The Canonical Forms on the Cotangent Bundle

73

Alternatively, let π : T ∗ Q −→ Q (q, p) −→ q and we claim that α(v) = p dπ(v) ∈ Ê p ∈ Tq∗ Q dπ(v) ∈ Tq Q. So for any manifold Q (which physically is the conﬁguration space in our classical mechanics exposition) the cotangent bundle T ∗ Q has a 1-form on it: α = pi dq i in local coordinates (q i , pi) on T ∗ Q coming from local coordinates on Q. Given any tangent vector v ∈ T(q,p) T ∗ Q, we can write v = ai ∂ ∂ + bi i ∂q ∂pi

thus this eats an element (q, p) of the cotangent space T ∗ Q and spits out a tangent vector belonging to T(q,p) . Then we declare α(v) = pi ai . This α is called the canonical 1-form on T ∗ Q because it’s coordinate independent. We have π : T ∗ Q −→ Q (q, p) −→ q and thus dπ : T(q,p) T ∗ Q −→ Tq Q and in fact α(v) = p dπ(v) , ∀v ∈ T(q,p) T ∗ Q this is the coordinate-free description of α. Let’s check that this is true, i.e., let’s show that α(v) = p(dπ(v)) = pi ai . Well, if v = ai ∂ ∂ + bi i ∂q ∂pi

74

Symplectic Geometry

then dπ(v) = ai

∂ ∂q i

∈ Tq Q

∂ where now { ∂qi } is a basis of tangent vectors on Q. We can then write out

p dπ(v) = (pj dq j )(ai = pi ai ,

∂ ∂ j ), and dq i = δi i i ∂q ∂q which is exactly as required.

5.2.2

**The Symplectic 2-form on the Cotangent Bundle
**

ω = dα.

Symplectic geometry focuses not on α but on the 2-form

We’ll write down Hamilton’s equations using ω and show that ω (unlike α) is invariant under the resulting time evolution maps T ∗ Q −→ T ∗ Q (or X −→ X).

If we isolate the key properties of ω that make this work, we can generalize classical mechanics to systems where the phase space X is not T ∗ Q or some open subset of T ∗ Q. For example, a classical spinning point particle has X = S 2 :

v njf rf J ∈ vr { h `TP ØÓ F j f B8 h41 $z (1 1 • 4 ( 8B $ FP T` ÒØ gr | v rv fjn

S2

Such phase spaces look locally but not globally like some cotangent bundle, and not in any canonical way. Here are the key properties of ω: 1. ω is a 2-form on X (in this case X = T ∗ Q) 2. ω is closed: dω = 0 (in this case exact, since ω = dα) 3. ω is nondegenerate:

∗ θ : Tx X −→ Tx X v −→ ω(v, −)

5.3 Hamilton’s Equations on a Symplectic Manifold

75

∂ ∂ is an isomorphism (it’s 1-1, hence onto). In our example if v = ai ∂qi + bi ∂pi , then

θ(v) = ω(v, −) = dpj ∧ dq j ai

∂ ∂ ,— + bi i ∂q ∂pi

= bj dq j − aj dpj which is nonzero if v is nonzero: ∂ −→ −dpi ∂q i ∂ −→ dq i θ: ∂pi θ: which should remind you of Hamilton’s equations! To summarize, we deﬁne a symplectic structure on a manifold X to be a nondegenerate closed 2-form ω on X, and we then call (X, ω) a symplectic manifold. Now let X be any symplectic manifold (e.g., T ∗ Q). Given any function H : X → Ê (like our Hamiltonian) we can deﬁne a vector ﬁeld vH on X by: vH = θ−1 (dH) and then Hamilton’s equations say that a “state” x(t) ∈ X evolves as follows: d x(t) = vH x(t) dt

PP vH PP P P PPPPPPPPP PPP• PP PP PP PP

dH

5.3

**Hamilton’s Equations on a Symplectic Manifold
**

∗ θ : Tx X −→ Tx X v −→ ω(v, −)

Let (X, ω) be a symplectic manifold: ω is a nondegenerate closed 2-form, so we get

76

Symplectic Geometry

**an isomorphism. For any H ∈ C ∞ (X) we thus get the Hamiltonian vector ﬁeld
**

y

vH = θ−1 (dH)

•

QQ Q

level sets of H

If X is the space of states of a classical system, we describe how a state x ∈ X evolves in time using the curve x(t) ∈ X satisfying Hamilton’s equations: d x(t) = vH x(t) dt with x(0) = x as its initial conditions. (Note that x is being used both for the initial state x ∈ X and the curve x : (a, b) → X and x(0) = x.) It’s easy to ﬁnd nonintegrable examples: X = T ∗Ê p2 H= − q4 2m Here the particle shoots oﬀ to ±∞ in ﬁnite time. There are even solutions in the 5-body Newtonian gravity problem where bodies shoot oﬀ to ∞ in ﬁnite time [Xia92]. Let’s see what these Hamilton’s equations look like when X = T ∗ Q (not always true, but it’s nice to see what this special type of system yields). Then ω = dpi ∧ dq i and we have θ : Tx Q −→ Tx ∗ Q ∂ −→ −dpi ∂q i ∂ −→ dq i ∂pi so

∗ θ−1 : Tx Q −→ Tx Q ∂ dq i −→ ∂pi ∂ dpi −→ − i ∂q

5.4 Darboux’s Theorem

77

so given H : T ∗ Q → Ê we get vH = θ−1 (dH) ∂H i ∂H dq + dpi = θ−1 ∂q i ∂pi ∂H ∂ ∂H ∂ = i − ∂q ∂pi ∂pi ∂q i and thus Hamilton’s equations say d x(t) = vH x(t) dt or d i ∂H q x(t) = vH (q i ) = − dt ∂pi ∂H d pi x(t) = vH (pi ) = i dt ∂q

and

Alas, these have a minus sign as compared with the usual Hamilton’s equations; to cure this we should deﬁne ω = −dα = dqi ∧ dpi and so get ∂H dq i , = dt ∂pi dpi ∂H = − i dt ∂q

We’ve just discovered that any symplectic manifold (M, ω) has a dual or time-reversed version (M, −ω).

5.4

Darboux’s Theorem

This condition seems special to the case of a cotangent bundle X = T ∗ Q, but in fact we have: Theorem 5.1 (Darboux’s theorem.). If (X, ω) is any symplectic manifold and x ∈ X, then you can ﬁnd symplectic coordinates pi , q i , (i = 1, . . . , n) in a neighborhood of x such that ω = dqi ∧ dpi A short way of stating Darboux’s theorem is to use some jargon and say every symplectic manifold of ﬁxed dimension is locally symplectomorphic, That is, every 2ndimensional symplectic manifold can be made to look locally like the linear symplectic

78

Symplectic Geometry

space C n with its canonical symplectic form. The above theorem is then just a deﬁnition of what we mean by a symplectomorphism: Let (M1 , ω1 ) and (M2 , ω2 ) be symplectic manifolds. A map φ : M1 → M2 is a symplectomorphism if it is a diﬀeomorphism and the pullback of ω2 under φ is equal to ω1 : φ∗ ω2 = ω1 . The Wikipedia entry contrasts symplectic and Riemannian manifolds: This result implies that there are no local invariants in symplectic geometry: a Darboux basis can always be taken, valid near any given point. This is in marked contrast to the situation in Riemannian geometry where the curvature is a local invariant, an obstruction to the metric being locally a sum of squares. It should be emphasized that the diﬀerence is that Darboux’s theorem states that ω can be made to take the standard form in an entire neighborhood around a point p ∈ M. In Riemannian geometry, the metric can always be made to take the standard form at any given point, but not always in a neighborhood around that point. Summarizing so far: 1. all symplectic manifolds are even-dimensional 2. all symplectic manifolds of the same dimension are locally alike—unlike Riemannian manifolds. We can think of any F ∈ C ∞ (X) as an observable: it assigns to any state x ∈ X a number F (x) ∈ Ê—namely the result of measuring something about x. We know how states evolve in time: d x(t) = vH x(t) . dt How do observables evolve in time? Given an observable F ∈ C ∞ (X) and a time t ∈ Ê we get a new observable Ft ∈ C ∞ (X) by Ft (x) = F x(t) . We then have, noting, d F x(t) = ∇F · x′ (t) = dF x′ (t) = vH F x(t) dt where we’ve used the deﬁnition of the 1-form df derived from any smooth function f (a 0-form) on a manifold M, that acts on vector ﬁelds v on M by df (v) = vf (the directional

5.4 Darboux’s Theorem

79

derivative of f in the direction of v). So (applying the above chain rule and Hamilton’s eqns.), d d Ft (x) = F x(t) dt dt = vH F x(t) = vH F (x) Abstracting this result: d Ft = vH Ft dt for the time evolution of an observable F .

Chapter 6 Poisson Brackets

(Week 9, May 23, 25, 27.) You can view a physical system as a point at some initial position in the phase space or conﬁguration space and describe it’s (the whole system!) time evolution by giving the motion of the point as a function of time. That’s like the Schr¨dinger picture in quantum o mechanics. Alternatively, you can view the same system as a bunch of objects that have certain properties (like mass, charge, position, velocity) and consider the time evolution as a change in those observable properties, which is like the Heisenberg picture in quantum mechanics. In this chapter we’ll begin with the analogous pictures for classical mechanics, then quickly move on to Poisson brackets and Noether’s theorem using Poisson brackets, and then take a look at Possion manifolds. All this will be in aid of proving Liouville’s theorem for symplectic manifolds. This will justify all the mathematics because the usual way that physicists derive Liouville’s theorem (“the distribution function is constant along any trajectory in phase space.”) uses physical assumptions like conservation of particles, whereas Liouville’s theorem, in the context of symplectic geometry, is a deeper result that generalizes the invariance of the volume measure on phase space to a theorem about the symplectic 2-form.

6.1

The Schr¨dinger Picture and the Heisenberg Pico ture in Classical Mechanics

In the Schr¨dinger picture observables are ﬁxed while states evolve in time; in the Heiseno berg picture states are ﬁxed while observables evolve in time: 80

6.1 The Schr¨dinger Picture and the Heisenberg Picture in Classical Mechanics o

81

Schr¨dinger picture o x ∈ X is a state x → x(t) where d x(t) = vH x(t) dt

and

Heisenberg picture F ∈ C ∞ (X) is an observable F → Ft where d Ft = vH Ft dt

F x(t) = Ft (x) is the result of measuring F in a state x after waiting a time t ∈ Ê. We deﬁne the Poisson bracket of F, G ∈ C ∞ (X) by {F, G} = vF G ∈ C ∞ (X) In this language, Hamilton’s equations become d Ft = {H, Ft } dt Example: If X = T ∗ Q then ω = dq i ∧ dpi (notice the diﬀerence with the initial development in §5.3 where we used dpi ∧ dq i, we’ve switched to dq i ∧ dpi in order to obtain agreement with the Euler-Lagrange equations). So we have θ : v −→ ω(, −) ∂ −→ dpi ∂q i ∂ −→ −dq i ∂pi so vF = θ−1 (dF ) ∂F ∂ ∂F ∂ = − i i ∂pi ∂q ∂q ∂pi so lastly {F, G} = vF G ∂F ∂G ∂F ∂G − = ∂pi ∂q i ∂q i ∂pi

82

Poisson Brackets

6.2

**The Hamiltonian Version of Noether’s Theorem
**

{H, F } = 0 ⇔ F is a conserved quantity

Suppose (X, ω) is a symplectic manifold, H ∈ C ∞ (X) has vH integrable. Then

where we say F is conserved (for the time evolution generated by H) if Ft = F, ∀t ∈ Ê. Suppose also that vF is integrable and let x −→ x(s) be the ﬂow generated by vF : d x(s) = vF x(s) . ds We say F generates symmetries of H if H(x) = H x(s) , Then we have a theorem. Theorem 6.1. F is a conserved quantity ⇔ F generates symmetries of H Proof. F is a conserved quantity ⇔ F = Ft , ∀t ∈ Ê d ⇔ Ft = 0, ∀t dt ⇔ {H, Ft } = 0, ∀t ⇒ {H, F } = 0 and (⇐) is true too: since Ft and F solve 1st order ODE we get F = Ft . Next,

d () dt

∀s ∈ Ê

= {H, ()}, so by uniqueness of solutions of

F generates symmetries of H ⇔ H(x) = H x(s) , ∀s ∈ Ê d ⇔ H x(s) = 0 dt ⇔ vF (H) = 0 ⇔ {F, H} = 0 by exactly the same argument as before. So to ﬁnish the proof we just need {F, G} = −{G, F }

6.2 The Hamiltonian Version of Noether’s Theorem

83

which is true in the previous example, but also in general, as we will now show: vF = θ−1 (dF ) θ(vF ) = dF ω(vF , −) = dF (−) ω(vF , vG ) = dF (vG ) = vG (F ) = {G, F } so {F, G} = vF G = ω(vG , vF ) = −ω(vF , vG ) = −{G, F }.

In short, it’s the antisymmetry of ω that says F generates symmetries of H ⇔ H generates symmetries of F In the next few exercises we’ll start to see how the Poisson bracket deﬁnes an algebra. After a few lemmas presented as examples we’ll prove a theorem and deﬁne a Poisson algebra. Example: Taking F = H, we see that energy is conserved! d Ht = 0 since {H, H} = 0 dt by antisymmetry of {−, −}. Example: If F and G are conserved quantities, then aF + bG, a, b ∈ Ê F G, and {F, G} are all also conserved quantities. First, aF + bG is a conserved quantity because {H, aF + bG} = vH (aF + bG) = avH F + bvH G = a{H, F } + b{H, G}.

84

Poisson Brackets

Next, F G is a conserved quantity because {H, F G} = vH (F G) = (vH F )G + F vH G = {H, F }G + F {H, G}. Finally, {F, G} is conserved since we have the Jacobi identity: {H, {F, G}} = {{H, F }, G} + {F, {H, G}} —but this is trickier! This last result uses the fact that ω is closed—and is equivalent to this fact. To prove the Jacobi identity we use the fact that ω is closed, and this formula d : Ωp (X) −→ Ωp+1 (X) which is deﬁned by (vi ∈ Vect(X), σ ∈ Ωp (X)), dσ(v0 , . . . , vp ) =

0≤i≤p

vi σ(v0 , . . . , vi , . . . , vp ) ˆ + (−1)i+j vi σ([vi , vj ], v0 , . . . , vi , . . . , vj , . . . , vp ) ˆ ˆ

0≤i<j≤p

(the second sum will vanish for coordinate vector ﬁelds). These imply 0 = dω(vF , vG , vH ) = vF ω(vG , vH ) − vG ω(vF , vH ) + vH ω(vF , vG ) − ω([vF , vG ], vH ) + ω([vF , vH ], vG ) − ω([vG , vH ], vF ) and note that vF ω(vG , vH ) = vF {H, G} = {F, {H, G}} and ω(vG , −) = dG(−) ω(vG , vH ) = dG(vH ) = vH G = {H, G} and so ω [vF , vG ], vH = −ω vH , [vF , vG ] = −dH [vF , vG ] = −[vF , vG ](H) = −vF vG H + vG vF H = −{F, {G, H}} + {G, {F, H}}

6.3 Poisson Algebras and Poisson Manifolds

85

Using the symmetry of {·, ·}, the nine terms we get from above reduce to three and we obtain the Jacobi identity: {H, {F, G}} = {{H, F }, G} + {F, {H, G}}. Now we can prove the theorem relating symplectic manifolds to a Poisson algebra.

6.3

Poisson Algebras and Poisson Manifolds

First a deﬁnition: a Poisson algebra denoted C ∞ (X), {·, ·} , has the following properties, 1. C ∞ (X) is a commutative associative algebra (with usual operations +, ·, etc.) 2. C ∞ (X), {·, ·} is a Lie algebra: (a) {F, aG + bH} = a{F, G} + b{F, H} (b) {F, G} = −{G, H} (c) {F, {G, H}} = {{F, G}, H} + {G, {F, H}} (Jacobi identity). 3. For any F ∈ C ∞ (X), {F, ·} is a derivation of C ∞ (X), meaning it’s linear and satisﬁes a Leibniz rule: {F, GH} = {F, G}H + G{F, H}. Theorem 6.2. If (X, ω) is a symplectic manifold then (X, {·, ·}) is a Poisson manifold, i.e., C ∞ (X), {·, ·} is a Poisson algebra. (Proof omitted. But we’ve already done most of the work for the proof.) If (X, {·, ·}) is a Poisson manifold and F ∈ C ∞ (X), then {F, ·} is a derivation, so (by a nice theorem) there’s a unique vector ﬁeld vF ∈ Vect(X) such that {F, G} = vF G. Theorem 6.3. If X, {·, ·} is a Poisson manifold then v : C ∞ (X) −→ Vect(X) F −→ vF is a Lie algebra homomorphism: meaning that it’s linear and v{F,G} = [vF , vG ]

86

Poisson Brackets

Proof. {F, −} and thus also vF are linear in F . Also, v{F,G} H = {{F, G}, H} = {F, {G, H}} + {{F, H G} = {F, {G, H}} − {G, {F, H}} = vF vG H − vG vF H = [vF , vG ]H.

6.3.1

A Poisson Manifold that is Not Symplectic

Here’s an example of a Poisson manifold that is not a symplectic manifold: Ê3 is odddimensional, hence not symplectic, but we can deﬁne {·, ·} on Ê3 by {x, y} = z {y, z} = x {z, x} = y where x, y, z are the coordinate functions Ê3 → Ê. The brackets are the Gibbs vector cross products in disguise. Using the fact that {F, −} is a derivation, one can calculate {F, G} for all polynomials. Then by some approximation argument one can deﬁne {F, G} for all F, G ∈ C ∞ (X) it has spheres as “symplectic leaves”

i f
sm u e w ~ d G Øq T 1 % 7 1 •

In fact, any Poisson manifold has a foliation of symplectic manifolds.

6.4

Liouville’s Theorem

Not to bore you with history, but we should at least recall that Liouville’s theorem is the fundamental result that forms the bedrock of both equilibrium and non-equilibrium statistical mechanics. In equilibrium statistical mechanics the stationary state solutions

6.4 Liouville’s Theorem

87

to Liouville’s equation for the density of states is satisﬁed by the Maxwell-Boltzmann distribution function. In non-equilibrium statistical mechanics one obtains the Vlasov equation for a collisionless system (a rare gas) which is incredibly useful in astrophysics and for describing collisionless plasmas. Generalized to collisional gases one obtains the Boltzmann transport equation (of fundamental importance to the physics of gases, ﬂuids, and for mass, heat & radiation transport). That’s the cursory history. Now for the modern treatment, though we’ll be stopping well short of applying the theory.

6.4.1

Phase Space Volumes

Let our conﬁguration space be Q = Ê and our phase space be X = T ∗ Q ∼ Ê2 ∋ (q, p), = and let 1 H = q 2 + p2 2 be the harmonic oscillator Hamiltonian. Then Hamilton’s equations say dq ∂H = =p dt ∂p ∂H dp =− = −q dt ∂q The drawing here depicts a phase space region at diﬀerent times (moving harmonically around the outer circle path), as well as the state at a point (q, p) ∈ X: p

@ @ hhh•(q,p) hhhh @hh@@@ hhhh h q @@ @@ @@ @@

vH

the phase space volume (in 2D it’ll be an area) around a point doesn’t change with time evolution. So the time evolution x → x(t) is rotation by t clockwise. As a region R ⊆ X evolves in time its area is preserved—or conserved! In fact this is completely general. For any symplectic manifold and any (integrable) Hamiltonian, the ﬂow x → x(t) given by Hamilton’s equations d x(t) = vH x(t) dt

88

Poisson Brackets

**preserves the symplectic structure ω. In our example ω = dq ∧ dp is the “area element”: Area(R) =
**

R

ω.

We’ve just examined a 1D example that had a 2D symplectic phase space. For a general 2n-dimensional symplectic manifold,

n

ω=

i=1

dqi ∧ dpi

in symplectic coordinates, and ω n = ω ∧ ω ∧ . . . ∧ ω, (n copies of ω get wedged here) = n! dq1 ∧ dp1 ∧ . . . dqn ∧ dpn is a 2n-form which measures volume, and ω n /n! is called the Liouville form. We can now state the modern version of Liouville’s theorem. Theorem 6.4 (Liouville’s theorem.). Let (X, ω) be any symplectic manifold and let H ∈ C ∞ (X) be such that vH is integrable, so that Hamilton’s equations determine the time evolution: φt X −→ X x −→ x(t). Then ω is preserved by φt : and thus φ∗ t φ∗ ω = ω t ωn ω = . n! n!

**Proof (sketch). In general if v is any integrable vector ﬁeld on any manifold X, we get a ﬂow
**

ee P

φt : X −→ X x −→ x(t) d x(t) = v x(t) dt x

•

mm T U oo

G F G

D yy 9 yy 9

•

x(t)

dd I

6.4 Liouville’s Theorem

89

and for any diﬀerential form µ ∈ Ωp (X) we have d ∗ φµ dt t where Lv : Ωp (X) −→ Ωp (X) is called the Lie derivative, which can be computed by Weil’s formula; Lv = iv d + div where d : Ωp (X) −→ Ωp+1 (X) is the exterior derivative, and iv : Ωp (X) −→ Ωp−1 (X) is the interior product, given by (iv µ)(v1 , . . . , vp−1) := µ(v, v1, . . . , vp−1 ) (Weil’s formula isn’t hard to prove, but we won’t do it now.) In our situation we note, dH(−) = ω(vH , −) = ivH ω(−) so, d ∗ φ ω dt t = Lv ω = ivH dω + divH ω, = divH ω = ddH = 0. One can similarly show but dω = 0 (ω is closed), so = Lv µ

t=0

t=0

d ∗ φ ω = 0 for any time t so dt t φ∗ ω = ω. t

Great, so ω is conserved for integrable vH . But we need to justify the use of Weil’s formula.

90

Poisson Brackets

6.4.2

Weil’s formula

d : Ωp −→ Ωp+1 grade 1 Lv : Ωp −→ Ωp grade 0 iv : Ωp −→ Ωp−1 grade -1

Why is Weil’s formula true? Well, in fact we have the following grade changing operators

(OK, so Lv “changes” the grade by zero), now all of these are superderivatives or graded derivatives. A superderivative of grade k is a linear operator D : Ωp → Ωp+k such that D(µ ∧ ν) = Dµ ∧ ν + (−1)pk µ ∧ Dν (for µ ∈ Ωp and ν ∈ Ωk ). If D, D ′ are superderivatives of grades k, k ′ , then their supercommutator is again a derivation: [D, D ′] = DD ′ − (−1)kk D ′ D. In particular, Weil’s formula says Lv = [d, iv ] = div + iv d, and we can reduce proving it to proving Lv and [d, iv ] agree on 1-forms, by the product rule. Before moving on to the next chapter let’s pause to consider what we’ve accomplished above. We’ve used almost purely geometric considerations to prove Liouville’s theorem, a theorem normally thought of by engineers and physicists as something very physical about conservation of particles. Yet the only physics we’ve touched upon is the symplectic structure of the phase space for “classical mechanics” systems, but there’s no need above to mention that we have a physical system, we could be talking about something totally diﬀerent, something of pure mathematical interest. This is not unusual in mathematical physics, and far from telling us that we are getting too far divorced from physics it is quite the opposite, in fact it tells us that physics is in some sense simpler (or deeper) than we might at ﬁrst imagine, in the sense that a lot of physics can be done by considering the geometric structure of things. In fancy philosophical lingo (which we make no real pretenses to understand!) one might even say that we’ve been “deconstructing” 19th and 20th century physics, taking out a lot of the turgid assumptions and traditional lore, and examining the remaining bare bones. In the process we’ve reconstructed classical mechanics more along the lines of Einstein’s model of general relativity, which was deeply about the geometry of spacetime.

′

Chapter 7 Introduction to Geometric Quantization

(Week 10, June 1, 3, 5.) In this chapter, especially in the ﬁnal section 7.3 we have condensed a lot of information for which glossary entries are provided as additional help to the reader, and hopefully as a spur to further independent study. Some recommendations for further study are given at the end of the chapter.

7.1

A Taste of Geometric Quantization

In Schr¨dinger’s approach to quantum mechanics, you take the conﬁguration space Q of o a classical system, give it a Riemannian metric, and then form a Hilbert space L2 (Q, vol) whose unit vectors are states in the corresponding quantum system, and where vol is the measure coming from the Riemannian metric. In geometric quantization we instead construct the Hilbert space from the phase space X, which is a symplectic manifold. De Broglie and then Bohr and Sommerfeld noticed a “quantization condition” which picked out the allowed “quantum orbits” among the classical ones in say, the harmonic 91

92

Introduction to Geometric Quantization

oscillator or hydrogen atom models: p

X

c c

q γ

**If the orbit is some loop γ : S ′ → X, it’s an allowed orbit if the phase e
**

i γ α/

=1

where α is the canonical 1-form on X (if X = T ∗ Q) and is Planck’s constant ( = h/2π). Here γ α is a term in the action for the path γ : [0, 1] → X × Ê in the extended phase ˜ iS space, and e is what appears in the wave approach to classical mechanics—the “eikonal approximation” relating particles to waves. So we’re demanding that we get complete constructive interference—that our particle’s “wavefunction” comes back into phase when the particle traverses the loop γ. When X = T ∗ Q we have α=

γ D

dα =

D

ω

**where D is any disk with boundary ∂D = γ. If Q = Ê, ω = dp ∧ dq so
**

c

D

γ

ω = area of D.

D

**The Bohr-Sommerfeld quantization condition says e or ω = 2πn = nh
**

D 1 which almost agrees with the modern formula (n + 2 )h. i D ω/

=1

7.1 A Taste of Geometric Quantization

93

**In modern terms, A = iα/ is a connection on some U(1) bundle over T ∗ Q, e
**

i γ α/

∈ U(1) iω

is the holonomy of this connection around the loop γ, and F =

is the curvature of A. These results are speciﬁc to the abelian U(1) bundle. So, in geometric quantization, we begin to quantize a system with phase space (X, ω) by ﬁnding a U(1) bundle P → X with a connection A such that the curvature F of A is iω/ . 2 We can do this if and only if ω/ is an integral closed 2-form, i.e., [ω/ ] ∈ HdeRham (X) is the image of i∗ H 2 (X, ) −→ H 2 (X, Ê) ∼ HdeRham (X) = 2 Given any U(1) bundle P , its ﬁrst Chern class c1 (P ) ∈ H 2 (X, ) is a complete invariant of U(1) bundles, and 2 i∗ c1 (P ) ∈ HDR (X) equals F 2πi where F is the curvature of any connection on P . The square brackets “[]” denote an equivalence class of curvatures derived from their respective available connections on P . If you need to review some of these concepts then the glossary at the end of this book may help, there are some entries for cohomology and Chern class. Example: Ê3 is a Poisson manifold with {x, y} = z {y, z} = x {z, x} = y This is the phase space of a spinning point particle—i.e., J ∈ Ê3 represents angular momentum. This is not symplectic, but it’s foliated by “symplectic leaves”—spheres centered at the origin

{|J| tt t tt f t t t n j tt u f s | i ttt
1 1 1 1 •t tt tt tt tt tt tt tt tt tt t

= j} = Xj

94

Introduction to Geometric Quantization

The sphere of radius j, Xj is the phase space for a particle of total angular momentum j. In fact Xj has an integral symplectic structure when 1 3 j = 0, , 1, , . . . 2 2 The phase space for a spinning point particle of total angular momentum j ∈ [0, ∞) is Xj = {J ∈ Ê3 : J = j}, a sphere with Poisson structure such that {x, y} = z {y, z} = x {z, x} = y.

One can check that this comes from a symplectic structure ω on Xj : {f, g} = ω(dg, df ) The 2-from ω is a multiple of the area 2-form on the sphere Xj , normalized so that ω = 4πj,

Xj

(note not 4πj 2 !)

We’d like to quantize the system if possible, so the question is, “when is ω integral?” There’s a theorem that says a closed 2-form ω on some manifold X is integral if for any 2D surface S mapped into X, ω∈ .

S

We’re using the symplectic leaves take note (an odd dimensional manifold like Ê3 is not itself symplectic). Here what we really want is

Xj

ω ∈ , 2π

(taking

= 1)

**This happens when
**

Xj

1 ω = 4πj = 2j ∈ 2π 2π

3 1 j = 0, , 1, , . . . 2 2 —the usual quantization condition for angular momentum! When this condition holds, there exists a U(1) bundle P → Xj with a connection A whose curvature F ∈ Ω2 (Xj , AdP ) = Ω2 Xj , µ(1) = Ω2 (Xj , iÊ) is U(1) F = iω,

•

that is,

µ(1) ∼ iÊ = 1

7.2 K¨hler Manifolds a

95

We then build the Hilbert space Hj of states of a quantum spin-j particle using P as follows. We form an associated vector bundle L = P ×U (1) using the God-given action of U(1) on . This sort of bundle, with a complex line bundle, or usually a “line bundle”

LX ∼ =

as ﬁbres, is called

•

ψ(x)

L

•

Xj

A section ψ of L looks locally like a complex function on the phase space. The Hilbert space Hj will consist of certain sections ψ of L. We shouldn’t use all L2 sections, since we want Hj to be ﬁnite dimensional! To get the right answer, we think of Xj as the Riemann sphere P 1 = ∪ {∞}. This lets us do complex analysis on Xj , and so deﬁne holomorphic sections of L. Then we let Hj be the space of these holomorphic sections. Then dim Hj = (2j + 1) as desired (for integral ω for geometric quantization). We make Hj into a Hilbert space as follows: U(1) acts on in a way that preserves the inner product on , so the ﬁbres Lx of L get an inner product. This lets us deﬁne ψ, φ =

Xj

ψ(x), φ(x) ω

So much for quantized spinning particles. We now want to know how to generalize the above.

7.2

K¨hler Manifolds a

So what’s going on in general? Our phase space X starts out as an integral symplectic manifold, which then acquires a line bundle L → X. To deﬁne holomorphic sections of L we need to equip X with extra structure—namely a K¨hler structure. a Deﬁnition 7.1. An almost K¨hler structure on a manifold X is a

96

Introduction to Geometric Quantization

**1. a linear operator Jx : Tx X −→ Tx X depending smoothly on x, such that
**

2 Jx = −1

**This makes Tx X into a complex vector space. 2. a complex inner product ·, ·
**

X:

Tx X × Tx X −→

depending smoothly on X. This makes Tx X into a (ﬁnite dimensional) Hilbert space. So a K¨hler manifold involves three ingredients, it is complex, it carries a Riemannian a metric and a symplectic form on it’s underlying real manifold. Given this, we have u, v

X

= ℜ u, v X + iℑ u, v = g(u, v) + iω(u, v)

X

where g is a Riemannian metric and ω is a nondegenerate 2-form Deﬁnition 7.2. An almost K¨hler manifold is K¨hler if ω is closed, i.e., a symplectic a a structure. If you have this then you can do geometric quantization. To recap: geometric quantization is a mathematical approach to obtain a quantum theory from a given classical theory, in the spirit of Bohr’s original correspondence principle but in modern form and in reverse! The idea is to carry out quantization while preserving certain analogies between the classical theory and the quantum theory (such as the time evolution given by the Hamiltonian, i.e., the analogy between Hamiltonian mechanics and the Heisenberg picture in QM).

7.3

General Geometric Quantization

With the above remarks in mind we can now give a terse summary of geometric quantization. Here is an eight-step synopsis (not exactly a mystical “eight-fold way”, but if you have not followed these lectures carefully it will seem mystical to you!). 1. We start with a classical phase space: mathematically, this is a manifold X with a symplectic structure

7.3 General Geometric Quantization

97

2. Then we do prequantization: this gives us an Hermitian line bundle L over X, equipped with a U(1) connection D, whose curvature equals iω. L is called the prequantum line bundle. Warning: we can only do this step if ω satisﬁes the Bohr-Sommerfeld condition, which says that ω/2π deﬁnes an integral cohomology class . If this condition holds, L and D are determined up to isomorphism, but not canonically. 3. The Hilbert space H0 of square-integrable sections of L is called the prequantum Hilbert space. This is not yet the Hilbert space of our quantized theory—it’s too big. But it’s a good step in the right direction. In particular, we can prequantize classical observables: there’s a map sending any smooth function on X to an operator on H0 . This map takes Poisson brackets to commutators, just as one would hope. The formula for this map involves the connection D. 4. To cut down the prequantum Hilbert space, we need to choose a polarization, say P . What’s this? Well, for each point x in X, a polarization picks out a certain subspace Px of the complexiﬁed tangent space at x. We deﬁne the quantum Hilbert space, H, to be the space of all square-integrable sections of L that give zero when we take their covariant derivative at any point x in the direction of any vector in Px . The quantum Hilbert space is a subspace of the prequantum Hilbert space. Warning: for P to be a polarization, there are some crucial technical conditions we impose on the subspaces Px . First, they must be isotropic: the complexiﬁed symplectic form ω must vanish on them. Second, they must be Lagrangian: they must be maximal isotropic subspaces. Third, they must vary smoothly with x. And fourth, they must be integrable. 5. The easiest sort of polarization to understand is a real polarization. This is where the subspaces Px come from subspaces of the tangent space by complexiﬁcation . It boils down to this: a real polarization is an integrable distribution P on the classical phase space where each space Px is Lagrangian subspace of the tangent space Tx X. 6. To understand this rigmarole, one must study examples! First, it’s good to understand how good old Schr¨dinger quantization ﬁts into this framework. Remember, o in Schr¨dinger quantization we take our classical phase space X to be the cotangent o bundle T ∗ M of a manifold M called the classical conﬁguration space. We then let our quantum Hilbert space be the space of all square-integrable functions on M. Modulo some technical trickery, we get this example when we run the above machinery and use a certain god-given real polarization on X = T ∗ M, namely the one given by the vertical vectors 7. It’s also good to study the Bargmann-Segal representation , which we get by taking X = n with its god-given symplectic structure (the imaginary part of the inner

98

Introduction to Geometric Quantization

product) and using the god-given K¨hler polarization. When we do this, our quana tum Hilbert space consists of analytic functions on n which are square-integrable with respect to a Gaussian measure centered at the origin. 8. The next step is to quantize classical observables, turning them into linear operators on the quantum Hilbert space H. Unfortunately, we can’t quantize all such observables while still sending Poisson brackets to commutators, as we did at the prequantum level. So at this point things get trickier and my brief outline will stop. Ultimately, the reason for this problem is that quantization is not a functor from the category of symplectic manifolds to the category of Hilbert spaces—but for that one needs to learn a bit about category theory and that’s for another book. If one is interested in the interplay between geometry and quantization then the online Quantum Gravity Seminar notes (http://math.ucr.edu/home/baez/QG.html) might be a good place to start. There are few introductory level books dealing with the correspondences between classical mechanics and quantum mechanics, but a good way to dig deeper into the geometry would be [BM94], for more on gauge theory try [O’R97], for a sweeping overview of the geometry of physics try [Pen05] for starters. All of these books cite numerous reference works that should keep you busy—at least until you’ve developed enough skills to undertake original research. For an in-depth exposition of classical mechanics theory try Arnol’d’s text [Arn78]—it’s Kung Fu is potent and stylistically unique. The traditional classical mechanics text by Goldstein [GPS02] provides a good background with plenty of example problems in case you want to develope speciﬁc mental muscles for mastering technical problem solving in classical mechanics without resorting to numerical methods. A ﬁnal word or two for students of physics is in order. At the end of the previous chapter we discussed the importance of reducing physics (in this case classical mechanics) to the simplest or deepest structure. We can probably never say that we’ve found the deepest description of nature’s physics, but the intention is still worthwhile. It’s worthwhile framing classical mechanics in the mathematical garb of this lecture note volume precisely because it makes it easier to see connections with other branches of physics, such as general relativity and quantum mechanics. If you haven’t had fun on this journey then perhaps at least we’ve shown that the mental pain was worth it! When we peel back the layers that most high school and undergraduate textbooks lather the subject with then we can more lucidly see that physics does seem to be a little more comprehensible as a whole than the disparate textbook treatments at ﬁrst seem to convey (most notably the initially baﬄing disparity between the formalisms of classical physics and quantum mechanics). We hope that some of this disparity has been removed and that the diligent student will be motivated to weave the same geometric themes of this small book into their own future research and study.

Glossary

adjoint Suppose H is a Hilbert space, with inner product ·, · . Then for a continuous linear operator A : H → H (this is the same as a bounded operator), using the Riesz representation theorem, one can show that there exists a unique continuous linear operator A∗ : H → H with the following property: A(x), y = x, A∗ y for all x, y ∈ H. This operator A∗ is the adjoint of A. More generally one can deﬁne an adjoint T ∗ : W → V for any linear operator T : V → W between vector spaces with an inner product by T v, w = v, T ∗w . The operation of taking the adjoint has some key properties: (1) A∗∗ = A, (2) if A is invertible, so is A∗ . (3) (A+B)∗ = A∗ +B ∗ , (4) (λA)∗ = λ∗ A∗ , where λ∗ denotes the complex conjugate of the complex number λ, ﬁnally (5) (AB)∗ = B ∗ A∗ is often used. (Source: Wikipedia) 98 Bargmann-Segal representation Also called the “coherant state representation” or “phase space representation”. Everyone bumps into the harmonic oscillator in quantum mechanics, but rather few get to see its full beauty. There are three main representations to learn: (1) the Schrdinger representation, (2) the Heisenberg (or Fock) representation, (3) the Bargmann-Segal representation. The ﬁrst should really be called the “wave” representation. In this representation we diagonalize the position operator, thinking of the state as a function on position space. Similarly, the second should be called the “particle” representation. In this we diagonalize the energy, thinking of the state as a linear combination of states |n > having n “quanta” of energy in them. The equivalence of these ﬁrst two representations is the basis of “wave-particle duality” in quantum ﬁeld theory. The third representation could be called the “complex wave” representation. At least that’s what it’s called in the book Introduction to Algebraic and Constructive Quantum Field Theory, where the ﬁrst representation is called the “real wave representation”. In some rough sense, this representation diagonalizes the creation operators. Of course, not being self-adjoint, the creation operators can’t be diagonalized in the usual sense. There is a nice substitute, however. In the complex wave 99

100

Glossary

representation, we think of phase space as a complex vector space using the trick Z = q + ip, and then think of states as analytic functions on phase space. Then the creation operator becomes multiplication by Z. Everyone who studies quantum mechanics learns about the ﬁrst two representations. The third, while in many ways the most beautiful, is somewhat less widely known. It’s lurking in the background whenever you relate quantum mechanics to the classical phase space and drag in the Z = q + ip trick. If you master these three basic viewpoints on the harmonic oscillator, it’s a snap to generalize to quantum ﬁeld theory, at least for free quantum ﬁelds, which are just big bunches of harmonic oscillators. (Source: Wikipedia) 97 category theory You don’t know category theory yet? If sets provide a foundation for mathematics based on formalising the idea of “ownership or membership” then categories can provide an alternative foundation formalising the notion of “relationship or map (arrow)”. A category (C, Hom) consists of the following three mathematical entities: (1). A class Ob(C) of objects, (2). A class Hom(C) of morphisms. Each morphism f has a unique source object X and target object Y . We write f : X → Y , and we say “f is a morphism from X to Y ”. We write Hom(X, Y ) to denote the hom-class of all morphisms from X to Y . (Some authors write Mor(X, Y ) or C(X, Y ).) (3). A binary operation ◦, called composition of morphisms, such that for any three objects X, Y , and Z, we have Hom(X, Y ) ◦ Hom(X, Z) ◦ Hom(X, Z). The composition of f : X → Y and g : Y → Z is written as g ◦ f or gf . (Some authors write f g.), and all of this stuﬀ is governed by two axioms: (Axiom.1) Associativity: If f : U → V , g : W → X and h : Y → Z then h ◦ (g ◦ f ) = (h ◦ g) ◦ f , and (Axiom.2) Identity: For every object X, there exists a morphism 1X : X → X called the identity morphism for X, such that for every morphism f : X → Y , we have 1Y ◦ f = f = f ◦ 1X . From these axioms, it can be proved that there is exactly one identity morphism for every object. Some authors deviate from the deﬁnition just given by identifying each object with its identity morphism. Relations among morphisms (such as f g = h) are often depicted using commutative diagrams, with “point” (corners) representing objects and “arrows” representing morphisms. The inﬂuence of commutative diagrams has been such that “arrow” and morphism are now synonymous. (Source: Wikipedia) 98 Chern class The kth Chern class, denoted ck (E) of the vector bundle E over a manifold M is the cohomology class of tr(F k )—the trace of the n-fold wedge product F k , also known as the nth Chern form—where F is the curvature of any connection on E. When properly normalized the integrals of the Chern classes over any compact oriented manifold mapped into M are integers, in which case they are integral cohomology

Glossary

101

classes. This is important in Yang-Mills theory and geometric quantization. (Source: [BM94], p.281) 93 cohomology In mathematics, speciﬁcally in algebraic topology, cohomology is a general term for a sequence of abelian groups deﬁned from a cochain complex. That is, cohomology is deﬁned as the abstract study of cochains, cocycles, and coboundaries. Cohomology can be viewed as a method of assigning algebraic invariants to a topological space that has a more reﬁned algebraic structure than does homology. Cohomology arises from the algebraic dualization of the construction of homology. In less abstract language, cochains in the fundamental sense should assign ‘quantities’ to the chains of homology theory. A particular type of cohomology theory (the deRham cohomology) follows from studying closed p-forms modulo the exact p-forms on a topological manifold. (Source: Wikipedia) 93 cohomology class (integral) Any closed p-form on a manifold M deﬁnes an element of the pth deRham cohomology of M. This is a ﬁnite-dimensional vector space, and it contains a lattice called the pth integral cohomology group of M. We say a cohomology class is integral if it lies in this lattice. Most notably, if you take any U(1) connection on any Hermitian line bundle over M, its curvature 2-form will deﬁne an integral cohomology class once you divide it by 2πi. This cohomology class is called the ﬁrst Chern class, and it serves to determine the line bundle up to isomorphism. (Source: Wikipedia) 96 cohomology, deRham A particular type of cohomology theory, deRham cohomology, arises when we want to study when closed forms are exact. All exact forms are closed by not vice versa. The pth deRham cohomology of a manifold M is a vector space, written H p (M), whose dimension is the number of ‘p-holes’ in M. To deﬁne this vector space, ﬁrst write Z p (M) for the set of closed p-forms on M. This is a vector space, since the sum of a closed form is again closed. Similarly, write B p (M) for the vector space of exact p-forms. The excat p-forms are a subspace of the closed p-forms, B p (M) ⊆ Z p (M), so the natural way to see how many closed forms there are that are not exact is to take the quotient space H p (M) = Z p (M)/B p (M) which is called the pth deRham cohomology group (actually more than a group since it’s a vector space). The concept arises, for example, in geometric quantization, and also in physics whenever one wants to integrate diﬀerential forms on a space that is not simply connected or modeled as such (as with wormholes, and with the Aharonov-Bohm eﬀect). (Source: [BM94]) 93

102

Glossary

complexiﬁcation We can tensor a real vector space with the complex numbers and get a complex vector space; this process is called complexiﬁcation. For example, we can complexify the tangent space at some point of a manifold, which amounts to forming the space of complex linear combinations of tangent vectors at that point. (Source: http://math.ucr.edu/home/baez/quantization.html) 97 distribution The word “distribution” means many diﬀerent things in mathematics, but here’s one: a “distribution” V on a manifold X is a choice of a subspace Vx of each tangent space Tp (X), where the choice depends smoothly on x. (Source: http://math.ucr.edu/home/baez/quantization.html) 97 Hamiltonian vector ﬁeld Given a manifold X with a symplectic structure ω, any smooth function f : X → R can be thought of as a “Hamiltonian”, meaning physically that we think of it as the energy function and let it give rise to a ﬂow on X describing the time evolution of states. Mathematically speaking, this ﬂow is generated by a vector ﬁeld v(f ) called the “Hamiltonian vector ﬁeld” associated to f . It is the unique vector ﬁeld such that ω(., v(f )) = df In other words, for any vector ﬁeld u on X we have ω(u, v(f )) = df (u) = uf The vector ﬁeld v(f ) is guaranteed to exist by the fact that ω is nondegenerate. (Source: http://math.ucr.edu/home/baez/quantization.html) 75 Hermitian Synonym for “self adjoint”. See the entries for Hermitian functions, matrix and operator 96 Hermitian function An “Hermitian function” is a complex function with the property that its complex conjugate, f ∗ , is equal to the original function with the variable changed in sign: f (x) = f (x)∗ . From this deﬁnition follows immediately that f is a Hermitian function, then the real part of f is an even function the imaginary part of f is an odd function. (Source: Wikipedia) 98

Glossary

103

Hermitian operator On a ﬁnite-dimensional inner product space, a self-adjoint operator is one that is its own adjoint, or, equivalently, one whose matrix is Hermitian, where a Hermitian matrix is one which is equal to its own conjugate transpose. By the ﬁnite-dimensional spectral theorem such operators have an orthonormal basis in which the operator can be represented as a diagonal matrix with entries in the real numbers. These can be generalized to operators on Hilbert spaces of arbitrary dimension. Self-adjoint operators are used in functional analysis and quantum mechanics. In quantum mechanics their importance lies in the fact that in the Dirac-von Neumann formulation of quantum mechanics, physical observables such as position, momentum, angular momentum and spin are represented by self-adjoint operators on a Hilbert space. 2 Of particular signiﬁcance is the Hamiltonian Hψ = − 2m ∆ψ + V ψ, which as an observable corresponds to the total energy of a particle of mass m in a potential ﬁeld V . Diﬀerential operators are an important class of unbounded operators. The structure of self-adjoint operators on inﬁnite dimensional Hilbert spaces essentially resembles the ﬁnite dimensional case, that is to say, operators are self adjoint if and only if they are unitarily equivalent to real-valued multiplication operators. With suitable modiﬁcations, this result can be extended to possibly unbounded operators on inﬁnite dimensional spaces. Since an everywhere deﬁned self adjoint operator is necessarily bounded, one needs be more attentive to the domain issue in the unbounded case. (Source: Wikipedia) 98 integrable distribution A distribution on a manifold X is “integrable” if at least locally, there is a foliation of X by submanifolds such that Vx is the tangent space of the submanifold containing the point x. (Source: http://math.ucr.edu/home/baez/quantization.html) 97 line bundle In algebraic topology and diﬀerential topology a line bundle is deﬁned as a vector bundle of rank 1. A line bundle expresses the concept of a line that varies from point to point of a space. For example a curve in the plane having a tangent line at each point determines a varying line: the tangent bundle is a way of organising these. (Source: Wikipedia) 95 square-integrable sections We can deﬁne an inner product on the sections of a Hermitian line bundle over a manifold X with a symplectic structure. The symplectic structure deﬁnes a volume form which lets us do the necessary integral. A section whose inner product with itself is ﬁnite is said to be square-integrable. Such sections form a Hilbert space H0 called the “prequantum Hilbert space”. It is a kind of preliminary version of the

104

Glossary

Hilbert space we get when we quantize the classical system whose phase space is X. (Source: http://math.ucr.edu/home/baez/quantization.html) 97 symplectic structure A symplectic structure on a manifold M is a closed 2-form ω which is nondegenerate in the sense that for any nonzero tangent vector u at any point of M, there is a tangent vector u at that point for which w(u, v) is nonzero. (Source: http://math.ucr.edu/home/baez/quantization.html) 59, 96 U(1) connection The group U(1) is the group of unit complex numbers. Given a complex line bundle L with an inner product on each ﬁber Lx , a U(1) connection on L is a connection such that parallel translation preserves the inner product. (Source: http://math.ucr.edu/home/baez/quantization.html) 96 vector bundle A vector bundle is a geometrical construct where to every point of a topological space (or manifold, or algebraic variety) we attach a vector space in a compatible way, so that all those vector spaces, “glued together”’, form another topological space (or manifold or variety). A typical example is the tangent bundle of a diﬀerentiable manifold: to every point of the manifold we attach the tangent space of the manifold at that point. Or consider a smooth curve in Ê2 , and attach to every point of the curve the line normal to the curve at that point; this yields the “normal bundle” of the curve. (Source: Wikipedia) 94 vertical vectors Given a bundle E over a manifold M, we say a tangent vector to some point of E is vertical if it projects to zero down on M. (Source: http://math.ucr.edu/home/baez/quantization.html) 52, 97

Bibliography

[Arn78] [BM94] [GPS02] [O’R97] [Pel94] [Pen05] V. I. Arnol’d. Mathematical methods of classical mechanics. Springer, New York, 1978. (Translated from Russian). J. C. Baez and J. P. Muniain. Gauge Fields, Knots and Gravity. World Scientiﬁc, Singapore, 1994. H. Goldstein, C. Poole, and J. Safko. Classical Mechanics. Addison-Wesley, Reading, Massachusetts, third edition edition, 2002. L. O’Raifeartaigh. The Dawning of Gauge Theory. Princeton University Press, Princeton, New Jersey, 1997. Peter Peldan. Actions for gravity, with generalizations: A review. Classical and Quantum Gravity, 11:1087, 1994. R. Penrose. The Road to Reality: A Complete Guide to the Laws of the Universe. Knopf, New York, 2005.

[WTM71] J. A. Wheeler, K. S. Thorne, and C. W. Misner. Gravitation. W. H. Freeman, New York, 1971. [Xia92] Zhihong Xia. The existence of noncollision singularities in Newtonian systems. Annals of Mathematics, 135(3):411–468, 1992.

105

Index

action, 4 charged particle, 38 deﬁnition, 4 Hamilton’s eqns, 59 least action, 4, 15 string action, 40 versus least time, 46 aﬃne, 55 angular momentum, 27 quantization, 93 Atwood machine, 28 Bargmann-Segal, 97 Bohr, 72, 91 Bohr-Sommerfeld, 72 condition, 97 boundary boundary terms, 5, 21 bundle abelian, 93 cotangent bundle, 51 ﬁber, 95 line bundle, 95 tangent bundle, 51 U(1), 93 canonical coordinates, 22 canonical forms, 72 catastrophe, 62 category theory, 98 centrifugal force, 13 charge charged particle, 38 charged particle general relativity, 45 Chern class, 93 Christoﬀel symbols, 45, 47 commutative algebra, 85 compact, 21 complexiﬁcation, 97 conﬁguration space, 14, 22 connection, 45 curvature, 93 U(1), 97 conservation conserved quantities, 25, 83 energy, of, 24 Jacobi identity, 83 conservative force, 3 coordinates canonical, 22 generalized, 17, 19, 21 cotangent bundle, 22, 51 canonical forms, 72 Darboux’s thm, 77 covariant derivative, 45 critical point, 4, 13, 15 geodesics, 48 D’Alembert, 8 d’Alembert’s principle, 9 Darboux’s theorem, 77 De Broglie, 91 De Rham cohomology, 93 derivation, 90 106

INDEX

107

derivative graded, 90 Lie derivative, 89 superderivative, 90 diﬀeomorphism, 34, 54 diﬀerential, 53 dynamics equilibrium, 12 eigenstates, 72 eikonal approximation, 63, 71 quantization, 92 electric ﬁeld, 38 electrodynamics, 35 electrostatics, 11 energy conservation, 24, 36, 42 Hamiltonian, 25 equilibrium, 7, 12 Euler-Lagrange and Hamilton’s eqns, 52, 57 equations, 14, 18 reparameterization invariance, 48 exterior derivative, 89 Fermat’s principle, 46 ﬁber bundle, 95 foliation, 86, 93 force centrifugal, 13 inertial, 12, 13 forms, 72 closed 2-form, 75 gauge symmetry, 35, 42 gauge transformation, 41 general relativity, 43 charged particle, 45 electromagnetic potential, 55 free particle, 44, 55 reparameterization invariance, 55 generalization, 12

generalized coordinates, 17, 19, 21 geodesic, 45, 46, 48 geometric optics, 61 geometric quantization, 91 geometrization of physics, 72 gravity 5-body problem, 76 GR, 43 Hamilton’s eqns, 51, 56 least action, 59 symplectic space, 75 Hamilton’s principle function, 64 Hamilton-Jacobi eqns, 61 Hamiltonian, 25 Hamiltonian vs Lagrangian, 6 harmonic oscillator, 87 mechanics, 51 relativistic, 36 vector ﬁeld, 76 harmonic oscillator, 87, 92 hase space symplectic, 59 Heisenberg QM picture, 80 Hilbert space, 91 Kahler, 96 prequantum, 97 holomorphic, 95 homomorphism, 85 Huygens, 61 hydrogen model, 92 inertial force, 13 inﬁnity 5-body, 76 inner product, 96 integrable

108

INDEX

polarization, 97 vector ﬁeld, 82 isotropic maximal subspace, 97 Jacobi least time, 46 Jacobi identity, 84 K¨hler manifold, 95 a Kaluza-Klein theory, 49 kernel, 53 Lagrangian choice of, 27 deﬁnition, 15 dynamics, 16 electrodynamics, 41 general relativity, 45 Hamiltonian vs Lagrangian, 6 regular, 55 string, 39 strongly regular, 54 leact action geodesics, 46 least action, 4, 6, 15 Hamilton’s eqns, 59 versus least time, 46 least energy, 9 least time, 10 versus least action, 46 Leibniz rule, 17 Levi-Civita connection, 45 Lie algebra, 85 line bundle, 95 prequantum, 97 linear transform, 55 Liouville’s theorem, 86 Lorentz force, 39 Lorentz transformation, 31 magnetic ﬁeld, 37

manifold, 9 compact, 21 cotangent space, 22 foliation, 86 Kahler, 95 Lorentzian, 44, 48 oriented, 38 Poisson, 85 Riemannian, 48, 54 smooth, 21 symplectic, 59, 75 vector manifold, 53 Maxwell-Boltzmann, 87 measure, 91 mechanics analogy to optics, 46 classical, 2 Hamiltonian, 6, 51 Lagrangian, 2 Newtonianian, 2 statistical, 87 metric Euclidean, 46 Lorentzian, 31, 44 nondegenerate, 54 Riemannian, 47 Minkowski spacetime, 33 momentum angular, 27 space translation, 26 nature beauty of, 12 Newton II, 19 Newtons laws, 2, 10 Noether’s theorem, 20, 23 Hamiltonian version, 82 nondegenerate, 54 one-parameter family of paths, 16, 22, 24, 34 operator

INDEX

109

grade changing, 90 linear, 90 optics analogy to mechanics, 46 eikonal approx., 63 geometric, 61 geometric approximation, 49 orbits periodic, 72 quantum, 91 oriented manifold, 38 oscillator harmonic, 92 parallel transport, 45 parameterization reparameterization, 34 path one-parameter family, 16, 22, 24, 34 variation, 20 pendulum double pendulum, 14 phase space, 54 extended, 65, 70 physics geometrization, 72 Poisson algebra, 83 Poisson bracket, 81 quantization, 98 Poisson manifold, 85 potential, 9 energy, 4 Riemannian case, 54 vector potential, 37, 49, 55 prequantization, 97 proper time, 33 quantization condition, 92 prequantization, 97 quantum mechanics, 19

Hamilton-Jacobi eqns, 69 quantization, 91 time evolution, 80 quasi-particle, 9, 15 refraction, 11 index of, 11 least time, 46 regular Lagrangian, 55 relativity general theory, 43 special theory, 31 reparameterization, 39 invariance, 48, 55 Riemannian manifold, 54 Schr¨dinger o QM picture, 80 scientiﬁc questions, 6 section holomorphic, 95 of a bundle, 95 Snell’s law, 11 Sommerfeld, 72, 91 space translation symmetry, 25 spacetime, 33 special relativity, 31 spin, 95 statics, 7 versus dynamics, 9 Stokes’ theorem, 72 string action, 40 Lagrangian, 39 strongly regular, 54 Lagrangian, 54 submanifold, 9 superderivative, 90 symemtry

110

INDEX

conserved quantity, 25 symmetry gauge symmetry, 35 rotational, 26 space translation, 25 spacetime translation, 42 time translation, 25 transformation, 22 symplectic 2-form, 74 geometry, 70 leaves, 86, 93 manifold, 75 structure, 59, 75 preservation, 88 symplectomorphism, 77 tangent bundle, 51 tangent vector, 15 time evolution, 78 proper time, 33 reversal invariance, 77 time translation, 20 symmetry, 25, 42 topology catastrophe, 62 variational derivative, 12 Hamilton-Jacobi, 65 path variation, 17, 20 vector cotangent vector, 44 tangent vector, 15 vector ﬁeld, 75 vertical vectors, 53 vector ﬁeld, 75 Hamiltonian ﬁeld, 76 vector potential, 37, 55 vector space, 53 vertical vectors, 53

virtual displacement, 9 virtual work, 8 waves eikonal approximation, 92 particles, 92 wave function, 61, 70 Weil’s formula, 89 WKB approximation, 71

- Letters to a German Princess on Physics
- Conformal field theory
- Little Things Make Human Divine
- Epol301 Course Notes - The Teacher in Context
- Personal Perspective on Harmony of Science and Religion
- Epsy302 Collective Notes - Teaching Models and Strategies
- Aleph Cardinality Paradox
- Sci-epol344 Course Notes - Science Curriculum Study
- Math-epol338 Course Notes - Mathematics Curriculum Study
- Epsy301 Collective Notes - The Learner in Context

Notes of lectures given by John Baez, UC Riverside

Notes of lectures given by John Baez, UC Riverside

- 2.4
- Lecture Slides W1 Clip 5
- Gabion Wall Design
- mechanics - chapter 4
- cburch-thesis 5-5.15.12
- Truss Critical Load
- Assessment
- Physics
- International Journal of Engineering Research and Development (IJERD)
- Http Www.inrisk.ubc.CA Process.php File=STRUCTURAL ANALYSIS Virtual Work and Variational Principles
- Experiment
- ch
- Energy
- Lecture 1- 1page
- Untitled
- 894-2436-1-PB
- 1_watersurface
- Centrifugal Force
- Einstein on the Ether 1924 Pb
- Euler and Runge Kutta
- Derivation of the Diffusivity Equation
- takehometest2013.pdf
- dot
- Advanced Mechanics of Solids (4)
- First Law
- The Mystery of Planck
- Numerical Methods in Heat, Mass
- Zero Point Energy is the Vacuum Constant in Time and Space
- save energy at home
- Bala ppt

Are you sure?

This action might not be possible to undo. Are you sure you want to continue?

We've moved you to where you read on your other device.

Get the full title to continue

Get the full title to continue reading from where you left off, or restart the preview.

scribd