A quick run through of 180 years of Classical Mechanics

– for better appreciation of Quantum Mechanics

Lagrangian and Hamiltonian Mechanics

For those who followed the course pattern of University Physics I and Classical Mechanics I or equivalents,
this will largely be a review. For those haven’t taken Lagrangian and Hamiltonian Mechanics, don’t worry –
just open up your mind, relax and absorb. 1
Second Course: Lagrangian and Hamiltonian Mechanics

Hinted at obtaining Equation of Motion by taking

derivatives of energy 2
 Q: Why partial
derivatives?

(A new function of the dimension of energy)

This is the essence of Lagrangian Mechanics

3
4
 Recipe of Lagrangian Mechanics
Lagrange (1736 – 1813) wrote the book Mécanique analytique (1787) that
defined the subject Analytical Mechanics

5
 That was a defining moment in physics! Let’s see why

Q: Why Lagrangian formulation?

Good Stuff #1:

• Energy (scalar) easier to handle from forces (vectors)

Good Stuff #2:

• Easy identification of coordinate-momentum pair(s)
• In QM, such pair becomes conjugated pair of
operators
• Systematic way of getting equation of motion

6
 • Easy identification of coordinate-momentum pair(s)

Also called generalized

momentum
Define momentum
How momentum
evolves in time?

Big idea: For every coordinate, there is an accompanying momentum

7
Side comment on Quantum Mechanics

Big idea from Lagrangian Mechanics:

For every coordinate, there is an accompanying momentum

Substitute into the Hamiltonian

8
Simplest Example

Harmonic Oscillator:

Systematic way of getting Equation of motion

9
“Simple” Pendulum: A not-so-simple example
[Physics is characterized by principles that can be generalized]

Coordinate is ϴ (doesn’t look like a coordinate in usual sense)

Key points: Angular momentum and torque appear by themselves, and systematic
way of getting equation of motion! 10
Good Stuff #3:
• Easy identification of conserved quantities
Simplest Example:

What’s being done?

• If L does not depend on a coordinate x, then the (generalized)
momentum px does not change in time (conserved)!

Deeper thought:
• Symmetry and Conservation Law

11
 Symmetry? What symmetry?
• Take something (an object such as a hexagon)
• Do something (operation) (rotation about the center through some angle)
• Has it been changed? If not, there is a symmetry of the object for the operation

Rotated by 10 degrees Changed!

Rotated by 60 degrees Not changed! There is a symmetry!

But what is the “object” in Mechanics that we want to

study its symmetry?
12
 The “object” is the Lagrangian L

Think like a physicist!

• Free particle => no force => V = 0 (or same constant) everywhere
• Thus, space x is homogeneous (meaning the same everywhere)
• Fancy way of saying something simple:
• Make an arbitrary translation in space x -> x + δ (called a
transformation)
• Inspect L and see if L is changed or not
• If L does not change under the transformation (translation here), then
obviously the same equation of motion follows (physics hasn’t been
changed)
• Accompanying this symmetry (L does not change and same eq of
motion), something is conserved (px here)
13
Bonus: Why bother and why so fancy?

• Big stuff in 20th century physics!

• Jumping to about 1930

• E.g. Requiring L for an electron not to change under a

transformation of the phase of the quantum mechanical
wavefunction (“phase (gauge) transformation”) of the
electron, the correct formula describing how an electron
interacts with EM fields appears automatically! (This is
QED and it is a gauge field theory (Yang and Mills 1954)).

14
Good Stuff #4:
• Principle of Least Action (completed by Hamilton)
A way to interpret the Lagrange equation

(This is the question)

15
 Harmonic oscillator (Configuration Space, displayed horizontally)

Lagrangian Mechanics
is about the
configuration space,
which is just x (a line
for 1D problem). The
Euler-Lagrange
equation (equation of
motion) governs how
the system evolves in
configuration space in
time, thus just x(t).

16
 Harmonic oscillator (Configuration Space, displayed vertically)

Lagrangian mechanics
is about the
configuration space,
which is just x (a line
for 1D problem). The
Euler-Lagrange
equation (equation of
motion) gives how the
system evolves in
configuration space in
time, thus just x(t),
here displayed
vertically.

17
 What is so special about the actual trajectory?

[Maupertuis (1747), d’Alembert (1743), Euler and Lagrange (1750’s), Hamilton (1834)]

(The Least Action Principle)

18
Why taken so long (90 years)? Need to wait for the maths (calculus of variation)
 An illustration: Harmonic oscillator (actions for correct & wrong motions)
[Meaning of Least Action Principle: If not following the right trajectory, Action will be higher]

• Action for 3 periods of oscillation

vs the weighting of the error Dashed line = actual path
function r

• Wrong Motion:
𝑘
𝑦 𝑡 = 𝐴 cos( 𝑡
𝑚

𝑘
+ 𝑟 sin 𝑡
𝑚

• r (the weighting of the error

function):
from -1 to 1
• r = 0 -> correct motion

[Animation credit: LEUNG Chun Hei (MPhil Student, CUHK)]

19
Turning Newton’s Law and Euler-Lagrange Equation
into an extremum principle is profound!
• Principle of least time in Optics (Fermat 1657) – path of
light from one point to another is one that takes the
shortest time

• Mechanics follows suit (around 1750) – actual trajectory

is one that minimizes the Action (inside action is the
Lagrangian L)

• Thermodynamics and Statistical Physics – Entropy in a

closed system is maximized as system approaches
equilibrium. At constant pressure, reactions go in a way
to minimize the Gibbs free energy G(T,p); etc.

• Formulating quantum field theories 20

 Key ideas/Summary: Lagrangian Mechanics
• Getting equation of motion systematically from L

• For every coordinate, there is a momentum

• Easy to explore conservative laws and their relation to

symmetry in L

• Put mechanics into an extremum principle (Least Action

Principle) [Feynman formulated QM based on the action in
1950’s]

• Bridging over to quantum mechanics: Coordinate-

momentum pair(s) become operators in a systematic way

• Bridging over to Hamiltonian Mechanics 21

 Let’s meet Euler and Lagrange
Leonhard Euler (1707 – 1783)

“The most beautiful equation of mathematics”

Graph Theory

[By Bogdan Giuşcă - Public domain (PD), based on the image, CC BY-SA 3.0,
Portrait by Jakob Emanuel Handmann (1753) https://commons.wikimedia.org/w/index.php?curid=112920]
[Taken from Wikipedia]

Rigid Body Motions

Euler’s equation in hydrodynamics

22
Leonhard Euler (1707 – 1783)

Euler is often regarded as the greatest mathematician of all time. He was born
1707 in Switzerland (29 years senior than Lagrange). At age 14, he entered the
University of Basel for religious studies, but soon he found that he was talented in
mathematics. He published many papers in mathematics when he was a student.
At age 19 (1726), he completed university studies and immediately after that he
was offered a position at the St. Petersburg Academy of Science in Russia. In St.
Petersburg, he was surrounded by many gifted scientists and he worked and
contributed to every branch of mathematics, pure and applied. In 1735 (28 years
old), Euler lost the vision in one eye due to a serous fever, but he remained
productive. In Russia, Peter the Great died in 1725 and Catherine the Great would
not become Empress until 1762. Russia was politically unstable in the 1730’s.
Euler moved to the Berlin Academy of Science in 1741 at the invitation of
Frederick the Great (King of Prussia) and worked there for 25 years. During those
years, he published close to 400 articles. During his stay in Berlin, Euler invited
Lagrange (29 years younger than him) to join him in Berlin in 1755 when Lagrange
was only 19 years old. But Lagrange turned down the offer and preferred to work
in his home town instead.

23
Leonhard Euler

In 1766, Russian had stabilized after Catherine the Great became Empress,
and Euler went back to St. Petersburg. A few years later, he lost the vision of
his another eye, but his mathematical works continued with the help of his
two sons and assistants. Euler made many contributions. In physics, you see
his Euler angles for rigid body, his equation in hydrodynamics, and in
mechanics. He invented the calculus of variations. In the least action
principle, his method helped when we varied the path from the actual
trajectory. This is what nowadays called functional derivatives. Euler,
D’Alembert and Lagrange defined the subject of Analytic Mechanics. Euler
also worked on vibrations of strings and membranes and he interpreted light
as waves. He died in 1783 in St. Petersburg. The Russian Academy of
Sciences continued to publish his completed works for almost 40 more years
after his death.

24
 Joseph Louis Lagrange (1736 – 1813)

(1787)

He participated in defining
the SI units at the time of
the French Revolution
25
Joseph-Louis Lagrange (1736 – 1813)

Lagrange was born 1736 in Turin (now in Italy). His family kept the French
spelling of the surname although they had lived in Italy for generations.
Lagrange wanted to be a lawyer, but his father lost his fortune and could not
support him to do that. He studied at the University of Turin and discovered
a talent in mathematics. At age 19, Lagrange become Professor of
Mathematics at the Turin Royal Artillery School. At that time, he was so good
that Euler invited him to Berlin to join him. But Lagrange preferred to work
alone in his home town. His early contributions to physics and mathematics
were about the theories on vibration of strings and propagation of sound. He
eventually moved to the Berlin Academy of Science in 1766 to occupy Euler’s
position, when Euler moved back to St. Petersburg. Lagrange worked in
Berlin for 20 years. When the political climate turned bad in Berlin, he moved
to the Academie des Sciences (Academy of Science) in Paris in 1787 where he
published his important books on Analytic Mechanics, which transformed
Mechanics into a branch of mathematical analysis, and Analytic Functions.
Later, he taught at Ecole Normale and Ecole Polytechnic in Paris.
You should have seen the Lagrange multipliers in problems for maximizing a
function under some constraints. It is again this technique that has led to the
Least Action Principle and the Euler-Lagrange Equation.
26
Joseph-Louis Lagrange

1790’s was an exceptional period in France. It was the end of the Age of
Enlightenment and the beginning of the French Revolution. A cause of the
French Revolution, believe it or not, was the inconsistency in the
measurement system. There were great scientists in France at that time –
Laplace and Legrendre for example. Lagrange’s personality avoided him to
get into the political conflicts at that time, and yet he was involved in the
effort of defining then new Metric System, which is still in use today. A
note on the SI units. The definition of Kg, which Lagrange worked on in
1790s, may be changing later this year to one that replies on the Planck
constant. Somehow quantum physics gets into the new definition of the
kilogram! This is a great example about the Nature of Science. The Eiffel
Tower in Paris has 72 scientists’ names engraved on it. Lagrange is one of
them. The next time you go to Paris, find Lagrange and Fourier there and
take a picture.

27
 Hamiltonian Mechanics
We saw the law of conservation of energy in
Newtonian Mechanics

From how L not depending on coordinates, we have

many conservation laws (often involving momenta).

Question: Where is conservation of energy in

Lagrangian Mechanics?

Answer: The key points are whether L depends on

time t explicitly and what energy really is
Hamilton (1805 – 1865) Lagrange (1736 – 1813)
28
Getting something from nothing: Let’s do 3 steps of math

Here enters the Hamiltonian.

The root of Hamiltonian in QM
29
 Getting something from nothing: Big Ideas

• Derived (identified) H
• H has dimension energy • H is constructed from
quantities in Lagrangian
• If L does not depend on Mechanics
time explicitly, then H is • Harmonic oscillator
conserved (does not
change in time)!
• This is a statement of
conservation of energy
• Energy conservation is
related to symmetry of L
in translation in time
• But what is energy?
Total energy
30
 Important Features of Hamiltonian H

• To get H, we need L (thus identified coordinates), then find

generalized momenta, then construct H (thus need to learn
Lagrangian Mechanics first)

• H(x,p) is a function of coordinate-momentum pair(s)

For students prefer

practicality, simply take
H(x,p) = total energy
= K.E. + P.E
and start doing QM

Why so? (next page)

Why is it important? The Hamiltonian (plus operators) is the starting
point of Quantum Mechanics
31
 Why so?
Vase or Faces?

Two ways of looking at the Same Picture

The point is: Student A and Student B are conveying

the same information, only in different ways!

This change in viewpoint is called a Legrendre (1752 –

1833) Transform. The same idea takes us from the
internal energy U(S,V) to Helmholtz free energy F(T,V)
or Gibbs free energy G(T,p) in thermodynamics. 32
 Why is it important?
Jumping to 1925-1926 (Quantum Mechanics)
Harmonic Oscillator
Go Quantum!

Schrodinger Equation (1926)

was built on the Hamiltonian

Do this for coordinate-

momentum pair

Schrodinger Equation of a Quantum Harmonic Oscillator that

gives the allowed energies E and wavefunctions
33
 Why is it important?

Go back to Hamilton’s time

H(x,p) describes a system by its coordinate and momentum as
it evolves in time

Hamiltonian Mechanics introduces the Phase Space (x,p) (2D

phase space) for a particle moving in 1D (x-direction)

Harmonic Oscillator
Amplitude A = 0.5
Spring constant k = 1
Mass m = 2
1
Energy = 2 𝑘𝐴2
Conservation of energy =>
System confined to an
ellipse in phase space

[Credit: LEUNG Chun Hei (MPhil student, CUHK)] 34

Equations governing motion in Phase Space - The Hamilton’s equations

Getting something from nothing again!

Lagrange
Equation

Two 1st order differential equations (c.f. Newton’s law which is one 2nd order
equation) giving how x updates and how p updates – they combine to give the
equation of motion. Just another way of doing Mechanics! 35
Why is it important?

• Concept of phase space is fundamental in developing

Statistical Mechanics (Boltzmann and Gibbs, end of 19th
century)

• Take the phase space with you to thermal physics &

statistical mechanics courses

36
 Further Development: More Classical Mechanics

Hamilton

Poisson (1781 – 1840)

In Heisenberg’s quantum mechanics (1925), quantum operators evolve in time following

an equation almost of the same form, with Bracket replaced by commutator.
37
 Further Development

(From F = x, G = p)

Dirac did his version of Quantum Mechanics (1925) starting from here!

38
 Key ideas/Summary: Hamiltonian Mechanics

• Energy conservation is related to symmetry of L in time translation

• Identified H(x,p)

• Phase space and how x and p moves (Hamilton’s equations)

• Led to different formulations of quantum mechanics in 1925-26

• Led to developments in Statistical Mechanics in late 1800’s

39
 Let’s meet Hamilton
William Hamilton (1805 – 1865)
Hamilton’s principle of least action

Hamilton’s canonical equations of motion

Hamilton’s theory of optics has an equation highly similar to the Schrodinger Equation

Hamilton Mathematics Institute, Trinity College Dublin

40
William Hamilton (1805 – 1865)

Hamilton was born in 1905, and lived in Dublin, Ireland, all his life. He would
be called a highly gifted person nowadays. His early talent was in linguistics.
He was fluent in English at age 3. At age 5, he translated texts from Latin,
Greek, and Hebrew. At age 13, he could read 13 languages. He then turned his
focus to mathematics. At age 16, he worked through Newton’s Principia and
Laplace’s Mecanique Celeste all by himself. At 18, he entered Trinity College in
Dublin and won outstanding awards in both the classics and the sciences. He
worked on mathematical optics in his undergraduate years. He presented his
results in Theory on Systems of Rays to the Royal Irish Academy in 1824 when
he was only 19 years old, although the paper in published form was delayed to
for 4 years and appeared in 1828. This early work was later regarded as a
masterpiece as it showed a way that optics and mechanics could be formulated
in the same way mathematically. This Hamilton’s optics has an equation that
looks very much like the Schrodinger equation in quantum mechanics
developed about a century later. In fact, Schrodinger referred to Hamilton’s
work in his QM papers.

41
William Hamilton

Hamilton was appointed Professor of Astronomy at Dublin before his

graduation at age 22, a position that he held to his death at age 60. He was
also appointed the Astronomer Royal of Ireland, and even better he was
allowed not to carry out the duties and left alone to work on his theories! In
1834 (age 29), he published the classic paper “On a general method on
dynamics”, which is 62 pages long. He formulated classical mechanics in a
way that would serve as one of the principal formulations of quantum
mechanics in the works of Heisenberg and Dirac (some 90 years later).
Therefore, his work influenced all the founders (Schrodinger, Heisenberg, and
Dirac) of quantum mechanics. There are many things under his name -- the
Hamiltonian, which we use a lot in quantum mechanics, his final version of
the principle of least action, and the Hamiltonian canonical equations. Later
in his life, he worked on turning complex numbers into a subject with solid
algebraic foundation. He started a related subject call Quaternions and wrote
a 800-page book on it, but the subject has not found its place in mathematics
so far. He also consumed too much alcohol in his later years.

42
 Summary: Classical Mechanics goes far beyond its domain!

Minimal learning outcomes for this discussion

An appreciation that classical mechanics had led a good

foundation for physicists around 1900 to pursue quantum
mechanics

Key ideas (think/talk like a physics student): Newton’s eq. of

motion -> Lagrangian + Euler equation also give eq. of motion ->
Lagrangian enters -> define momentum for each coordinate
(coordinate – momentum pair) [QM needs them] -> Hamiltonian
[QM starts with H(x,p)] -> Hamilton’s mechanics deals with phase
space -> motion in phase space implies Poisson bracket [closely
related to QM commutator]

[These are ideas to know, not the technical details!] 43

Practically:
All you need to know is to write down the
Hamiltonian
H(x,p) = KE + PE = T + U
for a given system,
and to realize that there is a classical mechanics
root of the Hamiltonian operator in QM

44
 Big Picture – Paths to perfection of Mechanics after Newton

Taken from: K. Simonyi, A Cultural History of Physics (CRC Press 2010) 45

I hope that you see what you learned in previous
courses is so profoundly uselessly (you might have
thought) useful!

46