Susskind Lecture 11 - Electric and Magnetic Forces
Lecture 11 closes classical mechanics by showing why magnetic forces force us to use gauge potentials. This is the classical doorway to quantum mechanics and field theory.
Fields and del operators
A scalar field assigns one number to each point, like temperature T(x, t). A vector field assigns a vector, like wind velocity, electric field, or magnetic field.
Useful derivative operations:
gradient: ∇V
divergence: ∇ · A
curl: ∇ × ATwo vector-calculus facts matter here:
∇ · (∇ × A) = 0
∇ × (∇V) = 0Magnetic field and vector potential
Magnetic fields have zero divergence:
∇ · B = 0Therefore they can be written as the curl of a vector potential:
B = ∇ × AThe vector potential is not unique. If s(x) is any scalar function, then:
A → A + ∇sleaves B unchanged. This is a gauge transformation.
Electric and magnetic forces
For a static electric field:
E = -∇Φ
F_electric = e E
potential energy = e ΦThe magnetic part of the Lorentz force is velocity-dependent:
F_magnetic = (e/c) v × BIt is perpendicular to v and B, so a static magnetic field changes the direction of velocity but does no work on the particle.
Lagrangian for a charged particle
To get the magnetic Lorentz force from an action principle, the Lagrangian must use the vector potential:
L = (1/2)m v^2 + (e/c) A(x) · v - e Φ(x)A gauge transformation changes the action only by an endpoint term. Since stationary-action variations keep endpoints fixed, the equations of motion are gauge invariant.
Mechanical versus canonical momentum
The conjugate momentum is:
p = ∂L/∂v = m v + (e/c) AThis is canonical momentum. It is not the same as mechanical momentum:
p_mechanical = m vMechanical momentum is gauge invariant and directly tied to motion. Canonical momentum is the variable used in Lagrangian/Hamiltonian mechanics, and it changes under gauge transformations.
Hamiltonian
Solve for velocity in terms of canonical momentum:
m v = p - (e/c) AThen the Hamiltonian is:
H = (1/2m) [p - (e/c) A]^2 + e ΦAlthough the magnetic vector potential appears, physical predictions depend only on gauge-invariant fields and mechanical motion.
Uniform magnetic field
For a constant B along z, the charged particle moves in circles in the x-y plane while moving uniformly along z. The full motion is a helix unless the z velocity is zero.
The circular motion is not caused by loss of energy. The magnetic force bends the velocity vector while preserving its magnitude.
Common pitfalls
- Thinking the vector potential is the magnetic field. B is the curl of A.
- Treating gauge choice as physical. Different A fields can describe the same B.
- Confusing canonical momentum with mechanical momentum.
- Thinking a static magnetic field changes kinetic energy. The magnetic force is perpendicular to velocity.
- Trying to write a magnetic-force Lagrangian using B directly. The action formulation needs A.