Electromagnetism Christopher R Prior Fellow and Tutor in Mathematics ASTeC Intense Beams Group Trinity College, Oxford Rutherford Appleton Laboratory Contents Maxwell’s equations and Lorentz Force Law Motion of a charged particle under constant Electromagnetic fields Relativistic transformations of fields Electromagnetic waves Waves in a uniform conducting guide Simple example TE01 mode Propagation constant, cut-off frequency Group velocity, phase velocity Illustrations Reading J.D. Jackson: Classical Electrodynamics H.D. Young and R.A. Freedman: University Physics (with Modern Physics) P.C. Clemmow: Electromagnetic Theory Feynmann Lectures on Physics W.K.H. Panofsky and M.N. Phillips: Classical Electricity and Magnetism G.L. Pollack and D.R. Stump: Electromagnetism What is electromagnetism? The study of Maxwell’s equations, devised in 1863 to represent the relationships between electric and magnetic fields in the presence of electric charges and currents, whether steady or rapidly fluctuating, in a vacuum or in matter. The equations represent one of the most elegant and concise way to describe the fundamentals of electricity and magnetism. They pull together in a consistent way earlier results known from the work of Gauss, Faraday, Ampère, Biot, Savart and others. Remarkably, Maxwell’s equations are perfectly consistent with the transformations of special relativity. Maxwell’s Equations Relate Electric and Magnetic fields generated by charge and current distributions. E = electric field D = electric displacement H = magnetic field B = magnetic flux density ρ= charge density j = current density µ0 (permeability of free space) = 4π 10-7 ε0 (permittivity of free space) = 8.854 10-12 c (speed of light) = 2.99792458 108 m/s r ∇⋅D = ρ r ∇⋅B = 0 r r ∂B ∇∧E = − ∂t r r r ∂D ∇∧H = j + ∂t r r r r In vacuum D = ε 0 E , B = µ 0 H , ε 0 µ 0 c 2 = 1 Maxwell’s 1st Equation Equivalent to Gauss’ Flux Theorem: r ρ r r 1 ∇⋅E = ⇔ ∫∫ E ⋅ dS = ε0 ε0 S ∫∫∫ ρ dV = V r ρ ∇⋅E = ε0 Q ε0 The flux of electric field out of a closed region is proportional to the total electric charge Q enclosed within the surface. A point charge q generates an electric field r E= r r ∫∫ E ⋅ dS = sphere q 4πε 0 r 3 r r q dS q ∫∫ r 2 = ε 0 4πε 0 sphere Area integral gives a measure of the net charge enclosed; divergence of the electric field gives the density of the sources. Maxwell’s 2nd Equation r ∇⋅B = 0 ⇔ r ∇⋅ B = 0 r r ∫∫ B ⋅ dS = 0 Gauss’ law for magnetism: The net magnetic flux out of any closed surface is zero. Surround a magnetic dipole with a closed surface. The magnetic flux directed inward towards the south pole will equal the flux outward from the north pole. If there were a magnetic monopole source, this would give a non-zero integral. Gauss’ law for magnetism is then a statement that There are no magnetic monopoles Maxwell’s 3rd Equation r r ∂B ∇∧E = − ∂t Equivalent to Faraday’s Law of Induction: r r r ∂B r ∫∫S ∇ ∧ E ⋅ dS = −∫∫S ∂t ⋅ dS r r r dΦ d ⇔ ⋅ = − ⋅ = − B d S E d l (for a fixed circuit C) ∫C dt dt ∫∫ S The electromotive force round a r r circuit ε = ∫ E ⋅ dl is proportional to the rate of change of flux of magnetic r r field, Φ = ∫ B ⋅ dl through the circuit. N S Faraday’s Law is the basis for electric generators. It also forms the basis for inductors and transformers. Maxwell’s 4th Equation r r r 1 ∂E ∇ ∧ B = µ0 j + 2 c ∂t r r ∇ ∧ B = µ0 j Originates from Ampère’s (Circuital) Law : r r r r r r ∫ B ⋅ dl = ∫∫ ∇ ∧ B ⋅ dS = µ0 ∫∫ j ⋅ dS = µ0 I C Ampère S S Satisfied by the field for a steady line current (Biot-Savart Law, 1820): r µ0 I B= 4π Biot ∫ r r dl ∧ r r3 For a straight line current µ0 I Bθ = 2πr Need for displacement current Faraday: vary B-field, generate E-field Maxwell: varying E-field should then produce a B-field, but not covered by Ampère’s Law. Apply Ampère to surface 1 (flat disk): line integral of B = µ0I Surface 2 Surface 1 Current I Closed loop Applied to surface 2, line integral is zero since no current penetrates the deformed surface. Q dQ dE In capacitor, E = , so I = = ε0 A ε0 A dt dt r r ∂E ε j = Displacement current density is d 0 ∂t r r r r r ∂E ∇ ∧ B = µ 0 ( j + jd ) = µ 0 j + µ 0ε 0 ∂t Consistency with charge conservation Charge conservation: Total current flowing out of a region equals the rate of decrease of charge within From Maxwell’s equations: Take divergence of (modified) Ampère’s equation the volume. ⇔ ⇔ r r d ∫∫ j ⋅ d S = − dt ∫∫∫ ρ dV r ∂ρ ∫∫∫ ∇ ⋅ j dV = − ∫∫∫ ∂ t dV r ∂ρ ∇⋅ j + =0 ∂t ( r r 1 ∂ ∇ ⋅ ∇ ∧ B = µ 0∇ ⋅ j + 2 ∇⋅E c ∂t r ∂⎛ρ⎞ ⇒ 0 = µ 0∇ ⋅ j + ε 0 µ 0 ⎜⎜ ⎟⎟ ∂t ⎝ ε 0 ⎠ r ∂ρ ⇒ 0 = ∇⋅ j + ∂t ) Charge conservation is implicit in Maxwell’s Equations Maxwell’s Equations in Vacuo In vacuum r r r r 1 D = ε 0 E , B = µ0 H , ε 0 µ0 = 2 c Source-free equations: r ∇⋅B = 0 v r ∂B ∇∧E+ =0 ∂t Source equations r ρ ∇⋅E = ε0 r v 1 ∂E r ∇∧B− 2 = µ0 j c ∂t Equivalent integral forms (sometimes useful for simple geometries) r v 1 ∫∫ E ⋅ dS = ε0 ∫∫∫ ρ dV r r ∫∫ B ⋅ dS = 0 r r r r d dΦ ⋅ = − ⋅ = − E d l B d S ∫ dt ∫∫ dt r r r r r r 1 d ∫ B ⋅ dl = µ0 ∫∫ j ⋅dS + c 2 dt ∫∫ E ⋅ dS Example: Calculate E from B r r d r ∫ E ⋅ dl = − dt ∫∫ B ⋅ dS z r ⎧ B0 cos ω t Bz = ⎨ 0 ⎩ Also from r < r0 r > r0 r r ∂B ∇∧E = − ∂t r < r0 2π rEθ = −π r 2 B0ω cos ω t ⇒ B0ω r Eθ = − cos ω t 2 r > r0 2π rEθ = −π r02 B0ω cos ω t ⇒ Eθ = − ω r02 B0 2r cos ω t r r r 1 ∂E then gives current density necessary ∇ ∧ B = µ0 j + 2 c dt to sustain the fields Lorentz force law Supplement to Maxwell’s equations, gives force on a charged particle moving in an electromagnetic field: r r r r f = q E+v ∧B r r r r For continuous distributions, have a force density f d = ρ E + j ∧ B Relativistic equation of motion ( ) r r ⎛v⋅ f dP F= ⇒ γ ⎜⎜ , dτ ⎝ c 4-vector form: r r⎞ ⎛ 1 dE dp ⎞ f ⎟⎟ = γ ⎜ , ⎟ ⎝ c dt dt ⎠ ⎠ 3-vector component: ( r r r r r d (m0γ v ) = f = q E + v ∧ B dt ) Motion of charged particles in constant electromagnetic fields ( r r r r r d (m0γ v ) = f = q E + v ∧ B dt Constant E-field gives uniform acceleration in straight line q d r ( ) γ v = E Solution of dt m0 2 ⎤ ⎡ ⎛ qEt ⎞ m0 c 2 ⎢ ⎟⎟ − 1⎥ x= 1 + ⎜⎜ ⎥ qE ⎢ m0 c ⎠ ⎝ ⎦ ⎣ 1 qE 2 ≈ t for qE << m0 c 2 m0 Energy gain = qEx ) Constant magnetic field gives uniform spiral about B with constant energy. r r v// = constant dv q r r = v∧B r dt m0γ x ⊥ = constant Relativistic Transformations of E and B r r According to observer O in frame F, particle has velocity v, fields are E and B and Lorentz force is ( r) r f ′ = q′E ′ r r r r f = q E+v ∧B In Frame F′, particle is at rest and force is Assume measurements give same charge and force, so r r r q = q′ and E ′ = E + v ∧ B r Point charge q at rest in F: E = See a current in F′, giving a field r r 1 r r B′ = B − 2 v ∧ E c Suggests r r r , B=0 3 4πε 0 r q Exact: r r r µ0 q v ∧ r 1 r r =− 2 v∧E B=− 3 4π r c ( ) r r r r r E⊥′ = γ E⊥ + v ∧ B , E//′ = E// r r r r ⎛ v∧E⎞ r B⊥′ = γ ⎜⎜ B⊥ − 2 ⎟⎟ , B//′ = B// c ⎠ ⎝ Electromagnetic waves Maxwell’s equations predict the existence of electromagnetic waves, later discovered by Hertz. No charges, no currents: r r ∂D ∇∧H = ∂t r ∇⋅D = 0 r r ∂B ∇∧E = − ∂t r ∇⋅B = 0 r r ∂B ∇ ∧ ∇ ∧ E = −∇ ∧ ∂t r r r 2 r ∂ ∇ ∧ ∇ ∧ E = ∇ ∇⋅E −∇ E =− ∇∧B r 2 ∂t = −∇ E r r ∂2D ∂2E = −µ = − µε 2 ∂ t 2 3D wave equation : ∂t ( ) ( ) ( ) ( ) r r r r 2 2 2 2 r ∂ E ∂ E ∂ E ∂ E ∇ 2 E = 2 + 2 + 2 = µε 2 ∂x ∂y ∂z ∂t Nature of electromagnetic waves A general plane wave r with angular frequency ω travelling in the direction of the wave vector k has the form r r r r r r r r E = E 0 exp[ j (ω t − k ⋅ x )] B = B0 exp[ j (ω t − k ⋅ x )] r r Phase ω t − k ⋅ x = 2π × number of waves and so is a Lorentz invariant. Apply Maxwell’s equations r r r ∇ ↔ − jk ∇⋅E = 0 = ∇⋅B ∂ r r& ↔ jω ∇ ∧ E = −B ∂t ↔ ↔ r r r r k ⋅E = 0 = k ⋅B r r r k ∧ E =ωB Waves are transverse to the direction of propagation, and r k are mutually perpendicular s r E, B and Plane electromagnetic wave Plane Electromagnetic Waves r r 1 ∂E ∇∧B = 2 c ∂t r r ω r ↔ k ∧B=− 2E c r r r Combined with k ∧ E = ω B r E ω kc 2 deduce that r = = ω B k 2π Wavelength λ = r k ω Frequency ν = 2π ⇒ speed of wave in vacuum is ω r =c k r r The fact that ω t − k ⋅ x is an invariant tells us that ⎛ω r⎞ Λ = ⎜ ,k ⎟ ⎝c ⎠ is a Lorentz 4-vector, the 4-Frequency vector. Deduce frequency transforms as ( ) r r c−v ω′ = γ ω − v ⋅ k = ω c+v Waves in a conducting medium r r For a medium of conductivity σ, j = σ E r r r& r r& Modified Maxwell: ∇ ∧ H = j + ε E = σ E + ε E Put r r r r E = E 0 exp[ j (ω t − k ⋅ x )] r r r r B = B0 exp[ j (ω t − k ⋅ x )] r r r r − j k ∧ H = σ E + jω ε E Dissipation factor σ D= ωε conduction current displacement current Copper : σ = 5.8 ×107 , ε = ε 0 ⇒ D = 1012 Teflon : σ = 3 ×10-8 , ε = 2.1ε 0 ⇒ D = 2.57 ×10 − 4 Attenuation in a Good Conductor r r r r − j k ∧ H = σ E + jω ε E r r r r& r Combine with ∇ ∧ E = − B ⇒ k ∧ E = ω µ H r r r r r r ⇒ k ∧ k ∧ E = ω µ k ∧ H = −ω µ (− jσ + ω ε )E ( ) k 2 = ω µ (− jσ + ω ε ) ⇒ σ >> ω ε For a good conductor D >> 1, Wave form is where δ = k 2 ≈ − jω µσ ⎡ ⎛ x ⎞⎤ ⎛ x⎞ exp ⎢ j ⎜ ω t − ⎟ ⎥ exp ⎜ − ⎟ δ ⎠⎦ ⎝ δ⎠ ⎣ ⎝ 2 ω µσ is the skin - depth copper.mov water.mov ⇒ k≈ ω µσ 2 (1 − j ) Maxwell’s Equations in a uniform perfectly conducting guide z Hollow metallic cylinder with perfectly conducting boundary surfaces Maxwell’s equations with time dependence exp(jωt)are: r r r r r r 2 ∂B ∇ E = ∇ ∇⋅E −∇ ∧ ∇ ∧ E ∇∧E = − = − jω µ H r ∂t ⇒ = jω µ ∇ ∧ H r r ∂D r r 2 ∇∧H = = jω ε E = −ω εµ E t ∂ 144444444444244 4444444443 r ⎧ ⎫ ⎛ ⎞⎪ E ⎪ 2 ⎪ ⎜ 2 ⎟⎪ ⎜⎜ ∇ + ω µε ⎟⎟ ⎨ r ⎬ = 0 ⎝ ⎠ ⎪⎪ H ⎪⎪ ⎩ ⎭ ( x y r r Assume E ( x, y, z , t ) = E ( x, y )e ( jω t −γ z ) r r H ( x, y, z , t ) = H ( x, y )e ( jω t −γ z ) γ is the propagation constant ) ( r ⎧E ⎫ Then ∇ t2 + (ω 2εµ + γ 2 ) ⎨ r ⎬ = 0 ⎩H ⎭ [ ] Can solve for the fields completely in terms of Ez and Hz ) Special cases Transverse magnetic (TM modes): Hz=0 everywhere, Ez=0 on cylindrical boundary Transverse electric (TE modes): Ez=0 everywhere, ∂H z = 0 on cylindrical boundary ∂n Transverse electromagnetic (TEM modes): Ez=Hz=0 everywhere requires γ 2 + ω 2εµ = 0 or γ = ± jω εµ A simple model with Ez=0 Transport between two infinite parallel conducting plates: r E = (0,1,0) E ( x) e ( jω t −γ z ) where E ( x) satisfies 2 d E ∇ 2t E = 2 = − K 2 E , K 2 = ω 2εµ + γ 2 dx ⎧ sin ⎫ i.e. E = A⎨ ⎬ Kx ⎩cos ⎭ y x To satisfy boundary conditions, E=0 on x=0 and x=a, so z E = A sin Kx, K = Kn = a x= 0 x= nπ , n integer a Propagation constant is γ = K n2 − ω 2εµ = nπ a ⎛ω ⎞ 1 − ⎜⎜ ⎟⎟ ⎝ ωc ⎠ 2 where ω c = Kn εµ Cut-off frequency, ωc 2 γ= ⎛ω⎞ nπ nπ nπx jω t −γ z e 1 − ⎜⎜ ⎟⎟ , E = A sin , ωc = a a a εµ ⎝ ωc ⎠ ω<ωc gives real solution for γ, so attenuation only. No wave propagates: cut-off modes. ω>ωc gives purely imaginary solution for γ, and a wave propagates without attenuation. γ = jk , k = εµ (ω − ω 2 2 c ) 1 2 ⎛ ω = ω εµ ⎜⎜1 − ⎝ ω ω only a finite number of modes can propagate. nπ aω ω > ωc = ⇒ n< εµ π a εµ 2 c 2 ⎞ ⎟⎟ ⎠ 1 2 For a given frequency For given frequency, convenient to choose a s.t. only n=1 mode occurs. Propagatedr electromagnetic fields From r ∂B ∇∧E =− , ∂t (assuming A is real) ⎧ Ak ⎛ nπ x ⎞ H = − sin ⎟ cos(ω t − kz ) ⎜ ⎪ x ωµ ⎝ a ⎠ r r ⎪ j H= Hy = 0 ∇∧E ⇒ ⎨ ωµ ⎪ A nπ ⎛ nπ x ⎞ cos⎜ ⎪H z = − ⎟ sin (ω t − kz ) ωµ a ⎝ a ⎠ ⎩ x z Phase and group velocities ∞ Plane wave exp j(ωt-kx) has constant phase ωt-kx at peaks ⇔ ω ∆ t − k∆ x = 0 ω ∆x vp = ∆t = k j [ω ( k ) t − kx ] A ( k ) e dk ∫ −∞ Superposition of plane waves. While shape is relatively undistorted, pulse travels with the group velocity dω vg = dk Wave packet structure Phase velocities of individual plane waves making up the wave packet are different, The wave packet will then disperse with time Phase and group velocities in the simple wave guide ( k = εµ ω − ω Wave number is so wavelength in guide λ = Phase velocity is v p = ω k > 2 2 c ) 1 2 < ω εµ 2π 2π > , the k ω εµ 1 εµ free-space wavelength , larger than free-space velocity Group velocity is less than infinite space value ( k = εµ ω − ω 2 2 2 c ) ⇒ dω k vg = = < dk ωεµ 1 εµ Calculation of wave properties If a=3 cm, cut-off frequency of lowest order mode is ωc 1 = ≅ 5 GHz 2π 2a εµ fc = At 7 GHz, only the n=1 mode propagates and ( k = εµ ω − ω λ= 2 c ) 1 2 ≈ 103 m −1 2π ≈ 6 cm k vp = vg = 2 ω k ≈ 4.3 ×108 ms −1 k ωεµ = 2.1× 108 ms −1 Waveguide animations TE1 mode above cut-off ppwg_1-1.mov TE1 mode, smaller ω ppwg_1-2.mov TE1 mode at cut-off ppwg_1-3.mov TE1 mode below cut-off ppwg_1-4.mov TE1 mode, variable ω ppwg_1_vf.mov TE2 mode above cut-off ppwg_2-1.mov TE2 mode, smaller ppwg_2-2.mov TE2 mode at cut-off ppwg_2-3.mov TE2 mode below cut-off ppwg_2-4.mov