Home / Learn / Exam prep / GATE Physics / Lesson 3.2
All lessons

Module 3: Electromagnetic theory · Lesson 3.2 · about 55 min

The method of images

Why images work, the sphere image derived from circles of Apollonius, cavities, spheres at fixed potential or carrying charge, line charges and cylinders with exact capacitances, dipoles near planes, wedges, the hemispherical boss, and infinite image series for parallel plates and a sphere near a plane.

Key points, meanings and formulas on one page: choose "Save as PDF" in the print box.

Module 3 · about 55 minutes. The image method replaces a conductor by a few fictitious charges placed where the conductor is. You met it for a grounded plane, a corner and a simple sphere in the IIT JAM lesson on conductors and images. Here we go further: where the sphere image comes from, cavities, spheres held at a potential or carrying charge, cylinders and two-wire lines with exact capacitances, dipoles, and the infinite image series that GATE likes to hide inside innocent-looking problems.

A conductor is just a boundary condition. Find a set of charges, all outside the region you care about, whose combined potential satisfies that boundary condition, and the uniqueness theorem guarantees their field is the real field inside the region.

1. The rules of the game

The method rests on the uniqueness theorem from boundary value problems:

  1. Images must sit outside the physical region (inside the metal, or beyond the boundary), so Poisson’s equation in the region is untouched.
  2. The boundary condition must hold: the right potential on grounded or fixed-potential conductors, or the right total charge on isolated ones.
  3. The answer is valid only in the physical region. Inside the metal the real field is zero, and the image charge is not the induced charge (for a sphere they are not even equal in size).

Forces, energies and induced charge densities then follow from the image picture, with one caution: the energy is half the naive qVimage, because the image moves and grows as the real charge is brought in.

2. Spheres

Where the sphere image comes from. Take a charge q at distance d from the centre and a trial image q′ at distance b on the same ray. A point on the sphere sees them at distances r1 and r2; the sphere is at zero potential if q/r1+q′/r2=0, i.e. if r2/r1 is the same for every point of the sphere. The locus of points with a fixed distance ratio from two points is a sphere (Apollonius), and it is our sphere of radius R exactly when b=R2/d (the two points are inverse in the sphere) and the ratio is R/d. So

q′=−qRd,b=R2d

The same formula works with the charge inside a grounded cavity (d<R): then b>R, the image lies in the metal and is larger than q. The wall still carries exactly −q (every field line ends on it), and the charge is pulled towards the nearest wall with F=kq2Rd/(R2−d2)2: the centre is an unstable equilibrium.

A charge near a conducting sphere. Field lines end on the sphere at right angles; show the images to see them continue (dashed) to the fictitious charges. Switch between a grounded and an isolated neutral sphere and slide the distance.

What follows from the image.

  • Force (grounded): F=kq2Rd(d2−R2)2, attractive. Close to the surface, with h=d−R≪R, it becomes kq2/4h2: the plane result.
  • Induced density: σ(θ)=−q(d2−R2)4πR(R2+d2−2Rdcos⁡θ)3/2, largest at the near point, −q(d+R)/4πR(d−R)2. It integrates to −qR/d; for d=2R over 80% of it is on the near hemisphere.
  • Energy: bringing q in from infinity takes W=−kq2R2(d2−R2).

Changing the sphere’s condition only adds a charge at the centre, because a central charge changes the sphere’s potential uniformly without disturbing the zero already built:

  • held at potential V0: add 4πϵ0RV0;
  • isolated and neutral: add +qR/d (total zero), giving an induced dipole qR3/d2 and a far force ≈2kq2R3/d5;
  • isolated with charge Q: add Q+qR/d.

Worked example: a like-charged sphere can attract. With charge Q on the sphere, the force on q (repulsion positive) is

F=kq[Q+qR/dd2−qRd(d2−R2)2]

The second term blows up as d→R, so any like charge that comes close enough is attracted. For Q=q the force vanishes where (u+1)3(u−1)2=u4 with u=d/R, whose root is the golden ratio u=1.618. This is the same physics that holds electrons inside a metal (the image force behind the work function).

Worked example: a sphere in a uniform field, by images. Model the field by +Q at z=−D and −Q at z=+D with D→∞, so E0=2Q/4πϵ0D2. Their images ∓QR/D at ∓R2/D form a dipole p=2QR3/D2=4πϵ0R3E0, the separation-of-variables result in three lines.

3. Line charges and cylinders

For line charges the potential is logarithmic, V=−λ2πϵ0ln⁡s, and a pair ±λ gives V=λ2πϵ0ln⁡s2s1. Its equipotentials are again circles of Apollonius, so the image of a line charge in a cylinder is an equal and opposite line at the inverse point R2/d. The cylinder then carries −λ per length and the force per length is λ2/2πϵ0(d−R2/d).

Turn the logic round and the image pair solves any problem whose conductors are circles of one Apollonius family:

Geometry Equivalent line charges Capacitance per length
wire (radius a) at height h above a plane ±λ at ±c, c2=h2−a2 2πϵ0cosh−1⁡(h/a)
two wires, radius a, axes D apart ±λ at ±D2/4−a2 πϵ0cosh−1⁡(D/2a)
wire off-axis inside a tube −λ outside the tube at R2/d (eccentric cylinders)

For thin wires cosh−1⁡x≈ln⁡2x, giving the familiar 2πϵ0/ln⁡(2h/a) and πϵ0/ln⁡(D/a). The equivalent lines sit slightly off the wire axes, nearer the other conductor.

Worked example. A wire of radius 1.0 mm with its axis 10 cm above ground: cosh−1⁡100=5.298, so C′=5.563×10−11/5.298=10.5 pF/m. A two-wire line with 1.0 mm wires 1.0 cm apart: cosh−1⁡5=2.292, C′=12.1 pF/m.

4. Dipoles, wedges and bosses

A dipole above a grounded plane. Image each charge: a dipole pointing straight away from the plane has an image pointing the same way (head to tail, attraction 3p28(4πϵ0)d4), while a dipole parallel to the plane has an image pointing the opposite way (side by side, attraction 3p216(4πϵ0)d4). In general the perpendicular component keeps its sign and the parallel component flips: a dipole at angle α to the normal has its image at −α.

Wedges. Two grounded half-planes meeting at angle α need images at every reflection of the charge in both planes. The set closes with the right signs only when 360∘/α is an even integer, giving 360∘/α−1 images (3 for 90∘, 5 for 60∘). For 120∘ or 72∘ an image would land inside the physical region with the wrong sign, and the method fails.

A hemispherical boss on a plane. Treat it as a whole sphere plus the plane. A charge q on the axis at height d needs three images: −qR/d at R2/d (the sphere), −q at −d (the plane), and +qR/d at −R2/d (the plane’s image of the sphere’s image). At d=2R the pull is 0.205kq2/R2.

5. When the images never stop

A charge between two grounded parallel plates. Each image must itself be imaged in the other plate, and so on: an infinite row with period 2L. The series converges, and two results come out cleanly without summing it:

  • Induced charges, by reciprocity. Smear the charge into a sheet at the same position: the plates take −q(L−x)/L and −qx/L.
  • Exponential screening. Far along the plates the potential is built from sin⁡(nπz/L)K0(nπρ/L) and dies as e−πρ/L: the alternating images cancel every multipole.

A sphere near a plane. Start with q0 at the centre of a sphere at potential V; image it in the plane, image that image in the sphere, and repeat. With cosh⁡α=h/a the charges inside the sphere form the series qn=q0sinh⁡α/sinh⁡nα, so

C=4πϵ0asinh⁡α∑n=1∞1sinh⁡nα

At h=2a the series gives C=1.34×4πϵ0a: the nearby plane raises the capacitance by a third.

Inversion. Behind the sphere image is Kelvin’s theorem: if V(r→) is harmonic, so is RrV(R2r→/r2). Inversion in a sphere maps planes to spheres and charges to image charges, which is why the plane and sphere images look so alike. Images in a dielectric half-space, where the image has the size −q(ϵ2−ϵ1)/(ϵ2+ϵ1), are treated in fields in matter.

Image charge
a fictitious charge outside the physical region that, together with the real charges, reproduces the boundary condition
Inverse point
for a sphere or circle of radius R, the point at R²/d on the same ray as the point at d
Circles of Apollonius
curves of fixed distance ratio from two points: the equipotentials of two opposite line charges, and (as spheres) of a charge and its image
Image force
the attraction of a charge to its own image: the force a conductor exerts through the charge it induces
Physical meaning
The induced surface charge has exactly the field outside the conductor that the images would have. The images are the simplest bookkeeping for a charge distribution that would be miserable to find directly.

“The image charge is the charge induced on the conductor.” Only for an infinite plane do they happen to be equal. A charge q inside a grounded cavity has an image −qR/d, larger than q, while the wall carries exactly −q. A charge outside a neutral sphere has two images summing to zero, matching the zero net induced charge, but they sit at points where the real charge is not. The images give the right field in the physical region and nothing else.

  • Images go outside the physical region, fix the boundary condition, and give the true field only in the region. Energies carry a factor 12.
  • Sphere: q′=−qR/d at R2/d, inside or outside. Add 4πϵ0RV0 (fixed potential), +qR/d (neutral) or Q+qR/d (charge Q) at the centre.
  • Grounded sphere force kq2Rd/(d2−R2)2; cavity kq2Rd/(R2−d2)2; neutral sphere far force ≈2kq2R3/d5. Like-charged sphere with Q=q: zero force at d=1.618R.
  • Line and cylinder: image −λ at R2/d. Wire to plane 2πϵ0/cosh−1⁡(h/a); wire to wire πϵ0/cosh−1⁡(D/2a).
  • Dipole image: perpendicular part kept, parallel part reversed. Wedge: works only when 360∘/α is even. Boss: three images.
  • Parallel plates: infinite images; reciprocity gives −q(L−x)/L; screening as e−πρ/L. Sphere near plane: C=4πϵ0asinh⁡α∑1/sinh⁡nα.

Practise: exam questions on this lesson

From warm-up to top-rank. Each question has a full solution and, where there is a classic slip, a warning. More on the GATE Physics practice page.

Next: multipole expansion explains why the far field of any of these image systems is a dipole or quadrupole. Back to boundary value problems for the series method, or try placing a charge and its image yourself in the field line plotter.

Check yourself

1A charge q sits 5 cm from the centre of a grounded spherical cavity of radius 10 cm. Where is its image?
2A line charge λ runs parallel to a long conducting cylinder that carries −λ per length. What is its image?
3An isolated neutral metal sphere is far from a point charge, at distance d. How does the force on the charge fall off?

Sources

  • J. D. Jackson, Classical Electrodynamics, 3rd ed. (1999), chapter 2
  • W. R. Smythe, Static and Dynamic Electricity, 3rd ed. (1968), chapter 5 (image series, spheres and cylinders)
  • D. J. Griffiths, Introduction to Electrodynamics, 4th ed. (2017), section 3.2
  • L. D. Landau and E. M. Lifshitz, Electrodynamics of Continuous Media, 2nd ed. (1984), section 3
  • GATE Physics (PH) syllabus, Electromagnetic theory section, published by the organising institute each year

Your ticks stay in this browser. Save them as a file on the course home.

Updated 2026-11-22. Not affiliated with the IITs, IISc or the GATE organising institute. The general aptitude part of GATE is not covered here. Check the official GATE website for the current pattern. All questions are original.