Field of an Infinite Line
Chapter 1 — Electrostatics
A long, straight, uniformly charged wire is the canonical cylindrically symmetric source. With Gauss's law the field drops out in two lines: it points radially outward from the wire and falls off as \(1/r\) — slower than a point charge, faster than an infinite plane.
Setup and symmetry
Take an infinite straight line of charge along the \(z\)-axis with uniform linear density \(\lambda\) (C/m). Symmetries:
- Translation along \(z\): \(\vec E\) doesn't depend on \(z\).
- Rotation about the wire: \(\vec E\) has no \(\hat\phi\) component.
- Reflection in any plane containing the wire: no \(\hat z\) component either.
Left with \(\vec E = E(r)\hat r\) in cylindrical coordinates.
Gauss's law cylinder
Coaxial cylinder of radius \(r\), length \(L\). Flux contributions:
- End caps: \(\vec E\cdot d\vec A = 0\) (field is radial, caps have normals along \(\hat z\)).
- Curved side: \(\vec E\parallel d\vec A\), magnitude \(E(r)\) constant, area \(2\pi r L\), flux \(E\cdot 2\pi r L\).
Enclosed charge: \(\lambda L\). Gauss:
\[E\cdot 2\pi r L = \frac{\lambda L}{\epsilon_0} \implies \boxed{E = \frac{\lambda}{2\pi\epsilon_0 r}}.\]
Direction: outward for \(\lambda > 0\), inward for \(\lambda < 0\).
Why \(1/r\) and not \(1/r^2\)?
Field lines from a line of charge spread into a cylinder, not a sphere. Total flux through any cylinder is constant (\(\lambda L/\epsilon_0\)), while the cylinder's curved area grows as \(r\), so \(E\propto 1/r\). You can check this by brute-force integrating Coulomb's law over the wire too — it's a calculus exercise, but Gauss is vastly easier.
Variants
Infinite cylindrical shell (radius \(R\), uniform \(\sigma\)): zero field inside; \(E = \lambda_{\text{eff}}/(2\pi\epsilon_0 r)\) outside, where \(\lambda_{\text{eff}} = 2\pi R\sigma\).
Solid infinite cylinder (radius \(R\), uniform volume density \(\rho\)): for \(r < R\), \(Q_{\text{enc}} = \rho \pi r^2 L\) gives \(E = \rho r/(2\epsilon_0)\). For \(r > R\), point-line behavior \(E = \rho R^2/(2\epsilon_0 r) = \lambda/(2\pi\epsilon_0 r)\).
Practice Problems
Hint
Solution
\(E = (2\times 10^{-6})/(2\pi\cdot 8.854\times 10^{-12}\cdot 0.1) \approx 3.6\times 10^5\) N/C.
Answer: \(\approx 3.6\times 10^5\) N/C, radially outward.
Hint
Solution
Each wire gives \(\lambda/(2\pi\epsilon_0 (d/2)) = \lambda/(\pi\epsilon_0 d)\). At the midpoint, the two contributions point in opposite directions (each away from its own wire) and cancel.
Answer: \(\vec E = 0\).
Hint
Solution
\(r < b\) (between inner wire and outer shell): \(Q_{\text{enc}} = \lambda L\), \(E = \lambda/(2\pi\epsilon_0 r)\). \(r > b\): \(Q_{\text{enc}} = 0\), \(E = 0\).
Answer: \(E = \lambda/(2\pi\epsilon_0 r)\) inside; \(E = 0\) outside.
Hint
Solution
Inside (\(r < R\)): \(E\cdot 2\pi r L = \rho\pi r^2 L/\epsilon_0\), so \(E = \rho r/(2\epsilon_0)\). Outside: \(E = \rho R^2/(2\epsilon_0 r)\). At \(r = R\), both give \(\rho R/(2\epsilon_0)\). Check.
Answer: \(E = \rho r/(2\epsilon_0)\) inside, \(E = \rho R^2/(2\epsilon_0 r)\) outside.