I wanted to see if the divergence of the lienard wiechert field follows Maxwell's equations (gauss' law):

$$\nabla \cdot \vec{E} = 0$$


$$ E(r,t)=\frac{e}{\gamma^2 R^2} \frac{n-\beta}{(1-n\cdot\beta)^3} + \frac{e}{cR} \frac{n \times ((n-\beta)\times \beta')}{(1-n\cdot\beta)^3}$$

But I get a non-zero answer! I calculated it for a very simple test case of a stationary observation point at $(x,y,z)$ and a source charge at some retarded time, t', is located at the origin $(0,0,0)$ with a velocity $\vec\beta=\beta<1,0,0>$ and acceleration $\beta'=\frac{\beta^2 c}{\rho} <0,-1,0>$.

This is simple circular motion. After computing the field, I compute the divergence ($\nabla = \frac{\partial}{\partial x} \hat{x}+ \frac{\partial}{\partial y} \hat{y}+ \frac{\partial}{\partial z} \hat{z}$), but again I do not get a zero answer!

Am I doing this correctly? Do I have to take the divergence of evaluated at some retarded time? I thought that the divergence of the field should be respective of the observation point and therefore be in the current time-space coordinates, not the retarded.

Any help would be greatly appreciated!

  • $\begingroup$ Do you get non-zero divergence even for the case where $\beta'$ is zero all the time (rectilinear motion)? $\endgroup$ Oct 9, 2018 at 9:50
  • $\begingroup$ shouldn't it satisfy $\nabla \cdot E = \delta(x-x_0)$, with $x_0$ the location of the source? $\endgroup$ Oct 12, 2018 at 16:40
  • $\begingroup$ Thats exactly what I thought! Afterall, this is how the LW field is solved for in the first place. But, is that only for the point $x_0$ where the charge is located? Kind of like the static case ($\nabla \cdot E =0$ except at $x_0$)? Im just kind of confused what to expect! $\endgroup$ Oct 13, 2018 at 1:03

1 Answer 1


The Lienard-Wiechert solution describes the electromagnetic fields emitted from a single point charge, ignoring backreaction of the emitted fields on the charge. That is, take a trajectory

$$\mathbb{R} \mapsto \mathbb{R}^4 $$ $$ \tau \mapsto x^\nu(\tau) $$

Then the current corresponding to a particle of charge $e$ moving along the trajectory is

$$ j^\nu(x) = e \int_{-\infty}^\infty \frac{ d x^\nu(\tau)}{ d \tau} \delta^{(4)}(x - x(\tau)) d \tau $$

Then the Lienard-Wiechert field is the solution to

$$ \frac{\partial}{\partial x^\mu} F^{\mu \nu}(x) = j^\nu(x) \ .$$

Note that a reasonable assumption on the trajectory of the particle is that it should move forward in time, i.e.

$$ \frac{ d x^0(\tau)}{ d \tau} > 0 $$

Then $x^0(\tau)$ is a strictly monotonous function of $\tau$, and thus, if $\tau_0$ is the unique solution to $x^0 = x^0(\tau_0)$ for fixed $x_0$,

$$ \delta(x^0 - x^0(\tau)) = \frac{\delta(\tau - \tau_0)}{\left| \frac{ d x^0}{ d \tau}(\tau_0) \right|} \ .$$

We may thus evaluate the current to

$$ j^0(x_0,\mathbf{x}) = e \delta(\mathbf{x} - \mathbf{x}(\tau_0)) \ , \ \ \ \mathbf{j}(x^0,\mathbf{x}) = e \frac{\frac{ d \mathbf{x}}{ d \tau}(\tau_0)}{\left| \frac{ d x^0}{ d \tau}(\tau_0) \right|} \delta(\mathbf{x} - \mathbf{x}(\tau_0)) \ . $$

Note that this depends on $x^0$, since $\tau_0$ is a function of $x^0$. For simplicity one could choose $\tau_0 = x^0$, but this is not necessary.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.