Suppose, in a 2D plane, there is an electric field: $$\vec{E}(x,y)=E_0\begin{pmatrix}-y \\ +x\end{pmatrix}=E_0(-y\vec{e}_x+x\vec{e}_y)$$ $$\vec{E}(r,\theta)=E_0 r \vec{e}_\theta$$ where $r,\theta$ are polar coordinates. This electric field goes around in a circle and closes in itself. What would the potential, defined through $\vec{E}=-\nabla V$, be as a function of $\theta$ at a given radius from the origin?

If we define the $-\infty<\theta<+\infty$, it would be a simple linear function in $\theta$: $V(r,\theta)=-E_0 r \theta$. But if we restrict $0\leq\theta<2\pi$, surely it would become a matter of convention, whether it is a single-valued function (therefore leading to a discontinuity in $\vec{E}$) or a multi-valued function?


1 Answer 1


I think you're missing something very important here. The electric field you definied , $\vec{E} = E_0(-y,x)$ is not conservative, for $\vec{E}$ to be conservative you need $\oint\vec{E}d\vec{l}$ to be $0$, which is not the case because $\frac{\partial(x)}{\partial{x}} \neq \frac{\partial(-y)}{\partial{y}}$ (Green's theorem). So you cannot assign a scalar potential V to $\vec{E}$, that's probably why you are having problems.

Also note that you would get a similar looking magnetic field $\vec{B}$ when there was a current along the z-axis. Ampère's circuital law states that: $$\oint_C\vec{B}.d\vec{l} = \mu_0I \neq 0$$ So you can't assign a scaler potential to the magnetic field as well.


Your Answer

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

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