5
$\begingroup$

In electromagnetism, we can write the Bianchi identity in terms of the field strength tensor $F_{\mu \nu}$ as,

$$ \partial_{\lambda} F_{\mu \nu} + \partial_{\mu} F_{\nu \lambda}+ \partial_{\nu} F_{\lambda \mu} = 0,\qquad \mu,\nu,\lambda=0,1,2,3. \tag{1}$$

Now, in a textbook I am reading (Classical Covariant Fields - Burgess), the Bianchi identity is given as,

$$ \sum_{j,k=1}^3\epsilon_{ijk} \partial_j E_k + \partial_t B_i = 0,\qquad i=1,2,3.\tag{2a}$$

and $$\sum_{i=1}^3\partial_i B_i = 0.\tag{2b} $$

However, I am struggling to see how these two forms are equivalent, i.e. starting from one equation (1), how can we arrive at the two others (2)?

$\endgroup$
3
  • 3
    $\begingroup$ Do you know the relationship between the field strength tensor $F_{\mu\nu}$ and the $E$ & $B$ fields? $\endgroup$
    – Qmechanic
    Commented Dec 2, 2016 at 12:10
  • $\begingroup$ Sure - see en.wikipedia.org/wiki/Electromagnetic_tensor Thanks for the edits to the question - could you explain these edits as I'm not sure I follow. Thanks! $\endgroup$ Commented Dec 2, 2016 at 13:00
  • $\begingroup$ The edits were made to make the statement more precise. $\endgroup$
    – Qmechanic
    Commented Dec 2, 2016 at 13:15

2 Answers 2

10
$\begingroup$

Let's start by contracting the first equation with the 4-dimensional totally antisymmetric tensor $\epsilon^{\alpha\lambda\mu\nu}$. Thanks to the properties of $\epsilon^{\alpha\lambda\mu\nu}$ we then have

$$ \epsilon^{\alpha\lambda\mu\nu} \partial_{\lambda} F_{\mu\nu} = 0 . $$

Next we separate the Faraday tensor into its temporal (0) and spatial (1,2,3) components. This gives us the electric and magnetic fields: $F_{00}=0$; $F_{k0}=-F_{0k}=E_k$; $F_{ij}=\epsilon_{ijk} B^k$. If we set $\alpha=0$ we get $$ \epsilon^{0ijk} \partial_{i} F_{jk} = \epsilon^{ijk} \partial_{i} (\epsilon_{jkp} B^p) = 2 \partial_{i} B^i = 2 \nabla\cdot\mathbf{B} = 0 , $$ because $\epsilon^{ijk} \epsilon_{jkp}=2\delta_p^i$. For $\alpha=i$ we have: $$ \epsilon^{ij0k} \partial_{j} F_{0k} + \epsilon^{ijk0} \partial_{j} F_{k0} + \epsilon^{i0jk} \partial_{0} F_{jk} = 0 . $$

$$ - 2 \epsilon^{ijk} \partial_{j} E_{k} - \epsilon^{ijk} \partial_{0} (\epsilon_{jkp} B^p) = 0 . $$

$$ \epsilon^{ijk} \partial_{j} E_{k} + \partial_{0} B^i = 0 . $$

$\endgroup$
2
  • $\begingroup$ That's great! Could you clarify for me how the initial contraction with the antisymmetric tensor works? Thanks! $\endgroup$ Commented Dec 2, 2016 at 16:22
  • $\begingroup$ Also, the form from Burgess refers to covariant magnetic field components, whilst your form seems to have a contravariant component? Thanks! $\endgroup$ Commented Dec 2, 2016 at 16:39
1
$\begingroup$

In reply to the additional query (posed in a comment): without loss of generality, assume i = 1. Then: $$ϵ^{ijk} ϵ_{jkp} = ϵ^{123}ϵ_{23p} + ϵ^{132}ϵ_{32p} = ϵ_{23p} - ϵ_{32p} = 2 ϵ_{23p} = 2 δ^1_p = 2 δ^i_p.$$

$\endgroup$

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.