经典电动力学中电磁张量
一般而言,如果存在麦克斯韦方程组:
\begin{align}
\nabla\cdot \bf{E}=\rho \\
\nabla \times \bf B – \frac 1 c \frac{\partial \bf E}{\partial t}=j \\
\nabla \cdot \bf B = 0 \\
\nabla\times \bf E + \frac 1 c \frac{\partial \bf B}{\partial t}=0
\end{align}
We would introduce a vector potential \(\bf A\) and scalar potential \(\phi\), which are defined as \(\bf B=\nabla\times\bf A\) and \(\bf E=-\nabla\phi – \frac{\partial \bf A}{\partial t}\). More generally, it can be extended to a rank four vector \(A^a=(\phi,\bf A)\).
This would remind us of a EM tensor \(F^{ab}=\partial^a A^b\). However, it is really annoying that F should have only six components.
Then we would like to transform it into a symmetric and traceless tensor \(F^{ab}=\partial^aA^b+\partial^bA^a-\frac 1 2\delta^{ab}F^{a}_b\). After a simple and straight forward calculation, I find the final expression of Maxwell Eqns are not beautiful enough because a lot of quantities can not be merged.
Another method is to make F a antisymmetric one, i.e., \(F^{ab}=\partial^aA^b-\partial^bA^a\). And this leads to good results. Here we reaches the standard definition of \(F^{ab}\).
A more advanced understanding is , it is better to define a quantity with a certain meaning. Here we would like to define F as a flux through a volume element \(\bf{\epsilon}\), which is in correspondence with the old quantities we use. That is to say, F is a 2-form. So it is definitely anti-symmetric tensor.
另一条路是通过相对论的Lagrangian推导四维形式的运动方程啊,会直接得到电磁反对称张量。
==,我只是想说,定义成对称无迹张量也可以……
无忌真是个好东西啊,quadrupole也是无忌,不过他用了球鞋函数来说明,没搞明白。。