Required math: algebra, calculus
Required physics: none
Reference: d’Inverno, Ray, Introducing Einstein’s Relativity (1992), Oxford Uni Press. – Section 6.3; Problem 6.5.
In the last post we derived the covariant derivative of a contravariant vector and found that
Unlike with the Lie derivative, we can’t derive the corresponding expression for a covariant vector by considering coordinate transformations, since the covariant derivative doesn’t rely on these transformations. What is done instead is to require that the covariant derivative obeys the product rule. To show how this works, we first need the covariant derivative of a scalar. Since the ordinary derivative of a scalar already behaves as a rank-one covariant tensor, we can just take the covariant derivative to be the same as an ordinary derivative:
Now we consider the quantity for two arbitrary vectors
and
. Because of the implied sum, this is a scalar quantity, so we get
| |
|
|
| |
|
|
Now if we require the covariant derivative to obey the product rule we must also have
We already have an expression for so if we plug this in and require the result equal to the ordinary derivative above, we get
| |
|
|
| |
|
|
| |
|
|
Since this must be true for any vector , the coefficients of
on each side must be equal and we get
Trackbacks
[...] seen the covariant derivative for the contravariant and covariant vector, but what about higher order tensors? For the special case where the higher order tensor can [...]