Kruskal Coordinates in Schwartzchild metric

  • Level: Graduate 
  • Thread starter mad mathematician
  • Start date
9 replies · 3K views · last reply
mad mathematician
Messages
151
Reaction score
26
I am reading the book by Birrell and Davies on QFT on curved spacetime.
On page 40 they write the following metric:
Equation (3.18): ds^2 = (1 - 2M/r)dt^2 - (1 - 2M/r)^-1 dr^2 - r^2(dtheta^2 + sin^2 theta dphi^2), the Schwarzschild line element.

and on the following page they provide the Kruskal Coordiantes and the changed metric:
Text showing Kruskal coordinates: ū = -4M e^(-u/4M) and v̄ = 4M e^(v/4M), with u = t - r*, v = t + r* and r* = r + 2M ln|(r/2M) - 1|, followed by the metric ds² = (2M/r)e^(-r/2M) dū dv̄ - r²(dθ² + sin²θ dφ²).

Now, I get something a little bit different than theirs.
According to my computations, if we change ##2M/r## to ##(1-2M/r)## and ##e^{-r/2M}## to ##e^{-r^*/2m}##, then I get that by Kruskal transformation that (3.20) is equivalent to (3.18).
Am I right?
 
If I understand you correctly: wouldn't that mean that your version of the Kruskal metric still has a coordinate singularity at r=2M?

So no, that doesn't seem right.
 
mad mathematician said:
Now, I get something a little bit different than theirs.
If you'll post your calculations, we'll be able to comment (and if you do so, please use LaTeX (which enables quoting)).
 
OK, here are my calculations. Let me know where did I get it wrong.
So we have the following: ##d\bar{u}=e^{-u/4M}du## and ##d\bar{v}=e^{v/4M}dv##.
Now, we should multiply ##d\bar{u}d\bar{v}=e^{(v-u)/4M}dudv##.
Now, ##du=dt-dr^* , dv=dt+dr^*##, so multiply and get:
##d\bar{u}d\bar{v}=e^{(v-u)/4M}(dt^2-(dr^*)^2)##, notice that: ##v-u=2r^*##, and ##dr^*=dr+dr/(r/2M-1)=dr/(1-2M/r)##, plug the last two equations and get: ##d\bar{u}d\bar{v}=e^{r^*/2M}(dt^2-dr^2/(1-2M/r)^2)##.
So as we can see, I hope, we get back the Schwarzchild metric (eq. 3.18), if we multiply ##d\bar{u}d\bar{v}## by ##e^{-r^*/2M}(1-2M/r)##.

Your comments are well appreciated, thanks!

Edit: changed what @JimWhoKnew rightly so remarked my mistake.
 
Last edited:
Shouldn't it be
##dr^*=dr+dr/(r/2M-1)=dr/(1-2M/r)##
?

I didn't read beyond that. I'll do, if necessary, after your reply.
 
JimWhoKnew said:
Shouldn't it be
##dr^*=dr+dr/(r/2M-1)=dr/(1-2M/r)##
?

I didn't read beyond that. I'll do, if necessary, after your reply.
good catch, I'll change it. you are right.
 
mad mathematician said:
good catch, I'll change it. you are right.
I lost track.
Can you see that after the correction, you have$$\frac{2M}r e^{-r/2M}d\bar{u}d\bar{v}=\frac{2M}r e^{-r/2M}e^{r^*/2M}(dt^2-dr^2/(1-2M/r)^2)$$which correctly yields the corresponding terms in Schwarzschild metric?
(when ##~r>2M~##)
 
JimWhoKnew said:
I lost track.
Can you see that after the correction, you have$$\frac{2M}r e^{-r/2M}d\bar{u}d\bar{v}=\frac{2M}r e^{-r/2M}e^{r^*/2M}(dt^2-dr^2/(1-2M/r)^2)$$which correctly yields the corresponding terms in Schwarzschild metric?
(when ##~r>2M~##)
There should be a difference between ##r## and ##r^*##, they aren't the same.
 
mad mathematician said:
There should be a difference between ##r## and ##r^*##, they aren't the same.
Of course they are not. That's how you can get the factors right.

Hint: use ##~e^{\ln{x}}=x~## in the equation in #7.
 
Last edited:
  • Like
Likes   Reactions: Nugatory and mad mathematician