The Kerr Metric: Notes For GR-I - CCD

The Kerr Metric
Notes for GR-I - CCD

While perfect spherical symmetry is a powerful assumption in considering many astrophysical bodies, virtually all astrophysical systems will involve some rotation. Thus it is important to consider solutions to the Einstein Field Equations which allow the introduction of rotation, that is they break the symmetry assumptions of the Schwarzschild solutions. In particular, such a rotating solution will be axially symmetric, and thus have fewer Killing vectors that the case of spherical symmetry. In 1963, a vacuum solution exhibiting dierential rotation of the spacetime was discovered by Roy Kerr. This was the rst exact vacuum solution to have this property. It has since been extended to include the case where there is an electric charge, etc. In Boyer-Lindquist coordinates (t, r, , ), the Kerr metric takes the form ds2 = dr2 + d2 + (r2 + a2 ) sin2 d2 2 2mr + a sin2 d dt dt2
(1)
where we have dened = r2 + a2 cos2 and = r2 2mr + a2 (2)
and we are using the signature ( + ++) here. For large r this metric approaches the Minkowski metric. It is useful to transform to Cartesian spatial coordinates at large r, dened by 2 x + y2 y 2 2 2 2 1 r = x + y + z , = tan (3) and = tan1 z x We then have g00 = 1 2m 2may 2max , g0 x = , g0 y = and g0z = 0 3 r r r3 (4)
These results indicate that at large r, the metric approaches the Schwarzschild metric, since the g00 falls o more slowly than the g0x and g0y terms, which fall o at least as fast as r2 . Further, the crossed nature of the g0x and g0y terms along with g0z = 0 indicates that there is rotation, proportional to the value of a, about the z axis.
c Charles C. Dyer
An indication of the presence of rotation is that if the direction of time is reversed, the suspected rotation should also reverse its direction. That rotation is present about the z axis, for a = 0, can thus be seen by noting that taking t t and together has no impact on the metric. Further, noting that neither taking t t and a a together, nor taking and a a together, alters the metric, it is clear that a is a measure of that rotation. When a = 0, neither t t nor has any eect on the metric, conrming that in this situation, there is no rotation. There is clearly a problem, in these coordinates, when = 0, since the coecient of dr2 diverges. This is similar to the coordinate problem at r = 2m in the Schwarzschild vacuum solution. The vanishing of leads to the two values r = m m2 a2 (5) when |a| < m. We will normally consider only this case, since for |a| > m the rotation is so rapid that it leads to eects that we do not expect to observe in the universe. We will nd that the surfaces r are event horizons like the surface r = 2m in the spherically symmetric case, since r+ 2m and r 0 as a 0. We shall refer to r+ and r as the outer and inner horizons, respectively. To handle the 0 problem, we proceed in analogy with the transition to EddingtonFinkelstein coordinates in the Schwarzschild vacuum metric. We choose a new coordinate, v , related to photons infalling from innity. In addition, we choose a new azimuthal angle, , to replace the azimuthal angle, , which takes us to a locally corotating frame. The transformations are given by dv = dt + r 2 + a2 a dr and d = d + dr (6)
In these new coordinates, (v, r, , ), called Kerr coordinates, the Kerr metric takes the form ds2 = d2 2a sin2 drd + 2drdv sin2 + (r2 + a2 )2 a2 sin2 d2 4m 2mr ra sin2 ddv 1 dv 2
(7)
The problem at = 0, that is where r = r , is no longer present. These coordinates provide a representation of the Kerr spacetime which extends right down to the ring singularity at r = 0 and = /2. To have a rst look at this singularity, we consider the Kretschmann scalar dened as K = Rabcd Rabcd , and after a fair bit of algebra yields: K = 48m2 (2 16a2 r2 cos2 )(r2 a2 cos2 )/6 (8)
Thus the Kretschmann scalar only diverges where goes to zero, and this requires that both r2 and cos2 vanish, that is where r = 0 and = /2, where we are taking a to be non-zero.
Examination of the Kerr solution indicates that there are two Killing vector elds. In either the Boyer-Lindquist coordinate system (t, r, , ), or the Kerr coordinate system (v, r, , ), the Killing vectors can be written as a = (0, 0, 0, 1) and a = (1, 0, 0, 0) (9)
as can be veried by noting that no components of the metric tensor depend on any of t, v , , or . Clearly a is associated with the rotational symmetry, so that no metric components depend on . Similarly, a is associated with the stationary nature of the metric, so that no metric components depend on t or v . Since we have already concluded that for a = 0, the solution is rotating, the solution is not static. The Killing vector a is space-like everywhere. The Killing vector a can change its nature, since 2mr a a = 1 (10) For very large r relative to m and a, we see that a a 1, so that a is time-like for large r. is null when 2mr = , ie. when r2 2mr + a2 cos2 = 0 which occurs when r =m m2 a2 cos2 (11) (12)
For most of our discussion, we will only be interested in the surface dened by the plus sign in this equation. We will refer to this surface as the stationary limit surface. It is important to note that this surface is not a sphere. In fact it touches the outer horizon at the poles, where r = r+ , but it bulges out at the equator. The region bounded by the stationary limit surface and the sphere r = r+ is called the ergo region. Outside this stationary limit surface, a is a time-like vector, so it is possible to dene c a a a time-like Killing observer whose 4-velocity is u = / c . While this observer will observe local rotation, he will see the same spacetime for all his time of observation. Inside this stationary limit surface, when a becomes space-like, there can be no such observer. This justies the reference to the stationary limit surface. Inside the stationary limit surface, while there will obviously be time-like observers, their world lines cannot also be the trajectories of the Killing vector eld a . Thus all physical observers in this region will see a dynamic world.
Axis of rotation Central ring singularity Stationary Limit Surface = 2mr r+ Ergo Region r
Inner horizon
Outer horizon
Figure 1: Cross-sectional view of Kerr Black Hole

The Kerr Metric: Notes For GR-I - CCD

Uploaded by

Copyright:

Available Formats

The Kerr Metric: Notes For GR-I - CCD

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Kerr Metric: Notes For GR-I - CCD

Uploaded by

Copyright:

Available Formats

The Kerr Metric

Notes for GR-I - CCD

where we have dened = r2 + a2 cos2 and = r2 2mr + a2 (2)

Figure 1: Cross-sectional view of Kerr Black Hole

You might also like