More about the physics
From QED
Contents |
RHIC
The RHIC facility at Brookhaven National Laboratory includes four experiments: BRAHMS, PHENIX, PHOBOS, and STAR.
The QCD coupling
Siggi Bethke is one of the modern experts on the running coupling of QCD. here is a talk he gave on the subject; see in particular the figure on this page, reproduced here, showing αs as a function of scale. An often-used value in RHIC physics is αs = 1 / 2.
Ian Hinchliffe has a summary page on αs, including world-average values at the Z peak and a Java script that tells you αs(Q) for any Q you enter.
Black hole temperature
A calculation which comes up often in black hole physics is the Hawking temperature. When the black hole metric has the form
where the
represents some additional terms that are static and radially symmetric, this is quite an easy computation. One first rotates to Euclidean time, upon which
and then demands that the horizon, where gtt and grr have simple zeroes (simple, at least, in the case where the temperature is non-zero), should become a non-singular point in the Euclidean geometry. The
part of the calculation doesn't affect the calculation because it's supposed to remain non-singular at the horizon. Assuming the horizon is at r = 0, and expanding gtt and grr to first order in r, one finds
where I have introduced near-horizon coordinates
Imposing that θ should be periodic with period 2π, and that tE should have period β = 1 / T, one immediately finds
For example, if − gtt = grr = 1 − 2M / r, as for the four-dimensional Schwarzschild black hole, then
It's also nice to know that T = κ / 2π, where κ is the surface gravity at the horizon. This allows for a calculation of T that doesn't involve Wick rotation, but it's usually not the fastest method. Finally, it's important to recall that T is computed with respect to a particular asymptotic Killing time: in the current context that time of course is t, and its normalization determines the normalization of T.
Conformal Fourier integrals
A useful result in dealing with conformal correlators is (in Euclidean signature)
Here
represent terms associated with regularization. For a limited range of Δ, the integral converges, and these terms aren't there. A typical way to deal with them is to say that 1 / x2Δ has some distribution added to it which is supported at x = 0; correspondingly, p2Δ − 4 has some terms analytic in p2 added to it. In short, these are contact terms.













