Black Strings, Branes, and Kaluza–Klein Horizons

Black strings and branes extend a black-hole horizon along one or more flat directions. The simplest examples are exact product metrics, yet their global physics differs from that of a localized hole: noncompact directions lead to charge densities, compact directions introduce physical lengths and gravitational tensions, and the horizon can become unstable at wavelengths that fit around the compact circle. These distinctions organize the Kaluza–Klein phase problem without confusing exactness, equilibrium mechanics, and dynamical stability.

Helpful background. Exact Solutions of Einstein’s Equations supplies product metrics and their coordinate domains; Global Structure supplies horizon topology and asymptotics; the Einstein equation supplies the Ricci-product check; Black-Hole Perturbation Theory supplies the Gregory–Laflamme eigenvalue problem; and Black-Hole Geometry and Mechanics supplies the charge and first-law framework.

Product horizons

Let $ds_d^2$ be a $d$-dimensional Ricci-flat black-hole metric and append $p$ Cartesian directions:

$$ds_D^2=ds_d^2+\sum_{i=1}^{p}(dz^i)^2, \qquad D=d+p.$$

The Levi–Civita connection has no mixed components, so

$$R_{\mu\nu}^{(D)}=R_{\mu\nu}^{(d)}, \qquad R_{ij}^{(D)}=R_{\mu i}^{(D)}=0.$$

Thus a Ricci-flat seed gives an exact Ricci-flat product. For a Tangherlini seed,

$$ds_D^2=-f(r)dt^2+\frac{dr^2}{f(r)} +r^2d\Omega_{d-2}^2+\sum_i(dz^i)^2,$$ $$f(r)=1-\left(\frac{r_0}{r}\right)^{d-3}.$$

If every $z^i$ is noncompact, a horizon section is $S^{d-2}\times\mathbb R^p$ and the natural extensive quantities are densities. If $z^i$ has period $L_i$, the corresponding factor is a circle and the section becomes $S^{d-2}\times T^p$. The local curvature is unchanged by the flat product, but the asymptotic geometry, topology, and charges are not.

In particular, the field falls as $r^{-(d-3)}$ in the transverse radius. It does not fall as $r^{-(D-3)}$, because flux does not spread along directions in which the solution is uniform. This is why a five-dimensional uniform string and a five-dimensional localized Tangherlini hole cannot share the same ADM normalization.

Mass density and compactification

Let

$$V_p=\prod_iL_i$$

be the compact worldvolume. Dimensional reduction gives $G_d=G_D/V_p$, and the seed mass formula becomes

$$\frac{M}{V_p} =\frac{(d-2)\Omega_{d-2}}{16\pi G_D} r_0^{d-3}.$$

The left side is the mass density of the noncompact brane or the total mass divided by the compact volume. Sending a period to infinity at fixed density makes the total mass diverge. Conversely, changing a compact period changes global boundary data even though the local metric coefficients near the horizon can remain unchanged.

For one compact string in total dimension $D$, the seed has dimension $d=D-1$. Its mass, area, and surface gravity are

$$M=\frac{(D-3)\Omega_{D-3}}{16\pi G_D} Lr_0^{D-4},$$ $$\mathcal A_H=\Omega_{D-3}Lr_0^{D-3}, \qquad \kappa=\frac{D-4}{2r_0}.$$

These formulas recover the lower-dimensional Tangherlini mechanics at fixed $L$.

Tension as a thermodynamic charge

The compact length is a physical modulus, and its conjugate is the gravitational tension $\mathcal T$. The first law is

$$dM=\frac{\kappa}{8\pi G_D}\,d\mathcal A_H +\mathcal T\,dL.$$

For the uniform neutral string, varying the formulas above at fixed area gives

$$\mathcal T=\frac{M}{(D-3)L}.$$

Thus the dimensionless relative tension is $\mathcal TL/M=1/(D-3)$. Scaling every length produces the Kaluza–Klein Smarr relation

$$(D-3)M=(D-2)\frac{\kappa\mathcal A_H}{8\pi G_D} +\mathcal TL.$$

Substitution supplies an independent check: the horizon term equals $(D-4)M/(D-3)$ before multiplication by $D-2$, and the tension term supplies the remainder required by homogeneity. Kastor and Traschen verify the uniform-string tension as the fixed-area derivative in Kastor and Traschen 2006, §3.1, pp. 3–4, eqs. (7)–(11). The general Kaluza–Klein scaling law, specialized here to zero momentum, gives the displayed Smarr relation Kastor, Ray, and Traschen 2007, §7, pp. 13–15, eqs. (51)–(53).

Tension is not ordinary material pressure placed on the horizon. It is an asymptotic gravitational charge conjugate to changing the compactification length. Different asymptotic prescriptions or several periods require a corresponding set of tensions and a carefully specified ensemble.

Kaluza–Klein branches and stability interface

With asymptotics $\mathbb R^{1,D-2}\times S^1$, at least three geometrically distinct stationary possibilities arise:

• a uniform string, invariant along the circle;
• a nonuniform string, whose horizon radius varies around the circle but whose horizon still wraps it;
• a localized black hole, whose horizon is spherical and does not wrap the circle.

The uniform product solves the exact vacuum equations for every $L$, but sufficiently long strings possess growing long-wavelength modes. Gregory and Laflamme established this instability by solving the linearized problem in Gregory and Laflamme 1993, pp. 2837–2840. The threshold is controlled by a dimensionless combination such as $L/r_0$ or $k_{\mathrm{GL}}r_0$ and depends on dimension.

This is a stability boundary, not a failure of background regularity. A regular exact horizon can be dynamically unstable. The nonlinear endpoint also cannot be inferred from the sign of a single mode: nonuniform strings, localized holes, topology-changing merger geometries, radiation, and possible pinch-off behavior require separate global and dynamical evidence.

Boosting a string before compactification adds momentum along the circle; dimensional reduction then produces lower-dimensional gauge and scalar charges. Those constructions are related but do not have the same charge content as the unboosted neutral product used here.

Check your understanding

Derive the uniform-string tension while holding the horizon area, rather than the radius, fixed.

Solution

At fixed area,

$$0=d\log\mathcal A_H =d\log L+(D-3)d\log r_0,$$

so

$$d\log r_0=-\frac{d\log L}{D-3}.$$

Since $M\propto Lr_0^{D-4}$,

$$d\log M =d\log L+(D-4)d\log r_0 =\frac{d\log L}{D-3}.$$

Therefore

$$\left(\frac{\partial M}{\partial L}\right)_{\mathcal A_H} =\frac{M}{(D-3)L}=\mathcal T.$$

Holding $r_0$ fixed instead would give $M/L$ and would not compute the first-law conjugate.

The next page asks which compact horizon topologies are ruled out before any explicit metric is constructed: Horizon Topology Theorems in Higher Dimensions.

References

• Gregory, Ruth, and Raymond Laflamme, “Black Strings and $p$-Branes Are Unstable,” Physical Review Letters 70, 2837–2840 (1993). DOI.
• Kastor, David, and Jennie Traschen, “Stresses and Strains in the First Law for Kaluza–Klein Black Holes,” Journal of High Energy Physics 2006(09), 022 (2006). DOI. Open PDF.
• Kastor, David, Sourya Ray, and Jennie Traschen, “The First Law for Boosted Kaluza–Klein Black Holes,” Journal of High Energy Physics 2007(06), 026 (2007). DOI. Open PDF.