Gravastars, Wormholes, and Thin-Shell Objects

Gravastars, wormholes, and thin-shell stars can share an exterior compactness while requiring very different interiors, topology, and stress–energy. The surface layer is physical data, not a bookkeeping device: its density and pressure follow from the jump in extrinsic curvature and may violate energy conditions or possess no known microphysical realization. Static junction balance also does not establish radial or nonradial stability. The comparison here is current through 2026-08-18.

Required background. The Einstein equations supply the geometry and distributional source relation.

Helpful background. Stress–Energy and Conservation supplies energy conditions and shell conservation; Relativistic Matter and Compact Objects supplies material-star comparisons; Black-Hole Geometry supplies horizons and maximal extensions; and Computational Relativity supplies perturbative and nonlinear checks.

Junction data are part of the model

Let a timelike hypersurface $\Sigma$ separate regions ${\cal M}_-$ and ${\cal M}_+$. Choose one unit normal $n^\mu$ pointing from minus to plus, and define the induced metric $h_{ab}$ and $K^\pm_{ab}=h_a{}^\mu h_b{}^\nu\nabla_\mu n_\nu|_\pm$. The normal is spacelike, $n^\mu n_\mu=1$; it points from the interior to the exterior on both sides of an ordinary star shell. Continuity of the first fundamental form is required, while a jump in the second gives the surface stress tensor:

$$8\pi G S_{ab} =-\left([K_{ab}]-h_{ab}[K]\right), \qquad [K_{ab}]=K^+_{ab}-K^-_{ab}.$$

Reversing the normal changes intermediate signs, so normal orientation must be declared before quoting surface density $\sigma$ or pressure $p_s$. Tangential conservation is

$$D_aS^a{}_b =-\left[T_{\mu\nu}n^\mu h^\nu{}_b\right],$$

under the same convention. Israel derives the distributional junction law and surface conservation in Israel 1966, §§ II–IV, with 1967 erratum. The spherical formulas below also follow from Poisson and Visser 1995, § II, Eqs. (3)–(13), PDF, with $G$ restored and the normal signs chosen for the stated gluing.

For a spherical shell at areal radius $a(\tau)$ between $ds_\pm^2=-f_\pm(r)dt_\pm^2+f_\pm(r)^{-1}dr^2+r^2d\Omega^2$, the junction conditions determine $\sigma(a,\dot a)$ and $p_s(a,\dot a,\ddot a)$. Specifying only $f_\pm$ and a static radius is incomplete unless the implied shell stress, equation of state, and response to displacement are accepted.

For a static ordinary star shell, where $r$ increases along the chosen normal on both sides, the nonzero mixed extrinsic curvatures are

$$K^\theta{}_{\theta\,\pm}=K^\varphi{}_{\varphi\,\pm} =\frac{\sqrt{f_\pm(a)}}{a}, \qquad K^\tau{}_{\tau\,\pm}=\frac{f_\pm'(a)}{2\sqrt{f_\pm(a)}}.$$

Writing $S^a{}_b=\operatorname{diag}(-\sigma,p_s,p_s)$ therefore gives

$$\begin{aligned} \sigma&=\frac{\sqrt{f_-}-\sqrt{f_+}}{4\pi Ga},\\ p_s&=\frac{1}{8\pi G} \left[\frac{f'}{2\sqrt f}+\frac{\sqrt f}{a}\right]. \end{aligned}$$

All quantities are evaluated at $a$, with $f_\pm(a)\gt0$. These two relations show why surface pressure cannot be guessed from surface density. For a wormhole joining two exterior regions, $r$ decreases toward the throat on one side and increases away on the other; the corresponding signs of $K^\theta{}_\theta$ change. Reusing the star-shell square-root difference would describe the wrong topology.

Gravastar and wormhole assumptions

A gravastar joins a de Sitter-like core to a Schwarzschild exterior across a material layer, seeking to avoid a horizon and singularity. In the original Mazur–Mottola model, the layer has finite thickness and stiff-fluid equation of state $p=\varepsilon$; its interfaces can also carry surface stresses. The compact thin-layer limit is developed in Mazur and Mottola 2004, preprint “The Vacuum Condensate Model,” Eqs. (7)–(16) and (20)–(21), PDF. A zero-thickness shell with specified $p_s(\sigma)$ is a further idealization. Changing the core, layer equation of state, or interface response requires recomputing regularity and stability for that particular model.

A traversable wormhole has two-sided topology and a throat at a minimum area. In static spherical general relativity, the flare-out condition together with the Einstein equation requires violation of the null energy condition near the throat under the standard assumptions:

$$T_{\mu\nu}k^\mu k^\nu\lt0$$

for some null $k^\mu$. Morris and Thorne give the geometric throat and traversability conditions in Morris and Thorne 1988, §§ III.C.3–III.D, Eqs. (31)–(37), p. 403; § III.F.1, Eqs. (52)–(56), p. 405. Modified gravity, quantum expectation values, nontraversable bridges, and distributional thin-shell wormholes change the hypotheses and cannot be merged into one claim that “wormholes require exotic matter”.

Energy conditions must be checked in the bulk and on the shell. There is a useful distinction between the intrinsic surface null condition and the full spacetime null condition. In an orthonormal shell frame, a spacetime null vector obeys $(k^{\hat\tau})^2=(k^{\hat n})^2+|k_\parallel|^2$, so its contraction with the distributional source has coefficient

$$S_{ab}k^ak^b =\sigma(k^{\hat n})^2+(\sigma+p_s)|k_\parallel|^2.$$

Testing only null vectors tangent to the shell gives $\sigma+p_s\ge0$. Testing all spacetime null vectors additionally requires $\sigma\ge0$: a ray crossing radially has $k_\parallel=0$. This is why a wormhole shell with negative $\sigma$ violates the distributional spacetime NEC even if $\sigma+p_s$ is positive. Curvature scalars on each side do not detect a delta-function shell, so “finite bulk curvature” does not prove a regular stress tensor.

Dated model matrix

This matrix records the principal formulation axes as of 2026-08-18.

Model class	Geometry and source	Evidence and result	Limitation or contrary requirement
Thin-shell gravastar	de Sitter-like interior, Schwarzschild exterior, timelike shell	Israel matching produces a horizonless static configuration at allowed radii.	Surface EOS and anisotropic stress are additional inputs; near-horizon stresses and radial/nonradial stability depend on the limiting procedure.
Finite-thickness gravastar	Smooth anisotropic transition between core and exterior	Can remove a distributional layer and make local energy conditions explicit.	The interpolation is a material model, not unique; causality, radial modes, and formation require separate checks.
Traversable wormhole	Two asymptotic regions joined at a flare-out throat	Static metrics can satisfy geometric traversability conditions.	In ordinary four-dimensional GR, standard throat assumptions force NEC violation; tidal forces, causal access, and stability further restrict parameters.
Thin-shell wormhole	Two exterior regions cut and joined at a shell	Junction conditions give explicit throat stress and a radial potential.	Orientation differs from a star shell and often yields negative surface energy; static balance alone does not imply survival.
Generic thin-shell star	Regular interior joined to vacuum exterior	A compact surface layer can be treated exactly as a distributional source.	Surface composition, sound speed, energy conditions, and coupling to nonspherical perturbations must be supplied.

Cardoso and Pani compare compact alternatives, their required stresses, and observable degeneracies in Cardoso and Pani 2019, §§2–4 and model-summary tables. Competing positions often arise because one analysis allows distributional stress or energy-condition violation while another requires a causal finite-thickness material. Those are different data classes, not conflicting solutions to identical assumptions.

Stability needs a response law

For radial shell motion, the junction equations can often be rearranged as

$$\dot a^2+V(a)=0.$$

A static radius satisfies $V(a_0)=V'(a_0)=0$; linear radial stability requires $V''(a_0)\gt0$. Indeed, the linearized acceleration equation is $\delta\ddot a+[V''(a_0)/2]\delta a=0$. The zero-energy constraint must also be satisfied by the perturbed initial data. Computing $V''$ generally needs a surface response such as $dp_s/d\sigma$, not only the equilibrium values. Nonradial shell modes, bulk fluid modes, and coupling to radiation are independent. For a vacuum bulk on both sides, surface conservation reduces to $d(4\pi a^2\sigma)/d\tau+p_s\,d(4\pi a^2)/d\tau=0$, hence $d\sigma/da=-2(\sigma+p_s)/a$ along a displaced family. Combining this identity with a specified $p_s(\sigma)$ determines the derivatives entering $V''$; see Poisson and Visser 1995, § III, Eqs. (14)–(29), PDF. For exotic matter, $dp_s/d\sigma$ is a response parameter and cannot be called a causal sound speed without a material perturbation model; see Poisson and Visser 1995, § IV, preprint pp. 3–4, PDF.

The unresolved question is whether a proposed stress tensor has a causal, stable microphysical realization and can form without fine-tuned boundary data. A corrected junction calculation can overturn a static branch, and a new unstable mode can invalidate a radial-only verdict. The interpretation could change with a corrected junction or perturbation analysis, a converged nonlinear instability, a microphysical realization of the required stress, or an observation that constrains a named model after its degeneracies and population prior are included.

Independent checks recompute the two extrinsic curvatures with one normal, verify surface conservation, test all claimed energy conditions, examine bulk and distributional curvature, and solve or bound the perturbation spectrum. Compactness by itself discriminates none of these models.

Check a static shell density

Join a flat interior, $f_-=1$, to a Schwarzschild exterior, $f_+=1-2GM/a$, with the standard outward orientation for a star shell. For a static shell,

$$4\pi G a\,\sigma=\sqrt{f_-}-\sqrt{f_+}.$$

Evaluate the dimensionless surface density at $a=3GM$ and interpret its sign.

Solution

At $a=3GM$, $f_+=1-2/3=1/3$, so

$$4\pi G a\,\sigma =1-\frac{1}{\sqrt3} \simeq0.423.$$

The surface energy density is positive in this orientation. This does not establish an energy condition or stability: the tangential pressure must also be computed, and radial stability needs a surface response law and $V''(a_0)>0$. A wormhole cut-and-paste orientation uses different normal data and cannot reuse this sign without recomputing the junction.

References

• Cardoso, Vitor, and Paolo Pani. “Testing the Nature of Dark Compact Objects: A Status Report.” Living Reviews in Relativity 22 (2019): 4. DOI.
• Israel, Werner. “Singular Hypersurfaces and Thin Shells in General Relativity.” Il Nuovo Cimento B 44 (1966): 1–14; erratum 48 (1967): 463. DOI.
• Mazur, Pawel O., and Emil Mottola. “Gravitational Vacuum Condensate Stars.” Proceedings of the National Academy of Sciences 101 (2004): 9545–9550. DOI.
• Morris, Michael S., and Kip S. Thorne. “Wormholes in Spacetime and Their Use for Interstellar Travel: A Tool for Teaching General Relativity.” American Journal of Physics 56 (1988): 395–412. DOI.
• Poisson, Eric, and Matt Visser. “Thin-Shell Wormholes: Linearization Stability.” Physical Review D 52 (1995): 7318–7321. DOI; Open PDF.