Exact Bianchi Solution Families

The Bianchi label fixes a spatial Lie algebra, not a unique spacetime. Each type contains families distinguished by the time-dependent invariant metric, matter, tilt, and global identifications. Representative exact solutions nevertheless reveal what each algebraic curvature term does: type I supplies free Kasner motion, type II a single curvature transition, type V the open isotropic limit, and type IX a closed curvature potential with Taub and isotropic subfamilies.

Required background. Einstein’s Equation supplies the homogeneous constraints and evolution; Continuum Kinematics and Balance Laws supplies perfect-fluid and tilted matter variables.

Helpful background. The Riemann Curvature Tensor supplies spatial curvature; Exact Solutions of Einstein’s Equations supplies the atlas context; Killing Fields and Isometry Groups supplies homogeneous group actions.

A common metric language

With invariant one-forms $\omega^i$, write an orthogonal homogeneous metric as

$$ds^2=-N^2(t)dt^2+\sum_{i=1}^3a_i^2(t)(\omega^i)^2.$$

Diagonal form is consistent for important non-tilted subclasses, not universal. It is useful to split the volume and anisotropy by

$$a_i=e^{\alpha+\beta_i}, \qquad \sum_i\beta_i=0.$$

Then $e^{3\alpha}$ is the spatial volume scale, while two independent $\beta_i$ describe shear. The structure constants generate a type-dependent spatial-curvature potential in the Hamiltonian constraint. Exact families correspond to potentials simple enough to integrate or to symmetry reductions that remove one anisotropy.

Types I, II, and V

Type I. Here $d\omega^i=0$, so the spatial curvature vanishes. Vacuum gives the Kasner family $a_i=t^{p_i}$. The isotropic type-I sector is spatially flat FLRW when supported by suitable matter or a cosmological constant; it is not a nontrivial point on the vacuum Kasner circle, because equal exponents cannot satisfy both $\sum_i p_i=1$ and $\sum_i p_i^2=1$. Perfect-fluid type-I families can interpolate between shear-dominated and fluid-dominated regimes according to the equation of state. Vacuum is not necessarily flat because the extrinsic curvature can carry Weyl curvature.

Type II. In a canonical basis, $d\omega^1=-\omega^2\wedge\omega^3$ and the other two exterior derivatives vanish. The general diagonal vacuum type-II family is triaxial: its three scale factors need not contain an equal pair. A locally rotationally symmetric vacuum Taub subfamily is obtained by imposing equality of the two transverse scale factors. Both the general family and that specialization exhibit one exact transition between two asymptotic Kasner regimes. In Hamiltonian language, one exponential curvature wall reverses the anisotropy momentum. Type II is therefore the elementary local model for a type-IX Kasner transition, although iterating such transitions is an asymptotic dynamical picture rather than one global type-II metric.

Type V. One may take

$$\omega^1=dx, \qquad \omega^2=e^xdy, \qquad \omega^3=e^xdz.$$

For $a_1=a_2=a_3=a(t)$, the spatial metric has constant negative curvature and gives open FLRW. The vacuum choice $a=t$ is Milne spacetime, a wedge of Minkowski space. Anisotropic type-V solutions retain the class-B algebra and must satisfy its momentum constraint; they cannot be obtained reliably by copying a class-A reduced Hamiltonian.

Type IX and locally rotationally symmetric sectors

Type IX has $SU(2)$ invariant forms, which in a fixed convention satisfy cyclic Maurer–Cartan equations such as $d\omega^1=-\omega^2\wedge\omega^3$. Equal scale factors produce closed $k=+1$ FLRW. Setting two scale factors equal gives a locally rotationally symmetric sector with an additional axial isotropy. The vacuum Taub region and its Taub–NUT analytic extensions lie in this sector; coordinate continuation, causal structure, and the meaning of the NUT parameter require a separate analysis of that spacetime.

Generic diagonal vacuum type IX has two anisotropy degrees of freedom and is not represented by a single elementary closed formula. Toward a spacelike singularity it can undergo a sequence of intervals approximated by Kasner solutions, connected by type-II-like curvature transitions. Exact Kasner and Taub metrics calibrate that description, but they do not by themselves prove every statement about the full oscillatory dynamics.

Ellis and MacCallum derive the homogeneous equations and representative class A/B structures in Ellis and MacCallum 1969, §§4–6, pp. 121–141. Exact metric families and parameter identifications are cataloged in Stephani et al. 2003, chs. 12–14.

Comparing families correctly

A useful comparison specifies: the invariant one-form convention; class A or B; diagonal or off-diagonal metric; matter and tilt; lapse and time range; spatial topology; and the limit being taken. Three checks are especially discriminating:

1. set all anisotropies to zero and verify the allowed FLRW curvature sign;
2. remove spatial structure constants and recover type-I equations only when the contraction is a legitimate algebra limit; and
3. compute the Hamiltonian and momentum constraints directly after imposing an LRS or matter ansatz.

The label “Taub” is not enough: type-II Taub dynamics and the type-IX Taub–NUT family are related historically and asymptotically but are not the same spacetime.

Exercise

Show that the isotropic type-V spatial metric is hyperbolic in upper-half-space coordinates.

Solution

The unit spatial metric is $dx^2+e^{2x}(dy^2+dz^2)$. Set $q=e^{-x}$, so $dx=-dq/q$ and $e^{2x}=q^{-2}$. Then $d\Sigma^2=q^{-2}(dq^2+dy^2+dz^2)$, the standard upper-half-space metric of sectional curvature $-1$. Multiplication by $a^2(t)$ gives physical curvature $-1/a^2$.

References

• Ellis, George F. R., and Malcolm A. H. MacCallum. “A Class of Homogeneous Cosmological Models.” Communications in Mathematical Physics 12 (1969): 108–141. DOI.
• Stephani, Hans, Dietrich Kramer, Malcolm MacCallum, Cornelius Hoenselaers, and Eduard Herlt. Exact Solutions of Einstein’s Field Equations. 2nd ed. Cambridge: Cambridge University Press, 2003. DOI.