1: %version 20-05-2005
2: \section{Geometric relations for hypersurfaces}
3: \label{geom}
4: The time-slicing (or ADM) formalism \cite{Wald:1984rg,MTW}, which we
5: employ in our analysis of Einstein's equations, makes essential use of
6: the geometry of hypersurfaces \cite{Eisenhart}. Therefore, we shall
7: begin with a review of the basic relations governing their
8: geometry.
9:
10:
11: A hypersurface in a space-time with coordinates $X^\mu$
12: ($\mu=0,\ldots, d$) and metric $\tg_{\mu\nu}$ is defined by a set of
13: $d+1$ functions, $X^\mu(x^i)$ ($i=1,\ldots, d$), where the $x^i$ are a
14: set of coordinates on the hypersurface (note the difference to appendix \ref{construnc}, where
15: $i$ labeled the internal coordinates of the 10-dimensional space-time).
16: The tangent vectors, $X^\mu_i \equiv
17: \partial_i X^\mu$, and the normal vector, $N^\mu$, of the hypersurface
18: can be chosen such that they satisfy the following orthogonality relations,
19: \begin{equation}
20: \label{geom:ortho}
21: \begin{split}
22: \tilde g_{\mu\nu}\, X^\mu_i X^\nu_j &= g_{ij}~,\\
23: X^\mu_i N_\mu &= 0~,\\
24: N^\mu N_\mu &=1~,
25: \end{split}
26: \end{equation}
27: where $g_{ij}$ represents the (induced) metric on the hypersurface.
28: Henceforth, a tilde will be used to label quantities
29: characterizing the ($d+1$)-dimensional space-time manifold, whereas
30: those of the hypersurface remain unadorned.
31:
32: The equations of Gauss and Weingarten define the second fundamental
33: form, $\K_{ij}$, of the hypersurface,
34: \begin{align}
35: \label{geom:gauss1}
36: \partial_i X^\mu_j + \tG{\mu}{\lambda\nu} X^\lambda_i X^\nu_j
37: - \Gamma^k_{\;\;ij} X^\mu_k &= \K_{ij} N^\mu~,\\
38: \label{geom:weingarten}
39: \partial_i N^\mu + \tG{\mu}{\lambda\nu} X^\lambda_i N^\nu
40: &= -\K^j_i X^\mu_j~.
41: \end{align}
42: The second fundamental form describes the extrinsic curvature of the
43: hypersurface, and is related to the intrinsic curvature by another
44: equation of Gauss,
45: \begin{align}
46: \label{geom:gauss2}
47: \tR_{\mu\nu\lambda\rho} X^\mu_i X^\nu_j X^\lambda_k X^\rho_l
48: &= R_{ijkl} + \K_{il} \K_{jk} - \K_{ik} \K_{jl}~.\\
49: \intertext{Furthermore, it satisfies the equation of Codazzi,}
50: \label{geom:codazzi}
51: \tR_{\mu\nu\lambda\rho} X^\mu_i X^\nu_j N^\lambda X^\rho_k
52: &= \nabla_i \K_{jk} - \nabla_j \K_{ik}~.
53: \end{align}
54: The symbol $\nabla$ denotes covariant derivatives with respect to the
55: induced metric $g_{ij}$.
56:
57: The above formulas simplify if (as in the familiar time-slicing
58: formalism), we choose space-time coordinates such that
59: \begin{equation}
60: \label{geom:Xdef}
61: X^0 = {\rm const.}~, \quad X^i=x^i~.
62: \end{equation}
63: Then, the tangent vectors are given by $X^0_i=0$ and $X^j_i
64: =\delta_i^j$. One conveniently splits up the space-time
65: metric as (shown here for Euclidean signature)
66: \begin{align}
67: \label{geom:split}
68: \tg_{\mu\nu} &= \begin{pmatrix} n_i n^i +n^2 & n_j \\
69: n_i & g_{ij} \end{pmatrix}~,\\
70: \intertext{whose inverse is given by}
71: \label{geom:splitinv}
72: \tg^{\mu\nu} &= \frac1{n^2} \begin{pmatrix} 1& -n^j \\
73: -n^i & n^2 g^{ij} +n^i n^j \end{pmatrix}~.
74: \end{align}
75: The matrix $g^{ij}$ is the inverse of $g_{ij}$, and is used to raise
76: hypersurface indices. The quantities $n$ and
77: $n^i$ are the lapse function and shift vector, respectively.
78:
79: The normal vector $N^\mu$ satisfying the orthogonality relations
80: \eqref{geom:ortho} is given by
81: \begin{equation}
82: \label{geom:normal}
83: N_\mu = (n,0)~, \qquad N^\mu= \frac1{n}(1,-n^i)~.
84: \end{equation}
85: Then, one can obtain the second fundamental form from the equation of
86: Gauss \eqref{geom:gauss1} as
87: \begin{equation}
88: \label{sli:Kij}
89: \K_{ij} = n \tG{0}{ij} = - \frac1{2n} \left(\partial_0 g_{ij} -
90: \nabla_i n_j - \nabla_j n_i \right)~.
91: \end{equation}
92:
93: We are interested in expressing all bulk quantities in terms of
94: hypersurface quantities. Using the equations of Gauss and Weingarten,
95: some Christoffel symbols can be expressed as follows,
96: \begin{align}
97: \label{geom:conn_kij}
98: \tG{k}{ij} &= \Gamma^k_{\;\;ij} - \frac{n^k}n \K_{ij}~,\\
99: \label{geom:conn_0i0}
100: \tG{0}{i0} &= \frac1n \partial_i n + \frac{n^j}n \K_{ij}~,\\
101: \label{geom:conn_ki0}
102: \tG{k}{i0} &= \nabla_i n^k - \frac{n^k}n \partial_i n
103: - n\K_{ij} \left( g^{jk}+ \frac{n^jn^k}{n^2} \right)~.
104: \end{align}
105: The remaining components, $\tG{0}{00}$ and $\tG{k}{00}$,
106: are easily found from their definitions using \eqref{geom:split}
107: and \eqref{geom:splitinv},
108: \begin{align}
109: \label{geom:conn_000}
110: \tG{0}{00} &= \frac1n \left( \partial_0 n +n^j \partial_j n +n^i n^j
111: \K_{ij} \right)~,\\
112: \label{geom:conn:k00}
113: \tG{k}{00} &= \partial_0 n^k + n^i \nabla_i n^k - n \nabla^k n -2n
114: \K^k_i n^i -n^k \tG{0}{00}~.
115: \end{align}
116:
117:
118:
119:
120:
121: