1: \sect{EE}{Einstein-Euler equations}

2: 

3: The Einstein-Euler equations or, in other words, the Einstein equations

4: coupled to a simple perfect fluid are given by the following system of equations

5: \lgath{EEeqn}{

6: G^{ij} = \frac{8\pi G}{c^4} T^{ij} \label{EEeqn.1} \\

7: \nabla_{i} T^{ij} = 0 \label{EEeqn.2} } where the stress-energy

8: tensor for the fluid is given by \lgath{EEdefs}{ T^{ij} = (\rho +

9: c^{-2} p)v^i v^j + p g^{ij} \,  } with $\rho$ the fluid density,

10: $p$ the fluid pressure, and $v$ the fluid four-velocity normalized

11: by $v^iv_i = -c^2$. Here $c$ is the speed of light and $G$ is the

12: Newtonian gravitational constant. We will only consider isentropic

13: flow where the pressure is related to the density by an equation

14: of state of the form $p = f(\rho)$

15: 

16: In formulating a symmetric hyperbolic system for the fluid

17: variables $\{\rho,v\}$, we need to deal with the well known

18: problem that the system becomes singular when $\rho + c^{-2} p =

19: 0$. This is a particular problem for fluid balls having compact

20: support. To get around this problem, we will follow Rendall

21: \cite{Ren92} and use a technique of Makino \cite{Mak} to

22: regularize the fluid equations so a class of gravitating fluid

23: ball solutions can be constructed. Thus as in \cite{Ren92}, we

24: assume an equation of state of the form \leqn{eos}{ p =

25: K\rho^{(n+1)/n} } where $K \in \Rbb_{>0}$, $n\in \Nbb$, and we

26: introduce a new ``density'' variable $\alpha$ via the formula

27: \leqn{dendef}{ \rho = \frac{1}{\bigl(4Kn(n+1)\bigr)^n}\alpha^{2n}

28: \, . } As discussed by Rendall, the type of fluid solutions

29: obtained by this method have freely falling boundaries and hence

30: do not not include static stars of finite radius  and so this

31: method is far from ideal. However, in trying to understand the

32: Newtonian limit and post-Newtonian approximations these solutions

33: are almost certainly general enough to obtain a comprehensive

34: understanding of the mathematical issues involved in the Newtonian

35: limit and post-Newtonian approximations. We would also like to

36: remark that the results contained in this article are largely

37: independent of the specific structure of the fluid equations. We

38: therefore expect that the analysis in this paper can be carried

39: over without much difficulty to other matter models whose

40: equations can be formulated as a symmetric hyperbolic system and

41: have a finite propagation speed for the matter density in the

42: limit $\ep\searrow 0$.

43: 

44: 

45: For the remainder of this article, we will assume that

46: spacetime $M$ is a manifold of the

47: form $M=S\times I\subset \Rbb^4$ where $S\cong \Rbb^3$ and $I\subset \Rbb$

48: is an open interval

49: containing $0$. We will always let capital Roman indicies, i.e. I,J,K etc.,

50: run from $1$ to $3$ and small Roman indicies, i,j,k etc., run from

51: $1$ to $4$. All equations will be formulated with respect

52: to a global set of Cartesian coordinates $(x^i,i=1,..4)$

53: on $M$ where $(x^I)$ are spatial coordinates that cover $S$

54: and $x^4$ is a ``time coordinate''

55: that covers the interval $I$.

56: