We present a spectral method for solving elliptic equations which arise in general relativity, namely three-dimensional scalar Poisson equations, as well as generalized vectorial Poisson equations of the type $\Delta \vec{N} + \lambda \vec{\nabla}(\vec{\nabla}\cdot \vec{N}) = \vec{S}$ with $\lambda \not= -1$. The source can extend in all the Euclidean space ${\bf R}^3$, provided it decays at least as $r^{-3}$. A multi-domain approach is used, along with spherical coordinates $(r,\theta,\phi)$. In each domain, Chebyshev polynomials (in $r$ or $1/r$) and spherical harmonics (in $\theta$ and $\phi$) expansions are used. If the source decays as $r^{-k}$ the error of the numerical solution is shown to decrease at least as $N^{-2(k-2)}$, where $N$ is the number of Chebyshev coefficients. The error is even evanescent, i.e. decreases as $\exp(-N)$, if the source does not contain any spherical harmonics of index $l\geq k -3$ (scalar case) or $l\geq k-5$ (vectorial case).

We present a numerical method to compute quasiequilibrium configurations of close binary neutron stars in the pre-coalescing stage. A hydrodynamical treatment is performed under the assumption that the flow is either rigidly rotating or irrotational. The latter state is technically more complicated to treat than the former one (synchronized binary), but is expected to represent fairly well the late evolutionary stages of a binary neutron star system. As regards the gravitational field, an approximation of general relativity is used, which amounts to solving five of the ten Einstein equations (conformally flat spatial metric). The obtained system of partial differential equations is solved by means of a multi-domain spectral method. Two spherical coordinate systems are introduced, one centered on each star; this results in a precise description of the stellar interiors. Thanks to the multi-domain approach, this high precision is extended to the strong field regions. The computational domain covers the whole space so that exact boundary conditions are set to infinity. Extensive tests of the numerical code are performed, including comparisons with recent analytical solutions. Finally a constant baryon number sequence (evolutionary sequence) is presented in details for a polytropic equation of state with gamma=2.

Exact models of uniformly rotating strange stars, built of self bound quark matter, are calculated within the framework of general relativity. This is made possible thanks to a new numerical technique capable to handle the strong density discontinuity at the surface of these stars. Numerical calculations are done for a simple MIT bag model equation of state of strange quark matter. Evolutionary sequences of models of rotating strange stars at constant baryon mass are calculated. Maximally rotating configurations of strange stars are determined, assuming that the rotation frequency is limited by the mass shedding and the secular instability with respect to axisymmetric perturbations. Exact formulae which give the dependence of the maximum rotation frequency, and of the maximum mass and corresponding radius of rotating configurations, on the value of the bag constant, are obtained. The values of T/W for rapidly rotating massive strange stars are significantly higher than those for ordinary neutron stars. This might indicate particular susceptibility of rapidly rotating strange stars to triaxial instabilities.

We present results of numerical computations of quasiequilibrium sequences of binary neutron stars with zero vorticity, in the general relativistic framework. The Einstein equations are solved under the assumption of a conformally flat spatial 3-metric (Wilson-Mathews approximation). The evolution of the central density of each star is monitored as the orbit shrinks in response to gravitational wave emission. For a compactification ratio M/R=0.14, the central density remains rather constant (with a slight increase, below 0.1%) before decreasing. For a higher compactification ratio M/R=0.17 (i.e. stars closer to the maximum mass configuration), a very small density increase (at most 0.3%) is observed before the decrease. This effect remains within the error induced by the conformally flat approximation. It can be thus concluded that no substantial compression of the stars is found, which would have indicated a tendency to individually collapse to black hole prior to merger. Moreover, no turning point has been found in the binding energy or angular momentum along evolutionary sequences, which may indicate that these systems do not have any innermost stable circular orbit (ISCO).

We report on general relativistic calculations of quasiequilibrium configurations of binary neutron stars in circular orbits with zero vorticity. These configurations are expected to represent realistic situations as opposed to corotating configurations. The Einstein equations are solved under the assumption of a conformally flat spatial 3-metric (Wilson-Mathews approximation). The velocity field inside the stars is computed by solving an elliptical equation for the velocity scalar potential. Results are presented for sequences of constant baryon number (evolutionary sequences). Although the central density decreases much less with the binary separation than in the corotating case, it still decreases. Thus, no tendency is found for the stars to individually collapse to black hole prior to merger.

We present spectral methods developed in our group to solve three-dimensional partial differential equations. The emphasis is put on equations arising from astrophysical problems in the framework of general relativity.

A multi-domain spectral method for computing very high precision 3-D stellar models is presented. The boundary of each domain is chosen in order to coincide with a physical discontinuity (e.g. the star's surface). In addition, a regularization procedure is introduced to deal with the infinite derivatives on the boundary that may appear in the density field when stiff equations of state are used. Consequently all the physical fields are smooth functions on each domain and the spectral method is absolutely free of any Gibbs phenomenon, which yields to a very high precision. The power of this method is demonstrated by direct comparison with analytical solutions such as MacLaurin spheroids and Roche ellipsoids. The relative numerical error reveals to be of the order of $10^{-10}$. This approach has been developed for the study of relativistic inspiralling binaries. It may be applied to a wider class of astrophysical problems such as the study of relativistic rotating stars too.

A general relativistic version of the Euler equation for perfect fluid hydrodynamics is applied to a system of two neutron stars orbiting each other. In the quasi-equilibrium phase of the evolution of this system, a first integral of motion can be derived for certain velocity fields of the neutron star fluid including the (academic) case of co-rotation with respect to the orbital motion (synchronized binaries) and the realistic case of counter-rotation with respect to the orbital motion. The velocity field leading to this latter configuration can be computed by solving three-dimensional vector and scalar Poisson equations.

Several applications of spectral methods to problems related to the relativistic astrophysics of compact objects are presented. Based on a proper definition of the analytical properties of regular tensorial functions we have developed a spectral method in a general sphericallike coordinate system. The applications include the investigation of spherically symmetric neutron star collapse as well as the solution of the coupled 2D-Einstein-Maxwell equations for magnetized, rapidly rotating neutron stars. In both cases the resulting codes are efficient and give results typically several orders of magnitude more accurate than equivalent codes based on finite difference schemes. We further report the current status of a 3D-code aiming at the simulation of non-axisymmetric neutron star collapse where we have chosen a tensor based numerical scheme.

We present a formalism well adapted to the numerical study of the encounter of an ordinary main sequence star with a massive black hole. Symmetry considerations, the use of a well adapted moving grid and a well adapted moving frame along with integration of the partial differential equations by means of pseudo-spectral methods result in a very powerful and accurate tool. The hydrodynamical equations are written in a moving frame which mimics the bulk of the movement of the fluid, resulting in very small relative velocities and a well suited spatial resolution throughout the calculations. Therefore, the numerical calculations are considerably simplified, smoothing in particular the Courant condition. Typical runs are performed within a few hours on a workstation with the high accuracy linked to the spectral methods. Predictions of the so-called affine are tested against this full numerical simulation.