Computational Astrophysics and Cosmology最新文献

A brief review of the origin of jets from disc-accreting rotating magnetized stars is given. In most models, the interior of the disc is characterized by a turbulent viscosity and magnetic diffusivity (‘alpha’ discs) whereas the coronal region outside the disc is treated using ideal magnetohydrodynamics (MHD). Extensive MHD simulations have established the occurrence of long-lasting outflows in the case of both slowly and rapidly rotating stars. (1) Slowly rotating stars exhibit a new type of outflow, conical winds. Conical winds are generated when stellar magnetic flux is bunched up by the inward motion of the accretion disc. Near their region of origin, the winds have a thin conical shell shape with half opening angle of ～30^°. At large distances, their toroidal magnetic field collimates the outflow forming current carrying, matter dominated jets. These winds are predominantly magnetically and not centrifugally driven. About 10-30% of the disc matter from the inner disc is launched in the conical wind. Conical winds may be responsible for episodic as well as long lasting outflows in different types of stars. (2) Rapidly rotating stars in the ‘propeller regime’ exhibit twocomponent outflows. One component is similar to the matter dominated conical wind, where a large fraction of the disc matter may be ejected in this regime. The second component is a high-velocity, low-density magnetically dominated axial jet where matter flows along the open polar field lines of the star. The axial jet has a mass flux of about 10% that of the conical wind, but its energy flux, due to the Poynting flux, can be as large as for the conical wind. The jet’s magnetically dominated angular momentum flux causes the star to spin down rapidly. Propeller-driven outflows may be responsible for protostellar jets and their rapid spin-down.

When the artificial requirement of symmetry about the equatorial plane is dropped, the conical winds are found to come alternately from one side of the disc and then the other, even for the case where the stellar magnetic field is a centered axisymmetric dipole.

Recent MHD simulations of disc accretion to rotating stars in the propeller regime have been done with no turbulent viscosity and no diffusivity. The strong turbulence observed is due to the magneto-rotational instability. This turbulence drives accretion in the disc and leads to episodic conical winds and jets.

We propose a modification for the tensor of artificial viscosity employable for generally comoving, curvilinear grids. We present a strong conservation form for the equations of radiation hydrodynamics for studying nonlinear pulsations of stars. However, the modification we propose is of general mathematical nature. We study a differential geometrically consistent artificial viscosity analytically and visualize a comparison of our approach to previous implementations by applying it to a simple self-similar velocity field which has a direct application in stars as the fundamental mode of pulsation is radial. We first give a general introduction to artificial viscosity and motivate its application in numerical computations. We then show how a tensor of artificial viscosity has to be designed when going beyond common static Eulerian or Lagrangian comoving rectangular grids. We derive and state the modified equations which include metrical terms that adjust the isotropic (pressure) part of the tensor of artificial viscosity.

The approximate computation of all gravitational forces between N interacting particles via the fast multipole method (FMM) can be made as accurate as direct summation, but requires less than $O\left(N\right)$ operations. FMM groups particles into spatially bounded cells and uses cell-cell interactions to approximate the force at any position within the sink cell by a Taylor expansion obtained from the multipole expansion of the source cell. By employing a novel estimate for the errors incurred in this process, I minimise the computational effort required for a given accuracy and obtain a well-behaved distribution of force errors. For relative force errors of ～10^?7, the computational costs exhibit an empirical scaling of $\propto N^{0.87}$. My implementation (running on a 16 core node) out-performs a GPU-based direct summation with comparable force errors for $N?{�10�}^{�5�}$.