.. under construction ..

On numerical simulations in astrophysics

Numerical simulations are useful tools for understanding astrophysical phenomena (as well as in many other fields). Important examples include: From the view of the code (which can be realized in every numerically capable language such as Fortran, C, C++, and others), two main types of dynamics have been developed: N-Body code is best suited to study the dynamics of systems consisted of a certain number of discrete bodies, e.g. stars or planets, which are in (usually gravitational) interaction. On the other hand, SPH is appropriate to simulate gasses or fluids, such as gaseous nebulae, stellar or planetary interiors, etc., by studying them in an approximation of smoothened interacting particles. Combinations of both are also commonly used, e.g. for model galaxies consisting of both stars and nebulae (dust and gas).

Traditionally, the interaction of mutually each pair of bodies is considered for calculations, leading to an increase of the CPU time required by N^2, if N is the number of particles. In 1986, however, this situation could be significantly improved when Joshua Barnes and Piet Hut introduced their TREE method (TREECODE). Here, the volume under consideration is devided in smaller and smaller cells, until each cell contains only zero or one particles. Then, the force on each particle is calculated by considering the interaction with the cells, taking into account large cells only for the remote (and thus weaker) interactions. This method has improved the increase of CPU time to a N*log N behaviour, so that larger particle numbers can be treated with this scheme, which has become quite common in the meantime.

Some useful links to numerical simulation code:

Some of the hottest papers in this field are available from here (xxx.lanl.gov, astro-ph preprint archive): Very preliminary collection of further links:
Hartmut Frommert [contact]
[SEDS]