Numerical simulations

Formation and evolution of structures in the universe
Through observational and theoretical progress made in recent decades, our knowledge of formation mechanisms of structures of the universe (galaxies, stars, planets) has been developed considerably. Despite the huge difference in spatial scales involved, the gravity and conservation of angular momentum play a central role. In galaxies, matter falls into the potential well made by dark matter and forms galactic disks by conservation of angular momentum.
At the heart of these galaxies, stars that form within giant molecular clouds result from the gravitational collapse of the densest parts of the clouds, the pre-stellar dense cores. During the collapse, the conservation of angular momentum again induces the formation of a disc in equilibrium that will constitute the primordial nebula in which planetary systems are forming. Of all the stars that form in a galaxy, the most massive of them have considerable impact on the evolution of galaxies because of their intense radiation and their explosion.
These supernovae maintain turbulence in the galactic disk and inject into the interstellar medium heavy elements such as iron, for example. This cycle of interstellar matter illustrates the complexity and the link between all these spatial and temporal scales.
Despite these advances, many questions of astrophysical kind are still subject of active research. How do galaxies evolve ? At what rate they form stars? What is the origin of the mass distribution of stars? How and how many planets form around a star? What radiation and magnetic environment a star provides planets of its system? How stars are exploding? With what consequences?
These issues of astrophysics nature involve the resolution of many other more basic types of questions such as: What is the role of the stellar feedback on galactic evolution? What processes control the formation and evolution of stars and circumstellar disks? How is the magnetic field generated in the stars? What is the impact of the magnetic activity of the stars on the atmosphere of the planets? What are the fundamental processes behind the explosion of supernovae ?
Multi-scale and multi-physics problems
Understanding all of these processes, all intertwined, is a considerable challenge. Indeed, the dynamics of spatial and temporal scales covers almost ten orders of magnitude. However, as recalled above, these scales are all interrelated. Either the biggest are initial or limits conditions for the smallest or the smaller feed back on the greatest.
Moreover, many fundamental physical processes are at work and must be considered in a realistic description. This includes the magneto-hydrodynamic turbulence, gravity, the radiation transfer, atomic and molecular physics or chemistry out of balance. In addition, the energy densities of these processes are often comparable, indicating that it is not relevant to neglect one or more. The simultaneous consideration of these processes is a major challenge and requires the development of appropriate algorithms to solve the equations of the problem.
At present, few science problems have such a combination of process and scales. Therefore, an innovative and original approach must be developed. It is based partly on observations (strong presence in Sap/Irfu) but also on numerical simulation that allows especially to interpret observations that are often degenerate. Moreover, time allocation of telescopes, highly competitive, is enhanced by the ability of applicants to predict and model the observed objects. So there is a strong synergy between the observational and theoretical approaches.
The set of skills required to understand such complex environments is so diverse that collaboration within a vast community appears as a necessity. In this respect, the COAST group has the expertise to produce and operate an extensive set of numerical simulations.
The numerical simulation: an essential tool increasingly specialized
The study and understanding of the problems described above necessarily requires the use of appropriate codes able to take into account a wide dynamic range of scales. This is the case for example of adaptive meshing codes, such as those developed in the COAST team, which allow to increase locally the resolution during the calculation. Furthermore, algorithms and adequate numerical schemes must be developed to address each of the phenomena studied. The diversity of processes involved results in an equivalent diversity of these methods.
Furthermore, these codes must be worn on massively parallel architectures and in addition more and more different (CPU, GPU, hybrid nodes).
Finally, an important step is the analysis and visualization of the results of simulations which itself constitutes an important and technical part of the work and requires more and more the development of appropriate tools.
Each of these steps involves extensive expertise and therefore teams organized in which all the necessary skills are represented. A wide variety of digital tools developed in our laboratory. Designed to irrigate several themes, their scopes cover areas as diverse as cosmology, the interstellar medium, protoplanetary disks and the physics of hot and dense plasmas created by lasers.
RAMSES code that simulates adaptive grid on the dynamics of gravitating magnetized fluids and massively parallel architectures on, FARGO the code to simulate fluid dynamics in systems in rapid rotation (planets and stars), HERACLES the code that allows simulaer fluid dynamics has submitted a radiation field.
These codes are primarily developed in the laboratory and are free to access for the scientific community. These tools are used by the laboratory to realize the theoretical studies on cosmology, galaxy formation, the formation of stars and planets, involving phenomena of accretion, and ejection of magneto-hydrodynamic instabilities.
Other laboratories AIM also use our tools to their own research. Our theoretical work as well as our IT developments are realized in collaboration with french and foreign teams.