Team: 6
School: Sandia High
Area of Science: Astrophysics & Cosmology
Interim:
The Problem
Astronomy and Cosmology have been historically limited by the aperture of telescopes and the quality of their instruments. While mathematical models can be constructed which make certain predictions regarding the evolution of galaxies and galaxy clusters, even the universe as a whole, the verification of these models has proved difficult, due to a variety of factors, including but not limited to: the physical limits of observatories, both on Earth and in Space, and the lack of remaining evidence for past events predicted by these models. The period between 300,000 to 300,000,000 years after the big bang is completely invisible to all current methods of detection, due to the lack of massive galaxies combined with the immense red shift of light emitted from those galaxies, caused by the expansion of the Universe. Individual Galaxy and Galaxy Cluster collisions take place over billions of years, and thus cannot be directly observed in real time. Thus, to understand the dynamics and evolution of these systems, computational methods must be used to verify these models.
Computational Solution
Because of the limitations of both ground and space-based observatories, scientists rely on simulations to validate models, using the cosmic microwave background from 300,000 years after the big bang as a starting point, we can simulate the evolution of the Universe based on model parameters over 13 billion years to the present. By comparing how the universe evolves with these different models, and comparing them to the physical observations which can be made, certain cosmological models can be ruled out or modified. SINGS serves as a tool by which these models can be verified, and much more. As a general gravitational simulator, SINGS provides a framework by which simulation scenarios can be constructed and then computed. It provides a large variety of numerical integrators and force codes, along with a variety of generation tools for constructing clouds and galaxies of arbitrary mass and density. The modularity of these systems allows for a high degree of customization when creating and running simulations with SINGS. The intended goal of the program being to compute astrophysical simulations of high accuracy to be useful for scientific research.
Progress to Date
As of the writing of this report, SINGS currently has 3 numerical integrators: Euler, Semi-Implicit Euler, and Runge-Kutta, and 3 force codes: Brute-Force Newtonian, Barnes-Hut, and a rudimentary Particle Mesh method which can be used for simulation computing. All integrators and 2 of the force codes have been optimized for parallel computing using OpenMP. A simple parameter file system has been created, which allows for the specification of simulation parameters, presently including the integration and force calculation methods to use, but also extending to threads to use, maximum memory allocation, physical time allotment, and the frequency of snapshot creation, among a variety of other factors. Currently, a simulation compiler is in the works, which creates and compiles the starting scenario for the simulation down to a binary.
Expected Results
By its completion, SINGS will allow for the creation of a wide array of simulation scenarios, from the collisions of galaxies and galaxy clusters to the large scale evolution of the universe. It will be able to utilize the high memory and thread count of supercomputers to perform simulations with extreme particle counts exceeding 1 million (with proper optimization on the part of the user), allowing users to investigate the dynamics of astrophysical systems to high accuracy. This is further aided by the relative simplicity of the language used for the configuration of simulation scenarios and parameters.
References
1. Barnes, J., & Hut, P. (1986). A hierarchical O(N log N) force-calculation algorithm. Nature, 324(6069), 446-440. doi:10.1038/324446a0
2. Springel, V., White, S. D. M., Jenkins, A., Frenk, C. S., Yoshida, N., Gao, L., ... Pearce, F. (2005). Simulations of the formation, evolution and clustering of galaxies and quasars. Nature, 435(7042), 629-636. doi:10.1038/nature03597
3. https://en.wikipedia.org/wiki/Semi-implicit_Euler_method
4. On the Clustering Tendancies among the Nebulae. II. A Study of Encounters between Laboratory Models of Stellar Systems by a New Integration Procedure, Holmberg, E. 1941, ApJ, vol.94, p.385.
Team Members:
Sponsoring Teacher: Bradley Knockel