In collaboration with X. Corap and E. Stein, we developed a galaxy collision simulation in Fortran with visualization in Python.

The main goal of this project was to study the evolution and effects of interactions between galaxies, which are dynamic events in the universe. Galaxies are dynamic objects rather than immutable structures; they are constantly evolving. Small galaxies form first and then merge to form larger galaxies. However, how does the structure of galaxies evolve during these collisions? To answer this question, the primary objective was to model an equilibrium distribution of stars, representing a typical galaxy (such as a disk or elliptical galaxy), and then perturb this structure by simulating the impact of a small galaxy (the impactor) colliding with the galaxy in place.

Parameters such as mass, velocity, and compactness of the impactor were varied to observe how they influence perturbations in the initial structure of the galaxy. If the disturbance is significant, the galaxy may not have time to regain a new equilibrium before another collision, potentially transforming initially balanced structures into completely unbalanced ones.

Due to the complexity of simulating the interaction of each star in a galaxy (approximately 10^{11} stars), we employed an N-body method known as particle-mesh, where each stellar particle represents a cloud of stars rather than a single star. This approach allows us to sample the phase space and subsequently solve the Vlasov-Poisson equation on a grid.

The code is available here (coming soon).