The simulation movies below were made for Federrath's MSc thesis, part of which was to implement tracer particles into the astrophysical simulation code ENZO. Tracer particles are passive Lagrangian particles, which move with the gas and collect Lagrangian information such as the gas density or temperature of a specific fluid/gas element.
In the following movie tracer particles were used to investigate baryonic matter flows in a cosmological simulation.
The following shows a simulation of driven supersonic molecular cloud turbulence including gravity, where I applied tracer particles to trace the flow of dense gas (published in Federrath et al. 2008, Physica Scripta, 132, 014025). After the development of the turbulent cascade, self-gravity is activated and tracer particles are placed in regions of density contrast greater than 10 times the mean density. After that the stochastic force field is deactivated leading to the decay of the turbulence due to the viscosity of the gas. We see that most of the tracer particles are mixed very efficiently. But as soon as the turbulence has decayed to a state of subsonic turbulence, some of the gas (including the tracer particles) is pulled into the potential wells of the densest regions (shown in dark red colour) by gravity. These tracer particles cannot mix with the surrounding medium and may carry characteristic chemical traces of the collapsing gas out of which new stars form inside the dense cores.
The following movie shows a test simulation with a shock propagating through an adiabatic medium at Mach 2. In the course of the propagation, the grid on which the hydrodynamic equations are solved is adaptively refined in the vicinity of the shock front (adaptive mesh refinement). Tracer Particles were added to the fluid, and advected with the shock. The colour of the tracer particles indicates the level of grid refinement at the position of the tracer particle (red: level 0; green: level 1).
The following movie shows the effect of adiabatic heating by kinetic energy dissipation in a simulation of driven turbulence (adiabatic index gamma=1.4). Tracer particles highlight the motion of the gas.
The Kelvin-Helmholtz instability (KHI) was used as a test case for my tracer particle implemention. The KHI is characterized by two flows running in opposite directions. Turbulent motions are excited at the interface of both flows because of shear. Tracer particles were added to highlight the dynamics involved in this process.
This shows another test simulation with the Kelvin-Helmholtz instability, but with fewer tracer particles and run for a longer time to see how the system evolves to a chaotic state.
For this simulation of the Kelvin-Helmholtz instability (KHI), three levels of refinement were used. All tracer particles are on the highest level of the grid hierarchy at all times, which requires processor-to-processor communication. Grid refinement was applied based on enstrophy (the curl of the velocity field), which results in de-refinement in some areas of low enstrophy.
Finally, here the tracer particles were injected dynamically during a simulation of the Kelvin-Helmholtz instability.