Example Setups
To get an idea of some of the different possible usages of GADGET-4,
and as a starting point for your own simulations and numerical
experiments, we include a small set of examples with the code
distribution. These can be found in the folder examples
. Each of the
examples has its own subdirectory containing the example's
configuration and parameterfile, and sometimes auxiliary files for the
specific runs.
To compile the code for one of the examples, you can either copy its
configuration file Config.sh
to the main code directory and run
make
there, or you execute make
by passing it the DIR
variable
with a value that points to the subdirectory of the example. The
executable for the example is then created right in the subdirectory
of the corresponding problem, allowing you to run it right there. This
is the recommended approach.
For convenience, the corresponding make-commands are contained in the
file make-examples.sh
, with one line with a suitable call of make
for each of the examples. Copying the corresponding line and pasting
it as a command into the main directory of the code will then build
the executable for the example. Executing the full shell script will
compile all the examples at once (provided there are no compilation
problems, of course).
Below, we describe each of the examples and also give a few suggestions for setup-variants that could be of interest. Note that for some of the examples, you need to obtain initial conditions that are available as part of a separate IC-package (183 MB) on the GADGET-4 web-site. This also contains the ICs for the examples adopted from the GADGET-2 code distribution (these ICs are identical to those distributed with GADGET-2).
Cosmological DM-only simulation with IC creation
The setup in DM-L50-N128
simulates a small box of comoving
side-length 50 Mpc/h using 128^3 dark matter particles. The initial
conditions are created on the fly upon start-up of the code, using
second order Lagrangian perturbation theory with a starting redshift
of z=63. The LEAN
option and 32-bit arithmetic are enabled to
minimize memory consumption of the code.
Gravity is computed with the TreePM algorithm at expansion order p=3. Three output times are defined, for which FOF group finding is enabled, and power spectra are computed as well for the snapshots that are produced. Also, the code is asked to compute a power spectrum for each output.
Aquarius Milky-Way Zoom
The directory DM-Zoom-Aq-C-5
contains a setup for a cosmological
zoom-simulation that follows the formation of a Milky Way-sized dark
matter halo. To carry out this example, you need access to the
Aq-C-5-dm
initial conditions files, which stem from the
Aquarius Project
and were also used as part of the
Aquila Project.
These IC files are available as part of the
IC-package
on the GADGET-4 web-site.
The example uses particle type 1 for the high-resolution dark matter particles, while types 2 to 4 are employed for more massive boundary particles of increasingly higher mass. For computing the gravitational forces, the TreePM algorithm is used.
The code is instructed to create 16 dumps, with output times that are specified through an input file. Before each snapshot is written, the FOF group finding algorithm is run on the high-resolution particles, followed by SUBFIND, and the snapshot dumps will be stored according to group order.
Interesting variants of this setup:
-
The default setup of this example does not enable a high resolution mesh, which could however be used in principle.
-
If the resolution is as high and spartially concentrated as here,
HIERARCHICAL_GRAVITY
can also be of interest for these types of zoom simulations. -
If a merger tree is desired, simulations of this kind could be run with the
MERGERTREE
option.
Colliding galaxies with star formation
This simulation with setup in the folder CollidingGalaxiesSFR
considers the collision of two compound galaxies made up of a dark
matter halo, a stellar disk and bulge, and cold gas in the disk that
undergoes star formation.
Radiative cooling due to helium and hydrogen is included. Star formation and feedback is modelled with a simple subgrid treatment. The simulation corresponds closely to the model of a galaxy collisions considered in the code paper.
Santa Barbara cluster
The Santa Barbara cluster is a hydrodynamical simulation of the formation of a galaxy cluster, which was introduced originally in the code comparison paper by Frenk et al. (1999).
In this example, we consider the problem at 2 x 64^3 resolution, using the pressure-based formulation of SPH (for the sake of change, and not because we think this is necessarily to be recommended for this problem).
Old examples from GADGET-2
Galaxy collision
This purely collisionless simulation in G2-galaxy
runs two disk
galaxies into each other, leading to a merger between the galaxies.
Each galaxy consists of a stellar disk, and a massive and extended
dark matter halo. This example uses plain Newtonian physics, with
20000 disk and 40000 halo particles in total.
The setup provided corresponds to a traditional tree algorithm with ordinary timestepping. Alternatives to this now possible with GADGET-4 include the use the FMM algorithm for gravity, and higher order multipole expansion. Note that this example from GADGET-2 is almost trivially small, and some of these alternatives will only show their real advantages in much larger simulations of higher resolution.
To get a first idea whether the example has worked you may check for
energy conservation by analysing the log-file energy.txt
. A simple
example for doing this is provided in the form of the IDL script file
plot_energy.pro
.
Adiabatic collapse of a gas sphere
This simulation in G2-gassphere
considers the gravitational collapse
of a self-gravitating sphere of gas which initially has a 1/r density
profile and a very low temperature. The gas falls under its own weight
to the centre, where it bounces back and a strong shock wave that
moves outwards develops. This common test problem of SPH codes has
first been described by Gus Evrard.
The simulation uses Newtonian physics in a natural system of units
(G=1). The setup corresponds to vanilla density-based SPH with the
entropy formulation introduced by
Springel & Hernquist (2002).
You can use the IDL-script plot_energy_gassphere.pro
to display the
evolution of thermal, kinetic and potential energy for the collapsing
gassphere. Note that this is a really tiny simulation of just 1472
particles.
Cosmological formation of a cluster of galaxies
This problem in G2-cluster
uses collisionless dynamics in an
expanding universe. It is a small cluster simulation that has been
set-up (a long time ago) with Bepi Tormen's initial conditions
generator ZIC using vacuum boundaries and a multi-mass technique. The
simulation has a total of 276498 particles. In a central
high-resolution zone there are 140005 particles, surrounded by a
boundary region with two layers of different softening, the inner one
containing 39616 particles, and the outer one 96877 particles.
Note that while this simulation is a cosmological simulation in comoving coordinates, it is unusual in that it doesn't use periodic boundary conditions but rather follows a sphere of matter around the origin with average density equal to the mean density. This technique is not very commonly used any more.
Large-scale structure formation including gas
This problem in G2-lcdm-gas
consists of 32^3 dark matter, and 32^3
gas particles, following structure formation in a periodic box of 50
Mpc/h on a side in a LCDM universe. Only adiabatic gas physics is
included, and the minimum temperature of the gas is set to 1000
K. This simple example uses grid initial conditions, where gas
particles are put at the centres of the grid outlined by the dark
matter particles. The simulation starts at z=10, and the code will
produce snapshot files at redshifts 5, 3, 2, 1, and 0. As in the
other old GADGET-2 examples, the SPH setup is density-based SPH based
on the entropy formation.