RADiation and Dust Affecting the GAs STate
RADAGAST is a local (0D) gas model for use in 3D dust radiative transfer simulations. Given a certain radiation field and dust model, the equilibrium state for various quantities of the gas are calculated self-consistently. Currently, the module treats the main species consisting of H (H+, H and H2), and their interactions with the radiation field and the dust grains.
Quantities calculated self-consistently
- Gas equilibrium temperature
- H+, H, and H2 abundance
- Level populations of H and H2
- Grain temperatures and charge distributions
A non-comprehensive list of processes included in this calculation
- Photoionization of H
- Radiative recombination of H+
- Photodissociation of H2
- Formation of H2 on dust grain surfaces
- Heating by the photoelectric effect on dust grains
- Cooling by collisions between gas and dust grains
- Net heating/cooling by collisional transitions of H2 (and H at high temperatures)
From the resulting state, the emissive properties and cross section of the gas can be calculated, so that a radiative transfer code can include these in its model.
- Cross sections for H photoionization, H2 photoionization, and H2 photodissociation
- Continuum emission of H (free-bound, free-free, and two-photon)
- Emission coefficients, opacities, and shapes for the lines of H and H2.
While RADAGAST was mainly developed for use in the 3D Monte Carlo Dust Radiative Transfer code SKIRT, it is designed as a stand-alone module. It is possible to use this module in any C++ code where the necessary details about the radiation field and the dust can be provided, without having to install anything related to SKIRT.
This code was developed during my PhD research, and contains many equations and data from other authors to implement the processes listed above. Consult my PhD thesis (to be published) for details, equations, and citations. Alternatively, one can build the documentation pages or read the comments for each function in the source code.
RADAGAST includes a release of the doctest C++ testing framework, to power a small set of unit tests.
It also uses Eigen3 for storing vectors and matrices, and performing linear algebra, and GSL for solving several non-linear equations. Currently, these libraries are dependencies, but the git history might contain a copy.
First read how to build the library.
Then try running the main binary produced at RADAGAST/../cmake_release/src/mains/main.
Experiment by giving a few numerical command line arguments to this program.
The command
main 1e5 1e4 1e3
will run an example model with a gas density of 1e5 cm-3, under the effect of a blackbody source located at a (fixed) distance of 1 pc with a color temperature of 1e4 K and a bolometric luminosity of 1e3 solar luminosities.
When this binary works without problems, copy the library built at RADAGAST/../cmake_release/src/core/libridge.a to your preferred location (or set CMAKE_INSTALL_PREFIX and run make install).
Then read the instructions on how to use the API, write a simple C++ script making the calls, and try building and linking your own C++ script to the library.
To store, use, and update gas properties, we make use of the newly developed dynamic state mechanisms for SKIRT. See pull request. The branch of SKIRT that integrates RADAGAST can be found on my fork, and is called gas-dev. This branch should be compatible with the latest commit of RADAGAST.
The versions of RADAGAST/SKIRT used for my PhD thesis are older. See the thesis release tag for SKIRT, and the thesis release tag for RADAGAST.