Skip to content

Dark Matter Simulation Code for Underground Scatterings - Crust Edition

License

Notifications You must be signed in to change notification settings

temken/DaMaSCUS-CRUST

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

30 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Build Status codecov License: MIT

DaMaSCUS - CRUST

ascl:1803.001 DOI arXiv arXiv

Dark Matter Simulation Code for Underground Scatterings - Crust Edition

DaMaSCUS-CRUST Version 1.2.0 02/03/2023

GENERAL NOTES

  • DaMaSCUS-CRUST is a MC simulator of dark matter (DM) particle trajectories through the Earth crust, atmosphere and/or additional shielding layers undergoing elastic scatterings on nuclei.
  • It is a tool to systematically determine the critical cross-section of strongly interacting DM, above which a given experiment loses detection sensitivity. As a final result it returns a precise exclusion band.
  • Users can easily change the layer structure in the config file and define layers by their density, thickness, and composition.
  • Both nuclear recoil and electron scattering experiments are included. User can also define a generic semiconductor experiment with a silicon or germanium target right in the configuration file.
  • There are a number of different DM-nucleus interaction types implemented, e.g. contact or light mediator interactions.
  • Two MC rare-event techniques can be invoked to speed up the simulations: Importance Sampling and Geometric Importance Splitting. Both can be (de-)activated and adjusted in the configuration file.
  • DaMaSCUS-CRUST is written in C++ and fully parallelized (openMPI).

For the underlying physics we refer to the corresponding papers, [arXiv:1802.04764] and [arXiv:1905.06348].

CONTENT

The included folders are:

  • bin/: This folder contains the configuration files and the executables after compilation.
  • build/: This folder contains all object files.
  • detectors/: Detector specific data, such as tabulated efficiencies or recoil energy data are stored here.
  • include/: All header files can be found here. Necessary 3rd party libraries can also be placed here.
  • results/: The resulting limits are saved in this folder.
  • src/: Here you find the source code of DaMaSCUS-CRUST.

INSTALLATION AND USAGE

Dependencies:

Before the code can be compiled, these dependencies need to be taken care of:

  • libconfig v1.6: For the input configuration files we use the libconfig library. Note that we need at least version 1.6. (Link)
  • MPI: For parallelization we use the Message Passing Interface (Link)

For the installation of the dependencies, it might be useful to take a look at the travis script.

Installation:

The code can be compiled using the makefile. It might be necessary to adjust the compiler lines and the path to the libraries

#Compiler and compiler flags
CXX := mpic++
CXXFLAGS := -Wall -std=c++11 
LIB := -lconfig++
INC := -I include
(...)

After that simply run

make

from the root directory in the terminal to compile DaMaSCUS-CRUST.

Running

make clean

deletes all objective files and executables.

Usage:

After successful compilation there is only one final step before running the program: Adjusting the configuration file in /bin/. This file specifies all the input parameter, the mass scan range, the experiment of interest, the shielding layer structure, the interaction type,... . This should be rather self-explanatory. However note that libconfig is not very forgiving with regards to the input parameter type. For example it might complain if an input parameter of type double is given as “1” instead of “1.0”.

Afterwards the program starts by

mpirun -n N DaMaSCUS-CRUST config.cfg

where N is the number of MPI processes.

After a successful run, the resulting constraints can be plotted with the included Mathematica notebook /Plot.nb.

New in version 1.1.1: Alternative, the user may also directly compute the underground DM speed distribution for a given mass, cross section, speed cutoff and sample size via command line arguments. The config file is still necessary to define the interaction, shielding layers, rare event techniques, etc..

Example: For a DM mass of 500 MeV, a nucleon cross section of 1e-31 cm^2, a minimum speed of 100 km/sec, and a sample size of 10000, the user can run

mpirun -n N DaMaSCUS-CRUST config.cfg 0.5 1e-31 100 10000  

Note: The minimum speed and sample size are optional command line arguments. The default value for v_Min is 0, whereas the default choice for the sample size is taken from the config file.

The resulting speed distribution is saved in the folder /results/Simulation_ID.

CITING DaMaSCUS-CRUST

If you decide to use this code, please cite

Emken, T. & Kouvaris, C., 2018, DaMaSCUS-CRUST, Astrophysics Source Code Library, record [ascl:1803.001]

BibTex entry:

@software{DaMaSCUScrust,
  author = {Emken, Timon},
  title = {{Dark Matter Simulation Code for Underground Scatterings - Crust Edition~(DaMaSCUS-CRUST) [Code, v1.2.0]}},
  year         = {2018},
  publisher    = {Zenodo},
  version      = {v1.2.0},
  doi          = {DOI:10.5281/zenodo.2846400},
  url          = {https://doi.org/10.5281/zenodo.2846400},
  howpublished={Astrophysics Source Code Library record \href{https://ascl.net/1803.001}{[ascl:1803.001]}. The code can be found under \url{https://github.com/temken/damascus-crust}. Version 1.2.0 is archived as \href{https://doi.org/10.5281/zenodo.2846400}{DOI:10.5281/zenodo.2846400}}
}

Please cite the original publications as well.

Emken, T. , Essig, R., Kouvaris, C., & Sholapurkar, M., Direct Detection of Strongly Interacting Sub-GeV Dark Matter via Electron Recoils, JCAP 1909 (2019) no.09, 070,[arXiv:1905.06348].

Emken, T.& Kouvaris, C., How blind are underground and surface detectors to strongly interacting Dark Matter?, Phys.Rev. D97 (2018) no.11, 115047 , [arXiv:1802.04764].

AUTHORS & CONTACT

The authors of DaMaSCUS-CRUST are Timon Emken and Chris Kouvaris.

For questions, bug reports or other suggestions please open an issue on GitHub.

LICENCE

This project is licensed under the MIT License - see the LICENSE file.