|Program name||Package id||Status||Status date|
|Package ID||Orig. computer||Test computer|
SoftWare for Optimization of Radiation Detectors (SWORD) is a framework to allow easy simulation and evaluation of radiation detection systems. It is targeted at system designers, who want to evaluate and optimize system parameters without actually building hardware first, at sponsors who need to evaluate proposed or actual system designs independent of the supplier, without having access to actual hardware, and at operators who want to use simulation to evaluate observed phenomena.
SWORD is vertically integrated and modular. It allows users to define their own radiation detection instruments by building them from basic geometric ‘objects’ and assigning those objects materials, detection, and/or radioactive emission properties. This process is accomplished by a CAD-like graphical user interface, in which objects may be defined, translated, rotated, grouped, arrayed, and/or nested to produce compound objects. In addition to providing the ability to build a detection system model from scratch, SWORD provides a library of ‘standard’ detector design objects that can be used ‘as is’ or modified by the user.
Using existing Monte Carlo radiation transport codes, SWORD supplies a vertically integrated framework for creating models, assigning emission spectra, running the MC code and analyzing the results. The user interface is independent of the chosen MC engine.
SWORD gives the user the option of running his simulation using one of two well-known known simulation engines: GEANT 4 from CERN (Version 10.1 included in package distribution) and MCNP. In general, the SWORD workflow consists of four steps:
Design the scenario. In this step, all the geometric elements of the simulation are defined, together with material properties, radioactive emission and detector properties. The tool used here is the SWORD geometry builder, a CAD-like graphical tool.
Configure the run. Here, all the simulation run parameters are defined. This includes simulated duration, which analysis processes will be run, and what outputs will be produced.
Run the simulation. Normally this is done from within the SWORD interface. However, SWORD can also be run in ‘batch’ mode without graphical interfaces. The latter is useful for running high compute time runs on high performance machines.
Examine the results. Spectra and images are output as ASCII text files. SWORD provides interactive viewers for spectra and images. Alternatively, the output files can be read and analyzed using a variety of software tools such as spreadsheet programs.
Improvements in SWORD6 over SWORD5
Geant v10.1 integration
Switched to Penelope for low energy EM physics
GDML output capability
Geant overlap checking
User control of minimum energy deposit recorded in Geant
LLNL fission package in MCNP
Improved array functionality
GUI right click functionality
Improved interface functionality
Mesh definition panel for ADVANTG/Denovo and MCNP mesh tally runs
VRML output from Geant
Silo output from MCNP mesh tally
Montage movie generation
KML heat map production
Improved backtrace output
Advanced run manager
Improved build system
Improved binary installers for Centos and Ubuntu
Requirement for Oracle Java eliminated
VM ships with Centos 6.7
Support for 64-bit platforms only
|Package ID||Status date||Status|
E. I. Novikova, M. S. Strickman, et al. (2006), ‘Designing SWORD - SoftWare for the Optimization of Radiation Detectors,’ IEEE Nuclear Science Symposium Conference Record 1: 607-612.
C. S. Gwon, E. I. Novikova, et al. (2007), ‘Interacting with the SWORD package (SoftWare for the Optimization of Radiation Detectors),’ IEEE Nuclear Science Symposium Conference Record 2: 1130-1133.
|Package ID||Computer language|
|CCC-0767/05||C++, JAVA, PYTHON|
Current distribution is available as a VMware virtual appliance, in addition to binaries for Centos and Ubuntu. It can run under free VMware server or player (player recommended) on a Windows or Linux host or under the VMware Fusion (http://www.vmware.com/, purchase only) on an Mac OSX host.
Installation instructions are included in the provided documentation. Users should be aware that current versions of the MCNP precompiled executables will not work with this version of SWORD. If the user desires to use MCNP with SWORD, the user will have to cross-compile MCNP.
Contributed by: Radiation Shielding Information Computational Center
Oak Ridge National Laboratory
Oak Ridge, Tennessee, USA
Developed by: Naval Research Laboratory, Washington, DC.
This package contains GEANT4 V10.1 which was developed at CERN (European Organization for Nuclear Research), Genève, Switzerland.
Keywords: Monte Carlo method, radiation detectors, simulation.