DEGAS 2

Welcome to the DEGAS 2 home page. DEGAS 2 is a Monte Carlo code for studying neutral transport in plasmas, with emphasis on fusion applications.

(Degas' The Rehearsal   appears courtesy of Nicolas Pioch. To learn more about Degas, the artist, check out his entry in the WebMuseum.)

DEGAS 2 User's Manual

NOTE: The reader is strongly encouraged to read the manual online as a PDF file. In general, however, the PDF file will be more up-to-date than the other versions. The HTML version is generated from the TeX source using TTH. If the equations in the HTML file are unintelligible, refer to this section of the TTH web page. Also, check that the browser is using "Western (ISO-8859-1)" encoding.

Source Code

DEGAS 2 usage is now governed by a licensing procedure. Once the licensing form has been submitted and approval given, the user will be able to download the DEGAS 2 source distribution (which includes the User's Manual and all other documentation) as a compressed tar file. Older versions of the code (labeled by their version number) can also be obtained.

Atomic Physics Data

Some of the atomic physics data files incorporated into DEGAS 2 have in the past been used elsewhere, either in other codes or for experimental and theoretical analyses. They are available directly via the links below. However, in the interest of simplifying maintenance, the detailed documentation on these files is contained in the User's Manual; the links here point to the corresponding page of the Manual.

Collisional Radiative Model for Hydrogen

This data file describes the "multi-step" ionization and recombination of hydrogen. Namely, at typical fusion plasma densities, the collisional and radiative processes that excite and de-excite the electron in the hydrogen atom occur at similar rates. A collisional radiative model provides effective ionization and recombination rates that account for the contributions of the excited states without them having to be considered explicitly in the transport code (i.e., DEGAS 2). For more on collisional radiative models in general, see Sec. 2.9 of the User's Manual. The data file consists of tables of these rates as functions of the electron density and temperature; the file contains only minimal documentation. A more comprehensive description can be found in Sec. 2.9.1 of the User's Manual. One can also find the version history of this file.

Collisional Radiative Model for Helium

The analogous data for the helium atom come from a code developed by M. Goto [see: M. Goto, J. Quant. Spectros. Radiat. Transfer 76, 331 (2003).]. For convenience, this data file has been formatted to largely resemble the one for hydrogen. The actual file that comes with the DEGAS 2 distribution is in a different format. For more detail on the model and the contents of the data file, see Sec. 2.9.2 of the User's Manual.

Elastic Scattering of Deuterium Atoms and Molecules on Deuterium Ions

(Since these data do not currently exist in a readable text format, this is just a placeholder.) These data describe both the elastic scattering and charge exchange processes and are based on detailed quantal calculations by the Controlled Fusion Atomic Data Center. More information on the calculations and the model used to treat these processes in DEGAS 2 can be found in Sec. 2.10.1 of the User's Manual.

Bateman Format for Plasma-Material Interaction Data

This format is referred to in the User's Manual (Sec. 3.9.2) and is used by other codes, including EIRENE. Since the original report is not readily available to the fusion community, a digitized verion has been posted here.

Verification and Validation

Code verification and validation are now being actively discussed by fusion plasma code developers. The objectives of V&V are to demonstrate, with some rigor, that a code works the way its author claims and that it is a physically reasonable description of reality. Public documentation of this effort allows readers of papers based on the code's results to evaluate these claims for themselves. Since journal publication of detailed V&V is problematic, an alternative approach, such as a presentation on the code's web site seems necessary. As an initial step towards that objective, we have a brief summary of the verification and validation of DEGAS 2 that was presented in a talk at the 2005 Transport Task Force Meeting (Napa, California). The other posters and talks listed elsewhere on this web site describe these V&V examples in more detail. Hopefully, this documentation will eventually be replaced with something more specific.

A first example of a specific verification test has been prepared for presentation at the 2006 Transport Task Force Meeting (Myrtle Beach, South Carolina). More precisely, the presentation made there was of the proposed format for "reference problems" and of the fluid neutral momentum transport reference problem; see the associated talk in the presentations section. The resulting text document may be included in a centralized database of such reference problems for others to use. Regardless, it also serves as part of the DEGAS 2 documentation; hence, its inclusion here. Even more pertinent is the actual "verification test" in which DEGAS 2 is compared with solutions from this reference problem. Although this is strictly speaking a work in progress, the text document describing its current state is still sufficiently useful to be posted here as well.

The comparison of DEGAS 2 with NSTX Gas Puff Imaging Experiments described in the paper presented at the 2006 PSI Conference in Hefei was, in the end, sufficiently quantitative and detailed that a description of it in terms of "validation" was warranted. This notion was explored further in a poster at the 2006 APS-DPP Meeting in Philadelphia. The upshot of this discussion is that a quantitative measure of validation is required and that corresponding metrics need to be established. Moreover, well characterized, repeatable (and repeated) validation experiments will be needed to achieve the best possible values for those metrics.

The subsequent talk presented at the 2007 Transport Task Force Meeting (San Diego, California) discusses validation metrics in more detail and contains an attempt at evaluating a couple of them using these data. However, the metrics used are not well suited to this situation, these evaluations are for demonstration purposes only.

The work presented at the 2010 Plasma Surface Interactions conference documents what is essentially a lower level, on the validation hierarchy, validation exercise. Since no plasma was involved in these lithium evaporation experiments, the prospects for being able to accumulate comprehensive diagnostic data from them were initially deemed good. Moreover, the components of the relatively simple model were thought to be well characterized. Both of these assumptions turned out to be inaccurate in the end, with signficant uncertainties arising from both experimental measurements and the input to the simulations. Moreover, comments from the PSI reviewers led us to conclude that our baseline model was perhaps too simple; the resulting changes were incorporated into the final version of the paper and the contemporaneous poster presentation at the 2010 APS-DPP meeting. The validation exercise was successful, however, in that it resulted in ideas for more discriminating experiments.

Analysis of 2010 NSTX GPI experiments by Chinese visitor Bin Cao resulted in comparisons of DEGAS 2 results even more detailed than those presented in 2006. Improved experimental methods and calibrations, as well as a simplified simulation procedure allowed four shots to be examined. The finite spatial resolution of the Thomson scattering electron density and temperature profiles remained a significant source of uncertainty, as did the effect of turbulent structures on the time averaged images. The peaks and widths of the radial profiles of the observed and simulated light emission matched to within the estimated errors (+/- 0.3 cm). Absolute calibrations of the GPI camera and gas puffing system allowed for the first time a quantitative comparison of the light emission rate. Because only the total number of deuterium molecules was known, the number of photons recorded by the camera per puffed deuterium atom was used as the basis for the comparison. The experimental result was 1/89 photons per atom (+/- 34%), while DEGAS 2 yielded 1/75 +/-18%; again, the results are within the estimated uncertainties. The paper describing this work can be obtained as a PPPL report or as a journal article (Fusion Science and Technology, to appear July 2013).

Presentations and Papers

This list has grown enough that it belongs on a separate page.

Links

Readers of this web page may find these other sites interesting and useful:
EIRENE
Is another Monte Carlo neutral transport code for simulating fusion plasmas. The two codes have some similarities and some differences. The web site also contains a large amount of the atomic and surface physics data used in EIRENE.
AMDIS
This is the home page of the IAEA Atomic and Molecular Data Unit. You can find not only data there, but also bibliographic information. The GENIE search tool allows you to search other available databases for data. There are also some codes available online.
CFADC
The Controlled Fusion Atomic Data Center is one of the other databases connected with the IAEA Atomic and Molecular Data Unit. Their web site contains some important data compiled for the U.S. fusion program, including neutral-ion and neutral-neutral elastic scattering data.
PPPL
Readers needing additional background on the need for DEGAS 2 and on magnetic fusion in general should consult the primary web site of the Princeton Plasma Physics Lab.
Stewart Zweben's Home Page
Stewart has led the development and application of the gas puff imaging (GPI) diagnostic. His web site contains additional details and presentations on the diagnostic as well as countless movies of edge turbulence in NSTX and Alcator C-Mod.
UTIAS
The University of Toronto Institute for Aerospace Studies Fusion Research Group focuses on edge physics (led by Peter Stangeby) and plasma-surface interactions (led by Tony Haasz). See especially their publications list.
UIUC PMI Group
David Ruzic's Plasma Material Interaction Group at the University of Illinois Urbana-Champaign has worked with and contributed to both DEGAS and DEGAS 2 over the years.
Center for Edge Physics Simulation (EPSI)
A SciDAC-3 collaboration targeted at developing first principles codes to simulate the edge and scrape-off layer of tokamak plasmas. DEGAS 2 is being coupled to kinetic plasma codes as part of this project.

Artwork

Effective visualization is essential to interpreting the results of a complex code like DEGAS 2. Along the way, some pretty pictures are bound to crop up. Questions, comments, suggestions, etc. to
Daren Stotler
Princeton Plasma Physics Laboratory
P.O. Box 451, MS 27
Princeton, NJ 08543
no spam