We are currently building a database of properties of dense plasmas to:
- gather and document results from disparate computational methods,
- allow the community to have quick access to results for direct comparisons between/among methods,
- quantify uncertainties across physical regimes (temperature, density, nuclear charge and stoichiometry for mixtures),
- motivate the need for new computational methods.
The database grew from three observations:
- nearly all published results are for unique conditions, obviating our ability to discover relative strengths and weaknesses of different approaches – increase the ease for obtaining data for comparison purposes,
- most high-fidelity physics codes model homogeneous and isotropic (0D) plasmas in spite of the need for simulations at hydrodynamic lengths scales (tens of picoseconds and cubic microns) – understand which physics simplifications allow us to build very fast methods,
- closely-related fields have begun to develop such databases; for example, the DFT-based materials database – move dense plasma physics into a data-driven field.
The simplest, and most common, item that will occur in the database is the radial distribution function g(r), which is a test of a computational method’s ability to correctly capture correlations, which in turn implies having accurate forces. Dynamical properties such as the velocity autocorrelation function (VACF) provide a much more stringent test; unfortunately, fewer results currently appear in the literature.
The current database is being constructed at GitHub here. There is no data yet! We are hoping to have the first datasets by mid-May 2017. We have presented our results and two recent workshops, and those talks will be posted soon.
If you would like to contribute to the database, please contact us through this website. All we need are:
- data files, preferably in CSV format,
- citation to published work (or, equivalent description),
- links you would like included.
This database was built primarily by: