University of Toronto, Canada
The OEDGE code
The OEDGE code suite is used for modeling the behavior of the edge plasma of magnetic fusion devices. The OEDGE code is composed of several elements. An Onion Skin Model (OSM) can be used to create a background plasma solution for the edge plasma. This includes solving for edge plasma density, temperature, flow velocity and electric field on a 2D grid representing the poloidal magnetic structure of the fusion device. OEDGE typically assumes toroidal symmetry. OEDGE can also load a plasma solution provided by other plasma solvers in use by the fusion modeling community. After an initial plasma solution is obtained, OEDGE runs the EIRENE Monte Carlo code to obtain the distribution and magnitude of hydrogenic source terms: ionization, recombination, power and momentum loss terms. These are then fed back to the OSM and the process is iterated until a stable plasma solution is found. OEDGE then uses the DIVIMP Monte Carlo plasma impurity code to model the impurity production, transport and deposition including such things as estimates of core impurity content, self-sputtering at surfaces and many other physics effects. The end result is a simulation of the fusion edge plasma including both hydrogenic and impurity behavior.
The input to these simulations is usually the density and temperature of the plasma at the target plates as measured by Langmuir probes. The simulation results are compared to measurements of hydrogenic and impurity spectral emissions, plasma condition measurements from Thomson scattering and reciprocating probes, neutral pressure measurements from in-vessel pressure gauges and many other diagnostics. Discrepancies between the simulation and experimental measurements are then investigated to determine whether physics not presently incorporated in the simulations could be the cause of the discrepancy. In this way, the OEDGE code is used to understand complex experiments with fusion edge plasmas and offers a tool giving insight into the physics underlying the experiments.
In terms of fundamental data used by the OEDGE code, there is a substantial body of data required by both EIRENE and DIVIMP in order to accomplish a simulation of this type. These codes require ionization and recombination rate data for hydrogen as well as the impurity species being modeled. The impurity species examined range from lithium to tungsten. These data are required for all charge states of the impurity species. Typically in DIVIMP, the code assumes a Maxwellian electron velocity distribution which then means that the code uses rate data as opposed to cross section data. The EIRENE code also uses data for hydrogen molecules as well as atoms and ions. In addition, these codes also need line emission data for both hydrogen and specific lines for each impurity species. The specific lines are those examined experimentally so they may vary from one fusion device to another depending on the diagnostic choices made. The spectral atomic data are combined with the impurity density calculated by the code to obtain a photon emission rate for each element in the simulation. These data are then combined with diagnostic geometry data loaded into the code to generate a simulated diagnostic measurement which takes into account such things as field of view of the instrument.
Finally, DIVIMP also makes use of surface sputtering data for both impurity species as well as hydrocarbon molecules. This includes both chemical and physical sputtering as well as primary (background ion) and self sputtering data. These data are used to calculate the source impurity particle distribution in the simulation. As a data consumer, OEDGE uses a broad range of atomic, molecular and surface data at a relatively processed level since the code only produces particle density estimates in the simulation and does not, at present, track the population of excited states.
As a result of the CCN meetings good information was acquired on capabilities and limitations of data available in the atomic physics community. In addition, the needs of a variety of consumer level codes were explained so the understanding of what is needed by these codes was clarified. This may result in improvement of the data currently available for fusion edge plasma modeling.