Speaker
Brendan Shanahan
Description
See the full Abstract at http://ocs.ciemat.es/EPS2018ABS/pdf/P5.1090.pdf
BSTING: modifying the BOUT++ framework for fluid simulations of
turbulence in non-axisymmetric geometries
B Shanahan1, B Dudson2, P Hill2, and P Helander1
1
Max-Planck Institute für Plasmaphysik, Teilinstitut Greifswald, Germany
2
York Plasma Institute, Department of Physics, University of York, UK
It is becoming increasingly important to simulate plasma turbulence in non-
axisymmetric configurations, especially in the edge where turbulence could become the
dominant transport process. The high collisionality of tokamak and stellarator edge plasmas
facilitates a fluid approach to turbulence simulations. While there are several fluid turbulence
codes for tokamak geometries, previous attempts to develop such a simulation framework for
stellarators have been unsuccessful. The recent implementation of the Flux Coordinate
Independent (FCI) [1] method for parallel derivatives in BOUT++ [2] has allowed for
simulations in nonaxisymmetric geometries [3,4]. Here we present the most recent results for
the BSTING project, which seeks to modify the BOUT++ framework to Simulate Turbulence
In Non-axisymmetric Geometries.
To allow for fully three dimensional turbulence simulations, the metric tensor
components in BOUT++ have been extended to vary in three dimensions. Following this
extensive modification, we present the results of several tests to ensure the accuracy and
stability of the framework have been maintained. Of particular importance are the tests of the
parallel derivatives and the associated parallel boundary conditions. These methods have
been examined qualitatively by tracing non-axisymmetric flux surfaces and quantitatively via
the Method of Manufactured Solutions [5], the results of which will be presented.
A fully three dimensional framework provides a very flexible test bed, and therefore
several new features to exploit this flexibility will be presented here: modifications to the
Zoidberg grid generator which allow for Wendelstein 7-X geometries and a newly-
implemented FCI curvilinear coordinate system are discussed in detail.
Finally, initial investigation of plasma filaments in non-axisymmetric geometries using
an isothermal model which evolves electron density, vorticity, electromagnetic potential and
parallel momentum is presented. The implications of these simulations for future experiments
will be explored.
[1] F Hariri and M Ottaviani, Computer Physics Communications, 184(11):2419 – 2429, 2013.
[2] B Dudson et al., Computer Physics Communications, 180: 1467-1480
[3] P Hill, B Shanahan and B Dudson, Computer Physics Communications, 213: 9-18, 2016
[4] B Shanahan et al., Journal of Physics: Conference Series 775 012012, 2016.
[5] P J Roache, Journal of Fluids Engineering 124 4, 2002.
