Scientific workflows for MHD stability chain analysis of tokamak plasmas
Poster (konferens), 2015
The onset of MHD activity in present tokamak plasmas plays a detrimental role in
the discharge performance and operational limits. A comprehensive characterisation of the
modes observed and stability boundaries often call for the comparison of the
experimentally measured toroidal/poloidal mode spectra and propagating frequencies
against the predictions from linear MHD codes. In addition, predicting the MHD stability
boundaries in operational plasma scenarios foreseen for ITER and under investigation for
future devices such as DEMO plays an essential role in fusion research. Thus, the
benchmark of high-resolution MHD equilibrium and stability solvers is an important step
in the verification and validation of the involved numerical codes. Such benchmarks are
greatly facilitated when all codes share the same physics and machine data ontology and
methods for reading and writing the data, as in the European Integrated Modelling
framework [1], presently maintained by the EUROfusion Code Development Work
Package. Under its auspices, a scientific workflow, coded in Kepler [2] and dedicated to
the analysis of MHD activity in tokamak plasmas, has been developed, that can be
coupled to equilibrium reconstruction codes or discharge simulator scientific workflows.
The workflow presently includes HELENA [3], CHEASE [4] and CAXE [5] equilibrium
codes and ILSA [6], MARS [7], MARS-F [8] and KINX [5] linear MHD stability codes.
This paper presents the benchmark of the codes integrated therein on custom-made
circular and elongated plasma equilibria as well as JET and AUG plasma equilibria
derived from experimental data. The study involved internal kink and global modes
extending to the vacuum region. Radial and poloidal convergence analysis and growth
rate dependence of global modes on the distance of the ideal conducting wall to the
plasma boundary are addressed. Examples of interchangeability between some of the
involved codes, an added value of the workflow user interface, are also included.