Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Initialising ...
Bierwage, A.; Di Troia, C.*; Briguglio, S.*; Vlad, G.*
Computer Physics Communications, 183(5), p.1107 - 1123, 2012/05
被引用回数:11 パーセンタイル:55.74(Computer Science, Interdisciplinary Applications)This work deals with the initial loading of phase space markers for global gyrokinetic particle-in-cell (PIC) simulations of plasmas that are magnetically confined in a toroidally axisymmetric configuration. A method is presented, which allows to prepare a marker distribution that is independent of time. This is achieved by loading markers on the toroidal surfaces of unperturbed guiding center orbits, where they are distributed uniformly in time. This allows to initialize global PIC codes with an accurate equilibrium distribution function and facilitates simulations with minimal noise-signal correlation. The concrete problem considered is the representation of energetic ions in tokamaks, which are characterized by large drifts across magnetic surfaces.
Bierwage, A.; Di Troia, C.*; Zonca, F.*
no journal, ,
A new software tool is presented which allows to initialize particle simulations of magnetically confined plasmas with an arbitrary particle distribution function. Besides facilitating simulations with (in principle) arbitrary spatial localization and velocity-space anisotropicity, the distinguishing feature of this tool is that it allows to construct an exact kinetic equilibrium for cases where magnetic drift orbits play an important role. One timely example is the interaction between energetic ions and magnetohydrodynamic modes in tokamaks used to study burning plasma conditions. In such cases, input data that consists of measurable (and physically meaningful) quantities is not sufficient to uniquely describe the equilibrium distribution function. The additional information required to design an exact equilibrium from incomplete input data is obtained by pre-computing the unperturbed drift orbits of all phase-space samples. Marker particles are distributed along these trajectories (excluding loss orbits) and appropriate weight factors are assigned. First simulation results with the new marker loading module will be presented as they become available.