scientific article; zbMATH DE number 3978721

From MaRDI portal
Publication:3743022





[For the entire collection see Zbl 0563.00023.] The substance of this article takes as its starting point the many different natural time scales that can be identified in the processes that occur within the plasma of a tokamak device. Under major simplifying assumptions, but still leaving problems of great physical and mathematical complexity, the author describes methods based on the existence of multiple time scales that can be used to tackle numerical simulation of phenomena occurring in tokamak discharges. In particular, he confines himself to the MHD equations, thus avoiding discussion of relatively high frequency phenomena, and, further, excludes any asymptotic method based on a small toroidal aspect ratio. After the basic MHD equations have been introduced, the different time scales are defined and illustrated on a diagram as having magnitudes ranging from \(O(10^ 2)\) to \(O(10^{-8})\) seconds. The author introduces his method by asserting some general principles (a) only try to model on a single time scale for a given case - faster time scale phenomena are averaged, slower ones ignored, (b) use a time- dependent coordinate system that continually aligns itself with the magnetic field, (c) use a combined stream function - potential representation for a two-dimensional velocity field, (d) modify freely phenomena occurring on faster time scales. These principles are applied in two specific cases which are discussed in detail in sections II (Ideal time scale MHD simulation) and III (Resistive time scale MHD simulation). In the former those MHD instabilities are considered which have a time scale much shorter than dissipative time scales and so the ideal MHD equations suffice. Magnetic flux coordinates are used, which are therefore time-dependent, and numerically the dynamic grid method is employed [the author, \textit{J. L. Johnson}, \textit{J. M. Greene} and \textit{R. C. Grimm}, J. Comput. Phys. 29, 101-126 (1978; Zbl 0401.76090)]. The method is illustrated in two-dimensions, the finite difference process described and application made to two examples whose results are discussed. In III the time scale is much larger than wave transit times and times associated with the diffusion of particles and energy parallel to the magnetic field. For axisymmetric configurations using the same coordinate transformation (b) as in II, two methods are examined, the zero-mass method [\textit{S. P. Hirshmann} and the author, Phys. Fluids 22, 731-742 (1979; Zbl 0424.76093)] and the large mass method. Again, details of the methods, including the numerical techniques, are given and examples presented by way of application. The paper ends with a short discussion of factors which have been left out of consideration, such as treatment of time scales intermediate between ideal and resistive MHD, and fully three-dimensional simulations. There are also some interesting remarks about the role of the numerical method in describing the underlying physical processes. The reviewer feels this work would be of great value to all those concerned in numerical simulation of plasma behaviour in a tokamak.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3743022)