Elsevier Science Home
Computer Physics Communications Program Library
Full text online from Science Direct
Programs in Physics & Physical Chemistry
CPC Home

[Licence| Download | New Version Template] acwb_v1_0.gz(25 Kbytes)
Manuscript Title: THALIA - a one-dimensional magnetohydrodynamic stability program using the method of finite elements.
Authors: K. Appert, D. Berger, R. Gruber, F. Troyon, K.V. Roberts
Program title: THALIA
Catalogue identifier: ACWB_v1_0
Distribution format: gz
Journal reference: Comput. Phys. Commun. 10(1975)11
Programming language: Fortran.
Computer: CDC CYBER 7326.
Operating system: SCOPE 3.4.2.
RAM: 25K words
Word size: 60
Peripherals: disc.
Keywords: Plasma physics, Ideal mhd, Stability, Variational principle, Finite elements, Spectrum, Instabilities, Eigenfunctions.
Classification: 19.6.

Subprograms used:
Cat Id Title Reference
ABUF_v1_0 OLYMPUS CPC 7(1978)245
ABUF_v2_0 OLYMPUS FOR IBM 370/165 CPC 9(1975)51
ABUF_v3_0 OLYMPUS FOR CDC 6500 CPC 10(1975)167
ACWC_v1_0 HYMNIA CPC 10(1975)30

Nature of problem:
The lifetime of a magnetically-confined plasma column depends critically on the growth rates of any unstable eigenmodes. The plasma is considered here as a one-component fluid described by the ideal one- dimensional MHD equations. Purely oscillating or purely growing and damped modes can be studied by linearising these equations and perturbing an equilibrium state of the column. Not only the purely growing modes, i.e. the unstable eigenfunctions, but also stable, purely oscillating modes can be of interest. THALIA has been written to find the whole spectrum from the variational principle.

Unusual features:
THALIA uses the CDC OLYMPUS package, and follows all prescriptions of the OLYMPUS system. The code is written in STANDARD FORTRAN, except for the use of the input facility NAMELIST which is available on most computers, and is optimized for speed and memory requirements. Memory requirements are reduced by storing only half the band-width for the symmetric band matrices A and B. For detecting a singular A or a degenerate eigenvalue problem in SIVI a machine dependent parameter EPSMAC has to be defined. EPSMAC is set to be 10**-12 for a CDC 6500.

Running time:
The time required is proportional to the number of intervals, to the number of wanted eigenvalues and to the number of iterations. A typical case with 100 intervals, 20 iteration steps takes about 10 s per eigenvalue.