State-of-the-art stellarator optimization code
toc The Drift Kinetic Equation Solver (DKES)<ref>W. I. van Rij and S. P. Hirshman, Variational Bounds for Transport Coefficients in Three-Dimensional Plasmas, Phys. Fluids B 1,563(1989)</ref> solves a set of 3D drift kinetic equations to obtain upper and lower bounds for the diffusion coefficients of a prescribed toroidal plasma equilibrium. The three dimensions are theta (poloidal angle), zeta (toroidal angle), and alpha (pitch angle). Straight-line flux coordinates are used to describe the equilibrium, which must satisfy the stellarator symmetry conditions.
Table of Contents
The DKES code is based on a variation principle for the linearized drift-kinetic, Fokker-Planck equation. The formulation neglects the effect of magnetic drifts on the orbits of the perturbed distribution function. This neglect precludes the code from being able to study resonant superbanana orbits, and systems with very low collisionality. Such effects must also be considered when the radial electric field potential satisfies \(e\Phi/T < 1/A\). The advantage of this formualtion is that it reduces the problem from 5 dimension to 3. The conservation of symmetry and conservative properties are also preserved allowing for a variation approach to be taken.
DKES is a component of the STELLOPT suite of codes.
The DKES code needs an ‘input_dkes.ext’ file which contains the DKES_INDATA
namelist such as
&DKES_INDATA
NZPERIOD= 3, ! Field Periods
MPOL=17, ! Number of poloidal modes in distribution
NTOR= 26, ! Number of toroidal modes in distribution
MPOLB = 6, ! Number of poloidal modes in equilibrium
NTORB = 8, ! Number of toroidal modes in equilibrium
LALPHA= 100, ! Number of polynom. to use in pitch angle
NRUN = 1, ! Number of cmul/efield pairs to be calculated
CMUL = 0.1000E-02, ! Array of Collisionality
EFIELD = 0.1000E+03, ! Array of Radial electric field values
CHIP = 0.1403, ! d(chi)/d(rho) Poloidal flux derivative
PSIP = 0.2783, ! d(psi)/d(rho) Toroidal flux derivative
BTHETA = 4.6719062102E-03, ! I Toroidal current
BZETA = 2.3228644337E+00, ! J the Poloidal current
IBBI = 1, ! 1: BORBI is B; 2: BORBI is 1/B
! BORBI is the of nonzero Fourier coefficients for B (or 1/B)
! B(theta,zeta) = borbi(n,m)*cos[(m-1)*theta - nzperiod*nvalsb(n)*zeta]
BORBI(0,0)= 0.15748E+01,
BORBI(0,1)= -.11885E+00,
BORBI(0,2)= -.19193E-01,
BORBI(1,2)= 0.88849E-02,
BORBI(0,3)= 0.48462E-02,
BORBI(2,3)= -.44171E-02,
BORBI(-1,1)= -.26255E-02,
BORBI(1,1)= -.23643E-02,
BORBI(2,2)= -.21185E-02,
BORBI(3,3)= 0.20004E-02,
BORBI(4,4)= -.13090E-02,
BORBI(1,0)= -.11995E-02,
BORBI(1,3)= 0.10712E-02,
BORBI(1,4)= -.99923E-03,
BORBI(2,0)= -.97714E-03,
BORBI(4,3)= -.95223E-03,
BORBI(5,2)= 0.87245E-03,
BORBI(3,4)= 0.86562E-03,
BORBI(-2,1)= -.86180E-03,
BORBI(4,2)= -.81482E-03,
BORBI(3,2)= 0.77844E-03,
BORBI(2,1)= -.71725E-03,
BORBI(-1,2)= 0.55397E-03,
BORBI(3,0)= 0.51978E-03,
BORBI(1,5)= 0.51622E-03,
! MMNN Matrix containing the distribution of toroidal mode numbers
! (expressed in units of nzperiod) in column 1, and the
! number of poloidal modes associated with each toroidal
! mode in column 2. the poloidal mode numbers associated
! with each toroidal mode are stored in the same row,
! starting in column 3. WARNING : this spectrum must
! encompass the spectrum of the input magnetic field.
mmnn(1,1)= 0, 14, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13,
mmnn(1,2)= 1, 15, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14,
mmnn(1,3)= 2, 17, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14, 15, 16,
mmnn(1,4)= 3, 16, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14, 15,
mmnn(1,5)= 4, 16, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14, 15,
mmnn(1,6)= 5, 16, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14, 15,
mmnn(1,7)= 6, 15, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14,
mmnn(1,8)= 7, 15, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14,
mmnn(1,9)= 8, 15, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13, 14,
mmnn(1,10)= 9, 14, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
13,
mmnn(1,11)= 10, 13, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
mmnn(1,12)= 11, 11, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
mmnn(1,13)= 12, 10, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12,
mmnn(1,14)= 13, 6, 4, 6, 7, 8, 9, 10,
mmnn(1,15)= 14, 3, 6, 7, 8,
mmnn(1,16)= 15, 1, 6,
mmnn(1,17)= -1, 10, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,
mmnn(1,18)= -2, 8, 1, 2, 3, 4, 5, 6, 7, 8,
mmnn(1,19)= -3, 8, 1, 2, 3, 4, 5, 6, 7, 8,
mmnn(1,20)= -4, 6, 1, 2, 3, 4, 5, 6,
mmnn(1,21)= -5, 5, 1, 2, 3, 4, 5,
mmnn(1,22)= -6, 4, 1, 2, 3, 4,
mmnn(1,23)= -7, 3, 1, 2, 3,
mmnn(1,24)= -8, 3, 1, 2, 3,
mmnn(1,25)= -9, 2, 1, 2,
mmnn(1,26)=-10, 2, 1, 2,
/
This data can be generated from a ‘boozmn’ Boozer Transformation file. The user can also provide DKES with such a file to ease implementation.
The simplest invokation of DKES is as follow
> xdkes input_dkes.example
where the file input_dkes.example contains the DKES_INDATA namelist
for that run.
Alternatively the user can invoke DKES as follows
> xdkes input.example 5 1.0E-3 1.0
where the files wout_example.nc (or wout.example) and boozmn_example.nc
are present. The next four numbers are the boozer radial surface,
collisionality (nu/v in m^-1), and radial electric field (E_s/v).
The user may provide a screen output flag after those values (T/F).
Additionally, the user may provide 3 additional command line arguments:
a filename modifer which is appended to the output files, the order of
the Legendre polynominals used in the code (100), and the number of iterations
for coupling (4). This invokation will automatically produce an input_dkes.example
file with the filename modifer appended if supplied.
The code will produce three files ‘dkesout’, ‘opt_dkes’, and ‘results’. Additionally, if called using the alternative invokation an ‘input_dkes’ file will be generated. The ‘results’ file contains a table of 19 values with a row for each collisionality/efield pair. The ‘opt_dkes’ file contains a simplified output used in the old version of STELLOPT. The ‘dkesout’ file contains a more detailed and nuanced version of the output in ‘results.’
The ‘results’ file may be read by any plotting package capable of
importing space delimited data. Note the columns for LXXm and LXXp
are the variational bounds for the coefficients.
Put links to tutorial pages here.