Running JOREK

Run the Code

  • Use the namelist input file as standard input for the binary:
./jorek_model303 < ./input
  • For OpenMP, e.g. with 16 threads, set the following before running the code:
export OMP_NUM_THREADS=16
  • Run with MPI parallelization, e.g. with 4 processes: 
mpirun -n 4 ./jorek_model303 < input

Namelist Input File

  • JOREK simulations are controlled by input parameters that are set in a namelist input file of the following form:
&in1
  restart=.false.
  nstep_n=0
  tstep_n=1.e3
  ...
/
  • Some input parameters are model-specific. Check models/modelXXX/initialise_parameters.f90 for details.
  • Default values are assigned to the input parameters by models/preset_parameters.f90.
  • The input parameters are read by MPI thread 0 and sent to all other MPI threads via trunk/communication/broadcast_phys.f90
  • Example input files can be found in the repository in the directories trunk/namelist/modelXXX/.
  • All input parameters are written out automatically at the start of a simulation. This allows checking that the correct parameters are used, especially in case default values are used.

General Parameters

Input parameter Explanation Comment
restart Restart an interrupted simulation Allows to restart a JOREK simulation from the jorek_restart.rst file. Default: .false.
rst_format Use this format number in the restart file. Valid options: 0, 1, >1 See export_restart.f90
regrid Realign the grid on flux surfaces Never used
nout Output a restart file every nout steps  

Physics Parameters

Input parameter Explanation Comment
tauIC Normalised inverse cyclotron frequency (extended MHD)  
eta Resistivity  
visco(_par) Viscosity of the perpendicular(parallel) velocity  
D_par Parallel particle diffusion = zero
D_perp(i) Perpendicular particle diffusion $D_{base} = D_1 \left(( 1 - D_2) + \frac{D_2}{2} \left(1 - \tanh \left(\frac{\psi - D_5}{D_4} \right) \right)\right)$ if model < 300 then $D = D_{base}$ else $D = D_{base} + \frac{D_6 D_2}{2} \left(1 - \tanh \left( \frac{D_5 + D_3 -\psi}{D_4} \right) \right)$ in the centre, (1-D(2)) in the pedestal, width of the pedestal region, width of transition, start of pedestal, diffusion outside of pedestal respectively (1-6)
ZK_par Parallel heat diffusion  
ZK_perp(i) Perpendicular heat diffusion See particle diffusion
F0 R*B_phi  
FF_0 FF' in the plasma centre  
FF_1 FF' at the plasma edge  
FF_coef(i) FF'(psi) = [(FF_0 - FF_1)*(1 + coef(1) * psi + coef(2)*psi**2 + coef(3)*psi**3) + coef(6)/cosh((psi - coef(7))/coef(8))**2 / (2.d0 * coef(8)) / (psi_1-psi_0) ]*0.5*(1- tanh((psi-coef(5))/coef(4))) + FF_1  
heatsource Amplitude of energy source heating, in normalized units
particlesource Amplitude of particle source  
T_0, T_1 Initial temperature profile, on axis (0) and at boundary  
T_coef(i) T(psi) = (T_0 - T_1)*(1 + coef(1) * psi + coef(2)*psi**2 + coef(3)*psi**3) *0.5*(1- tanh((psi-coef(5))/coef(4))) + T_1  
rho_(0,1,coef(i)) Initial density profile (see temperature)  
V_(0,1,coef(i)) Initial toroidal velocity (rotation frequency if normalized_velocity_profile = .t.) profile (see temperature)  
central_density Central number density Used in normalization of other quantities
central_mass Central average ion mass Deuterium mass if not defined

Numerics Parameters

Input parameter Explanation Comment
tstep_n Size of the time step $\Delta t$ Up to seven values possible (corresponding to several nstep_n values)
nstep_n Number of time steps Up to seven values possible (corresponding to several tstep_n values)
tstep Timestep when not using nstep_n  
nstep Number of tstep steps to make  
eta_num Numerical hyper-resistivity It is recommended to use small values according to eta_num \lesssim eta^2 and perform a scan of eta_num to ensure it does not influence results.
visco(_par)_num Numerical hyper-viscosity It is recommended to use small values according to visco_num \lesssim visco^2 and perform a scan of visco_num to ensure it does not influence results.
D_perp_num Hyper-diffusion (particles, numerical stabilization)  
ZK_perp_num Hyper-diffusion (heat, numerical stabilization)  
tgnum(i) Factor for Taylor-Galerkin stabilization in equation i See information about using the stabilization.
iter_precon redo factorisation if gmres iterations > iter_precon GMRES parameters
gmres_m gmres dimension  
gmres_4 scaling of error see source code
gmres_max_iter maximum number of gmres iterations  
gmres_tol convergence when error < gmres_tol  

Geometry Parameters

Input parameter Explanation Comment
mf Number of harmonics in fbnd if mf=0 use ellipticity etc
fbnd(i) Radius of plasma boundary (Fourier series) in polar angle has mf terms
ellip Ellipticity of initial polar grid  
tria_u/l Upper/Lower triangularity  
quad_u/l Upper/Lower quadrangularity  
xampl Analytic definition of psi(theta) on the boundary amplitude
xsig    
xwidth width of perturbation psi(theta) in theta  
xtheta position of x-point in angle  
xleft Shifts the radial position of the plasma  
xshift Shifts the vertical plasma position  
xpoint Plasma has x-point  
R_begin Square Bezier grid minimum R  
R_end Square Bezier grid maximum R  
Z_begin Square Bezier grid minimum Z  
Z_end Square Bezier grid maximum Z  
R_geo R-coordinate of the centre of the polar grid  
Z_geo Z-coordinate of the centre of the polar grid  
amin The minor radius of the polar grid  
n_R The number of horizontal nodes in the square Bezier grid Leads to n_R-1 elements in that direction
n_Z The number of vertical nodes in the square Bezier grid Leads to n_Z-1 elements
n_radial The number of radial nodes in the polar Bezier grid (grid 1)  
n_pol The number of poloidal nodes in the polar Bezier grid (grid 1)  
n_flux The number of radial nodes on the flux aligned grid inside the plasma boundary (grid 2)  
n_tht The number of poloidal nodes (grid 2)  
n_open The number of radial nodes, from magnetic axis to separatrix (grid 2)  
n_leg The number of nodes along the separatrix legs, from x-point to divertor wall (grid 2)  
n_private The number of radial nodes in the private region, region below the separatrix legs (grid 2)