XGC1
|
The module "ptb_3db" contains the framework for handling perturbed 3D magnetic fields in XGCa and XGC1. More...
Public Member Functions | |
subroutine | read_3db (grid, psn) |
read_3db allocates memory for the perturbed field and reads the perturbed field from a file. Supported file formats are (R,Z)-ASCII and M3D-C1. More... | |
subroutine | read_3db_test (grid, psn) |
Sets up an analytical magnetic field perturbation with fixed toroidal mode number and (currently) 2 poloidal mode numbers. The number of poloidal mode numbers can be adjusted with few code changes. The most important parameters are the value of the poloidal mode numbers, the centers of the Gaussian envelopes of each mode number, the width of the Gaussian, and the amplitude. More... | |
subroutine | read_3db_m3dc1_mesh (grid) |
Allocates memory for and reads perturbed field data from M3D-C1 file. For M3D-C1 input, the fusion-io library by N. Ferraro is used. The interpolation used in the fusion-io library keeps the field divergence free. More... | |
subroutine | read_3db_m3dc1_rz (grid) |
Allocates memory for and reads perturbed field data from M3D-C1 file. This version of the routine reads the vector potential A on an (R,Z) grid. The perturbed field on the XGC mesh can be evaluated in a separate subroutine. Eventually, the perturbed magnetic field is to be calculated directy from A using spline interpolation. For M3D-C1 input, the fusion-io library by N. Ferraro is used. More... | |
subroutine | extract_resonant_vector_potential_from_m3dc1 (grid, psn, itor) |
Extension of RZ reader for EM Use RZ reader to read the full vector potential from M3D-C1. For EM, proeceed to extract the (near) resonant components of \(A_\parallel\) (A_parallel) on the XGC mesh. Interpolate those back to the RZ grid and subtract from the raw vector potential from M3D-C1. Then re-initialize the spline interpolation for A. This enables pushing particles with the full M3D-C1 field while still calculating the plasma response to the near resonant modes. As a stop-gap measure, the such filtered vector potential is used to calculate the perturbed field on the XGC mesh because the spline interpolation of the vector potential on the RZ grid is not available in C++ yet. This currently works only for single toroidal mode number input. Extending it to multiple mode numbers seems straightforward. More... | |
subroutine | read_3db_m3dc1_eem (grid, psn) |
Allocates memory for and reads perturbed field data from M3D-C1 file for use with the explicit EM method. This reader reads only the parallel component of the perturbed vacuum vector potential for a single toroidal mode number. More... | |
subroutine | init_apar_3db (grid, psn) |
Sets the symplectic vector potential equal to the vacuum RMP vector potential. More... | |
subroutine | psn_mem_alloc_3db (psn, n) |
Allocates memory for the perturbed current in a psn-object. More... | |
real(kind=8) function | ptb_3db_screening_factor (r, z, psi) |
Evaluates the screening factor for artificial exponential damping of the vacuum perturbed field towards the magnetic axis. The screening shape function is simply exp((psi-psi0)/w). Since psi decreases towards the magnetic axis, psi-psi0 becomes negative at some point. More... | |
subroutine | ptb_3db_calc_db (grid, dA_phi, dB_re, dB_im) |
Evaluates the perturbed field from the perturbed vector potential on the XGC grid. More... | |
subroutine | update_3db () |
Updates the ramp-up factor of the vacuum RMP field. The external (or vacuum) RMP field can be ramped up linearly instead of a hard switch from off to on. The ramp-up factor ptb_3db_rampup_fac determines the amplitude of the external RMP field throughout XGC. This routine is used only in electrostatic simulations. In electromagnetic simulations, push_As updates the ramp-up factor. More... | |
subroutine | ptb_3db_damping_factor_em (grid, Apar, dApar) |
Calculates an adaptive damping factor to limit the time rate of change of the perturbed vector potential The routine calculates the poloidal Fourier spectrum of the current increment of the vector potential (dApar –> effectively the Hamiltonian vector potential Ah) and the vector potential itself (Apar). Then it compares the amplitude of the increment and the current value and computes a damping factor for each poloidal mode number such that the increment is limited to be a fraction of the current value. Using this damping factor, the increment is transformed back to real space and returned to the calling routine. More... | |
Public Attributes | |
integer | ptb_3db_mode =0 |
Mode of operation: (0) static RMP field (ES), or fully self-consistent RMP field (EM version) (2) self-consistent RMP field with time-averaged field-equations (EM version) More... | |
character(len=250) | ptb_3db_filename ='C1.h5' |
File containing the perturbed field data. More... | |
integer | ptb_3db_file_format =1 |
Format of input file with perturbed field data: (0) toroidal Fourier coefficients in R,Z, (1) read from M3D-C1 file using fusion-io (2) generic test data (no file) More... | |
logical | ptb_3db_single_mode_input =.true. |
For M3D-C1 input; If .true., M3D-C1 file contains only one toroidal mode number. More... | |
integer | ptb_3db_single_mode_ntor =3 |
The mode number in the M3D-C1 file in case of single mode input. More... | |
real(kind=8) | ptb_3db_mult_factor =1D0 |
Factor for scaling the perturbed field. More... | |
integer | ptb_3db_m3dc1_timeslice =1 |
Time slice that is to be read from M3D-C1 file. More... | |
logical | ptb_3db_em_full_spec =.false. |
If false, only modes |m/q-n|<=sml_mode_select_mres_q are retained, if true, the remaining modes are retained in toroidal Fourier representation but without RMP response (EM only). More... | |
real(kind=8) | ptb_3db_em_inpsi =-1D0 |
Inner boundary of RMP representation in A_s. More... | |
real(kind=8) | ptb_3db_em_outpsi =1.1D0 |
Outer boundary of RMP representation in A_s. More... | |
real(kind=8) | ptb_3db_em_bd_width =0.05D0 |
Blending width between A_s and full A representation. More... | |
integer | ptb_3db_dasdt_opt =1 |
Selector for how to compute dAs/dt for RMP penetration. More... | |
integer | ptb_3db_mstep_es =0 |
Number of ES time steps in RMP penetration calculation with ptb_3db_mode==2. More... | |
integer | ptb_3db_mstep_em =1 |
Number of EM time steps in RMP penetration calculation with ptb_3db_mode==2. More... | |
integer | ptb_3db_es_to_em_dt_ratio =1 |
Ratio of ES to EM time step size in RMP penetration calculation with ptb_3db_mode==2 (also: the number of time steps over which the perturbed current is averaged) More... | |
real(kind=8) | ptb_3db_update_alpha =1.0D0 |
Damping factor for update of the perturbed field. More... | |
integer | ptb_3db_start_time =1 |
Time step in which perturbed field is switched on. More... | |
logical | ptb_3db_rampup_vac =.true. |
(.true.) Ramp up perturbed field slowly, (.false.) turn on perturbed field abruptly More... | |
integer | ptb_3db_rampup_time =1000 |
Number of time steps over which the perturbed field is ramped up. More... | |
real(kind=8) | ptb_3db_rampup_fac =0D0 |
The current relative amplitude of the vacuum vector potential (if ptb_3db_mode==2) More... | |
real(kind=8) | ptb_3db_rampup_fac0 =0D0 |
The previous (backup) relative amplitude of the vacuum vector potential (if ptb_3db_mode==2) More... | |
logical | ptb_3db_screening_on =.false. |
Artificial screening for static RMP field: (.false.) no screening, (.true.) exponential damping towards the axis. More... | |
real(kind=8) | ptb_3db_screening_width1 =0.02D0 |
Decay length of the exponential damping function in units of normalized poloidal flux. More... | |
real(kind=8) | ptb_3db_screening_width2 =0.02D0 |
Width of tanh for damping in the core in units of normalized poloidal flux. More... | |
real(kind=8) | ptb_3db_screening_psi1 =0.975D0 |
Center of the tanh function for screening in the core (<1) More... | |
real(kind=8) | ptb_3db_screening_fac1 =0.9D0 |
Reduction of deltaB at the separatrix (<=1) More... | |
real(kind=8) | ptb_3db_screening_fac2 =0.8D0 |
Fraction of vacuum RMP in the core (<=1) More... | |
integer | ptb_3db_num_ntor =1 |
Number of toroidal mode numbers (<= sml_nphi_total) More... | |
integer, dimension(:), allocatable, target | ptb_3db_ntor |
Array to store the toroidal mode numbers. More... | |
integer | ptb_3db_ntor_min =3 |
smallest toroidal mode number More... | |
integer | ptb_3db_num_mpol =30 |
Number of poloidal mode numbers. More... | |
integer, dimension(:), allocatable | ptb_3db_mpol |
Array to store the poloidal mode numbers. More... | |
integer | ptb_3db_nr =1000 |
(R,Z) grid size in R-direction More... | |
integer | ptb_3db_nz =1000 |
(R,Z) grid size in Z-direction More... | |
real(kind=8) | ptb_3db_r_min |
real(kind=8) | ptb_3db_r_max |
Range of R in (R,Z) grid. More... | |
real(kind=8) | ptb_3db_z_min |
real(kind=8) | ptb_3db_z_max |
Range of Z in (R,Z) grid. More... | |
real(kind=8) | ptb_3db_dr |
real(kind=8) | ptb_3db_dz |
grid spacing of (R,Z) grid More... | |
real(kind=8), dimension(:,:,:), allocatable, target | ptb_3db_bfield_re_vac |
perturbed vacuum field on XGC mesh, real part, dimensions: (grid vertex, R-Z-phi components, tor. mode number) More... | |
real(kind=8), dimension(:,:,:), allocatable, target | ptb_3db_bfield_im_vac |
perturbed vacuum field on XGC mesh, imaginary part More... | |
real(kind=8), dimension(:,:,:,:), allocatable | ptb_3db_vecpot_re_vac_rz |
perturbed vacuum vector potential on (R,Z) mesh, real part More... | |
real(kind=8), dimension(:,:,:,:), allocatable | ptb_3db_vecpot_im_vac_rz |
perturbed vacuum vector potential on (R,Z) mesh, imaginary part More... | |
real(kind=8), dimension(:), allocatable | ptb_3db_rgrid |
real(kind=8), dimension(:), allocatable | ptb_3db_zgrid |
R and Z grid in case of (R,Z) input data. More... | |
type(ezspline2_r8), dimension(:,:), allocatable | ptb_3db_vecpot_spl_rz_re |
Spline for interpolation of the real part of (R,Z) data to XGC mesh. More... | |
type(ezspline2_r8), dimension(:,:), allocatable | ptb_3db_vecpot_spl_rz_im |
Spline for interpolation of the imaginary part of (R,Z) data to XGC mesh. More... | |
type(ezspline2_r8) | ptb_3db_spl_tmp |
Dummy 2D spline used for setting up the 2D interpolations with bicub_mod. More... | |
type(bicub_type), dimension(:,:,:), allocatable | ptb_3db_bicub_da |
Spline for interpolation of the perturbed vector potential with bicub_mod. More... | |
The module "ptb_3db" contains the framework for handling perturbed 3D magnetic fields in XGCa and XGC1.
subroutine ptb_3db_module::extract_resonant_vector_potential_from_m3dc1 | ( | type(grid_type), intent(in) | grid, |
type(psn_type), intent(inout) | psn, | ||
integer, intent(in) | itor | ||
) |
Extension of RZ reader for EM Use RZ reader to read the full vector potential from M3D-C1. For EM, proeceed to extract the (near) resonant components of \(A_\parallel\) (A_parallel) on the XGC mesh. Interpolate those back to the RZ grid and subtract from the raw vector potential from M3D-C1. Then re-initialize the spline interpolation for A. This enables pushing particles with the full M3D-C1 field while still calculating the plasma response to the near resonant modes. As a stop-gap measure, the such filtered vector potential is used to calculate the perturbed field on the XGC mesh because the spline interpolation of the vector potential on the RZ grid is not available in C++ yet. This currently works only for single toroidal mode number input. Extending it to multiple mode numbers seems straightforward.
[in] | grid | XGC mesh object, type(grid_type) |
[in,out] | psn | XGC field object, type(psn_type) |
[in] | itor | Index of targeted mode number, integer |
subroutine ptb_3db_module::init_apar_3db | ( | type(grid_type), intent(in) | grid, |
type(psn_type), intent(inout) | psn | ||
) |
Sets the symplectic vector potential equal to the vacuum RMP vector potential.
[in] grid type(grid_type), XGC grid information
[in,out] | psn | Field data object, type(psn_type) |
subroutine ptb_3db_module::psn_mem_alloc_3db | ( | type(psn_type) | psn, |
integer | n | ||
) |
Allocates memory for the perturbed current in a psn-object.
psn | (in) type(psn_type), field data |
n | (in) integer, the number of mesh vertices |
subroutine ptb_3db_module::ptb_3db_calc_db | ( | type(grid_type), intent(in) | grid, |
real (kind=8), dimension(grid%nnode,2,ptb_3db_num_ntor), intent(in) | dA_phi, | ||
real (kind=8), dimension(3,grid%nnode,ptb_3db_num_ntor), intent(out) | dB_re, | ||
real (kind=8), dimension(3,grid%nnode,ptb_3db_num_ntor), intent(out) | dB_im | ||
) |
Evaluates the perturbed field from the perturbed vector potential on the XGC grid.
grid | (in) type(grid_type), grid data |
dA_phi | (in) real(8), perturbed vector potential (dA_phi, covariant phi-component) |
dB_re | (out) real(8), real (cos) part of the perturbed mag. field |
dB_im | (out) real(8), imaginary (sin) part of the perturbed mag. field |
subroutine ptb_3db_module::ptb_3db_damping_factor_em | ( | type(grid_type), intent(in) | grid, |
real(kind=8), dimension(grid%nnode), intent(in) | Apar, | ||
real(kind=8), dimension(grid%nnode), intent(inout) | dApar | ||
) |
Calculates an adaptive damping factor to limit the time rate of change of the perturbed vector potential The routine calculates the poloidal Fourier spectrum of the current increment of the vector potential (dApar –> effectively the Hamiltonian vector potential Ah) and the vector potential itself (Apar). Then it compares the amplitude of the increment and the current value and computes a damping factor for each poloidal mode number such that the increment is limited to be a fraction of the current value. Using this damping factor, the increment is transformed back to real space and returned to the calling routine.
[in] | grid | XGC mesh object, type(grid_type) |
[in] | Apar | Current symplectic vector potential As, double |
[in,out] | dApar | Increment of the vector potential (=Ah), double |
real (kind=8) function ptb_3db_module::ptb_3db_screening_factor | ( | real (kind=8), intent(in) | r, |
real (kind=8), intent(in) | z, | ||
real (kind=8), intent(in) | psi | ||
) |
Evaluates the screening factor for artificial exponential damping of the vacuum perturbed field towards the magnetic axis. The screening shape function is simply exp((psi-psi0)/w). Since psi decreases towards the magnetic axis, psi-psi0 becomes negative at some point.
r | (in) real(8), major radius |
z | (in) real(8), Z coordinate |
psi | (in) real(8), normalized poloidal flux corresponding to r and z |
subroutine ptb_3db_module::read_3db | ( | type(grid_type), intent(in) | grid, |
type(psn_type), intent(inout) | psn | ||
) |
read_3db allocates memory for the perturbed field and reads the perturbed field from a file. Supported file formats are (R,Z)-ASCII and M3D-C1.
[in] | grid | type(grid_type), XGC grid information |
[in,out] | psn | XGC field data object, type(psn_type) |
subroutine ptb_3db_module::read_3db_m3dc1_eem | ( | type(grid_type), intent(in) | grid, |
type(psn_type), intent(inout) | psn | ||
) |
Allocates memory for and reads perturbed field data from M3D-C1 file for use with the explicit EM method. This reader reads only the parallel component of the perturbed vacuum vector potential for a single toroidal mode number.
[in] | grid | type(grid_type), XGC grid information |
[in,out] | psn | XGC field data object, type(psn_type) |
subroutine ptb_3db_module::read_3db_m3dc1_mesh | ( | type(grid_type), intent(in) | grid | ) |
Allocates memory for and reads perturbed field data from M3D-C1 file. For M3D-C1 input, the fusion-io library by N. Ferraro is used. The interpolation used in the fusion-io library keeps the field divergence free.
grid | (in) type(grid_type), XGC grid information |
subroutine ptb_3db_module::read_3db_m3dc1_rz | ( | type(grid_type), intent(in) | grid | ) |
Allocates memory for and reads perturbed field data from M3D-C1 file. This version of the routine reads the vector potential A on an (R,Z) grid. The perturbed field on the XGC mesh can be evaluated in a separate subroutine. Eventually, the perturbed magnetic field is to be calculated directy from A using spline interpolation. For M3D-C1 input, the fusion-io library by N. Ferraro is used.
grid | (in) type(grid_type), XGC grid information |
subroutine ptb_3db_module::read_3db_test | ( | type(grid_type), intent(in) | grid, |
type(psn_type), intent(inout) | psn | ||
) |
Sets up an analytical magnetic field perturbation with fixed toroidal mode number and (currently) 2 poloidal mode numbers. The number of poloidal mode numbers can be adjusted with few code changes. The most important parameters are the value of the poloidal mode numbers, the centers of the Gaussian envelopes of each mode number, the width of the Gaussian, and the amplitude.
[in] | grid | XGC grid data object, type(grid_type) |
[in,out] | psn | XGC field data object, type(psn_type) |
subroutine ptb_3db_module::update_3db | ( | ) |
Updates the ramp-up factor of the vacuum RMP field. The external (or vacuum) RMP field can be ramped up linearly instead of a hard switch from off to on. The ramp-up factor ptb_3db_rampup_fac determines the amplitude of the external RMP field throughout XGC. This routine is used only in electrostatic simulations. In electromagnetic simulations, push_As updates the ramp-up factor.
real (kind=8), dimension(:,:,:), allocatable, target ptb_3db_module::ptb_3db_bfield_im_vac |
perturbed vacuum field on XGC mesh, imaginary part
real (kind=8), dimension(:,:,:), allocatable, target ptb_3db_module::ptb_3db_bfield_re_vac |
perturbed vacuum field on XGC mesh, real part, dimensions: (grid vertex, R-Z-phi components, tor. mode number)
type (bicub_type), dimension(:,:,:), allocatable ptb_3db_module::ptb_3db_bicub_da |
Spline for interpolation of the perturbed vector potential with bicub_mod.
integer ptb_3db_module::ptb_3db_dasdt_opt =1 |
Selector for how to compute dAs/dt for RMP penetration.
real (kind=8) ptb_3db_module::ptb_3db_dr |
real (kind=8) ptb_3db_module::ptb_3db_dz |
grid spacing of (R,Z) grid
real (kind=8) ptb_3db_module::ptb_3db_em_bd_width =0.05D0 |
Blending width between A_s and full A representation.
logical ptb_3db_module::ptb_3db_em_full_spec =.false. |
If false, only modes |m/q-n|<=sml_mode_select_mres_q are retained, if true, the remaining modes are retained in toroidal Fourier representation but without RMP response (EM only).
real (kind=8) ptb_3db_module::ptb_3db_em_inpsi =-1D0 |
Inner boundary of RMP representation in A_s.
real (kind=8) ptb_3db_module::ptb_3db_em_outpsi =1.1D0 |
Outer boundary of RMP representation in A_s.
integer ptb_3db_module::ptb_3db_es_to_em_dt_ratio =1 |
Ratio of ES to EM time step size in RMP penetration calculation with ptb_3db_mode==2 (also: the number of time steps over which the perturbed current is averaged)
integer ptb_3db_module::ptb_3db_file_format =1 |
Format of input file with perturbed field data: (0) toroidal Fourier coefficients in R,Z, (1) read from M3D-C1 file using fusion-io (2) generic test data (no file)
character(len=250) ptb_3db_module::ptb_3db_filename ='C1.h5' |
File containing the perturbed field data.
integer ptb_3db_module::ptb_3db_m3dc1_timeslice =1 |
Time slice that is to be read from M3D-C1 file.
integer ptb_3db_module::ptb_3db_mode =0 |
Mode of operation: (0) static RMP field (ES), or fully self-consistent RMP field (EM version) (2) self-consistent RMP field with time-averaged field-equations (EM version)
integer, dimension(:), allocatable ptb_3db_module::ptb_3db_mpol |
Array to store the poloidal mode numbers.
integer ptb_3db_module::ptb_3db_mstep_em =1 |
Number of EM time steps in RMP penetration calculation with ptb_3db_mode==2.
integer ptb_3db_module::ptb_3db_mstep_es =0 |
Number of ES time steps in RMP penetration calculation with ptb_3db_mode==2.
real (kind=8) ptb_3db_module::ptb_3db_mult_factor =1D0 |
Factor for scaling the perturbed field.
integer ptb_3db_module::ptb_3db_nr =1000 |
(R,Z) grid size in R-direction
integer, dimension(:), allocatable, target ptb_3db_module::ptb_3db_ntor |
Array to store the toroidal mode numbers.
integer ptb_3db_module::ptb_3db_ntor_min =3 |
smallest toroidal mode number
integer ptb_3db_module::ptb_3db_num_mpol =30 |
Number of poloidal mode numbers.
integer ptb_3db_module::ptb_3db_num_ntor =1 |
Number of toroidal mode numbers (<= sml_nphi_total)
integer ptb_3db_module::ptb_3db_nz =1000 |
(R,Z) grid size in Z-direction
real (kind=8) ptb_3db_module::ptb_3db_r_max |
Range of R in (R,Z) grid.
real (kind=8) ptb_3db_module::ptb_3db_r_min |
real (kind=8) ptb_3db_module::ptb_3db_rampup_fac =0D0 |
The current relative amplitude of the vacuum vector potential (if ptb_3db_mode==2)
real (kind=8) ptb_3db_module::ptb_3db_rampup_fac0 =0D0 |
The previous (backup) relative amplitude of the vacuum vector potential (if ptb_3db_mode==2)
integer ptb_3db_module::ptb_3db_rampup_time =1000 |
Number of time steps over which the perturbed field is ramped up.
logical ptb_3db_module::ptb_3db_rampup_vac =.true. |
(.true.) Ramp up perturbed field slowly, (.false.) turn on perturbed field abruptly
real (kind=8), dimension(:), allocatable ptb_3db_module::ptb_3db_rgrid |
real (kind=8) ptb_3db_module::ptb_3db_screening_fac1 =0.9D0 |
Reduction of deltaB at the separatrix (<=1)
real (kind=8) ptb_3db_module::ptb_3db_screening_fac2 =0.8D0 |
Fraction of vacuum RMP in the core (<=1)
logical ptb_3db_module::ptb_3db_screening_on =.false. |
Artificial screening for static RMP field: (.false.) no screening, (.true.) exponential damping towards the axis.
real (kind=8) ptb_3db_module::ptb_3db_screening_psi1 =0.975D0 |
Center of the tanh function for screening in the core (<1)
real (kind=8) ptb_3db_module::ptb_3db_screening_width1 =0.02D0 |
Decay length of the exponential damping function in units of normalized poloidal flux.
real (kind=8) ptb_3db_module::ptb_3db_screening_width2 =0.02D0 |
Width of tanh for damping in the core in units of normalized poloidal flux.
logical ptb_3db_module::ptb_3db_single_mode_input =.true. |
For M3D-C1 input; If .true., M3D-C1 file contains only one toroidal mode number.
integer ptb_3db_module::ptb_3db_single_mode_ntor =3 |
The mode number in the M3D-C1 file in case of single mode input.
type (ezspline2_r8) ptb_3db_module::ptb_3db_spl_tmp |
Dummy 2D spline used for setting up the 2D interpolations with bicub_mod.
integer ptb_3db_module::ptb_3db_start_time =1 |
Time step in which perturbed field is switched on.
real (kind=8) ptb_3db_module::ptb_3db_update_alpha =1.0D0 |
Damping factor for update of the perturbed field.
real (kind=8), dimension(:,:,:,:), allocatable ptb_3db_module::ptb_3db_vecpot_im_vac_rz |
perturbed vacuum vector potential on (R,Z) mesh, imaginary part
real (kind=8), dimension(:,:,:,:), allocatable ptb_3db_module::ptb_3db_vecpot_re_vac_rz |
perturbed vacuum vector potential on (R,Z) mesh, real part
type (ezspline2_r8), dimension(:,:), allocatable ptb_3db_module::ptb_3db_vecpot_spl_rz_im |
Spline for interpolation of the imaginary part of (R,Z) data to XGC mesh.
type (ezspline2_r8), dimension(:,:), allocatable ptb_3db_module::ptb_3db_vecpot_spl_rz_re |
Spline for interpolation of the real part of (R,Z) data to XGC mesh.
real (kind=8) ptb_3db_module::ptb_3db_z_max |
Range of Z in (R,Z) grid.
real (kind=8) ptb_3db_module::ptb_3db_z_min |
real (kind=8), dimension(:), allocatable ptb_3db_module::ptb_3db_zgrid |
R and Z grid in case of (R,Z) input data.