PRISM Common

The common/ directory contains all backend-independent modules shared by the CPU and FNL backends of PRISM. Every type described here is aggregated into prism_common_object, which serves as the base type for both backends.

Source Layout

File	Module	Description
adam_prism_parameters.F90	adam_prism_parameters	Physical constants, eigenvalue and eigenvector matrices
adam_prism_physics_object.F90	adam_prism_physics_object	Physical model selection, variable counts, eigenvector dispatch
adam_prism_numerics_object.F90	adam_prism_numerics_object	Temporal/spatial scheme selection, divergence correction config
adam_prism_common_object.F90	adam_prism_common_object	Base aggregate type for all backends
adam_prism_bc_object.F90	adam_prism_bc_object	Boundary conditions handler
adam_prism_ic_object.F90	adam_prism_ic_object	Initial conditions handler
adam_prism_io_object.F90	adam_prism_io_object	I/O configuration and output routines
adam_prism_time_object.F90	adam_prism_time_object	Time integration control
adam_prism_coil_object.F90	adam_prism_coil_object	Current-carrying coil current density sources
adam_prism_fWLayer_object.F90	adam_prism_fWLayer_object	Far-wave absorbing boundary layer
adam_prism_external_fields_object.F90	adam_prism_external_fields_object	Externally prescribed electromagnetic fields
adam_prism_rk_bc_object.F90	adam_prism_rk_bc_object	RK sub-step integrator for boundary conditions
adam_prism_pic_object.F90	adam_prism_pic_object	Particle-in-Cell configuration and particle-grid weighting
adam_prism_particle_injection_object.F90	adam_prism_particle_injection_object	Initial particle space and velocity distribution
adam_prism_leapfrog_pic_object.F90	adam_prism_leapfrog_pic_object	Leapfrog (Boris) PIC particle time integrator
adam_prism_rk_pic_object.F90	adam_prism_rk_pic_object	Runge-Kutta PIC particle time integrator
adam_prism_riemann_library.F90	adam_prism_riemann_library	Maxwell Riemann solvers (LLF, HLL) and convective flux routines
adam_prism_common_library.F90	adam_prism_common_library	Barrel re-export of all common modules

---

adam_prism_parameters — Physical Constants

Module-level parameters (all save):

Constant	Value	Description
NV_MAX	11	Maximum supported variable count (static array sizing)
MU0	1.256637e-6	Vacuum magnetic permeability (H/m)
EPS0	8.854187e-12	Vacuum permittivity (F/m)
C0	1/√(MU0·EPS0)	Vacuum speed of light (m/s)
E_CHARGE	-1.6e-19	Electron charge (C)
E_MASS	9.11e-31	Electron mass (kg)
K_B	1.380649e-23	Boltzmann constant (J/K)
PI	3.141592653589793	π

Derived constants: MU0_SQ = √MU0, MU0_SQ_I2 = 1/(2√MU0), EPS0_SQ = √EPS0, EPS0_SQ_I2 = 1/(2√EPS0).

Eigensystem Arrays

The Maxwell system has wave speeds $\pm c_0$ and two degenerate zero speeds. The module provides target arrays used by prism_physics_object to set eigenvector pointers:

Symbol	Shape	Description
EV(6)	[0, 0, c0, c0, −c0, −c0]	Eigenvalues of the 6-variable system
ER(6,6,3)	real, target	Right eigenvectors for x, y, z directions
EL(6,6,3)	real, target	Left eigenvectors for x, y, z directions
IEV(6)	all ones	Identity eigenvalues (conservative reconstruction)
IERL(6,6,3)	identity	Identity right/left eigenvectors (conservative)
IEV_D(7), IERL_D(7,7,3)	7-variable	With D divergence-cleaning scalar φ
IEV_B(7), IERL_B(7,7,3)	7-variable	With B divergence-cleaning scalar ψ
IEV_D_B(8), IERL_D_B(8,8,3)	8-variable	With both φ and ψ

---

adam_prism_physics_object — Physical Model

The prism_physics_object reads the physical model from the [physics] INI section and assembles the variable layout and eigensystem pointers.

INI Keys ([physics])

Key	Values	Description
physical_model	electromagnetic / PIC	Select EM or Particle-in-Cell model
chi	real > 0	Divergence-cleaning propagation speed coefficient (hyperbolic only); evmax = chi · c₀

Accepted aliases for electromagnetic: Maxwell, EM, em, maxwell. For PIC: pic, ParticleInCell.

Data Members

Member	Default	Description
physical_model	—	Selected model string
nv_c	6	Conservative variable count (D, B — grows by nv_cl)
nv_s	3	Source variable count (J, always zero flux)
nv_cl	0	Divergence-cleaning variable count (0, 1, or 2)
nv_pic	0	PIC charge density variable count (0 or 1)
nv	nv_c + nv_s + nv_cl + nv_pic	Total variable count
var_Jx/Jy/Jz	7/8/9 (no cleaning)	Q-vector indices of J components; shift with cleaning
chi	0	Cleaning speed coefficient
evmax	c₀ or chi·c₀	Maximum eigenvalue (maximum signal speed)
erw, elw	pointer	Right/left eigenvectors for WENO characteristic reconstruction

Variable Index Constants

VAR_DX = 1,  VAR_DY = 2,  VAR_DZ = 3
VAR_BX = 4,  VAR_BY = 5,  VAR_BZ = 6
! J indices shift depending on cleaning: var_Jx = 7, 8, or 9

Eigenvector Dispatch

At initialize, the pointers erw and elw are set based on the combination of reconstruction_vars and div_corr_var:

reconstruction_vars	Div correction	erw / elw
CONSERVATIVE	any	IERL / IERL_D / IERL_B / IERL_D_B (identity)
CHARACTERISTICS	none / POISSON	ER(6,6,3) / EL(6,6,3)
CHARACTERISTICS	HYPERBOLIC, D only	ER_D(7,7,3) / EL_D(7,7,3)
CHARACTERISTICS	HYPERBOLIC, B only	ER_B(7,7,3) / EL_B(7,7,3)
CHARACTERISTICS	HYPERBOLIC, D+B	ER_D_B(8,8,3) / EL_D_B(8,8,3)

For hyperbolic divergence cleaning the enlarged eigensystem uses the cleaning speed ch = chi·c₀: eigenvalues include 0, ±ch, ±c₀.

---

adam_prism_numerics_object — Numerical Schemes

Configured from the [numerics] INI section.

INI Keys ([numerics])

Key	Values	Description
scheme_time	RUNGE_KUTTA / LEAPFROG / BLANES_MOAN / COMMUTATOR_FREE_MAGNUS	Temporal integration scheme
scheme_space	WENO / FD_CENTERED / FV_CENTERED	Spatial discretisation scheme
fdv_order	integer	Order of centered FD/FV schemes
reconstruction_variables	CONSERVATIVE / CHARACTERISTICS	WENO reconstruction basis
constrained_transport	NO / D / B / DB	Enable Constrained Transport on D and/or B
divergence_correction	POISSON / HYPERBOLIC	Divergence error control strategy (absent = none)

The half-stencil width is derived as fdv_half_stencil = fdv_order / 2. Per-derivative stencil widths in fdv_half_stencils(6) are incremented progressively for higher-order derivatives.

---

adam_prism_common_object — Base Aggregate Type

prism_common_object aggregates every framework and PRISM sub-object and owns all simulation field arrays. Both backends (prism_cpu_object, prism_fnl_object) extend this type.

Data Members

Framework objects:

Member	Type	Description
mpih	mpih_object	MPI handler
adam	adam_object	Top-level ADAM framework handle
field	field_object pointer	Field storage and metrics
grid	grid_object pointer	Grid geometry
amr	amr_object	AMR marker handler
ib	ib_object	Immersed Boundary handler
slices	slices_object	Slice output handler
leapfrog	leapfrog_object	Leapfrog field integrator
blanesmoan	blanesmoan_object	Blanes-Moan field integrator
cfm	cfm_object	Commutator-Free Magnus integrator
rk	rk_object	Runge-Kutta field integrator
weno	weno_object	WENO reconstructor
flail	flail_object	Linear algebra methods (Poisson solver)

PRISM sub-objects:

Member	Type	Description
io	prism_io_object	I/O configuration
numerics	prism_numerics_object	Scheme selection
physics	prism_physics_object	Physical model and variable layout
ic	prism_ic_object	Initial conditions
bc	prism_bc_object	Boundary conditions
rk_bc	prism_rk_bc_object	RK sub-step BC integrator
time	prism_time_object	Time control
fWLayer	prism_fWLayer_object	Far-wave absorbing layer
coil	prism_coil_object	Coil current density sources
external_fields	prism_external_fields_object	Prescribed external fields
pic	prism_pic_object	PIC configuration and weighting
particle_injection	prism_particle_injection_object	Initial particle distribution
leapfrog_pic	prism_leapfrog_pic_object	Leapfrog PIC integrator
rk_pic	prism_rk_pic_object	RK PIC integrator

Convenience pointers (bound to field/physics members at init):

ngc, ni, nj, nk, nb, blocks_number, nv, nv_c, nv_s, nv_cl

Field arrays (allocated in allocate_common):

Array	Shape	Description
q	(nv, 1-ngc:ni+ngc, …, nb)	Conservative state vector (D, B, J, ±φ/ψ, ±ρ)
dq	same as q	Residual right-hand side
curl	same as q	Curl of field components (optional output)
divergence	same as q	Divergence of field components (optional output)
q_pic	(8, particle_number)	PIC particle state (PIC model only)
pic_fields	(8, particle_number)	Field values at particle locations (PIC model only)

Energy diagnostics:

Member	Description
energy_D(:)	Time history of electric field energy
energy_B(:)	Time history of magnetic field energy
rms_energy_error_D/B	RMS relative energy error

initialize_common Sequence

The 34-step initialization sequence called by both backends:

1. mpih%initialize — MPI setup
2. io%initialize(filename) — load and parse INI file
3. bc%initialize — read boundary condition types
4. numerics%initialize — read scheme selections and divergence correction
5. physics%initialize — configure physical model and eigensystem (uses numerics outputs)
6. adam%grid%initialize — initialize grid from INI (with BC type array)
7. adam%compute_blocks_number — estimate blocks from available memory
8. adam%initialize — allocate tree, maps, and field storage for nv variables
9. Associate convenience pointers (field, grid, ngc, ni, …, nv_cl)
10. pic%initialize — PIC configuration and particle count (PIC model only)
11. particle_injection%initialize — particle distribution config (PIC model only)
12. adam%refine_uniform — uniform mesh refinement to configured levels
13. adam%prune — prune tree regions defined by ijkl_prune
14. amr%initialize — AMR marker configuration
15. field%compute_metrics — cell geometry and metrics
16. time%initialize — read it_max, time_max, CFL
17. ic%initialize — read initial conditions type and parameters
18. fWLayer%initialize — compute far-wave layer damping function
19. coil%initialize — compute coil current density fields
20. ib%initialize — immersed boundary setup
21. slices%initialize — slice output configuration
22. external_fields%initialize — external field type and parameters
23. rk%initialize — RK coefficients (if RUNGE_KUTTA)
24. rk_bc%initialize — RK-BC stage arrays (if RUNGE_KUTTA)
25. leapfrog%initialize — leapfrog weights (if LEAPFROG)
26. blanesmoan%initialize — Blanes-Moan weights (if BLANES_MOAN)
27. cfm%initialize — CFM weights (if COMMUTATOR_FREE_MAGNUS)
28. weno%initialize — WENO stencil configuration (if WENO)
29. leapfrog_pic%initialize — Boris leapfrog PIC (if PIC + LEAPFROG)
30. rk_pic%initialize — RK PIC (if PIC + RUNGE_KUTTA)
31. flail%initialize — linear algebra methods
32. check_ngc_number — verify ngc ≥ weno%S and ngc ≥ fdv_half_stencil; stop on failure
33. allocate_common — allocate dq, divergence, curl, and PIC arrays
34. io_initialize — register auxiliary output fields with the I/O layer

io_initialize registers auxiliary fields conditionally:

Condition	Registered field	Names
io%save_residual_fields	dq	res_Dx … res_Jz [± res_ph/res_ps/res_rh]
io%save_divergence_fields	divergence	div_D, div_B, div_J, totvar, div_05, div_06, div_09…div_11 [± div_07/div_08/div_12]
coil%total_coils_number > 0	coil%j_vec, coil%coil_flag	j_vec_1/2/3, f_Gauss, coil_flag
io%save_curl_fields	curl	curlD_x/y/z, curlB_x/y/z, curlJ_x/y/z

Methods

Method	Description
allocate_common	Allocate dq, divergence, curl, and PIC arrays
initialize_common	Full 34-step initialization (called by backend initialize)
load_restart_files	Load q and time from HDF5 restart files
save_energy_error	Write energy error history if save frequency triggered
save_restart_files	Save HDF5 restart files and an XH5F snapshot
save_xh5f	Save simulation data in XDMF+HDF5 format with variable names

---

adam_prism_bc_object — Boundary Conditions

Reads one BC type per face from six INI sections.

INI Sections

[bc_x_min], [bc_x_max], [bc_y_min], [bc_y_max], [bc_z_min], [bc_z_max]

Each section requires the key type:

type value	Integer constant	Description
extrapolation	BC_EXTRAPOLATION = 1	Zero-gradient extrapolation
Neumann	BC_NEUMANN = 2	Neumann (zero normal derivative)
Dirichlet	BC_DIRICHLET = 3	Dirichlet (prescribed field values)
Silver_Muller	BC_Silver_Muller = 4	Silver-Müller absorbing BC
periodic	BC_PERIOD = 5	Periodic
radiative	BC_radiative = 6	Radiative (first-order absorbing)

The object stores bc_type(6) (integer codes) and q(9,6) (field values for Dirichlet BCs, one set per face).

---

adam_prism_ic_object — Initial Conditions

Configured from the [initial_conditions] INI section.

IC Types

type value	Description
vacuum	All fields set to zero
riemann-problem	Piecewise constant regions; each region in [initial_conditions_region_N]
plane_wave	Sinusoidal electromagnetic plane wave
rmf_field	Rotating Magnetic Field (reads from [external_fields])
uniform_field	Uniform background B and/or D
magnetic_nozzle	Magnetic nozzle configuration (parameters TBD)
rmf_magnetic_nozzle	Combined RMF + magnetic nozzle

INI Keys ([initial_conditions])

Key	Description
amr_iterations	Number of AMR iterations during IC application
type	IC type string (see table above)
regions_number	Number of regions (Riemann problem only)
kx, ky, kz	Wave number direction cosines (plane wave)
lambda	Wavelength (plane wave)
B0	Background magnetic field amplitude (plane wave)
B_x, B_y, B_z	Uniform background B components

Region sections (riemann-problem): [initial_conditions_region_N] with keys Dx, Dy, Dz, Bx, By, Bz, Jx, Jy, Jz, emin_x/y/z, emax_x/y/z.

---

adam_prism_io_object — I/O Handler

Configured from the [IO] INI section.

INI Keys ([IO])

Key	Default	Description
output_basename	—	Prefix for all output files
it_save	100	Field output save frequency (iterations)
restart	.false.	Enable restart from existing files
restart_basename	—	Prefix for restart files
restart_save	100	Restart save frequency
residuals_save	10	Residuals log save frequency
save_memory_status	.false.	Log memory usage during allocations
save_residual_fields	.false.	Include dq in field output
save_curl_fields	.false.	Include curl field in output
save_divergence_fields	.false.	Include divergence field in output
save_gradient_fields	.false.	Include gradient field in output
save_laplacian_fields	.false.	Include Laplacian field in output

Output Files

File	Columns	Trigger
<basename>-energy_error.dat	it blocks time error_D error_B rms_D rms_B	Every energy_error_save iterations
<basename>-residuals.dat	it time blocks rq1 … rqN	Every residuals_save iterations

---

adam_prism_time_object — Time Handler

Configured from the [time] INI section.

INI Keys ([time])

Key	Default	Description
it_max	-1	Maximum iteration count (−1 = terminate by time)
time_max	1.0	Maximum physical time
CFL	0.3	CFL number for time-step control

Termination Logic (is_to_save)

Condition	Meaning
mod(it, it_save) == 0	Periodic save
it == it_max	Final iteration (when it_max > 0)
it_max ≤ 0 .and. time ≥ time_max	Physical time limit reached
it_max > 0 .and. it ≥ it_max	Iteration limit reached

print_progress prints it, time-step dt, current time, and a percentage completion based on whichever termination criterion is active.

---

adam_prism_coil_object — Coil Sources

Models current-carrying coils as volumetric current density J sources.

INI Section ([coils_input])

Key	Values	Description
rectangular_coils_number	integer	Number of rectangular coils
circular_coils_number	integer	Number of circular coils

Per-coil configuration (section [coil_N]):

Key	Values	Description
coil_type	rectangular / circular	Loop geometry
current_type	DC_current / AC_current	Current waveform
normal	+x / -x / +y / -y / +z / -z	Coil plane normal
A	real	Current amplitude (A)
f	real	Frequency (Hz) — AC only
phase	real	Initial phase (rad) — AC only
x_center, y_center, z_center	real	Coil centre coordinates
lx, ly	real	Rectangular loop dimensions
r_coil	real	Circular loop radius
sigma	real	Gaussian current distribution width

Internal Arrays

Array	Shape	Description
J_vec	(total_coils_number, 4, ni, nj, nk, nb)	Per-coil current density vector and Gaussian weight
coil_flag	(ni, nj, nk, nb)	Integer flag: which coil passes through each cell

J_vec(:,1:3,…) holds current density direction components; J_vec(:,4,…) holds the Gaussian distribution value f_Gauss.

---

adam_prism_fWLayer_object — Far-Wave Absorbing Layer

Implements a sponge-like damping layer at domain boundaries to absorb outgoing electromagnetic waves, following Barbas notation.

INI Section ([fWLayer])

Key	Description
C	Layer width in cells (0 = layer disabled)
layer_xm, layer_xp, …	Boolean flags for each of the six sides (−x, +x, −y, +y, −z, +z)

Damping Function

The damping factor is not stored: compute_fwl_factor evaluates it per cell, on demand, from the cell's distance to the boundary and the face's profile geometry. It is declared !$acc routine seq / !$omp declare target, so one function serves both the CPU and the device backend.

For a face spanning profile_cells cells over a physical profile_extent, a cell at center_distance from the boundary-cell center gets the profile-shape coefficient

$$f_i=\{\begin{array}{ll}\frac{1}{150}(-7C_r^2+255C_r+250)& C_r<40\\ 25& C_r\ge 40\end{array}$$

with C_r = profile_cells, and from it the damping factor

$$f=\frac{1}{f_i}{\log }_{10}[r({10}^{f_i}-1)+1],r=\mathrm{clamp}(\frac{\mathtt{\text{center\_distance}}}{\mathtt{\text{profile\_extent}}},0,1)$$

A face with profile_cells <= 0 or profile_extent <= 0 returns f = 1 (no damping). profile_cells and profile_extent are derived per face at initialization from the requested physical width (see compute_absorbing_face_range); on a uniform mesh profile_extent equals C*ds, which is what the former cell-count C input expressed directly.

7e39318b removed the f array from prism_fWLayer_object together with its HDF5 save block, so fWL_x, fWL_y and fWL_z are no longer emitted as output fields.

---

adam_prism_external_fields_object — External Fields

Prescribed electromagnetic fields added to (and subtracted from) the state vector around each time step.

INI Section ([external_fields])

Key	Values	Description
ef_type	RMF / magnetic_nozzle / RMF_and_magnetic_nozzle / none	External field type
RMF_frequency	real	Rotating magnetic field frequency (Hz)
RMF_B_amplitude	real	Rotating magnetic field amplitude (T)
RMF_rotation_axis	X / Y / Z	Rotation axis

At initialization, the add_external_fields and sub_external_fields procedure pointers are dispatched to the corresponding implementation (add_external_fields_rmf, etc.) based on ef_type.

---

adam_prism_rk_bc_object — RK-BC Integrator

Stores the per-stage boundary condition state during Runge-Kutta sub-stepping.

Data Members

Member	Description
q_bc_rk(:,:,:,:,:,:)	Per-stage BC field values
dq_bc_rk(:,:,:,:,:)	Per-stage BC residuals
ark, brk, crk	Low-storage RK coefficients
alph, beta, gamm	SSP RK coefficients
ssa, ssb	Symplectic-splitting RK coefficients
nrk	Number of RK stages

Methods

Method	Description
initialize	Mirror the parent rk_object coefficient set
initialize_stages	Zero the stage arrays
assign_stage	Copy current q to the stage buffer
compute_stage	Advance the BC state for one RK sub-step

---

adam_prism_pic_object — Particle-in-Cell

Manages macro-particle configuration, grid-particle mapping, and weighting.

INI Section ([PIC])

Key	Values	Description
problem_type	plasma / single_particle	Simulation scenario
plasma_density	real	Initial plasma number density (m⁻³) — plasma only
neutral_fraction	real	Fraction of neutral particles — plasma only
particle_weighting_model	CIC / NGP / TSC	Charge deposition scheme
current_weighting_model	CIC / NGP / TSC	Current deposition scheme
field_weighting_model	0D / 1D	Field interpolation order at particle positions
scheme_time	LEAPFROG / RUNGE_KUTTA	Particle time integrator

Particle Count

For plasma: particle_number = plasma_density × domain_volume. The population is split into ions, electrons, and neutrals according to neutral_fraction. For single_particle: particle_number = 1.

Weighting Models

Scheme	Method	Description
NGP	Nearest Grid Point	Particle charge/current deposited to the containing cell only
CIC	Cloud-in-Cell	Trilinear interpolation to 3×3×3 neighbourhood
TSC	Triangular Shaped Cloud	Quadratic weight function over 3×3×3 neighbourhood

Field interpolation models:

Model	Description
0D	Zero-order: field value at the nearest grid point
1D	First-order: piecewise-linear field interpolation at particle position

At initialization, the three procedure pointers particle_weighting, current_weighting, and field_weighting are dispatched to the selected method implementations.

neighbour_list(4, particle_number)

Stores the current grid indices (block, i, j, k) for each particle, updated each time step by particle_cartesian_grid_index.

---

adam_prism_particle_injection_object — Particle Injection

Sets the initial spatial and velocity distributions for PIC particles.

INI Section ([particle_injection])

Key	Values	Description
space_distribution	Uniform_domain_space_distribution / Uniform_boxes_space_distribution / Uniform_cell_space_distribution / Space_random_number_generator / Space_layered_number_generator	Initial spatial distribution
velocity_distribution	Uniform_Maxwellian / Non_uniform_Maxwellian / Velocity_random_number_generator / Velocity_layered_number_generator	Initial velocity distribution
T_i, T_e, T_n	real	Ion/electron/neutral temperature (uniform Maxwellian)
T_i_x/y/z, T_e_x/y/z, T_n_x/y/z	real	Per-axis temperatures (non-uniform Maxwellian)
v_drift_x/y/z	real	Drift velocity components
box_number	real	Number of boxes for charge-neutral injection
space_pairing	logical	Pair ions and electrons in space
velocity_pairing	logical	Pair ion/electron velocities
v_av_correction	logical	Correct mean velocity after sampling
x_position, y_position, z_position	real	Initial position (single particle)
charge, mass	real	Charge and mass (single particle)

At initialization, the procedure pointers particle_space_injection and particle_velocity_injection are dispatched to the selected distribution subroutines.

---

adam_prism_leapfrog_pic_object — Leapfrog PIC Integrator

Implements the Boris leapfrog particle pusher, integrating particle equations of motion staggered in time relative to the field update. The Boris method is energy-conserving and second-order accurate for the Lorentz force equation.

---

adam_prism_rk_pic_object — Runge-Kutta PIC Integrator

Implements explicit Runge-Kutta time integration for PIC particle trajectories, sharing RK coefficients with the parent field integrator rk_object.

---

adam_prism_riemann_library — Riemann Solvers

Provides the Maxwell convective flux routines and Riemann solvers used in all spatial schemes.

Public Routines

Routine	Description
compute_riemann_maxwell_llf	LLF (Rusanov) solver
compute_riemann_maxwell_hll	HLL (Harten–Lax–van Leer) solver (internal)
compute_convective_fluxes_maxwell	Base flux for the 6-variable system
compute_convective_fluxes_maxwell_div_d	Flux extended with D-cleaning scalar φ
compute_convective_fluxes_maxwell_div_b	Flux extended with B-cleaning scalar ψ
compute_convective_fluxes_maxwell_div_d_b	Flux with both φ and ψ
compute_eigenvalues_vector	Maximum wave speed for time-step control

LLF Solver

$$F_{\text{LLF}}(q_L,q_R)=\frac{1}{2}[F(q_L)+F(q_R)-c_0(q_R-q_L)]$$

applied to the 6 conservative variables (D, B). The J source variables have zero flux and are not passed through the Riemann solver.

HLL Solver

$$F_{\text{HLL}}=\frac{{\lambda }^{+}{F}_L-{\lambda }^{-}{F}_R+{\lambda }^{+}{\lambda }^{-}(q_R-q_L)}{{\lambda }^{+}-{\lambda }^{-}},{\lambda }^{+}=c_0,{\lambda }^{-}=-c_0$$

compute_convective_fluxes_interface

Abstract interface for flux routines, allowing procedure pointer dispatch in backend implementations:

subroutine compute_convective_fluxes_interface(sir, q, f, chi)
  real(R8P), intent(in)    :: sir(3)  ! directional unit increment (x, y, or z)
  real(R8P), intent(in)    :: q(1:)   ! state vector
  real(R8P), intent(inout) :: f(1:)   ! computed fluxes
  real(R8P), intent(in)    :: chi     ! divergence-cleaning speed coefficient
end subroutine

---

adam_prism_common_library — Barrel Re-export

A single convenience module that re-exports all 17 common modules. Both backends use only this one use statement:

use adam_prism_common_library

Modules re-exported: adam_prism_parameters, adam_prism_physics_object, adam_prism_numerics_object, adam_prism_common_object, adam_prism_bc_object, adam_prism_ic_object, adam_prism_io_object, adam_prism_time_object, adam_prism_coil_object, adam_prism_fWLayer_object, adam_prism_external_fields_object, adam_prism_rk_bc_object, adam_prism_pic_object, adam_prism_particle_injection_object, adam_prism_leapfrog_pic_object, adam_prism_rk_pic_object, adam_prism_riemann_library.

---

License

PRISM is part of the ADAM framework, released under the GNU Lesser General Public License v3.0 (LGPLv3).

Copyright (C) Andrea Di Mascio, Federico Negro, Giacomo Rossi, Francesco Salvadore, Stefano Zaghi.