NEOPAX._boundary_conditions¶

Modular boundary condition (BC) abstraction for NEOPAX, inspired by torax.

Attributes¶

BC_REGISTRY

Classes¶

`BoundaryCondition`
`DirichletBC`
`NeumannBC`
`RobinBC`
`BoundaryConditionModel`	Left/right radial BC for 1D arrays with optional species-wise values.

Functions¶

`register_bc`(name, bc_class)
`get_bc`(name, args, *kwargs)
`build_boundary_condition_model`(bc_cfg, dr[, ...])
`_as_like_template`(value, template)	Broadcast or resize scalar/species-wise values to match `template` shape.
`_boundary_cell_spacing`(face_centers, *, side)
`right_constraints_from_bc_model`(bc_model, default_value)	Translate right BC model to `CellVariable` right-face constraints.
`left_constraints_from_bc_model`(bc_model, default_value)	Translate left BC model to `CellVariable` left-face constraints.
`apply_cell_centered_boundary_state`(profile, bc_model, ...)	Project cell-centered profiles so boundary cells satisfy the configured BCs.

Module Contents¶

class NEOPAX._boundary_conditions.BoundaryCondition¶

abstractmethod apply(flux, state=None, axis=True, edge=True, **kwargs)¶

class NEOPAX._boundary_conditions.DirichletBC(axis_value, edge_value)¶

Bases: BoundaryCondition

axis_value¶

edge_value¶

apply(flux, state=None, axis=True, edge=True, **kwargs)¶

class NEOPAX._boundary_conditions.NeumannBC(grad_axis, grad_edge, dr)¶

Bases: BoundaryCondition

grad_axis¶

grad_edge¶

dr¶

apply(flux, state=None, axis=True, edge=True, **kwargs)¶

class NEOPAX._boundary_conditions.RobinBC(alpha_axis, beta_axis, gamma_axis, alpha_edge, beta_edge, gamma_edge, dr)¶

Bases: BoundaryCondition

alpha_axis¶

beta_axis¶

gamma_axis¶

alpha_edge¶

beta_edge¶

gamma_edge¶

dr¶

apply(flux, state, axis=True, edge=True, **kwargs)¶

NEOPAX._boundary_conditions.BC_REGISTRY¶

NEOPAX._boundary_conditions.register_bc(name, bc_class)¶

NEOPAX._boundary_conditions.get_bc(name, *args, **kwargs)¶

class NEOPAX._boundary_conditions.BoundaryConditionModel¶

Left/right radial BC for 1D arrays with optional species-wise values.

dr: float¶

left_type: str = 'dirichlet'¶

right_type: str = 'dirichlet'¶

left_value: jax.numpy.ndarray | None = None¶

right_value: jax.numpy.ndarray | None = None¶

left_gradient: jax.numpy.ndarray | None = None¶

right_gradient: jax.numpy.ndarray | None = None¶

left_decay_length: jax.numpy.ndarray | None = None¶

right_decay_length: jax.numpy.ndarray | None = None¶

reference_profile: jax.numpy.ndarray | None = None¶

reference_profiles: jax.numpy.ndarray | None = None¶

static _as_jnp_or_none(value, species_names=None)¶

static _pick_for_row(value, row_index: int)¶

_infer_gradient(arr: jax.numpy.ndarray) → tuple[jax.numpy.ndarray, jax.numpy.ndarray]¶

_infer_log_gradient_coeff(boundary_value: jax.numpy.ndarray, boundary_grad: jax.numpy.ndarray) → jax.numpy.ndarray¶

_infer_decay_length(boundary_value: jax.numpy.ndarray, boundary_grad: jax.numpy.ndarray) → jax.numpy.ndarray¶

_apply_ghost_row(arr: jax.numpy.ndarray, row_index: int, reference_row: jax.numpy.ndarray | None) → jax.numpy.ndarray¶

apply_ghost(arr: jax.numpy.ndarray) → jax.numpy.ndarray¶

apply_ghost_all(arr2d: jax.numpy.ndarray) → jax.numpy.ndarray¶

NEOPAX._boundary_conditions.build_boundary_condition_model(bc_cfg: dict, dr: float, reference_profile=None, reference_profiles=None, species_names=None)¶

NEOPAX._boundary_conditions._as_like_template(value, template)¶: Broadcast or resize scalar/species-wise values to match template shape.

NEOPAX._boundary_conditions._boundary_cell_spacing(face_centers, *, side: str)¶

NEOPAX._boundary_conditions.right_constraints_from_bc_model(bc_model, default_value, profile=None, face_centers=None)¶

Translate right BC model to CellVariable right-face constraints.

Returns:: (right_face_constraint, right_face_grad_constraint) where one of the entries is typically None and the other is a JAX array.
Return type:: tuple

NEOPAX._boundary_conditions.left_constraints_from_bc_model(bc_model, default_value, profile=None, face_centers=None)¶: Translate left BC model to CellVariable left-face constraints.

NEOPAX._boundary_conditions.apply_cell_centered_boundary_state(profile, bc_model, face_centers)¶: Project cell-centered profiles so boundary cells satisfy the configured BCs.