tud_lbm.config.simulation_config

Validated, serialisable simulation configuration for TUD-LBM.

SimulationConfig is a frozen Python dataclass used for parsing, validation, and serialisation. It never enters a JIT boundary.

Usage:

from config.simulation_config import SimulationConfig

cfg = SimulationConfig(
    grid_shape=(128, 128, 1),
    tau=0.8,
    nt=5000,
    collision_scheme="bgk",
)

Attributes

CONFIG_SECTION
ARRAY_ELIGIBLE
NESTED_SWEEPABLE
MIN_GRID_DIMENSIONS
MIN_TAU_VALUE

Classes

SimulationConfig

Validated, serialisable simulation configuration for TUD-LBM.

Functions

array_field(→ dataclasses.field)	Field factory for array-eligible SimulationConfig fields.
get_array_eligible_fields(→ frozenset[str])	Get names of fields eligible for array expansion in configuration sweeps.
get_nested_sweepable_fields(→ frozenset[str])	Return field names whose dict sub-keys may carry list sweep values.
get_fields_for_section(→ frozenset[str])	Get field names belonging to a specific configuration section.

Module Contents

tud_lbm.config.simulation_config.CONFIG_SECTION: str = 'config_section'

tud_lbm.config.simulation_config.ARRAY_ELIGIBLE: str = 'array_eligible'

tud_lbm.config.simulation_config.NESTED_SWEEPABLE: str = 'nested_sweepable'

tud_lbm.config.simulation_config.MIN_GRID_DIMENSIONS: int = 2

tud_lbm.config.simulation_config.MIN_TAU_VALUE: float = 0.5

tud_lbm.config.simulation_config.array_field(*, default: object = dataclasses.MISSING, default_factory: object = dataclasses.MISSING, section: str | None = None, nested_sweepable: bool = False, **kwargs: object) → dataclasses.field

Field factory for array-eligible SimulationConfig fields.

Parameters:

• default – Default value for the field.

• default_factory – Default factory function for the field.

• section – Config section name for serialisation routing.

• nested_sweepable – If True, sub-keys inside this dict field will also be inspected for list values during Cartesian-product expansion (e.g. gravity_force, wetting_config).

• **kwargs – Additional keyword arguments passed to the field factory.

Returns:

A dataclass field with array-eligible metadata.

class tud_lbm.config.simulation_config.SimulationConfig

Validated, serialisable simulation configuration for TUD-LBM.

This frozen dataclass is used for parsing, validation, and serialisation. It never enters a JIT boundary and serves as the configuration container for all simulation parameters including grid, collision, boundary conditions, and output settings.

sim_type: Literal['single_phase', 'multiphase', 'multiphase_wetting', 'multiphase_hysteresis', 'multiphase_hysteresis_chemical_step'] = 'single_phase'

simulation_name: str | None = None

lattice_type: str = 'D2Q9'

grid_shape: tuple[int, Ellipsis]

nt: int

tau: float

collision_scheme: str

k_diag: tuple[float, Ellipsis] | None

bc_config: dict[str, Any] | None = None

wetting_config: dict[str, Any] | None

hysteresis_config: dict[str, Any] | None

chemical_step_config: dict[str, Any] | None

gravity_force: dict[str, Any] | None

electric_force: dict[str, Any] | None

gravity_masked_force: dict[str, Any] | None

init_type: str = 'standard'

init_dir: str | None = None

initialisation: dict[str, Any]

results_dir: str = ''

save_interval: int = 0

skip_interval: int = 0

save_fields: list[str] | None = None

plot_fields: list[str] | None = None

animate_fields: list[str] | None = None

output_format: str | list[str] | None = 'numpy'

output_dir: str | None = None

eos: str | None

kappa: float | None

rho_l: float | None

rho_v: float | None

interface_width: int | None

g: float | None

extra: dict[str, Any]

__post_init__() → None

Validate and normalize configuration after initialization.

property is_single_phase: bool

Check if simulation is single-phase.

property is_multiphase: bool

Check if simulation is multiphase.

property force_enabled: bool

Check if any force field is populated.

to_dict() → dict[str, Any]

Convert configuration to dictionary format.

__repr__() → str

Return string representation of SimulationConfig.

tud_lbm.config.simulation_config.get_array_eligible_fields() → frozenset[str]

Get names of fields eligible for array expansion in configuration sweeps.

tud_lbm.config.simulation_config.get_nested_sweepable_fields() → frozenset[str]

Return field names whose dict sub-keys may carry list sweep values.

tud_lbm.config.simulation_config.get_fields_for_section(section: str) → frozenset[str]

Get field names belonging to a specific configuration section.