GaPFlow.Problem

GaPFlow.Problem#

class GaPFlow.Problem(options: dict, grid: dict, numerics: dict, prop: dict, geo: dict, gp: dict | None = None, database: Database | None = None, extra_field: ndarray[tuple[Any, ...], dtype[_ScalarT]] | None = None)#

Bases: object

Problem driver for GaPFlow simulations.

Sets up field collections, constitutive models (pressure, wall stress, bulk stress), optional Gaussian-process surrogate databases, time-stepping parameters, and I/O.

Notes

Calling the constructor __init__() directly expects properly formatted input dictionaries. It is recommended to use the from_yaml() or from_string() class methods, which automatically sanitize the simulation input.

Examples

>>> from GaPFlow import Problem
>>> myProblem = Problem.from_yaml('my_input_file.yaml')
__init__(options: dict, grid: dict, numerics: dict, prop: dict, geo: dict, gp: dict | None = None, database: Database | None = None, extra_field: ndarray[tuple[Any, ...], dtype[_ScalarT]] | None = None) → None#

Constructor.

Parameters:

  • options (dict) – general simulation options.

  • grid (dict) – Parameters controlling spatial discretization.

  • numerics (dict) – Time integration parameters.

  • prop (dict) – Material properties.

  • geo (dict) – Geometry settings.

  • gp (dict or None) – Parameters controlling the GP surrogate models.

  • database (GaPFlow.db.Database or None) – A database object, handling the GP training data with an attached MD runner.

  • extra_field (numpy.ndarray or None) – An additional field, whose entries can be used as GP features (besides the solution itself and the topography).

Methods

__init__(options, grid, numerics, prop, geo)

Constructor.

animate([save, seconds])

Create an animation of the solution time series.

from_string(ymlstring)

Create a Problem instance from a YAML string.

from_yaml(fname)

Create a Problem instance from a YAML file.

plot([ax])

Plot a snapshot of the solution and the current stress state.

plot_topo([show_defo, show_pressure])

Plot the gap topography, optionally in deformed state and with pressure profile.

run([keep_open])

Run the time-stepping loop until convergence, maximum iterations, or until a termination signal is received.

update()

Performs a single time step using the MacCormack [1] predictor corrector scheme.

write([scalars, fields])

Write scalars, fields and hyperparameters to disk as configured.

Attributes

cfl

Current CFL number.

converged

Return True if residuals in the buffer are below tolerance.

dt_crit

Critical timestep determined by grid spacing and sound speed.

kinetic_energy

Total kinetic energy (scalar).

mass

Total mass integrated over domain (scalar).

q

Full density field

q_has_nan

Check for NaNs in the solution field.

q_has_negative_density

Check for negative densities in the solution field.

q_is_valid

Validity flag for the solution field.

v_max

Maximum speed in the domain (scalar).
animate(save: bool = False, seconds: float = 10.0) → None#

Create an animation of the solution time series.

Checks if simulation has run already and if output has been generated. For 1D elastic simulations, height and deformation are included.

Parameters:

  • save (bool, optional) – Whether to save the animation as an .mp4 file, by default False.

  • seconds (float, optional) – Duration of the animation in seconds (if saved), by default 10.0

property cfl: floating#

Current CFL number.

property converged: bool#

Return True if residuals in the buffer are below tolerance.

property dt_crit: floating#

Critical timestep determined by grid spacing and sound speed.

classmethod from_string(ymlstring: str) → Self#

Create a Problem instance from a YAML string.

Parameters:

ymlstring (str) – YAML content as a string.

Returns:

Instantiated Problem object.

Return type:

Problem

classmethod from_yaml(fname: str) → Self#

Create a Problem instance from a YAML file.

Parameters:

fname (str) – Path to YAML configuration file.

Returns:

Instantiated Problem object.

Return type:

Problem

property kinetic_energy: floating#

Total kinetic energy (scalar).

property mass: floating#

Total mass integrated over domain (scalar).

plot(ax=None) → None#

Plot a snapshot of the solution and the current stress state.

Parameters:

ax (matplotlib.pyplot.axis, optional) – An axis to plot into, if None or wrong shape a new axis is created

plot_topo(show_defo=False, show_pressure=False) → None#

Plot the gap topography, optionally in deformed state and with pressure profile.

Parameters:

  • show_defo (bool) – Flag for showing deformation, default is False

  • show_pressure (bool) – Flag for showing pressure, default is False

property q: ndarray[tuple[Any, ...], dtype[floating]]#

Full density field

property q_has_nan: bool#

Check for NaNs in the solution field.

property q_has_negative_density: bool#

Check for negative densities in the solution field.

property q_is_valid: bool#

Validity flag for the solution field.

run(keep_open: bool = False) → None#

Run the time-stepping loop until convergence, maximum iterations, or until a termination signal is received.

Parameters:

keep_open (bool, optional) – If True, keeps files open after run for following runs to be written in the same files, by default False.

update() → None#

Performs a single time step using the MacCormack [1] predictor corrector scheme.

References

property v_max: floating#

Maximum speed in the domain (scalar).

write(scalars: bool = True, fields: bool = True) → None#

Write scalars, fields and hyperparameters to disk as configured.