# This code is part of KQCircuits
# Copyright (C) 2024 IQM Finland Oy
#
# This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public
# License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later
# version.
#
# This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied
# warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License along with this program. If not, see
# https://www.gnu.org/licenses/gpl-3.0.html.
#
# The software distribution should follow IQM trademark policy for open-source software
# (meetiqm.com/iqm-open-source-trademark-policy). IQM welcomes contributions to the code.
# Please see our contribution agreements for individuals (meetiqm.com/iqm-individual-contributor-license-agreement)
# and organizations (meetiqm.com/iqm-organization-contributor-license-agreement).
from dataclasses import dataclass, field
from typing import ClassVar
from kqcircuits.simulations.export.solution import Solution
[docs]
@dataclass(kw_only=True, frozen=True)
class ElmerSolution(Solution):
"""
A Base class for Elmer Solution parameters
Args:
percent_error: Stopping criterion in adaptive meshing.
max_error_scale: Maximum element error, relative to percent_error, allowed in individual elements.
max_outlier_fraction: Maximum fraction of outliers from the total number of elements
maximum_passes: Maximum number of adaptive meshing iterations.
minimum_passes: Minimum number of adaptive meshing iterations.
is_axisymmetric: Simulate with Axi Symmetric coordinates along :math:`y\\Big|_{x=0}` (Default: False)
mesh_levels: If set larger than 1 Elmer will make the mesh finer by dividing each element
into 2^(dim) elements mesh_levels times. Default 1.
mesh_size: Dictionary to determine the mesh refinement. The keys (string) denote the entities where to apply the
refinement and values (double) denote the maximal length of the mesh elements.
Optionally, the values can be set as a lists of doubles. Then, value[0] is the maximal mesh element length
inside at the entity and its expansion, value[1] is expansion distance in which the maximal mesh element
length is constant (default=value[0]), and value[2] is the slope of the increase in the maximal mesh element
length outside the entity.
The key can be a single layer name, or it can consist of multiple layer names separated with the & symbol,
meaning the entity will be an intersection of listed layers. Optionally, one can use a pattern for layer
names with the * symbol representing any string in a layer name. The ! symbol before the layer name or
pattern means that the complement of layer(s) is used instead.
The key 'global_max' is reserved for setting global maximal element length.
For example, if the dictionary is {'substrate*': 10, 'substrate*&vacuum': [2, 5], 'global_max': 100}, then
the maximal mesh element length is 10 inside the substrates and 2 on region which is less than 5 units away
from any substrate-vacuum interface. Outside these regions, the mesh element size can increase up to 100.
mesh_optimizer: Dictionary to determine mesh optimization, or None (default) to ignore optimization. The
dictionary can contain keywords 'method', 'force', 'niter' and 'dimTags'. See Gmsh manual
(gmsh.model.mesh.optimize) for details. The default value for 'method' is 'Netgen'.
vtu_output: Output vtu files to view fields in Paraview.
Turning this off will make the simulations slightly faster
save_elmer_data: Save the full Elmer model after simulation. This can be used to restart the simulation
or extract result field values as a post-processing step.
"""
tool: ClassVar[str] = ""
percent_error: float = 0.005
max_error_scale: float = 2.0
max_outlier_fraction: float = 1e-3
maximum_passes: int = 1
minimum_passes: int = 1
is_axisymmetric: bool = False
mesh_levels: int = 1
mesh_size: dict = field(default_factory=dict)
mesh_optimizer: dict | None = None
vtu_output: bool = True
save_elmer_data: bool = False
[docs]
def get_solution_data(self):
"""Return the solution data in dictionary form."""
sol_dict = {**self.__dict__, "tool": self.tool}
sol_dict["solution_name"] = sol_dict.pop("name")
return sol_dict
[docs]
@dataclass(kw_only=True, frozen=True)
class ElmerVectorHelmholtzSolution(ElmerSolution):
"""
Class for Elmer wave-equation solution parameters
Args:
frequency: Units are in GHz. Give a list of frequencies if using interpolating sweep.
frequency_batch: Number of frequencies calculated between each round of fitting in interpolating sweep
sweep_type: Type of frequency sweep. Options "explicit" and "interpolating".
max_delta_s: Convergence tolerance in interpolating sweep
london_penetration_depth: Allows supercurrent to flow on the metal boundaries within a layer
of thickness `london_penetration_depth`
quadratic_approximation: Use edge finite elements of second degree
second_kind_basis: Use Nedelec finite elements of second kind
use_av: Use a formulation of VectorHelmHoltz equation based on potentials A-V instead of electric field E.
For details see https://www.nic.funet.fi/pub/sci/physics/elmer/doc/ElmerModelsManual.pdf
WARNING: This option is experimental and might lead to poor convergence.
conductivity: Adds a specified film conductivity on metal boundaries. Applies only when `use_av=True`
nested_iteration: Enables alternative nested iterative solver to be used. Applies only when `use_av=True`
convergence_tolerance: Convergence tolerance of the iterative solver. Applies only when `use_av=True`
max_iterations: Maximum number of iterations for the iterative solver.
Applies only when `use_av=True` and only to the main solver (not to calc fields or port solver)
"""
tool: ClassVar[str] = "wave_equation"
frequency: float | list[float] = 5
frequency_batch: int = 3
sweep_type: str = "explicit"
max_delta_s: float = 0.01
london_penetration_depth: float = 0
quadratic_approximation: bool = False
second_kind_basis: bool = False
use_av: bool = False
conductivity: float = 0
nested_iteration: bool = False
convergence_tolerance: float = 1.0e-10
max_iterations: int = 2000
def __post_init__(self):
"""Cast frequency to list. Automatically called after init"""
if isinstance(self.frequency, (float, int)):
# hack to modify the attributes of frozen dataclass
object.__setattr__(self, "frequency", [float(self.frequency)])
elif not isinstance(self.frequency, list):
object.__setattr__(self, "frequency", list(self.frequency))
[docs]
@dataclass(kw_only=True, frozen=True)
class ElmerCapacitanceSolution(ElmerSolution):
"""
Class for Elmer capacitance solution parameters
Args:
p_element_order: polynomial order of p-elements
linear_system_method: Options: 1. Iterative methods "mg" (multigrid), "bicgstab" or any other iterative
solver mentioned in ElmerSolver manual section 4.3.1. 2. Direct methods "umfpack", "mumps", "pardiso" or
"superlu". Note that the use of other methods than "umfpack" requires Elmer to be explicitly compiled
with the corresponding solver software. If a direct method is used the parameters
"convergence_tolerance", "max_iterations" and "linear_system_preconditioning" are redundant
integrate_energies: Calculate energy integrals over each object. Used in EPR simulations
convergence_tolerance: Convergence tolerance of the iterative solver.
max_iterations: Maximum number of iterations for the iterative solver.
linear_system_preconditioning: Choice of preconditioner before using an iterative linear system solver
electric_infinity_bc: effectively extend the model domain to infinity using spherical boundary conditions
"""
tool: ClassVar[str] = "capacitance"
p_element_order: int = 3
linear_system_method: str = "mg"
integrate_energies: bool = False
convergence_tolerance: float = 1.0e-9
max_iterations: int = 500
linear_system_preconditioning: str = "ILU0"
electric_infinity_bc: bool = False
[docs]
@dataclass(kw_only=True, frozen=True)
class ElmerCrossSectionSolution(ElmerSolution):
"""
Class for Elmer cross-section solution parameters.
By default both 2D Capacitance and 2D Inductance simulation will be run when using this
Args:
p_element_order: polynomial order of p-elements
linear_system_method: Options: 1. Iterative methods "mg" (multigrid), "bicgstab" or any other iterative
solver mentioned in ElmerSolver manual section 4.3.1. 2. Direct methods "umfpack", "mumps", "pardiso" or
"superlu". Note that the use of other methods than "umfpack" requires Elmer to be explicitly compiled
with the corresponding solver software. If a direct method is used the parameters
"convergence_tolerance", "max_iterations" and "linear_system_preconditioning" are redundant
integrate_energies: Calculate energy integrals over each object. Used in EPR simulations
boundary_conditions: Parameters to determine boundary conditions for potential on the edges
of simulation box. Supported keys are `xmin` , `xmax` ,`ymin` and `ymax`
Example: `boundary_conditions = {"xmin": {"potential": 0}}`
convergence_tolerance: Convergence tolerance of the iterative solver.
Applies only to capacitance part of the simulation
max_iterations: Maximum number of iterations for the iterative solver.
Applies only to capacitance part of the simulation
run_inductance_sim: Can be used to skip running the inductance simulation and just do 2D capacitance.
No impendance can then be calculated but useful for making EPR simulations faster
linear_system_preconditioning: Choice of preconditioner before using an iterative linear system solver
voltage_excitations: Can be used to excite signals with arbitrary voltages, instead of 1V. If this parameter is
used, no capacitances will be computed.
electric_infinity_bc: effectively extend the model domain to infinity using spherical boundary conditions
"""
tool: ClassVar[str] = "cross-section"
p_element_order: int = 3
linear_system_method: str = "mg"
integrate_energies: bool = False
boundary_conditions: dict = field(default_factory=dict)
convergence_tolerance: float = 1.0e-9
max_iterations: int = 500
run_inductance_sim: bool = True
linear_system_preconditioning: str = "ILU0"
voltage_excitations: list[float] | None = None
electric_infinity_bc: bool = False
[docs]
@dataclass(kw_only=True, frozen=True)
class ElmerEPR3DSolution(ElmerSolution):
"""
Class for Elmer 3D EPR simulations. Similar to electrostatics simulations done with ElmerCapacitanceSolution,
but supports separating energies by PartitionRegions. Always reports energies for each layer.
Args:
p_element_order: polynomial order of p-elements
linear_system_method: Options: 1. Iterative methods "mg" (multigrid), "bicgstab" or any other iterative
solver mentioned in ElmerSolver manual section 4.3.1. 2. Direct methods "umfpack", "mumps", "pardiso" or
"superlu". Note that the use of other methods than "umfpack" requires Elmer to be explicitly compiled
with the corresponding solver software. If a direct method is used the parameters
"convergence_tolerance", "max_iterations" and "linear_system_preconditioning" are redundant
convergence_tolerance: Convergence tolerance of the iterative solver.
max_iterations: Maximum number of iterations for the iterative solver.
linear_system_preconditioning: Choice of preconditioner before using an iterative linear system solver
voltage_excitations: Can be used to excite signals with arbitrary voltages, instead of 1V. If this parameter is
used, no capacitances will be computed.
electric_infinity_bc: effectively extend the model domain to infinity using spherical boundary conditions
"""
tool: ClassVar[str] = "epr_3d"
p_element_order: int = 3
linear_system_method: str = "mg"
convergence_tolerance: float = 1.0e-9
max_iterations: int = 1000
linear_system_preconditioning: str = "ILU0"
voltage_excitations: list[float] | None = None
electric_infinity_bc: bool = False
[docs]
def get_elmer_solution(tool="capacitance", **solution_params):
"""Returns an instance of ElmerSolution subclass.
Args:
tool: Determines the subclass of ElmerSolution.
solution_params: Arguments passed for ElmerSolution subclass.
"""
for c in [ElmerVectorHelmholtzSolution, ElmerCapacitanceSolution, ElmerCrossSectionSolution, ElmerEPR3DSolution]:
if tool == c.tool:
return c(**solution_params)
raise ValueError(f"No ElmerSolution found for tool={tool}.")