crp/OpenMC
1
0
Fork 0
mirror of https://github.com/openmc-dev/openmc.git synced 2026-07-28 22:26:08 -04:00
Code Issues Projects Releases Packages Wiki Activity Actions 5

address comments by @drewejohnson

Browse source
This commit is contained in:
church89 2024-03-14 17:30:58 +01:00
parent f3dfdf8a45
commit 9d6c5be32d
2 changed files with 96 additions and 74 deletions
Download patch file Download diff file Expand all files Collapse all files

2
openmc/deplete/abc.py
View file

@ -691,7 +691,7 @@ class Integrator(ABC):
self._solver = func
@property
def get_batchwise(self):
def batchwise(self):
return self._batchwise
def _timed_deplete(self, n, rates, dt, matrix_func=None):

168
openmc/deplete/batchwise.py
View file

@ -3,6 +3,7 @@ from copy import deepcopy
from numbers import Real
from warnings import warn
import numpy as np
from math import isclose
import openmc.lib
from openmc.mpi import comm
@ -40,7 +41,7 @@ class Batchwise(ABC):
operator : openmc.deplete.Operator
OpenMC operator object
bracket : list of float
Bracketing interval to search for the solution, always relative to the
Initial bracketing interval to search for the solution, relative to the
solution at previous step.
bracket_limit : list of float
Absolute bracketing interval lower and upper; if during the adaptive
@ -68,6 +69,7 @@ class Batchwise(ABC):
search_for_keff_output : Bool, Optional
Print full transport logs during iterations (e.g., inactive generations, active
generations). Default to False
Attributes
----------
burn_mats : list of str
@ -75,6 +77,7 @@ class Batchwise(ABC):
local_mats : list of str
All burnable material IDs being managed by a single process
List of burnable materials ids
"""
def __init__(
@ -165,15 +168,34 @@ class Batchwise(ABC):
OpenMC parametric model
"""
@abstractmethod
def search_for_keff(self, x, step_index):
"""Perform the criticality search on parametric material variable.
The :meth:`openmc.search.search_for_keff` solution is then used to
calculate the new material volume and update the atoms concentrations.
Parameters
----------
x : list of numpy.ndarray
Total atoms concentrations
Returns
-------
x : list of numpy.ndarray
Updated total atoms concentrations
root : float
Search_for_keff returned root value
"""
def _search_for_keff(self, val):
"""Perform the criticality search for a given parametric model.
If the solution lies off the initial bracket, this method iteratively
adapt it until :meth:`openmc.search.search_for_keff` return a valid
solution.
The adapting bracket algorithm takes the ratio between the bracket and
the returned keffs values to scale the bracket until a valid solution
is found.
It calculates the ratio between guessed and corresponding keffs values
as the proportional term to move the bracket towards the target.
A bracket limit pose the upper and lower boundaries to the adapting
bracket. If one limit is hit, the algorithm will stop and the closest
limit value will be used.
@ -190,7 +212,7 @@ class Batchwise(ABC):
targeted value
"""
# make sure we don't modify original bracket
bracket = deepcopy(self.bracket)
bracket = deepcopy(self.brack½et)
# Run until a search_for_keff root is found or out of limits
root = None
@ -240,7 +262,7 @@ class Batchwise(ABC):
)
# if the difference between keffs is smaller than 1 pcm,
# the grad calculation will be overshoot, so let's set the root
# the slope calculation will be overshoot, so let's set the root
# to the closest guess value
if abs(np.diff(keffs)) < 1.0e-5:
arg_min = abs(self.target - np.array(keffs)).argmin()
@ -252,9 +274,10 @@ class Batchwise(ABC):
)
root = guesses[arg_min]
# Calculate gradient as ratio of delta bracket and delta keffs
grad = abs(np.diff(bracket) / np.diff(keffs))[0].n
# Move the bracket closer to presumed keff root.
# Calculate slope as ratio of delta bracket and delta keffs
slope = abs(np.diff(bracket) / np.diff(keffs))[0].n
# Move the bracket closer to presumed keff root by using the
# slope
# Two cases: both keffs are below or above target
if np.mean(keffs) < self.target:
@ -265,7 +288,7 @@ class Batchwise(ABC):
dir = -1
bracket[np.argmin(keffs)] = bracket[np.argmax(keffs)]
bracket[np.argmax(keffs)] += (
grad * (self.target - max(keffs).n) * dir
slope * (self.target - max(keffs).n) * dir
)
else:
if guesses[np.argmax(keffs)] > guesses[np.argmin(keffs)]:
@ -274,14 +297,14 @@ class Batchwise(ABC):
dir = 1
bracket[np.argmax(keffs)] = bracket[np.argmin(keffs)]
bracket[np.argmin(keffs)] += (
grad * (min(keffs).n - self.target) * dir
slope * (min(keffs).n - self.target) * dir
)
#check if adapted bracket are within bracketing limits
if max(bracket) + val > self.bracket_limit[1]:
bracket[np.argmax(bracket)] = self.bracket_limit[1] - val
if min(bracket) + val < self.bracket_limit[0]:
bracket[np.argmin(bracket)] = self.bracket_limit[0] - val
if bracket[1] + val > self.bracket_limit[1]:
bracket[1] = self.bracket_limit[1] - val
if bracket[0] + val < self.bracket_limit[0]:
bracket[0] = self.bracket_limit[0] - val
else:
# Set res with closest limit and continue
@ -413,7 +436,7 @@ class BatchwiseCell(Batchwise):
Specific classes for running batch wise depletion calculations are
implemented as derived class of BatchwiseCell.
.. versionadded:: 0.13.4
.. versionadded:: 0.14.1
Parameters
----------
@ -422,7 +445,7 @@ class BatchwiseCell(Batchwise):
operator : openmc.deplete.Operator
OpenMC operator object
bracket : list of float
Bracketing interval to search for the solution, always relative to the
Initial bracketing interval to search for the solution, relative to the
solution at previous step.
bracket_limit : list of float
Absolute bracketing interval lower and upper; if during the adaptive
@ -455,9 +478,10 @@ class BatchwiseCell(Batchwise):
----------
cell : openmc.Cell or int or str
OpenMC Cell identifier to where apply batch wise scheme
cell_materials : list of openmc.Material
Depletable materials that fill the Cell Universe. Only valid for
translation or rotation attributes
universe_cells : list of openmc.Cell
Cells that fill the openmc.Universe that fills the main cell
to where apply batch wise scheme, if cell materials are set as
depletable.
lib_cell : openmc.lib.Cell
Corresponding openmc.lib.Cell once openmc.lib is initialized
axis : int {0,1,2}
@ -503,8 +527,9 @@ class BatchwiseCell(Batchwise):
check_value("axis", axis, (0, 1, 2))
self.axis = axis
# list of material fill the attribute cell, if depletables
self.cell_materials = None
# Initialize container of universe cells, to populate only if materials
# are set as depletables
self.universe_cells = None
@abstractmethod
def _get_cell_attrib(self):
@ -544,26 +569,37 @@ class BatchwiseCell(Batchwise):
val : openmc.Cell
Openmc Cell
"""
all_cells = [
cell_bundles = [
(cell.id, cell.name, cell)
for cell in self.geometry.get_all_cells().values()
]
if isinstance(val, Cell):
check_value("Cell exists", val, [cell[2] for cell in all_cells])
check_value("Cell exists", val,
[cell_bundle[2] for cell_bundle in cell_bundles])
elif isinstance(val, str):
if val.isnumeric():
check_value("Cell id exists", int(val), [cell[0] for cell in all_cells])
val = [cell[2] for cell in all_cells if cell[0] == int(val)][0]
check_value("Cell id exists", int(val),
[cell_bundle[0] for cell_bundle in cell_bundles])
val = [cell_bundle[2] for cell_bundle in cell_bundles \
if cell_bundle[0] == int(val)][0]
else:
check_value("Cell name exists", val, [cell[1] for cell in all_cells])
val = [cell[2] for cell in all_cells if cell[1] == val][0]
check_value("Cell name exists", val,
[cell_bundle[1] for cell_bundle in cell_bundles])
val = [cell_bundle[2] for cell_bundle in cell_bundles \
if cell_bundle[1] == val][0]
elif isinstance(val, int):
check_value("Cell id exists", val, [cell[0] for cell in all_cells])
val = [cell[2] for cell in all_cells if cell[0] == val][0]
check_value("Cell id exists", val,
[cell_bundle[0] for cell_bundle in cell_bundles])
val = [cell_bundle[2] for cell_bundle in cell_bundles \
if cell_bundle[0] == val][0]
else:
ValueError(f"Cell: {val} is not supported")
@ -572,6 +608,7 @@ class BatchwiseCell(Batchwise):
def _model_builder(self, param):
"""Builds the parametric model to be passed to the
:meth:`openmc.search.search_for_keff` method.
Callable function which builds a model according to a passed
parameter. This function must return an openmc.model.Model object.
@ -586,16 +623,13 @@ class BatchwiseCell(Batchwise):
Openmc parametric model
"""
self._set_cell_attrib(param)
# At this stage it not important to reassign the new volume, as the
# nuclide densities remain constant. However, if at least one of the cell
# materials is set as depletable, the volume change needs to be accounted for
# as an increase or reduction of number of atoms, i.e. vector x, before
# solving the depletion equations
return self.model
def search_for_keff(self, x, step_index):
"""Perform the criticality search on the parametric cell coefficient and
update materials accordingly.
The :meth:`openmc.search.search_for_keff` solution is then set as the
new cell attribute.
@ -633,7 +667,7 @@ class BatchwiseCell(Batchwise):
# if at least one of the cell materials is depletable, calculate new
# volume and update x and number accordingly
# new volume
if self.cell_materials:
if self.universe_cells:
volumes = self._calculate_volumes()
x = self._update_x_and_set_volumes(x, volumes)
@ -648,7 +682,7 @@ class BatchwiseCellGeometrical(BatchwiseCell):
:meth:`openmc.deplete.Integrator.add_batchwise` method from an integrator
class, such as :class:`openmc.deplete.CECMIntegrator`.
.. versionadded:: 0.13.4
.. versionadded:: 0.14.1
Parameters
----------
@ -662,7 +696,7 @@ class BatchwiseCellGeometrical(BatchwiseCell):
Directional axis for geometrical parametrization, where 0, 1 and 2 stand
for 'x', 'y' and 'z', respectively.
bracket : list of float
Bracketing interval to search for the solution, always relative to the
Initial bracketing interval to search for the solution, relative to the
solution at previous step.
bracket_limit : list of float
Absolute bracketing interval lower and upper; if during the adaptive
@ -742,20 +776,22 @@ class BatchwiseCellGeometrical(BatchwiseCell):
check_value("attrib_name", attrib_name, ("rotation", "translation"))
self.attrib_name = attrib_name
# check if cell is filled with 2 cells
check_length("fill materials", self.cell.fill.cells, 2)
# check if cell is filled with a universe containing 2 cells
check_type("fill universe", self.cell.fill, openmc.Universe)
# check if universe contains 2 cells
check_length("universe cells", self.cell.fill.cells, 2)
# Initialize vector
self.vector = np.zeros(3)
self.cell_materials = [
cell.fill for cell in self.cell.fill.cells.values() if cell.fill.depletable
self.universe_cells = [
cell for cell in self.cell.fill.cells.values() if cell.fill.depletable
]
check_type("samples", samples, int)
self.samples = samples
if self.cell_materials:
if self.universe_cells:
self._initialize_volume_calc()
def _get_cell_attrib(self):
@ -788,9 +824,10 @@ class BatchwiseCellGeometrical(BatchwiseCell):
def _initialize_volume_calc(self):
"""Set volume calculation model settings of depletable materials filling
the parametric Cell.
"""
ll, ur = self.geometry.bounding_box
mat_vol = openmc.VolumeCalculation(self.cell_materials, self.samples, ll, ur)
mat_vol = openmc.VolumeCalculation(self.universe_cells, self.samples, ll, ur)
self.model.settings.volume_calculations = mat_vol
def _calculate_volumes(self):
@ -806,9 +843,9 @@ class BatchwiseCellGeometrical(BatchwiseCell):
volumes = {}
if comm.rank == 0:
res = openmc.VolumeCalculation.from_hdf5("volume_1.h5")
for mat_idx, mat in enumerate(self.local_mats):
if int(mat) in [mat.id for mat in self.cell_materials]:
volumes[mat] = res.volumes[int(mat)].n
for cell in self.universe_cells:
mat_id = cell.fill.id
volumes[mat_id] = res.volumes[mat_id].n
volumes = comm.bcast(volumes)
return volumes
@ -821,7 +858,7 @@ class BatchwiseCellTemperature(BatchwiseCell):
:meth:`openmc.deplete.Integrator.add_batchwise` method from an integrator
class, such as :class:`openmc.deplete.CECMIntegrator`.
.. versionadded:: 0.13.4
.. versionadded:: 0.14.1
Parameters
----------
@ -832,7 +869,7 @@ class BatchwiseCellTemperature(BatchwiseCell):
attrib_name : str
Cell attribute type
bracket : list of float
Bracketing interval to search for the solution, always relative to the
Initial bracketing interval to search for the solution, relative to the
solution at previous step.
bracket_limit : list of float
Absolute bracketing interval lower and upper; if during the adaptive
@ -941,7 +978,7 @@ class BatchwiseMaterial(Batchwise):
Specific classes for running batch wise depletion calculations are
implemented as derived class of BatchwiseMaterial.
.. versionadded:: 0.13.4
.. versionadded:: 0.14.1
Parameters
----------
@ -953,7 +990,7 @@ class BatchwiseMaterial(Batchwise):
Dictionary of material composition to parameterize, where a pair key value
represents a nuclide and its weight fraction, respectively.
bracket : list of float
Bracketing interval to search for the solution, always relative to the
Initial bracketing interval to search for the solution, relative to the
solution at previous step.
bracket_limit : list of float
Absolute bracketing interval lower and upper; if during the adaptive
@ -1025,7 +1062,7 @@ class BatchwiseMaterial(Batchwise):
for nuc in mat_vector:
check_value("check nuclide exists", nuc, self.operator.nuclides_with_data)
if round(sum(mat_vector.values()), 2) != 1.0:
if not isclose(sum(mat_vector.values()), 1.0, abs_tol=0.01):
# Normalize material elements vector
sum_values = sum(mat_vector.values())
for elm in mat_vector:
@ -1070,24 +1107,6 @@ class BatchwiseMaterial(Batchwise):
return val
@abstractmethod
def _model_builder(self, param):
"""Builds the parametric model to be passed to the
:meth:`openmc.search.search_for_keff` method.
Callable function which builds a model according to a passed
parameter. This function must return an openmc.model.Model object.
Parameters
----------
param :
Model material function variable
Returns
-------
_model : openmc.model.Model
Openmc parametric model
"""
def search_for_keff(self, x, step_index):
"""Perform the criticality search on parametric material variable.
@ -1110,7 +1129,8 @@ class BatchwiseMaterial(Batchwise):
# even though they are also re-calculated when running the search_for_kef
self._update_materials(x)
# Solve search_for_keff and find new value
# Set initial material addition to 0 and let program calculate the
# right amount
root = self._search_for_keff(0)
# Update concentration vector and volumes with new value
@ -1128,7 +1148,7 @@ class BatchwiseMaterialRefuel(BatchwiseMaterial):
:meth:`openmc.deplete.Integrator.add_batchwise` method from an integrator
class, such as :class:`openmc.deplete.CECMIntegrator`.
.. versionadded:: 0.13.4
.. versionadded:: 0.14.1
Parameters
----------
@ -1142,7 +1162,7 @@ class BatchwiseMaterialRefuel(BatchwiseMaterial):
Dictionary of material composition to parameterize, where a pair key value
represents a nuclide and its weight fraction, respectively.
bracket : list of float
Bracketing interval to search for the solution, always relative to the
Initial bracketing interval to search for the solution, relative to the
solution at previous step.
bracket_limit : list of float
Absolute bracketing interval lower and upper; if during the adaptive
@ -1220,8 +1240,10 @@ class BatchwiseMaterialRefuel(BatchwiseMaterial):
def _model_builder(self, param):
"""Callable function which builds a model according to a passed
parameter. This function must return an openmc.model.Model object.
Builds the parametric model to be passed to the
:meth:`openmc.search.search_for_keff` method.
The parametrization can either be at constant volume or constant
density, according to user input.
Default is constant volume.