Useful for benchmarking and other code to code comparisons.
The data is pulled right from the C api but now the uncertainties
are preserved and written to the depletion results file
The predictor saves BOS number densities, eigenvalues, and
other results before the depletion event, unlike all
other schemes, which have at least one depletion event between
the BOS transport solution and the depletion solve.
This causes the first value in the resulting depletion process
time vector to be zero, and the last value to be non-zero,
opposite to other schemes.
openmc.deplete.tally_helers.ChainFissTallyHelper is responsible
for populating the fission_q vector, building the reaction rate
tallies, and updating the energy produced by fission per material.
This abstraction will be helpful as more methods for pulling
fission q values are introduced, namely as issue #1238, indirect
energy release, gets resolved
Designed to facilitate building universe that represent
pins, or concentric cylinders of materials. Can be built
by passing surfaces and materials, or radii and materials.
Places last material spanning region out to infinity from
the last surface.
Supports subdividing rings into equal volume slices of
unique [cloned] materials, useful for depletion.
Unit tests added for failure modes as well as basic operation
A dictionary of constant fission q values [eV] can be passed
into the initialization of deplete.Chain, deplete.Operator,
and model.Model objects. This can be used for direct
comparisons with other codes with readily available
constant Q-values.
Tests have been added on the Chain and Operator side.
The additional fiss_q_values passed into Model is forwarded
directly on to the Operator call.
Due to where the results are saved for the predictor routine,
at the top of the for loop vs. the end, the positions
are incremented by one in the depletion time. This causes
a value of zero to be returned as the first item in the array
from get_depletion_time.
This save was moved in commit 87aed7a0b0
of PR #1014 to make the restart feature easier.
Needed because ./tests/regression_tests/deplete/test_reference.h5
does not contain this dataset. Now, a check is performed
to see if "depletion file" is present in the root of the hdf5 file.
Otherwise, the proc_time is set to the array of nans
The index - 1 was causing values to be written in a correctly sized
matrix, but in an offset pattern, e.g. first value written
to last position, second value to first position, etc.
The ResultsList.get_depletion_time method will already
trim this array in the retrieval process.
