Advanced workflows
Use these workflows after you have run and checked a single-material problem. They share the solver but add their own inputs, coverage requirements, and output formats.
Transport flux and mesh calculations
Import supported OpenMC statepoints, MCNP MESHTAL/MCTAL files, or FISPACT fluxes into actinv-flux-1 NDJSON. Supply a tally identity and source-rate normalization where required. Review energy ordering, units, cell volume, and rebinning before solving.
actinv mesh calculates independent cells and writes actinv-mesh-result-1 NDJSON. Cells do not exchange material or flux and have no transport or thermal feedback. Workload grouping can reuse identical spectra; output selection, a post-hoc memory guard, and checkpoint resume are described in the mesh specification.
Photon exports turn computed sources into transport inputs. Ordinary exports use a point at the origin; mesh exports require cell bounds. Check spatial interpretation in the receiving transport code.
export-source converts a banded actinv-r2s-source-1 document into the format a transport code or activation-adjacent R2S tool reads directly: OpenMC settings-XML, an MCNP SDEF deck, a Serpent src card block, or — for ALARA-based R2S tools such as nucleide and PyNE — one ALARA .photonSrc file per cell plus a provenance index (export-source alara, requires --shutdown-t-s). See the command reference, docs/INTERCHANGE_TRANSPORT.md for the pinned format contract, and contrib/nucleide_r2s/demo.py for a working round-trip through nucleide’s own ALARA reader.
Build libraries and response tables
build-library converts supported evaluated ENDF-6 data into an activation library and adjacent index. Projectile, group structure, temperature, product-state identity, and builder options determine compatibility.
build-covariance creates an MF=33 sidecar for a specific activation library. build-shielding creates a neutron self-shielding table. build-damage creates a damage-energy table. These inputs are requested separately in the problem; an activation-library build alone does not enable those calculations.
See the specification reference for commands and formats. Keep source data and attribution outside the repository.
Uncertainty and self-shielding
The uncertainty section selects response bands and coverage policy. MF=33 cross sections are the default channel; half-life and independent fission-yield uncertainties are optional. Results explains how to read the bands.
Since 1.4.0, channels: ["flux"] is available with groupwise spectrum.relative_error, or each mesh cell’s own supplied errors. This propagates tally statistics and can run without covariance when it is the only requested channel. It excludes systematic transport errors and is omitted from measurement-design rankings. See the uncertainty reference.
The self_shielding section selects a hash-pinned neutron table and composition-derived or fixed background dilution. It applies bounded Bondarenko factors over the resolved and unresolved ranges; it does not shield damage observables. Use the self-shielding reference for the supported table contract and completeness option.
Studies
An actinv-study-1 document expands cases over declared material, schedule, data, and calculation choices. study validate, build, and run separate document checking, deterministic case generation, and execution.
The study schema also describes refinement and seeded robustness sampling. Consult the study reference for scalar-response restrictions, sampling channels, and execution limits.
Reverse calculation
actinv reverse estimates a common flux multiplier or per-irradiation-step multipliers from measured activities. It requires a linear trace-regime problem and rejects unsupported or unidentifiable cases. It estimates normalization; it does not unfold the spectrum shape.
reverse-qualified adds nuclear-data covariance to the measurement model and reports posterior covariance and identifiability. Its forward problem must request the relevant banded activity responses. See the command reference and reverse example.
Design search and impurity budgets
optimize searches declared composition fractions, flux scale, and schedule durations with a seeded optimizer. Constraints can use propagated uncertainty edges. Banded candidates can take minutes each; use it as a batch workflow. Read the optimization schema and steel example.
decide evaluates declared response constraints and ranks measurement targets. budget estimates and verifies impurity limits against a supplied clearance calculation. These workflows inherit the selected response table’s applicability; they do not choose a jurisdiction or establish regulatory acceptance. See the budget schema.
For an explicitly selected U.S. Part 61 rule pack, waste reports nominal
component classes and waste budget verifies class-specific impurity limits.
waste bounds classifies user-declared absolute whole-component activity boxes
under the bundled pack and returns a conservative class superset. These boxes
are deterministic enclosures, not probabilities or propagated uncertainty.
waste composition projects caller-declared fixed-rate affine pure-element
responses over bounded natural-element weight percentages that sum to 100%; it
does not generate responses or establish physical linearity. Separate nuclide
projections produce a conservative class superset. See
component waste classes and impurity limits for geometry, inventory
coverage, external-tritium declarations, and both range input schemas.
waste composition solve generates those pure-element responses from one
prepared native ACTINV run per component, checks every distinct projected
extremum composition plus a canonical feasible reference against full native
runs, and retains audit evidence. It verifies selected points and does not
bound solver or nuclear-data error. The fixed-rate affine model remains
qualified only within its declared domain; see the P109 section of the public
waste handbook.