Keyboard shortcuts

Press ← or → to navigate between chapters

Press S or / to search in the book

Press ? to show this help

Press Esc to hide this help

Move a FISPACT-II deck to ACTINV

actinv import-fispact reads a FISPACT-II input deck (.i) and writes an ACTINV problem specification. It converts the problem description: the material, the neutron (or charged-particle) spectrum, the irradiation and cooling schedule, and the few output settings that have an ACTINV counterpart. It does not read FISPACT-II’s nuclear data. The activation library and decay data of the converted problem are ACTINV’s, so results differ from a FISPACT-II run by the difference between the two data sets as well as by the solvers.

actinv import-fispact DECK.i OUT.json [--fluxes FLUXES] [--files FILES]
                      [--library LIB.npz] [--decay DECAY]
                      [--covariance COV.npz] [--strict]

The command writes OUT.json (the problem) and OUT.import.json (the import report, below). Four further options supply data that a deck does not name; each file is hashed into the spec, and a deck that does not use an option imports exactly as it would without it:

OptionUsed byFills
--fission-yields FILE...USEFISSIONfission_yields.files: ENDF neutron-induced yield files, one per parent. The parent of each file is read from its own header.
--resonance FILE...SSFGEOMETRYself_shielding.geometry.resonance_files: ENDF evaluations with resolved resonance parameters
--self-shielding-table FILEPROBTABLE, SSFDILUTIONself_shielding.table: a Bondarenko table
--damage TABLE (optional)ALLDISPEN, ATDISPENdamage.table: a damage table, for example from actinv build-damage --method kinematic. Without it a neutron deck uses the catalog’s tendl-2025-neutron-damage. A table that records its displacement energies (an actinv-damage-table-2 file) is compared with the deck’s.

A file that cannot be read stops the import. An option the deck never uses is listed under options_unused in the report. Run and check the converted problem in the usual way:

actinv import-fispact TENDL-2017_1996exp_5min.i fe_5min.json
actinv validate fe_5min.json
actinv run fe_5min.json fe_5min_result.json

The desktop imports a deck from the toolbar (Import FISPACT-II deck…) and shows the same per-keyword report as actinv import-fispact. A deck that irradiates, imported without a fluxes file, gets a flat placeholder spectrum flagged as a required input. Replace it before running, because the result means nothing until you do.

Coming from FISPACT-II

This map is by capability, not by keyword; the keyword translation below says what the importer does with each deck line. Supported means ACTINV has an equivalent that does the same job. Supported differently means the job gets done, but the input, the output or the numerical treatment is not FISPACT-II’s. Not supported means ACTINV has nothing equivalent. Not needed marks a keyword that exists only for FISPACT-II’s own printing, files or solver. A capability can be supported in ACTINV and still be absent from the importer, in which case you add the spec section by hand; the last column says so. “Imported” means the importer maps the deck keywords.

FISPACT-II capability (keywords)ACTINV equivalentStatusImporter and notes
Material and irradiation: MASS, FUEL, DENSITY, FLUX, ZERO, TIME, PULSE, PROJECTILEmaterial, spectrum.total, schedule, projectile; the Material and Irradiation & cooling pagesSupportedImported. Gamma, triton and helion need --library; triton and helion have not been run on real data
Wall loading and power: WALL, POWERstep wall_loading_MW_m2, power_W_cm3 or power_W_per_g; a kerma table; the step Drive on Irradiation & coolingSupported differentlyImported, but a power step needs a kerma table the deck cannot supply. ACTINV holds the power on sub-intervals within a step; max_substeps: 1 reproduces FISPACT-II’s single normalization
Elemental partition: PARTITIONschedule[].partitionSupportedImported. Applied at the start of the next step
Group conversion: GRPCONVERT, CNVTYPE, LIMGRPspectrum.rebinSupportedImported for FISPACT-II standard structures
Cross-section and half-life changes: OVER with ACROSS, ALAMoverridesSupported differentlyImported. ADCROSS, ADLAM (error factors) are not imported; overrides are not step-scoped
Change of library mid-run: GETXS after the first TIMEstep librarySupportedNot imported: it needs the second library’s path
Cooling from decay data alone: GETDECAY without GETXSdecay-only runSupportedImported
Library choice: GETXS, GETDECAY, ARRAY, COLLAPSElibrary, decay; actinv data fetchSupported differentlyACTINV uses its own libraries, not the deck’s. Library-preparation switches (LIBVERSION, COMPRESSXS, FULLXS, SAVELINES) have no counterpart and are noted
Inventory listing, totals, elements, dominant nuclides: ATOMS, HALF, NOCOMP, NOSORT, SORTDOMINANT, MIND, NOSTABLEsummary section; Step summary on What to computeSupportedImported (MIND, NOSTABLE, SORTDOMINANT)
Tab and plot files: TAB1 to TAB4, GRAPH, NUCGRAPHactinv report; the Nuclides pageSupported differentlyNoted. ACTINV writes one result JSON and renders tables from it afterwards
Depletion rates: DEPLETIONsummary.depletion; actinv report --tables depletionSupportedNot imported (reported as unsupported); add summary.depletion by hand
Point-source dose and bremsstrahlung: DOSE 2, BREMSSTRAHLUNGphoton.point_source, photon.bremsstrahlung; Photons & doseSupported differentlyImported. The dose is an air kerma rate, and bremsstrahlung is a separate block; both need the photon response described below
Photon groups: GROUP, READGGphoton.group_structure (fispact-24, steiner-22, or custom)SupportedGROUP is imported. READGG is imported as a custom structure when the files file has a ggbins entry whose file holds the group count and then that many plus one ascending energies in eV, as the FISPACT-II manual describes; a missing, unreadable or malformed ggbins file is reported as unsupported with the reason
Approximate gamma spectra for nuclides without spectral data: SPEKphoton.missing_spectra: "estimate"Supported differentlyImported as approximated. ACTINV’s estimate is an exponential shape set by each nuclide’s mean energy per decay mode, not FISPACT-II’s spectrum, and the result carries certified lower and upper dose bounds
Transport index and hazards: ATWO, HAZARDS, CLEARradiological with bundled: tables; actinv report --tables regulatory; Regulatory tables on What to computeSupported differentlyATWO is imported. HAZARDS and CLEAR are noted: ACTINV chooses no regulation for you. 49 CFR 173.435/173.436 and Euratom 2013/59 Annex VII are bundled; ICRP 119 coefficients are not and come from a local script. These are the regulations’ numbers, not FISPACT-II’s data files
Geometry self-shielding: SSFGEOMETRY, SSFCHOOSE, SSFFUEL, SSFMASSself_shielding.geometrySupportedNot imported (reported as unsupported); the section needs hash-pinned ENDF resonance files the deck does not name
Dilution override: SSFDILUTION, PROBTABLEself_shielding.dilution_overrides, self_shielding.tableSupported differentlyNot imported. ACTINV needs a hash-pinned Bondarenko table; tables are built from evaluations with actinv build-shielding
Uncertainty: UNCERTAINTY, NOERRORuncertainty (MF=33 covariance)Supported differentlyImported. The default covariance is the TENDL-2025 MF=33 data, not the deck’s library. ACTINV propagates full-covariance bands rather than FISPACT-II’s pathway sums
Uncertainty controls: COVARIANCE, UNCTYPE, ERRORuncertainty.covariance, uncertainty.channels, overridesSupported differentlyNot imported
Monte Carlo sampling and sensitivity: MCSAMPLE, SENSITIVITYstudy.robustness.statisticsSupported differentlyNot imported (reported as unsupported). The statistics are those of an ACTINV robustness study, not a FISPACT-II sampling run
Pathways: PATH, ROUTES, LOOKAHEAD, PATHRESET, GENERIC, INDEXPATH, RESULTpathways; Pathways pageSupported differentlyPATH and ROUTES are imported as targets with default thresholds; the others are noted. The result reports a closure for each target
Fission: READSF, FISYIELD, CUMFYLD, USEFISSION, NOFISS, FISCHOOSEfission_yields.files, fission_yields.spontaneous, options.fission, actinv inspect yieldsSupported differentlyNot imported. Which actinides fission is set by the yield files you supply, with no keyword switch
Displacement damage: ALLDISPEN, ATDISPENdamageSupported differentlyImported against --damage TABLE, or against the catalog table tendl-2025-neutron-damage for a neutron deck without it; a per-element ATDISPEN value overrides ALLDISPEN in any order. They are mapped when the table was built at the deck’s energies (or records none, as a table 1 does) and reported as unsupported otherwise, with the actinv build-damage --method kinematic --displacement-energy command that builds a matching table. An element with no energy in the deck takes the table’s energy, else FISPACT-II’s default (25 eV outside its table 8). With --damage and neither keyword the section uses those defaults, as FISPACT-II computes displacement rates by default. With the catalog table installed, the energies are compared at import. If it is not installed, the deck’s energies are written as given, an element the deck omits takes the table’s energy at run time (reported as approximated), and the run refuses a mismatch. A deck for another projectile needs --damage and is unsupported without it, naming the option
High-energy data: USESPALLATION, USEXSEXTRAgroup structures to 1 GeV; build-library splice optionSupported differentlyNot imported. You build the library with the splice table
Library inspection: PRINTLIB, EXTRACTXSactinv inspectSupportedNoted. Works without a run
X-ray switch: NOXRAYphoton.exclude_xraysSupportedImported
Cross-section threshold switch, iron background removal: XSTHRESHOLD, IRONnoneNot supportedXSTHRESHOLD is noted (ACTINV propagates the full covariance); IRON is reported as unsupported
Run control and print switches: CLOBBER, LOGLEVEL, MONITOR, TOLERANCE, NOT1 to NOT4, CULTAB, SPLIT, JSONnoneNot neededNoted. ACTINV uses CRAM, which has no solver tolerance, and writes only the files you name

After the import, edit the JSON problem rather than the deck: ACTINV runs the problem, not the deck. A converted problem does not reproduce FISPACT-II’s numbers by construction, because the data differ. See Understand the competitive benchmark for the recorded FNS comparison on common data.

Import in the desktop

The desktop’s Import FISPACT-II deck… (Ctrl/Cmd+I) runs the same importer and shows the same per-keyword report. It differs from the CLI in one case: a deck that irradiates but comes without a fluxes file converts with a flat placeholder spectrum, flagged as a required input, so you can open the problem and fill in the spectrum. The CLI refuses such a deck. Replace the placeholder before running. See Import a FISPACT-II deck.

What is read

InputUsed for
DECK.imaterial, flux, schedule, projectile, and the keywords listed below
--fluxes FLUXESthe group spectrum: a FISPACT-II fluxes file (values highest energy first, first-wall loading, title)
--files FILESa FISPACT-II files file. Its fluxes entry names the spectrum when --fluxes is absent. Without --files, DECK_files and then files beside the deck are tried.
--library, --decaythe ACTINV activation library and decay data for the problem
--covariancean MF=33 covariance sidecar paired with --library, used when the deck asks for uncertainties

The xs_endf, dk_endf and other data entries of a files file are listed in the report for reference and never read. Without --library and --decay the problem uses the catalog references that actinv new uses (catalog:tendl-2025-neutron-709g, catalog:endfb-viii-0-decay with the JEFF-3.3 fallback, and for charged particles the matching 162-group library); install them with actinv data fetch. A gamma-projectile deck needs --library.

The group count of a fluxes file is taken from the file itself. ACTINV reads 709-group neutron spectra and 162-group charged-particle spectra directly; a GETXS 1 N line must agree with the file. A file in another FISPACT-II standard structure (66, 69, 100, 172, 175, 211, 315, 351, 586, 616 or 1102 groups) is imported when the deck says GRPCONVERT: the importer takes the file’s boundaries from that structure and writes a spectrum.rebin section, so ACTINV converts the spectrum to the library’s groups inside the run. A file with any other group count stops the import; convert it to 709 groups first.

How the deck is read

The importer follows the FISPACT-II input rules:

  • Keywords are free format and case-insensitive. Only the leading characters matter: a word of three or more characters that begins exactly one keyword is that keyword (TOL, HAZA, UNCERT, TOLER are TOLERANCE, HAZARDS, UNCERTAINTY, TOLERANCE), and an ambiguous abbreviation is unknown. The deprecated spellings NOHEAD/NOHEADER, COLLAPSE (GETXS 1), ARRAY (GETDECAY 0) and AINPUT (GETXS 0 and GETDECAY 0), and the legacy switches MCSEED, CONV, DOMINANT, EAFVERSION, ENFA, LEVEL, LINA, LOOPS, NEWFILE, SEQNUMBER, SEQUENTIAL and TAPA are known. A word that matches nothing is reported as a separate unsupported entry (“unknown to this importer”), including when it follows a keyword that takes a list, such as PATH or OVER.
  • The deck is read as bytes. A UTF-8 byte-order mark is ignored, lines end at \n, \r\n or a bare \r, and bytes that are not UTF-8 (a Latin-1 micro sign in a comment, for example) are replaced rather than stopping the import; the report records deck.lossy_utf8.
  • << ... >> comments may appear anywhere, including inside a line, and may span lines. A line whose first character is * is text (the title after FISPACT, and after END); << has no meaning inside it.
  • The library phase ends at FISPACT, the initial conditions at the first TIME, and the deck at END. Nothing after END is read.
  • TIME t [SECS|MINS|HOURS|DAYS|YEARS] sets an interval; the next ATOMS, STEP or SPECTRUM runs it. Each run becomes one step of the ACTINV schedule, at the flux in force when the output keyword is read. A word after the number that is not one of these units is an unknown keyword, and the interval is in seconds. A bare number is seconds (the UKAEA tutorial deck says so).
    • Assumption. That TIME only sets up an interval and the output keyword runs it is the reading of the UKAEA FISPACT-II tutorial decks, whose comments say that ATOMS “tells FISPACT-II to solve the rate equations”. It is not confirmed against a FISPACT-II output for a deck with two TIME lines before one ATOMS; the pypact reader treats every TIME as a step. The importer therefore keeps the deferral but never hides its cost: a TIME that a later TIME replaces, or that no output keyword follows, is unsupported (so --strict stops), and its entry gives the lost interval and flux. Put an ATOMS after each TIME to make the intent unambiguous.
  • FLUX sets the flux for the steps that follow. The first positive value becomes spectrum.total, and each step’s flux multiplier is its flux divided by that value, so a step at FLUX 0 is a cooling step. ZERO also sets the flux to zero, as in FISPACT-II. The ACTINV schedule carries elapsed time through ZERO; FISPACT-II restarts the elapsed time printed after it (the cooling time restarts at zero, as in the pypact reference test31.out). The interval durations agree; absolute times in the two outputs differ by the irradiation time. FISPACT-II is documented to allow ZERO once and to refuse irradiation after it; the importer does neither check and simply sets the flux to zero (both statements are unverified here).
  • PULSE n ... ENDPULSE runs its body n times in total (PULSE 3 gives three executions, not four), expanded into explicit steps (at most 100000). Loops nest to 64 levels, and a deck whose loops expand to more than two million keyword executions is refused before anything runs. Keywords in the body are reported once, at their first execution. The body of PULSE 0 does not run, but its keywords are still listed (marked as skipped), so an unsupported keyword there is visible to --strict. This reading of n has not been checked against the FISPACT-II manual.

Keyword translation

FISPACT-IIACTINV
MASS kg n and ELEMENT percent linesmaterial with mass_g and basis wt_percent
FUEL n and NUCLIDE atoms linesmaterial with basis atoms_per_g and explicit nuclide keys; the mass is taken from mass numbers (the atom counts are exact)
FLUX, ZERO, TIME, ATOMS, STEP, SPECTRUM, PULSEspectrum.total and schedule
PROJECTILE nprojectile (1 neutron, 2 deuteron, 3 proton, 4 alpha, 5 gamma, 6 triton, 7 helion). Gamma, triton and helion have no catalog library: pass --library.
FISPACT title linetitle
MIND, NOSTABLE, SORTDOMINANTthe summary block: min_atoms, include_stable, dominant_top
UNCERTAINTY 1 or 2uncertainty with an MF=33 covariance (see below)
UNCTYPE 1, 2, 3uncertainty.channels (cross_section_mf33, decay_constants, both); type 2 is approximated
SENSITIVITY SIGMAuncertainty.voi, approximated
ERRORper-reaction overrides of kind cross_section_uncertainty, approximated
USEFISSION with --fission-yieldsfission_yields.files
NOFISS, CUMFYLD, FISYIELDno yield files; fission_yields.yields: "cumulative"; fission_yields.parents
FISCHOOSEpathways.fission_parents
SSFGEOMETRY (and SSFCHOOSE) with --resonanceself_shielding.geometry
PROBTABLE, SSFDILUTION with --self-shielding-tableself_shielding.table (approximated), self_shielding.dilution_overrides
NOERRORno uncertainty block
GETXS, GETDECAYthe library and decay references; a GETXS 1 N group count is checked against the spectrum
POWER p n mt…schedule power_W_cm3, with material.density_g_cm3 from DENSITY (see below)
WALL wschedule wall_loading_MW_m2, with spectrum.wall_loading_MW_m2 taken from the fluxes file (see below)
PARTITION element/fraction pairsschedule partition on the next step
OVER nuclide with ACROSS, ALAMoverrides entries of kind cross_section and half_life
GRPCONVERT, CNVTYPE, LIMGRPspectrum.rebin with convention, emin_eV, emax_eV
DOSE 2 distancephoton.point_source
GROUP 0, GROUP 1photon.group_structure fispact-24, steiner-22
BREMSSTRAHLUNG nuclidesphoton.bremsstrahlung.nuclides
PATH, ROUTESa pathways section, the nuclides named as targets
ATWOradiological with the bundled us-49cfr173-a2 table
HALF, HAZARDS, CLEAR, GRAPH, NUCGRAPH, TAB1 to TAB4, NOT1 to NOT4, TOLERANCE, PRINTLIB, MONITOR, CLOBBER and similarrecorded as noted: they select FISPACT-II printing, files or solver controls and do not change the calculated physics. The tables and plot data come from actinv report (below). HAZARDS and CLEAR name a regulation, which ACTINV does not choose for you: add a radiological section with the table you want (see Bundled regulatory tables).

A deferred setting that appears twice (UNCERTAINTY, GRPCONVERT, CNVTYPE, LIMGRP) takes its last value, and the earlier entries say they were superseded.

UNCERTAINTY needs a covariance sidecar. With --covariance FILE it is hashed and used. With the default neutron library the catalog covariance sidecar is used and the entry is approximated, because its uncertainties are the TENDL-2025 MF=33 data, not those the deck’s FISPACT-II library carried. With another --library and no --covariance the entry is unsupported. Level 3 (pathways only) produces no band.

DENSITY is written to material.density_g_cm3 only when a POWER step needs it. Otherwise it is recorded in the report and left out, since it only adds per-cm3 summary totals and the converted problem stays identical to a hand-written one; add it yourself if you want them. The summary block carries the summary options; a build without them rejects a spec that contains it.

Keywords that need more than the deck

  • POWER. POWER p n mt_1 … mt_n gives a power density p in W/cc normalized on the kerma of the listed MT numbers (usually 1 301). It is written as power_W_cm3 on every step that follows until the next FLUX, WALL, ZERO or POWER, and the deck’s DENSITY becomes material.density_g_cm3. The list is read as n followed by exactly n MT numbers; ACTINV normalizes on the kerma table’s total (MT=301), and another list, a missing list or words after the list are named in the entry. A power-driven step needs a kerma section with a hash-pinned actinv-kerma-table-1 file, which no deck provides, so the entry is approximated: the importer writes the spec without it, lists it (and the density, when the deck has no DENSITY) under required_additions, and says so in its summary line. The spec does not validate until you add them. Build the table with scripts/build_kerma_table.py.

  • WALL. The first-wall loading of the fluxes file becomes spectrum.wall_loading_MW_m2, and a step’s multiplier is wall divided by it, as in section 4.91 of the FISPACT-II manual. Because the multiplier then applies to the whole file spectrum, spectrum.total is the sum of the file’s group values and any FLUX in the same deck is expressed against it. A file that records no positive loading falls back to ACTINV’s 4.44e13 n/cm2/s per MW/m2 equivalence, and the entry is approximated.

  • PARTITION. Element/fraction pairs, with or without a leading pair count, give the fraction of each element’s atoms kept. FISPACT-II applies the split when it reads the keyword, ACTINV at the start of the next step, so a PARTITION with no step after it is unsupported.

  • OVER. Each ACROSS, ALAM, ADCROSS and ADLAM is reported under its own name; OVER itself is the worst of them. ACROSS target daughter value becomes a cross_section override on the ENDF reaction that takes the target to the daughter, found from the change in Z and A: (n,g), (n,2n), (n,3n), (n,p) and (n,a) only, for neutron projectiles. FISPACT-II sets the channel to that daughter; ACTINV sets the whole collapsed reaction and keeps its product branching, so the entry is approximated. A daughter that is an isomer, a change that two reactions share (for example (n,d) and (n,np)) and a negative value are unsupported. ALAM value [unit] becomes a half_life override. ADCROSS and ADLAM are unsupported: ACTINV takes a relative standard uncertainty, FISPACT-II an error factor, and the spec does not define the conversion. Cross-section overrides cannot be combined with uncertainty, so the importer drops the uncertainty block and marks UNCERTAINTY as unsupported.

  • GRPCONVERT, CNVTYPE, LIMGRP. CNVTYPE 0 is equal_lethargy, CNVTYPE 1 is equal_energy. LIMGRP low high is read as two energies in eV, so the entry is approximated. The group counts of the spectrum come from the fluxes file, not from the arguments of GRPCONVERT.

  • DOSE 2, GROUP, BREMSSTRAHLUNG. DOSE 2 distance writes photon.point_source; a distance below 0.3 m is raised to 0.3 m, as FISPACT-II does. The dose values are null until you add a photon.response with an air attenuation curve. The shipped photon response has none, so an imported DOSE 2 gives null doses by default: rebuild a response with scripts/build_photon_response.py --air-attenuation and declare its new SHA-256, or set photon.point_source.air_attenuation to false for the unattenuated dose. The dose is the air kerma rate (numerically Gy/h), not an ambient dose equivalent, and it needs options.outputs to include photons or dose when the deck sets outputs. BREMSSTRAHLUNG is approximated: ACTINV reports the bremsstrahlung photons and their dose in a separate block and does not add them to the decay-photon dose. The electrons per decay come from each nuclide’s beta-minus decay branching and their energy is the STYP 1 average energy divided by that branching. A deck that lists no nuclides, or an element, is unsupported.

  • PATH, ROUTES. The nuclides named become pathways.targets and the route-search thresholds keep their defaults; numeric arguments are listed in the entry as not interpreted. ACTINV refuses pathways together with power-driven steps or PARTITION; the importer then drops the section and marks the entry unsupported.

  • ATWO. The A2 values of 49 CFR 173.435 (the IAEA SSR-6 Rev.1 table) are written as bundled:us-49cfr173-a2, with one response per uranium absorption type. They are not FISPACT-II’s own A2 file, so the entry is approximated. Nuclides with an unlimited A2 (Mn-53, Ni-59, Zr-93, …) count as covered with zero contribution, so require_complete: true accepts an activation inventory that holds them.

  • Fission. USEFISSION needs --fission-yields; without it the entry is unsupported and names the option. With it fission_yields.files lists the files, and the energy setting stays at the spec default. NOFISS forces no yield files, whatever else is given. CUMFYLD sets fission_yields.yields to cumulative. FISYIELD n selects the parents: a positive count includes the listed actinides, a negative count excludes them and zero excludes every parent that has a yield file. Without FISYIELD the FISPACT-II default of U-235, U-238 and Pu-239 applies, as an include list cut down to the parents that have a yield file. A named parent with no yield file does not fission in FISPACT-II either, so it is dropped and the entry names it. FISCHOOSE restricts only the pathway analysis, so it becomes pathways.fission_parents when the deck has PATH or ROUTES and is noted otherwise.

  • Uncertainty. UNCTYPE 1, 2 and 3 set the uncertainty channels. ACTINV propagates the MF=33 covariance whenever a covariance sidecar is given, so type 2 cannot leave the cross-section contribution out: the entry is approximated and says the band is wider. SENSITIVITY SIGMA asks for the ranked first-order variance shares (uncertainty.voi) in place of sampled sensitivity coefficients, and adds the uncertainty block when no UNCERTAINTY keyword did; SENSITIVITY LAMBDA is unsupported. MCSAMPLE and MCSEED configure the sampling that the analytic result replaces: they are noted beside a mapped SENSITIVITY and unsupported without one. ERROR n with parent, daughter and a fractional error becomes a cross_section_uncertainty override on the reaction that takes the parent to the daughter (Fission or 0 is MT 18). The error is read as the relative standard uncertainty of that reaction, which is an interpretation, so the entry is approximated; -1 asks for the library value and is noted. The override needs the uncertainty block with the cross-section channel, so without one the entry is unsupported. COVARIANCE is noted: ACTINV has no collapse step.

  • Self-shielding. SSFGEOMETRY type length [height] (1 foil, 2 wire, 3 sphere, 4 cylinder, lengths in cm) with --resonance becomes self_shielding.geometry, with the density from DENSITY (when the deck has none the density is listed under required_additions), and SSFCHOOSE becomes its choose list. SSFMASS and SSFFUEL give a composition for the self-shielding calculation, but ACTINV shields the problem’s own material, so they are approximated. PROBTABLE with --self-shielding-table becomes self_shielding.table; the Bondarenko factors stand in for FISPACT-II’s probability tables, so it is approximated. SSFDILUTION with the table becomes dilution_overrides. FISPACT-II numbers the energy groups from the highest energy (group 1), and ACTINV’s index is 0-based and ascending in energy, so group g of the 709-group library is index 709 - g. FISPACT-II does not allow SSFGEOMETRY beside PROBTABLE, and neither does ACTINV.

  • Noted and corrected. XSTHRESHOLD and DEPLETION are noted: the first is a cost control of FISPACT-II’s pathway-based uncertainty estimate, the second an output request (set summary.depletion for the table). RESULT is noted: a converted deck always defines a material and runs a schedule, whose inventory the pathways use.

A power-driven step is refused beside uncertainty, pathways and step libraries, and the importer follows those refusals instead of writing a spec ACTINV would reject.

The import report

Nothing is dropped silently. OUT.import.json (actinv-fispact-import-report-1) lists every keyword with its line, the section of the deck it appeared in, its arguments, and one status:

StatusMeaning
mappedtranslated into a spec field
approximatedtranslated with a stated caveat, given in detail
notedno effect on the calculated physics (printing, files, solver tolerances)
unsupportedwould change the physics and has no ACTINV equivalent, or is unknown to this importer (a misspelling, for instance)

The report also holds the counts, options_unused (present only when an option was given that no keyword of the deck used), the spectrum source and total, the library chosen, the content of the files file, the SHA-256 of the deck, required_additions (spec sections that the deck implies and no deck can supply, such as kerma for POWER; empty for most decks), and a schedule_map giving, for each ACTINV step, the deck line that defined it, whether it is irradiation or cooling, and the FISPACT-II output keyword that ran it. ACTINV reports every step in full, so the choice between ATOMS, STEP and SPECTRUM does not change what is computed.

Check every approximated and unsupported entry before trusting a converted problem. --strict makes the command fail, after writing the report, when any entry is unsupported.

Not translated

These are reported as unsupported, each with a note saying what to do:

  • The self-shielding keywords PROBTABLE, SSFGEOMETRY, SSFCHOOSE, SSFFUEL, SSFMASS and SSFDILUTION, unless the option that supplies their data is given (--resonance or --self-shielding-table, above). Each entry then names the missing option. A target type that ACTINV has no shape for, a cylinder without its height and SSFGEOMETRY beside PROBTABLE stay unsupported.
  • A GETXS after the first TIME. ACTINV expresses it as a step library override, which needs the second library’s path and the spectrum in its group structure. The entry names the schedule step to edit. Step libraries are refused beside uncertainty, kerma, power-driven steps, partition, pathways and cross-section overrides.
  • ADCROSS, ADLAM (error factor against relative uncertainty), and ACROSS daughters that are isomers or ambiguous.
  • Displacement energies that differ from the damage table’s (the entry gives the build-damage command for a matching table), displacement damage in a non-neutron deck without --damage, USEFISSION without --fission-yields, SENSITIVITY LAMBDA, IRON (not a keyword of the 2023 manual), READSF, USESPALLATION, USEXSEXTRA and the other library-selection keywords.
  • A GRPCONVERT spectrum whose group count is not a FISPACT-II standard structure stops the import, since its boundaries are not in the deck.

A material defined after the first TIME is not representable.

Behaviour to know about

  • MASS percentages. Percentages that do not sum to 100 are written as given and the spec does not renormalize them; the importer says the same reading applies in FISPACT-II, which has not been confirmed from the manual or an output (unverified).
  • MIND. The UKAEA tutorial decks comment that 1e5 atoms is the FISPACT-II default threshold. A deck with no MIND has that implicit threshold; the importer neither writes nor mentions it, so set summary.min_atoms if the nuclide lists should match (unverified default).
  • GRPCONVERT a b. The two numbers are not read. The group counts come from the fluxes file and the destination structure is always the library’s, whatever b says.
  • Isotopic abundances. A MASS composition is resolved through ACTINV’s natural-abundance table. FISPACT-II’s differs slightly, and the report does not list the difference.
  • OVER after the first step. ACTINV overrides are not step-scoped, so an OVER that follows a scheduled step applies to the whole run, including the earlier steps. Its entries are approximated and say so.
  • DOSE. DOSE 1 after a DOSE 2 and before any step cancels the point source (the DOSE 2 entry is marked superseded). After a step the point source stays for the whole run and the DOSE 1 entry is approximated.
  • --files. A fluxes entry that names a missing file stops the import only when the deck irradiates; a cooling-only deck imports and the report records the unresolved entry (fluxes_file_unresolved).
  • --strict. After a failure the report is written, and an OUT.json from an earlier run is removed so that it cannot be mistaken for the result. OUT may not be the deck itself.

Tables and plot data from a result

FISPACT-II writes its tables and graph files while it runs. ACTINV writes one result JSON, and actinv report renders it afterwards, so no tab file is needed and the tables can be regenerated for any step. The report is a pure function of the result file; it runs no solver and reads no data.

actinv report RESULT.json [--tables LIST] [--format csv|tsv|markdown|text]
              [--out DIR] [--step N|all] [--nuclides Co60,Mn54]
FISPACT-II--tablesContent
TAB1tab1atoms and grams per nuclide, per step
TAB2tab2activity (Bq) and heat (W, where FISPACT-II prints kW) per nuclide, per step; also the alpha, beta and gamma heat split and the half-life
TAB3tab3ingestion and inhalation dose responses, per step. These are totals over nuclides: the result holds the response value and its activity coverage, not the per-nuclide coefficients
TAB4tab4decay-photon source per energy group, per step (photons_s_g, power_W_g, absolute and per gram)
GRAPH, NUCGRAPHtimeseriesone row per step: total activity, alpha, beta and gamma heat, the photon dose rates and radiological responses when present, and --nuclides columns (atoms and Bq per gram)
SORTDOMINANT listingsdominant, elements, totalsthe summary dominant lists, the element table and the step totals (see the results page)

Grams, the heat split, half-lives, absolute (whole-sample) columns, the element table and the dominant lists come from the result’s summary; a deck imported with MIND, NOSTABLE or SORTDOMINANT already has the block. Without it tab1 and tab2 still give atoms and Bq per gram from the base fields, the other tables are reported as unavailable, and standard error names every column that was left out. Nothing is filled in. tab3 is unavailable unless the problem declares radiological with a dose response.

Quantities keep the result’s basis. A _per_g column is per gram of initial material; the unsuffixed atoms, mass_g, activity_Bq and heat_W columns scale by material.mass_g. csv and tsv hold exact values, formatted so they parse back to the same number; text and markdown round to six significant digits. See Command line for the options.

Convert a directory of decks

controls/p132_import_all.py imports every */*.i deck under a directory with the release binary and prints the status counts per keyword:

cargo build --release -p actinv-cli
python3 controls/p132_import_all.py ~/nuclear-data/conderc-fns/fns --out /tmp/p132-imports

It passes the library, decay and covariance files named by ACTINV_LIBRARY, ACTINV_ENDF_DECAY and ACTINV_COVARIANCE, when set, and otherwise lets the importer use the catalog references.