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

Validation evidence

ACTINV carries numerical controls, reader and data-processing checks, and comparisons with measured activation histories. Evidence establishes behavior for the recorded inputs, implementation, and acceptance bounds. Assess applicability to your own material, spectrum, data, time range, and response separately.

What has been checked

EvidenceWhat it assessesPractical limit
Analytic and independent CRAM controlsTime integration, network accounting, and numerical comparisonsDoes not establish evaluated-data accuracy
IAEA FNS decay-heat experimentsRecorded irradiation/cooling cases across 73 materials and 132 experimentsResults depend on evaluation choice and the benchmark metric
IRDFF-II spectrum-averaged cross sections120 measured reaction cross sections in the Cf-252 and U-235 fission spectra (IRDFF-II Tables 18–20), with TENDL-2025: median absolute log error 0.014 to 0.024 per tableTests cross sections in fission spectra, not 14 MeV fields or irradiation histories
NPI Řež quasi-monoenergetic fields29 measured activation reaction rates in two 7Li(p,n) neutron fields peaking near 10–12.5 MeV, an independent facility from FNS; see the benchmarkTests activation cross sections folded with a published spectrum, not inventories or decay heat
Fission decay-heat experimentsFission-product heat after thermal fission of U-233, U-235 and Pu-239, measured at five laboratories (Oak Ridge, Los Alamos, IRT, UC Berkeley, CEA Cadarache), from 0.75 s to 1.5 × 10⁵ s; see the benchmarkTests fission yields and decay data, not activation cross sections; thermal fission only
FNG/ITER cell-620 historySupplied activation history and selected nuclides over recorded timesDoes not validate complete shutdown-dose transport
Photon and response controlsReader agreement, source conservation, unit conversions, and screening responsesLimited reference cases; contact dose remains a slab proxy
Import and mesh controlsSupported tally subsets, normalization, rebinning, and independent-cell outputDoes not qualify the originating transport calculation
Covariance and sensitivity controlsRetained uncertainty propagation and coverage reportingDoes not include all uncertainty sources

Compare the same data

Engine comparisons and library comparisons answer different questions. A calculation with EAF-2010 and one with TENDL-2025 can disagree because their evaluated inputs differ. Record the activation and decay data for both sides before assigning a difference to a solver.

The older validation record begins with a historical EAF-2010 run. Its figures are tied to that run and are not the current default-library score. Understand the competitive benchmark explains the later comparisons, metrics, and data-choice limits, with links to the full report and evidence.

Follow the evidence

The repository stores protocols, amendments, compact results, and independent checker scripts. Some verdicts are conditional or failed; those records are retained with their stated scope. A later software release does not convert a failed evaluation comparison into a pass.

Use the recorded validation controls, data disclosure, and relevant release notes together. Scope and qualification states the boundary between these checks and approval of an analysis.