Round 1pending pipeline run
VERA Problem 1A — single PWR pin (HZP)
Single fuel pin in square unit cell, pitch 1.26 cm (Westinghouse 17×17 design). HZP / 565 K / 1300 ppm B.
What this page validates
Round 1 tests nucdata.tools' ACE pipeline — the ACER stage of our generator on the ENDF/B-VIII.1 evaluations — by feeding the produced ACE into OpenMC continuous-energy and comparing against OpenMC's own official ENDF/B-VIII.1 ACE for the same problem. If our pipeline emits the right cards, the two k-inf values must agree within Monte-Carlo statistics + a few tens of pcm. The CASL VERA reference is shown for external context but is on a different evaluation (pre-VIII.1) and therefore is not the validation criterion.
What this page does not validate: the SHEM-281 multigroup library + DRAGR draglib used by deterministic codes (DRAGON5 / OpenMC-MG). That codepath goes through GROUPR + DRAGR, not ACER, and gets its own page in round 2.
Result — k-inf comparison
pendingThe Python pipeline that runs OpenMC twice (once on nucdata.tools-built ACE, once on openmc.data official ACE) and writes back into
/data/vera_problem_1a_result.json hasn’t been kicked off yet. Spec below is locked; numbers in the table will populate after the first run.| column | library / code | k-inf | ±1σ (pcm) | role |
|---|---|---|---|---|
| A | nucdata.tools ENDF/B-VIII.1 ACE → OpenMC CE | — | — | device under test |
| B | openmc.data official ENDF/B-VIII.1 ACE → OpenMC CE | — | — | consistency baseline |
| C | CASL VERA reference (KENO-VI CE, ENDF/B-VII.0) | — | — | external anchor — different evaluation, NOT a pipeline-bias signal |
A − B (pipeline bias) — round 1 criterion
— pcm
A − C (external anchor) — informational only
— pcm
Includes ENDF/B-VII.x → VIII.1 evaluation drift; not a test of the pipeline.
boron sweepΔ(A−B) is spectrum-dependent. At 0 ppm B the same libraries agree within MC uncertainty (≈ +32 ± 33 pcm); at 1300 ppm B the offset is amplified by U-238 resolved-resonance capture weighting under the harder spectrum. A B-10-only swap test had no k effect, ruling out B-10 processing as the driver. See
scripts/vera/diagnostics/NOTES_step16_boron.md.T sweepΔ(A−B) shrinks at higher fuel temperature: ≈ +23 ± 31 pcm at 900 K (PASS) vs the +60 ± 31 pcm at 565 K shown above. Consistent with Doppler broadening flattening the U-238 RRR (n,γ) peaks where the bias lives. T processing is consistent between the two NJOY recipes. See
scripts/vera/diagnostics/NOTES_step17_temperature.md.Geometry
| pitch (square unit cell) | 1.26 cm |
| fuel pellet OR | 0.4096 cm |
| clad IR | 0.418 cm |
| clad OR | 0.475 cm |
| boundary (4 sides) | reflective |
| axial | infinite (2D) |
Materials
Fuelρ = 10.257 g/cm³ · T = 565 K
UO₂ — 3.1 wt% U-235 (CASL Table P1-3 fabrication blend)
| isotope / element | wt % |
|---|---|
| U234 | 0.02318 |
| U235 | 2.733 |
| U236 | 0.01260 |
| U238 | 85.38 |
| O16 | 11.85 |
Stoichiometric UO₂ (O/U = 2). U-vector matches CASL Table P1-3 verbatim — including U-236 trace (0.0143 wt% of U). Atom densities verified against CASL Table P1-4 to ≤ 0.001% relative.
Gapρ = 0.000178 g/cm³ · T = 565 K
He fill gas
| isotope / element | atom % |
|---|---|
| He4 | 100.0 |
Cladρ = 6.56 g/cm³ · T = 565 K
Zircaloy-4 (CASL Table P1-2 Appendix A mixing table)
| isotope / element | wt % |
|---|---|
| Zr | 98.23 |
| ↳ Zr90 | 49.8 |
| ↳ Zr91 | 11.0 |
| ↳ Zr92 | 17.0 |
| ↳ Zr94 | 17.6 |
| ↳ Zr96 | 2.89 |
| Sn | 1.450 |
| ↳ Sn112 | 0.0133 |
| ↳ Sn114 | 0.00918 |
| ↳ Sn115 | 0.00477 |
| ↳ Sn116 | 0.206 |
| ↳ Sn117 | 0.110 |
| ↳ Sn118 | 0.349 |
| ↳ Sn119 | 0.125 |
| ↳ Sn120 | 0.477 |
| ↳ Sn122 | 0.0689 |
| ↳ Sn124 | 0.0876 |
| Fe | 0.2100 |
| ↳ Fe54 | 0.0119 |
| ↳ Fe56 | 0.193 |
| ↳ Fe57 | 0.00454 |
| ↳ Fe58 | 0.000614 |
| Cr | 0.1000 |
| ↳ Cr50 | 0.00417 |
| ↳ Cr52 | 0.0837 |
| ↳ Cr53 | 0.00967 |
| ↳ Cr54 | 0.00245 |
| Hf | 0.01000 |
| ↳ Hf174 | 0.0000156 |
| ↳ Hf176 | 0.000519 |
| ↳ Hf177 | 0.00184 |
| ↳ Hf178 | 0.00272 |
| ↳ Hf179 | 0.00137 |
| ↳ Hf180 | 0.00354 |
Full Zircaloy-4 per CASL Table P1-2 — 5 elements (Zr/Sn/Fe/Cr/Hf), no O. The earlier round-1 spec carried 0.125 wt% O16 as an ASTM-B350-11 surrogate; that's not what CASL uses for the benchmark, and removing it closes the third material-fidelity gap (alongside U-236 fuel and Hf-trace clad).
Moderatorρ = 0.743 g/cm³ · T = 565 K
H₂O + 1300 ppm B (borated coolant)
| isotope / element | wt % |
|---|---|
| H1 | 11.18 |
| O16 | 88.69 |
| B | 0.1300 |
| ↳ B10 | 0.0240 |
| ↳ B11 | 0.106 |
S(α,β): H2O on H1 @ 565 K
TSL evaluated at exactly 565 K. ENDF/B-VIII.1 H-in-H₂O is tabulated every 5 K from 285 K to 650 K (94 T points) — 565 K is on the grid, no snap or interpolation needed.
Isotope list (processed through nucdata.tools generator for column A)
U234U235U236U238O16H1B10B11Zr90Zr91Zr92Zr94Zr96Sn112Sn114Sn115Sn116Sn117Sn118Sn119Sn120Sn122Sn124Fe54Fe56Fe57Fe58Cr50Cr52Cr53Cr54Hf174Hf176Hf177Hf178Hf179Hf180He4
38 isotopes total. Sn natural composition (10 stable isotopes) included for Zr-4 alloy fidelity. He-4 for the gap. Plus the H-in-H₂O TSL at the temperature noted in the moderator card.
Reference + scope
- Spec source: Godfrey 2014 — "VERA Core Physics Benchmark Progression Problem Specifications" (CASL-U-2012-0131-004 Rev 4), ORNL.
- Reference evaluation: KENO-VI CE on ENDF/B-VII.0 (Godfrey 2014 Table P1-5). Our pipeline emits ENDF/B-VIII.1; absolute Δk vs CASL absorbs that evaluation drift.
- Pipeline: nucdata.tools NJOY generator (CE-targeted ACER workflow — `openmc_ce` preset) → ACE files → OpenMC continuous-energy pin run with reflective square BC → k-inf pulled from the statepoint and written into
/data/vera_problem_1a_result.json. - Out of scope (round 1): temperature sweeps, boron sweeps, enrichment variants (1B / 1C), assembly extension (Problem 2), depletion, multigroup library validation. Each gets a follow-up page.
- Provenance (when populated): — · OpenMC — · NJOY-generator commit — · — active particles per column.