The citable record

What's in the Zenodo deposit.

Everything on these pages is a demonstration of one artifact: the open validation package at DOI 10.5281/zenodo.21248916, published under CC BY 4.0. It holds the complete 81-analysis programme behind the MetroVolt design paper — the code, the input decks, the computed outputs, and the adverse results, deposited on purpose. If you cite anything, cite the DOI.

81
analyses S01–S81
48 + 24 + 4
Tier-1 · Tier-2 · gates (= 76)
+5
design-space studies S77–S81
43 / 0 / 4
run_all.py OK · fail · skip

Two tiers, stated honestly

Tier-1 — computed (48 of 76). A shared reduced-order engine — Bosch–Hale reactivities, NRL relativistic bremsstrahlung, IPB98(y,2) confinement, FreeGS free-boundary equilibria, analytic Stix/Spitzer kinetics — reported with measured values and pass/fail, fully reproducible from the deposited Python. This is the engine these Live Physics pages run in your browser.

Tier-2 — deck-deposited (24 of 76). Runnable input decks for the flagship codes — CGYRO, TGLF, OpenMC, ASCOT5, DREAM — carrying the frozen parameters and acceptance criteria, pending HPC execution. The decks are in the archive for anyone to run independently.

Four decision gates. S18 (confinement), S26 (hot-ion balance), S51/S65 (DEC exhaust) — the go/no-go adjudicators for the design's stated feasibility gaps.

Five design-space studies (S77–S81). Forward-looking scans deposited outside the validation set because they explore alternatives to the frozen point — including S77's negative result (deeper triangularity does not buy confinement margin) and S79's adverse finding (Ip is the least forgiving parameter). Null results are part of the record here, not a filing drawer.

The archive, mapped

PathContents
MANIFEST.mdThe authoritative index — every analysis S01–S81 mapped to its files and status.
run_all.pyRuns the entire Tier-1 suite in one command; prints OK / FAIL / SKIP per script.
REPRODUCE.mdPer-script runtimes, a smoke-test flag, and the expected 43 / 0 / 4 sweep counts.
*.pdfThe bound validation report — the human-readable companion to the code.
common/The shared physics engine (kronos_physics.py, KX1_burn_engine.py) imported by every part.
paper1/ – paper4/S01–S64 by part: systems & equilibrium · core physics · fuel cycle, neutronics & magnets · DEC, BOP & economics.
extensions/S65–S76 — referee-requested, manuscript-update, and review-response analyses.
design_space/S77–S81 forward-looking studies (δ-scan, field lever, robustness, operating window, low-neutron dividend).
data/ · figures/The computed CSV outputs and deposit figures the scripts regenerate.

Reproduce it locally

Python 3.9+, minimal dependencies (numpy, scipy, matplotlib, freegs — no licensed software, no network, materials data bundled as curated CSV). Download the archive from the DOI, then:

# from the unzipped deposit
pip install -r requirements.txt
python run_all.py
# expected: 43 OK / 0 FAIL / 4 SKIP  # the 4 SKIPs are the external flagship-code decks

The in-browser engine on these pages is a numpy-free port of the same files, with the coefficients verbatim — and it proves the equivalence by re-deriving the frozen numbers every time a page loads. For the full-fidelity record (FreeGS equilibria, the complete 400-point profiles, every CSV), run the deposit itself.

Cite it

Ford, P. I. (2026). Kronos MetroVolt Validation Manifest: An 81-Simulation Data Package for a
Negative-Triangularity Spherical-Tokamak D-3He Conceptual Power Plant (v1.0) [Data set].
Zenodo. https://doi.org/10.5281/zenodo.21248916   · License: CC BY 4.0

The design paper itself — a single comprehensive manuscript in four parts — posts to arXiv (physics.plasm-ph) in July 2026 and is submitted to IOP Nuclear Fusion; the Zenodo record links both. A browsable GitHub mirror of the deposit is linked from the Zenodo page.

Why we publish this way A conceptual design lives or dies on whether its numbers survive contact with someone else's computer. The typical concept paper deposits nothing, or a "data available on reasonable request" line. This one deposits the generator — and the two governing feasibility gaps (required confinement H₉₈ ≈ 1.8–2.2, and the hot-ion two-fluid balance) are quantified and gate-adjudicated in the same archive, not blended away. Read the plain-language version in WP33 “The Open Validation Manifest” and WP34 “Run It Yourself.”
The writeup

The plain-language paper behind this instrument: 81 Simulations, Zero Secrets: The Open Validation Manifest · Run It Yourself: 43 Scripts, One Command, Same Numbers.

KRONOS FUSION ENERGY · MetroVolt Live Physics Zenodo · Whitepapers · Simulations