Program · reference

About this record.

How to cite the record, who built it, and the reference layer — publications, FAQ, glossary, datasets and the changelog.

Kronos Fusion Energy

About & How to Cite

Kronos Fusion Energy is developing two complementary fusion machines — a breeder (Hyperion) that makes tritium, neutrons and helium-3, and a burner (Aegis · MetroVolt) that makes net electricity. This interactive 3-D model is the public face of that work, built on fully open, reproducible physics.

What this is — and isn't
What it is
A physics-grounded, interactive 3-D model of both machines — geometry, live in-browser solving, and a full engineering-assurance stack.
What is honest
Every surface is tagged MODEL · COMPUTES NOW · AWAITS HARDWARE. Nothing is built yet; nothing is overstated.
What it becomes
A live digital twin once a machine exists and sensors come online — the model is built ready to receive it.
Open science — cite the record

Every number in this model reproduces from five open Zenodo deposits (CC BY 4.0). Please cite the deposit you draw on:

DepositDOI
Breeder — Hyperion (D–T spherical tokamak)10.5281/zenodo.21746157
Burner — D–³He tandem mirror10.5281/zenodo.21746479
REBCO magnet & tape system10.5281/zenodo.21842514
Direct Energy Conversion (DEC)10.5281/zenodo.21842864
AI + ML + Quantum reproducibility10.5281/zenodo.21842371
Interactive model — deep links

Link straight to the interactive view for each deposit — drop these into the papers and the Zenodo "related links". Copyable; they open the model on the right machine + tab.

Patents & intellectual property
StatusReferenceScope
GRANTEDU.S. Patent 12,009,112Foundational method — issued.
FILED · PAT. PEND.U.S. Application No. 64/128,097 (Omnibus)Engineering implementations — magnets, DEC, fuel cycle, control — Patent Pending.

Kronos maintains a clean, structured patent portfolio — granted patents plus filed applications — as its intellectual-property source of truth. Kronos®, Hyperion™, Aegis™ and MetroVolt™ are trademarks of Kronos Fusion Energy, Inc.

No license granted. Access to this model conveys no license to any patent, trademark, copyright, trade secret or other intellectual property. Reverse engineering, extraction of proprietary design parameters, or the creation of derivative machine designs from this model is prohibited. Open datasets are licensed CC BY 4.0 for the data and code as deposited; the underlying machine designs and marks remain proprietary.

Legal notice & terms of use
Contact
Webkronosfusionenergy.com
Open recordZenodo · Kronos Fusion Energy community
Enquiriesvia kronosfusionenergy.com
ReviewersIndependent reproduction is welcome — request an issued reviewer account, or report an error, using the template below.
© 2026 Kronos Fusion Energy, Inc. All rights reserved. Model figures and open data are CC BY 4.0 via the deposits above; designs, marks and implementations are proprietary and protected. This model is a design & assurance tool — see Future Capability ↗ for what becomes a live digital twin.
Kronos Fusion Energy

Team & Experience

The Kronos design rests on a rare depth of experience — the people, founders and advisors whose work runs from the first high-field tokamaks through ITER, NIF and the leading stellarators. Roles and tenures vary; this is the lineage the two-machine architecture is built on.

150+
Combined years in fusion
40+
Senior scientists, engineers & leaders
6
Decades of programs
Alcator→ITER→NIF
Programs the team has built
Six decades, one bench
1960s – 1970s
The first high-field tokamaks
Dr. Carl Weggel
MIT Francis Bitter Magnet Lab — designed the Alcator tokamaks (Guinness-record field)
Dr. Robert J. Weggel
Co-designed Alcator C magnets; a career of record-setting magnets
Dr. Steven O. Dean
Directed DOE/AEC Confinement Systems — the US fusion program takes shape
1980s
Mirrors, D–³He & compact fusion
Dr. Gerald Kulcinski
UW–Madison — D–³He tandem mirror with direct conversion (1987); steady-state He-3 fusion
Dr. Carl Weggel
Ultra-high-field magnets for the RIGGATRON compact-fusion program (INESCO)
Dr. Robert J. Weggel
Brookhaven — 20 T hybrid magnets & cryogenic superconducting systems
Dr. David A. Hammer
Cornell — founded the X-pinch pulsed-power program
1990s – 2000s
ITER, NIF & the codes that run fusion
Dr. Ruben Fair
Head of US ITER projects (PPPL); world-first HTS machines
Dr. Siegfried Glenzer
Led the first inertial-confinement fusion experiments at NIF
Dr. Donald A. Spong
ORNL — STELLOPT; NCSX / QPS stellarator design
Dr. Wilfred A. Cooper
EPFL — VMEC / TERPSICHORE; Wendelstein 7-X optimization
Dr. Patrick H. Diamond
Plasma turbulence & transport — Hannes Alfvén Prize
Dr. Philippe Lebrun
CERN — led the LHC cryogenics, the largest ever built
Dr. Jack J. Dongarra
LINPACK / TOP500 — HPC backbone of fusion simulation (Turing Award)
2010s
SPARC / CFS, SMRs & the materials frontier
Patrick Schweiger
Chief Engineer, Commonwealth Fusion Systems; Gen-IV at TerraPower
Dr. Robert J. Weggel
Consulted for CFS on large-scale HTS field coils
Dr. Nasr Ghoniem
Directed the UCLA Fusion Science Center — materials under radiation
Dr. Gary Was · Dr. Ahmed Hassanein
Radiation materials & plasma–material interactions (Michigan · Argonne)
Martin Owens
Directed the BWRX-300 SMR program (GE-Hitachi); LANL capital projects
Dr. Ray Sedwick
Direct energy conversion & bremsstrahlung recovery (Maryland / MIT)
2020s
Kronos — the two-machine architecture
Priyanca Ford & the founding team
Converge six decades of experience into a breeder-first design: HYPERION + the D–³He burner
The full advisory bench
Physics, magnets, materials, computation, safety & delivery
Programs are placed by era; roles and tenures vary. Current roles and dates are maintained on the company leadership page. This model does not assert an active advisory relationship beyond what that page states.
Open science

Publications & Talks

Every headline number in this model reproduces from open, citable deposits. Below are the live records (published, CC BY 4.0) and the peer-review track that is in preparation.

Published — open deposits (live, CC BY 4.0)
DepositDOILicense
Breeder — Hyperion (D–T spherical tokamak)10.5281/zenodo.21746157CC BY 4.0
Burner — D–³He tandem mirror10.5281/zenodo.21746479CC BY 4.0
REBCO magnet & tape system10.5281/zenodo.21842514CC BY 4.0
Direct Energy Conversion (DEC)10.5281/zenodo.21842864CC BY 4.0
AI + ML + Quantum reproducibility10.5281/zenodo.21842371CC BY 4.0
Peer review & venues — in preparation
IN PREPARATION
Preprint — arXiv (physics.plasm-ph)
Breeder and burner design papers, cross-linked from the Zenodo records.
TARGET
Journal — IOP Nuclear Fusion (or equivalent)
Full design & validation manuscript.
TARGET
Conference — IAEA Fusion Energy Conference
Two-machine architecture & breeder-first strategy.
The open deposits are the citable source of record now. Journal and conference versions are forward-looking — listed as targets, not as accepted publications. Where a venue is confirmed, its status will change here first.
Straight answers

Frequently Asked Questions

The honest answers — including the hard ones. Where something isn't proven, we say so.

The basics
What is the Kronos 3D Model?
An interactive, physics-grounded 3-D model of two fusion machines. You can rotate the reactor, inspect any component, run a real solver in your browser, and check every headline number against open datasets. It is a design-and-assurance tool — a twin-ready platform.
Is this a working fusion reactor?
No. Nothing is built yet. This is a validated design model. Each surface is tagged so you always know what it is: MODEL (specification), COMPUTES NOW (a real calculation), or AWAITS HARDWARE (a structure waiting for a real machine).
Is it a "digital twin"?
Not yet — an operational digital twin needs a physical machine feeding it live sensor data. This is the model, physics and assurance layer, built ready to receive that machine. The Future Capability page explains the path: 3-D model → digital shadow → digital twin.
For skeptics — the fast answers
"Fusion is always 30 years away."
Fair — which is why nothing here claims a working machine. This is a design-and-assurance model with every headline number reproducible from open data. Judge the evidence, not the promise.
"Is this aneutronic?"
No. Low-neutron, not aneutronic — about a 5.44% neutron fraction on the burner. We never claim zero neutrons.
"Does it actually make net power?"
The burner projects net +104 / +850 / +2832 MWe — but only if a plug-to-central density ratio of ~16 is achieved. That is a requirement, not a result, and it is labelled that way everywhere it appears.
"What's the catch?"
Three, stated openly: breeder confinement (H₉₈) and the ~90% divertor radiated fraction; burner plug density and helium-3 supply; the high-field plug magnet needs a full bore-resolved structural analysis — a gate, not a solved result.
"Why should I believe any of it?"
You don't have to. Five open Zenodo deposits (CC BY 4.0) plus a runbook let you regenerate every headline figure yourself; 182 / 182 integrity checks pass.
"Is this the sim you actually use internally?"
No — this is the public, browser-weight version. Kronos runs the same physics at far higher fidelity on internal HPC (full FEA / CFD / MCNP). See Methodology & limitations.
The two machines
What are the two machines?
A BREEDER (Hyperion) — a D–T spherical tokamak whose product is tritium, neutrons and helium-3 — and a BURNER (Aegis · MetroVolt) — a D–³He tandem mirror whose product is net electricity.
Why two machines instead of one?
Fuel follows purpose. The breeder concentrates the unavoidable radiological load at one isolatable, remotely-sited machine; the burner is low-neutron and clean enough to site at the point of use. One makes the fuel and isotopes; the other makes the power.
What are Aegis and MetroVolt?
The same burner in two housings — a fixed-site installation and a data-centre generator — across three central-cell lengths (55 / 440 / 1400 m) giving net +104 / +850 / +2832 MWe.
Is it real — and how would I know?
How do I know the numbers are real?
Every headline number reproduces from five open Zenodo datasets (CC BY 4.0). The Reproducibility page lists five pre-registered tests and exactly how to run them. You don’t have to trust us — you can regenerate the results.
What is actually computed vs. assumed?
It’s labelled throughout: FROZEN (from the deposit), SCREENING (a reduced-order model), or UNCOMPUTED (needs a full FEA / CFD / MCNP run). We name what we haven’t computed rather than estimate past the evidence.
Is the science peer-reviewed?
The work is released as open, reproducible deposits and manuscripts, and independent reproduction is invited. Treat every figure as auditable rather than certified — that honesty is deliberate.
The physics — the hard questions
Is the burner aneutronic?
No — it is low-neutron, not aneutronic. Residual D–D and D–T burn-up reactions give a neutron fraction of about 5.44%, roughly 10–15× cleaner than a mainstream D–T tokamak — but not zero. We never call it aneutronic.
What is the fusion gain?
The breeder reaches fusion gain Q ≈ 3.076 (fusion power ÷ auxiliary power) at its frozen point. The burner reaches an engineering Q_E ≈ 1.309 at its closing point.
Does the burner really make net electricity?
At the design point, yes — net +104 / +850 / +2832 MWe by housing length. Important caveat, stated openly: this is requirement-class — it depends on achieving a plug-to-central density ratio of 16. Below that it does not close, and we show that.
What is "negative triangularity (δ −0.30)"?
It describes the plasma’s cross-section — the "D" opens outward toward the wall rather than inward. It is the value the published equilibrium is built on, and the 3-D model shows it faithfully.
What are the biggest open risks?
Breeder: the divertor’s required ~90% radiated-power fraction. Burner: the plug-density requirement and helium-3 supply. Magnets: the high-field plug needs a bore-resolved structural analysis — currently flagged as a requirement-class gate, not a solved result.
Safety & environment
Can a fusion machine melt down?
No. There is no chain reaction and no meltdown path — the plasma simply stops if disturbed. Decay heat after shutdown is orders of magnitude below fission and is passively removable.
What about radioactive waste?
No high-level waste and no long-lived actinides. Activated structure is designed to reach low-level (Class C) waste; the low-neutron burner is cleaner still.
Is it radioactive at all?
The breeder handles tritium and activates its structure — addressed by multiple confinement barriers and low-activation materials. The burner is low-neutron by design, which is what makes point-of-use siting defensible.
Fuel, magnets & computing
Where does the helium-3 come from?
This is an open supply risk, stated as such. The breeder co-produces helium-3; longer-term sourcing (including lunar and tritium-decay routes) is discussed — as a risk to solve, not a solved problem.
What are the magnets made of?
REBCO high-temperature superconductor. The breeder’s peak on-conductor field is 16.84 T (within the 20.1 T demonstrated limit); the burner’s plug reaches 26.49 T — whose structure is flagged for a full bore-resolved analysis.
Does quantum computing give you an advantage?
Not this decade — and we say so. The deposit’s own verdict is "no crossover this decade." We ship a reproducible quantum proof-of-concept and an honest resource estimate; classical methods win today.
Open science & using this
Can I reproduce the results myself?
Yes. Five open deposits (code + data, CC BY 4.0) and a step-by-step runbook. The Reproducibility page shows the five acceptance tests and the environments that regenerate them.
How do I cite this?
Cite the relevant Zenodo deposit by DOI — all five are listed on the About & Cite page.
Can I download the data and code?
The full code and data are in the open Zenodo deposits. Downloads and exports inside this model itself are reserved for the account owner; the science is public on Zenodo.
The company & timeline
When will there be a real machine?
The roadmap runs physics freeze → component demonstration → a prototype around 2030 → a first product around 2036. These are targets tied to readiness gates, shown honestly on the Roadmap.
Who is Kronos Fusion Energy?
A fusion company developing the breeder + burner strategy, built on open, reproducible science. See About & Cite for how to reach us and how to reference the work.
Can't find your question? See About & Cite ↗ or reproduce any result via Reproducibility ↗.
Plain language

Glossary

The symbols and terms used across this model, in one place — so a non-specialist can follow every panel.

TermPlain-language meaning
δ — triangularityThe shape of the plasma cross-section. The Kronos breeder runs negative triangularity δ = −0.30 — the outer edge bows inward, which can improve edge stability.
Q — fusion gainFusion power out ÷ heating power in. Q = 1 is scientific breakeven. Hyperion computes Q ≈ 3.42.
Q_E — engineering gainNet electricity out ÷ all electricity the plant recirculates. Q_E > 1 means a net power source. The burner is ≈ 1.31 at the closing point.
TBR — tritium breeding ratioTritium atoms bred per tritium atom burned. TBR > 1 means the machine makes more fuel than it uses.
f_n — neutron fractionShare of fusion energy carried by neutrons. Lower is cleaner and easier to shield — the burner is ≈ 5.4%.
dpa — displacements per atomHow many times, on average, each atom in a material is knocked out of place by radiation — the standard measure of material damage and lifetime.
H₉₈ — confinement factorHow well the plasma holds heat versus the standard scaling law (H₉₈ = 1 is baseline). It is the breeder's single biggest uncertainty.
β — betaPlasma pressure ÷ magnetic pressure — how efficiently the field confines the plasma.
ST — spherical tokamakA compact, low-aspect-ratio tokamak — cored-apple shaped rather than a fat doughnut. The breeder architecture.
Tandem mirrorA linear machine that traps plasma between two magnetic "plugs." The burner architecture.
DEC — direct energy conversionTurning charged-particle energy straight into electricity, skipping a steam cycle — key to the low-neutron burner.
MW vs MWeMegawatts of fusion (thermal) power vs. MWe — megawatts of net electricity delivered.
Definitions are deliberately plain. The exact formulae live on the Governing Physics ↗ page.
The evidence base

Research library & datasets

Everything in this model traces back to a dataset or manuscript. This is the full open evidence base — physics scans, uncertainty studies, the validation suite, the magnet & tape program, direct-energy conversion, and the AI/ML/quantum reproducibility record.

Open datasets & what's inside
Breeder — Hyperion
10.5281/zenodo.21746157
0-D evaluator + 25,201-config design scan, operating window, 23-check validation suite. Powers the Physics Engine, Design Space, UQ and Validation pages.
Burner — D–³He tandem mirror
10.5281/zenodo.21746479
Fuel trajectory / closure window, power ledger, 45-check validation (H41–H80). Powers the burner physics, length/fuel interactivity and net-power figures.
REBCO magnet & tape
10.5281/zenodo.21842514
16-track magnet study + tape conductor: J_c(B,T) envelope, winding-pack λ-stack, bore-resolved stress, quench protection, neutron lifetime, fatigue, and the field ladder vs demonstrated coils. Powers the Materials pages.
Direct Energy Conversion (DEC)
10.5281/zenodo.21842864
Ion-only DEC stage model and efficiency basis for the burner — turning charged-particle energy into electricity without a steam cycle.
AI + ML + Quantum reproducibility
10.5281/zenodo.21842371
The reproducibility record for the AI/ML stack and a quantum proof-of-concept + honest resource estimate ("no crossover this decade").
Manuscripts
IN PREPARATION
Preprint — arXiv (physics.plasm-ph)
Breeder and burner design papers, cross-linked from the Zenodo records.
TARGET
Journal — IOP Nuclear Fusion (or equivalent)
Full design & validation manuscript.
TARGET
Conference — IAEA Fusion Energy Conference
Two-machine architecture & breeder-first strategy.
Public sees the catalogue and DOIs; the largest raw archives and internal working data unlock for signed-in reviewers. Everything published is CC BY 4.0.
Transparency

Changelog

What has changed in the model and the underlying record, and when. A maintained model is a trustworthy one.

Aug 2026
Design freeze
Publication design freeze established — the frozen point every panel reproduces (breeder Q 3.076 / 85.0 MW; burner Q_E 1.309).
Aug 2026
Five open deposits published
Breeder, burner, REBCO, DEC and AI/ML/Quantum records live on Zenodo (CC BY 4.0).
Aug 2026
Reproducibility verified
182 / 182 integrity checks pass; Tier-1 byte + Tier-2 tolerance reproduction.
Aug 2026
3-D model — public build
Interactive model with live in-browser breeder solver, maturity tags, open-risk dashboard, and public teaser + reviewer detail.
Ongoing
Peer-review track
arXiv → journal/IAEA versions in preparation; status updates land here first.
Physics changes are tied to a design freeze; the current baseline is the Aug 2026 publication freeze. Presentation and feature updates are listed separately.
KRONOS FUSION ENERGY · Kronos Model Conceptual design and simulation study; no machine has been built