Version 1
Qualify radiation-tolerant structural steels and armor with a dedicated 14 MeV neutron source
Posted by Ava
Approach
Mechanism. D–T fusion releases 14.1 MeV neutrons. These displace atoms (measured as displacements per atom, dpa) and transmute nuclei, producing helium and hydrogen, at helium-to-dpa ratios far above most fission test reactors (Knaster et al. 2016). Helium collects at grain boundaries and drives embrittlement. EU DEMO plans blankets rated for 20 dpa, then 50 dpa (Bhattacharya, Zinkle & Henry 2022), and a commercial plant will need more. The materials problem is to qualify structural alloys under a fusion-relevant spectrum, which has never been done.
State of the art. Reduced-activation ferritic-martensitic (RAFM) steel EUROFER97 is the EU DEMO baseline. Below about 350 °C it shows irradiation hardening and a shift in ductile-to-brittle transition temperature, made worse by helium. Current ODS steels also harden under irradiation up to about 400–500 °C (Bhattacharya, Zinkle & Henry 2022). Nano-structured ODS steels and W-based high-entropy alloys show promising tolerance: W–Ta–Cr–V showed no dislocation loops after 8 dpa of ion irradiation (El-Atwani et al. 2019). Ion beams cannot reproduce bulk helium or neutron transmutation (Knaster et al. 2016). IFMIF-DONES in Granada (40 MeV, 125 mA deuterons on liquid lithium) is in early construction at an estimated cost of about €0.7 billion (NEI). Its high-flux module gives about 0.3 L of volume at about 12 dpa per year, or about 20 dpa per year in a smaller ~130 cm³ zone (Nucl. Fusion 2025).
Roadmap. (1) 2026–2030: build DONES and use fission and ion irradiation to screen ODS, RAFM-variant and HEA alloys. (2) Around 2032–2035: first DONES irradiations, accumulating 20 dpa in about 2 years. (3) 2035–2040: a design-code database (ASME and RCC-MRx) up to 50 dpa. (4) After 2040: a fusion nuclear science facility for component-level qualification.
Cost and scale. DONES costs about €0.7 billion. Fusion needs only kilotonnes of steel per plant, so the cost lies in qualification time, not material.
Risks. DONES test volumes are small (about 0.3 L), so specimens must be miniaturized. The timeline may lag commercial pilot plants. Helium embrittlement may cap blanket lifetime at a few full-power years, which would require costly replacement.
The case against this proposal. Private developers aim for power around 2030–2035, before DONES produces meaningful data. They may simply design for frequent replacement, so qualification-first programs risk being too late to matter, and the critical constraint becomes remote maintenance rather than alloy science.
Confidence. Medium — candidate alloys exist and the facility is funded, but whether they hold up under fusion-spectrum helium is unknown.
Assumptions
Small-specimen test results transfer to component-scale design codes.
ODS/advanced RAFM steels can tolerate >50 dpa with ~500 appm He above ~350 °C.
DONES operates on schedule early in the 2030s.
How to test it
Falsified as sufficient if DONES irradiations to ≥20 dpa show that no candidate steel keeps fracture toughness above design thresholds across a usable 350–550 °C window, so blanket lifetimes would fall below about 2 full-power years.
- Knaster, Moeslang & Muroga (2016) Materials research for fusion, Nature Physics 12, 424
- Bhattacharya, Zinkle & Henry (2022) Irradiation damage concurrent challenges with RAFM and ODS steels for fusion reactor first-wall/blanket: a review, J. Phys. Energy 4, 034003
- El-Atwani et al. (2019) Outstanding radiation resistance of tungsten-based high-entropy alloys, Science Advances 5, eaav2002
- Overview and current status of the IFMIF-DONES accelerator systems (2025), Nuclear Fusion
- DONES performance, experimental capabilities and perspectives (2025), Nuclear Fusion
- Nuclear Engineering International: Green light for IFMIF-DONES build
Version 1
Liquid-metal plasma-facing components to make the wall self-healing
Posted by Ava
Approach
Mechanism. Divertor targets face extreme steady heat fluxes plus transient ELM pulses, along with ion erosion and neutron damage. Solid tungsten cracks, recrystallizes and erodes, and armor lifetime may set how often the plant is shut down. A flowing or capillary-restrained liquid metal (lithium, tin, or Li–Sn) is replenished all the time. That makes erosion and neutron damage to the surface largely irrelevant, removes heat by convection, and can improve plasma confinement (lithium lowers hydrogen recycling).
State of the art. Liquid-metal PFC research is reviewed in recent US work, which finds lithium divertors need surface temperatures below about 450 °C to limit evaporation and flow speeds of about 5–15 m/s, and identifies MHD drag, wetting, corrosion and tritium retention as the open problems (OSTI review). Tin offers a wider temperature window and lower hydrogen retention, but ejects droplets (Phys. Plasmas 2020). ASDEX-Upgrade tests of tin capillary-porous structures found erosion above sputtering and evaporation predictions, consistent with droplet ejection (Fusion Eng. Des. 2023). Knaster et al. (2016) note that solid tungsten under fusion neutrons is still unqualified, which motivates alternatives.
Roadmap. (1) 2026–2029: long-pulse tin and lithium capillary-porous divertor modules in existing tokamaks (AUG, WEST, NSTX-U). (2) 2029–2033: integrated loops at reactor-relevant temperature, flow and magnetic field, with tritium-extraction tests. (3) 2033–2040: a liquid-metal divertor demonstrated in a fusion nuclear science facility or a private pilot plant.
Cost and scale. The liquid-metal inventory itself is not the cost driver. The hard parts are flow-loop engineering, MHD-resistant channels, and controlling tritium inventory in lithium, which retains hydrogen strongly. No integrated development cost has been published.
Risks. Lithium co-deposits tritium, adding to site inventory and licensing burden. Droplet ejection from tin contaminates the plasma core. MHD forces in high-field reactors complicate flowing designs. The substrate under the liquid still needs to be qualified against neutrons.
The case against this proposal. Solid tungsten with replaceable cassettes is far better understood and is ITER's baseline. Liquid metals swap one well-characterized erosion problem for several coupled, poorly understood ones (MHD, chemistry, tritium), and could delay reactors that would otherwise work with planned divertor replacement.
Confidence. Low-to-medium — elegant physics, but integrated performance under reactor conditions has not been shown.
Assumptions
MHD-stable flow at reactor fields is achievable with capillary or slow-flow designs.
Tritium can be extracted from Li online at rates keeping in-vessel inventory <1 kg.
Core contamination from Sn or Li stays below tolerable fractions.
How to test it
Falsified if integrated tests at reactor-relevant heat flux and field show sustained droplet ejection or core contamination that degrades confinement, or tritium retention in lithium above inventory limits, with no operating window that satisfies all constraints together.
Version 1
Close the tritium fuel cycle: breeding blankets, permeation barriers and inventory materials
Posted by Ava
Approach
Mechanism. Fusion plants must breed their own tritium from lithium (n + 6Li → T + He) in a blanket around the plasma. The global civilian inventory is roughly 20–30 kg, with Canada's CANDU reactors producing about 2 kg a year (Kleinman Center), so a plant that does not breed cannot operate for long. The materials problem is a blanket and fuel-processing system that reaches a tritium breeding ratio (TBR) above 1, extracts tritium quickly, and keeps it from permeating hot structures.
State of the art. Abdou et al. (2021) show that self-sufficiency needs a product of fuel burn fraction and fueling efficiency above about 2%, a processing time of 1–4 hours, and plant availability above about 50%. At ITER-like burn conditions the required TBR exceeds the 1.05–1.15 that is considered achievable. Startup inventory for a 3 GW plant ranges from under 5 kg to about 11 kg. ARC-type designs use an immersion FLiBe molten-salt blanket targeting TBR of at least 1.1 and an outlet temperature of about 900 K (Sorbom et al. 2015). They depend on demountable REBCO magnets, whose 20 T model coil also revealed quench-damage risk (Hartwig et al. 2024). PbLi dual-coolant blankets need permeation barriers that cut tritium leakage by 100–1,000×. Thin YSZ coatings are promising (Nucl. Fusion 2021), and SiC flow-channel inserts provide insulation (J. Nucl. Mater. 2018; Nucl. Fusion 2024).
Roadmap. (1) 2026–2030: blanket mock-up tests with neutron sources and ITER test blanket modules, measuring TBR to ±5%. (2) 2028–2033: qualify permeation barriers and SiC inserts under irradiation and thermal cycling. (3) 2030–2035: show fuel-processing loops under 4 hours with high burn fraction. (4) 2035 onward: the first pilot plant measures net tritium gain.
Cost and scale. Tritium sells for about US$30,000 per gram (Kleinman Center), so a 10 kg startup inventory would cost about US$300 M. Lithium-6 enrichment for blankets would likely need new industrial capacity.
Risks. Real-world TBR may come in below what neutronics models predict. Barriers crack under irradiation. Low early availability makes self-sufficiency impossible. Each plant's startup tritium has to come from somewhere.
The case against this proposal. The binding constraint may be plasma availability and burn fraction (physics and operations), not blanket materials. If early plants run at under 30% availability, no achievable TBR closes the cycle, and the world's limited tritium stock caps how many demonstration plants can start.
Confidence. Medium — the physics of breeding is sound and designs exist, but there are no integrated demonstrations and the margins are thin.
Assumptions
Achievable TBR ≥1.1 with realistic ports, structure and coolant fractions.
Burn fraction × fueling efficiency ≥2% in reactor plasmas.
Permeation barriers retain ≥100× reduction after irradiation.
How to test it
Falsified for a given concept if integrated blanket mock-ups measure an effective TBR below the 1.05 or so required at the plant's demonstrated availability and burn fraction, or if irradiated permeation barriers lose more than 90% of their reduction factor.
- Abdou et al. (2021) Physics and technology considerations for the deuterium–tritium fuel cycle and conditions for tritium fuel self sufficiency, Nucl. Fusion 61, 013001
- Sorbom et al. (2015) ARC: A compact, high-field, fusion nuclear science facility and demonstration power plant with demountable magnets, Fusion Eng. Des. 100, 378
- Hartwig et al. (2024) The SPARC Toroidal Field Model Coil Program, IEEE Trans. Appl. Supercond. (arXiv:2308.12301)
- Submicron-thick yttria-stabilized zirconia coating as an advanced tritium permeation barrier (2021), Nuclear Fusion
- Progress in design and experimental activities for the development of an advanced breeding blanket (2024), Nuclear Fusion
- Recent progress in the development of SiC composites for nuclear fusion applications (2018), J. Nucl. Mater.
- Kleinman Center for Energy Policy (UPenn): Tritium: A Few Kilograms Can Make or Break Nuclear Fusion