Case study #4 (all three tools, aspirational, not yet built): ODS FeCrAl cladding design for LBE-cooled fast reactor with integrated coating and structural qualification

A Gen IV developer will want to evaluate ODS FeCrAl variants alongside conventional FeCrAl candidates for the same LBE cladding application — same 500 to 550°C service, same oxygen-controlled LBE coolant, same experimental exposure programme — because ODS variants are expected to outperform conventional alloys under neutron irradiation by suppressing void swelling and retaining creep strength at dose. The question is whether the ODS processing route can be specified tightly enough that the predicted property advantages are actually realised in the fabricated coupons, and whether the integrated system — ODS substrate plus LBE environment plus potential coating — survives the full service envelope better than the conventional alternative.

The HEA tool would run first on both ODS and conventional FeCrAl composition space simultaneously, generating the Pareto front across creep strength, oxidation resistance, and cost as before. The ODS candidates would enter with an additional processing parameter vector — Y2O3 dispersion size and number density, which the FM tool’s ODS submodule expects as inputs — and the HEA tool flags these compositions with elevated Mahalanobis distance because the oxide dispersion microstructure is not captured in its standard composition-property relationships. That flag would be expected and appropriate: the HEA tool hands off to the FM ODS submodule precisely because the strengthening mechanism is dispersion-controlled rather than γ’-controlled and requires the separate physics.

The FM steel tool’s ODS submodule would then run on the shortlisted ODS candidates. It takes the composition, the Y2O3 dispersion parameters, and the service conditions and returns creep life with the dispersion strengthening contribution separated from the solid solution contribution this decomposition is what would make the submodule valuable, telling the developer whether a shortfall in predicted creep life is recoverable by tightening the dispersion processing window or requires a composition change. It would also run the radiation damage assessment: ODS alloys suppress void swelling through oxide-interface sink density, and the submodule’s sink strength calculation should give a dose-dependent remaining life estimate that conventional FeCrAl cannot provide. HAZ risk scoring would be flagged as high-uncertainty for ODS weld joints because solid-state joining is typically required — fusion welding destroys the dispersion — and the submodule should say so explicitly rather than returning a conventional HAZ score that would be physically misleading.

Austenitic would then run the LBE corrosion module on the ODS candidates using the same logic as Case Study 2. The key difference is that the oxide dispersion modifies the effective Al activity at the surface and therefore the alumina-forming kinetics — the protection factor for ODS FeCrAl in LBE is not identical to conventional FeCrAl at the same bulk composition, and if Austenitic’s LBE module has a dispersion correction term this is where it earns its value. If it does not, that interface is the honest gap to flag: the corrosion lifetime estimate for ODS candidates carries additional uncertainty beyond the standard LBE module’s acknowledged order-of-magnitude band, because the surface oxide formation kinetics are dispersion-modified and unanchored.

The integrated output would rank ODS against conventional FeCrAl candidates on a joint probability of surviving the full service envelope, decomposed by failure mode. The expected result — and the blind prediction to lodge before experiments — would be that ODS variants show superior radiation-dosedependent remaining life and comparable or better corrosion resistance if the dispersion is wellcontrolled, but carry wider joint probability uncertainty because the processing-property relationship introduces an additional uncertainty source that neither the HEA tool nor Austenitic currently models. That honest decomposition is the deliverable: not a recommendation to use ODS, but a quantified statement of which failure mode ODS solves, which uncertainty it adds, and exactly what the experimental coupon programme needs to measure to resolve that uncertainty dispersion stability post-irradiation and post-LBE exposure, characterised by TEM on the postexposure cross-section, mapped against the submodule’s sink strength prediction.

The strategic point: the ODS submodule should not just add one more alloy class to the FM tool. It would make the entire pipeline an integrated computational framework that can compare ODS and non-ODS candidates on a consistent uncertainty-quantified basis for the same service environment. That comparison is exactly what a Gen IV developer needs when deciding whether the manufacturing complexity of ODS is justified by the performance margin — and right now they are making that decision with experimental data alone, slowly and expensively.