First VULCANO test in SARNET2
In the hypothetical case of a nuclear reactor severe accident, the core could melt, form a magma called corium and interact with the containment concrete.
The 1st SARNET2 MCCI test (VB-ES-U2) was successfully performed in June in the CEA Cadarache VULCANO facility. It is
the first milestone of a joint program of activities on this issue with 17 European partners combining experimental
activities, modelling and numerical simulation.
Previous 2D tests within the VULCANO national program or an OECD project at ANL (USA) had shown a difference in the
ablation behaviours of siliceous or limestone concretes:
- Isotropic ablation of limestone concretes;
- Anisotropic ablation of siliceous concretes with a preferential ablation of lateral walls.
Six major differences can be found between these two types of concretes. In order to determine the cause of the
observed ablation behaviour, it has been decided to conduct more analytic tests in which the concrete are specially
designed to have intermediate properties between siliceous and limestone concretes. The VBES-U2 concrete has been
realized using clinker (calcinated calcium silicate used to produce cement) aggregates. The furnace has been loaded
with 90 kg of a mixture of depleted uranium, zirconium, iron, calcium and silicon oxides. Once melted, the corium
has been poured in a clinker-concrete hemicylindrical crucible. The radioactive decay heat has been simulated by
induction during 3 hours.
Scheme of the VULCANO facility
View of the corium surface after 2h30 of test during a corium flowover the pool surface
An anisotropic ablation, with a strong lateral one, has been observed in a similar way to silica-rich concretes. The
cause of anisotropy may be restricted to the three following effects: gas flow rate, gas injection topology,
concrete shrinkage. The interpretation of the on-line measurements, observations during the test section dismounting
and material analyses should precise the above-mentioned first conclusions. Two new tests are planned to determine
which of the three listed effects effectively causes these anisotropies. It is then expected to derive a predictive
modelling, which is necessary to transpose these results to the reactor scale.
View of the final test Section
Contacts: C. Journeau, L Ferry, P. Piluso