[en] The Phebus FP project in an international reactor safety project. Its main objective is to study the release, transport and retention of fission products in a severe accident of a Light Water Reactor (LWR). The FPT4 test was performed with a fuel debris bed geometry, to look at late phase core degradation and the releases of low volatile fission products and actinides. Post Test Analyses results indicate that releases of noble gases (Xe, Kr) and high-volatile fission products (Cs, I) were nearly complete and comparable to those obtained during Phebus tests performed with a fuel bundle geometry (FPT1, FPT2). Volatile fission products such as Mo, Te, Rb, Sb were released significantly as in previous tests. Ba integral release was greater than that observed during FPT1. Release of Ru was comparable to that observed during FPT1 and FPT2. As in other Phebus tests, the Ru distribution suggests Ru volatilization followed by fast redeposition in the fuelled section. The similar release fraction for all lanthanides and fuel elements suggests the released fuel particles deposited onto the plenum surfaces. A blockage by molten material induced a steam by-pass which may explain some of the low releases. The revaporisation testing under different atmospheres (pure steam, H₂/N₂ and steam /H₂) and up to 1000 .deg. C was performed on samples from the first upper plenum. These showed high releases of Cs for all the atmospheres tested. However, different kinetics of revaporisation were observed depending on the gas composition and temperature. Besides Cs, significant revaporisations of other elements were observed: e.g. Ag under reducing conditions, Cd and Sn in steam-containing atmospheres. Revaporisation of small amounts of fuel was also observed in pure steam atmosphere

[en] The ternary oxide ceramic system UO₂-ZrO₂-FeO is a refractory system that is of great relevance to the nuclear industry as it represents one of the main systems resulting from the interaction of the Zircaloy cladding, the UO₂ fuel and the structural elements of a nuclear reactor. It is particularly the high temperature properties that require investigation; that is, when substantial overheating of the nuclear core occurs and interactions can lead to its degradation, melting and result in a severe nuclear accident. There has been much work on the UO₂-ZrO₂ system and also on the ternary system with FeO but there is still a need to examine 2 further aspects; firstly the effect of sub-oxidized systems, the UO₂-Zr and FeO-Zr systems, and secondly the effect of Fe/Zr or Fe/U ratios on the melting point of the U-Zr-Fe oxide system. Samples of UO₂-Zr and UO₂-ZrO₂-FeO were fabricated at ITU and then characterized by optical microscopy (OM) and X-ray diffraction to determine the ceramic's structure and verify the composition. Thereafter the samples are to be melted by laser flash heating and their liquidus and solidus temperatures determined by pyrometry. This programme is currently ongoing. The frozen samples, after testing, were then sectioned, polished and the molten zone micro-analytically examined by OM and SEM-EDS in order to determine its structure and composition and to compare with the existing phase diagrams. Examples of results from these systems will be given. Finally, a reacted Zr-FeO thermite mixture was examined, which had been used to generate high temperatures during tests of reactor melt-concrete interactions. The aim was to assess the reaction and estimate the heat generation from this novel technique. These results allow verification or improvement of the phase diagram and are of primary importance as input to models used to predict materials interactions in a severe nuclear accident.