SEMPER FI – EMPIRE : Study of fission rate and fission density effects

Salvato Daniele

Promoter

Detavernier Christophe, (UGent), christophe.detavernier@ugent.be

SCK•CEN Mentor

Van den Berghe Sven
sven.van.den.berghe@sckcen.be
+32 14 33 24 35

SCK•CEN Co-mentor

Leenaers Ann
ann.leenaers@sckcen.be
+32 14 33 30 44

Expert group

Microstructural and Non-destructive Analysis

PhD started

2016-11-01

Short project description

For the development high-density, low-enriched research reactor (RR) fuels, uranium alloys (eg. U-Mo) or intermetallics (eg. U3Si2) are used. These are most often dispersed in an Al based matrix and sandwiched as a thin layer in an Al alloy cladding to form a fuel plate by hot rolling. These fuels are used up to very high burnups (>75% or fission densities >5E21 f/cc) and operated at fairly low temperatures (<250°C fuel temperature).  Since each fission causes 2 ions (the fission products) to be propelled at high energy (~80-100 MeV each) through the fuel and eventually become implanted in the fuel, a lot of damage cascades are created in the lattice, causing it to accumulate important amounts of deformation energy (high defect concentration) over time.  Since the temperature is too low for these defects to anneal out, the accumulated energy will eventually reach a level where athermal restructuring of the fuel is possible.  This effect, known from other fuel systems, is given different names, such as recrystallisation, grain refinement, High Burnup Structure (HBS) formation, etc. The phenomenon is called the RIM effect in power reactor fuels such as UO2 and MOX, as it occurs in the outer, colder part of a fuel pellet. In those fuels, it has been studied thoroughly and, although the exact mechanism behind it is still under debate, it is known to be associated to a subdivision of the fuel grains with formation of a large number of new grain boundaries. It affects the lattice parameter, hardness and porosity of the fuel. It is also associated to a precipitation (athermally) of fission products, particularly fission gas (Xe-Kr) and results in a swelling of the fuel. In RR fuel, such fuel swelling is directly translated in swelling of the fuel plates, since the fuel plates contain no open volume to accommodate the volume increase.  Fuel plate swelling is the technologically most important phenomenon, since the water cooling gaps between RR fuel plates in an assembly are very narrow.

It has been found that the qualification of the U7Mo fuels at a dispersal density of 8gU/cc in an Al-Si matrix is hampered by interaction phase formation between fuel and matrix and an associated marked acceleration of the plate (and thus fuel) swelling at fission densities above 4.5x1021 f/cc, eventually leading to plate failure in some cases. SCK•CEN has performed experiments with ZrN coated fuel particles in a pure Al matrix (SELENIUM project) that show a much improved behaviour in terms of fuel-matrix interaction.  However, a very similar swelling acceleration is still observed at the same fission density, even if it does not lead to plate failure in that case.  It is hypothesized that this acceleration may be related to a restructuring of the U7Mo and thus be intrinsic to this particular fuel with this microstructure. It is important to develop further understanding of the recrystallisation phenomenon and its associated effects, as well as identify the material properties that can influence the recrystallisation thresholds or its effects.

Objective

Within this PhD project, the goal is to separate out the phenomenon of recrystallisation and provide a scientific description.  All efforts to ‘simulate’ the effect of fission on the fuel matrix have important limitations, so to achieve a representative study, it is necessary to use irradiation.  Within this PhD, the ultimate goal is to irradiate small disks of U7Mo in the BR2 reactor at representative temperatures and up to appropriate fission densities.  Possibilities for such an irradiation have been verified, but the candidate will need to work out the details of the experiment.  Collaborations with PNNL, INL and ANL in the USA are possible for this.  By such an experiment, a systematic matrix of materials irradiated at different temperatures and up to different fission densities will be generated and the irradiated disks can be examined with a variety of techniques, particularly microhardness measurements, SEM-TEM for microstructure work, laser flash (collaboration with ITU) for thermal conductivity, X-ray diffraction for lattice parameter, etc. This will allow a deeper understanding of the mechanisms underlying the recrystallisation, help in defining thresholds for recrystallisation in terms of fission density, temperature and the influence of the microstructure on the recrystallisation effect.  In the end, this data will provide input for mechanistic modeling (not within the scope of this work) of fuel behaviour.

Preparative work can be accomplished by the irradiation of U7Mo disks with ion beams.  Although this only partially simulates the situation in pile and limits the effect to a thin layer on the surface of only a few µm thick, it is the closest possible simulation and a very useful way to provide important first indications.  Furthermore, there are important advantages since the samples do not get activated and temperature control is easier. Collaborations with the University of Munich or the Argonne National Lab in Chicago are immediately available for the ion irradiations, but possibilities also exist in France (GANIL).

The candidate will be involved in the preparation of the UMo discs needed for the irradiations (ion and neutron), the preparation and design of the in-pile experiment (based on existing technology), will be directly involved in the ion irradiation experiments and post-irradiation examinations of the ion-irradiated specimens. As part of an ongoing irradiation and PIE program that will enter the hot cells in 2015, the candidate will have the opportunity to study irradiated U7Mo fuel at different levels of recrystallisation. Although in that case different parameters are harder to separate, it will provide a good basis to maintain the technological relevance of the scientific approach with the disc irradiation. It can also provide backup materials in case the irradiation project suffers delays, which is always a risk in the nuclear field. Immediate opportunities for publication exist as the data is technologically and scientifically highly relevant.