HTR fuel modelling with the ATLAS code. Thermal mechanical behaviour and fission product release assessment

To support AREVA NP in its design on HTR reactor and its HTR fuel R and D program, the Commissariat a l'Energie Atomique developed the ATLAS code (Advanced Thermal mechanicaL Analysis Software) with the objectives: - to quantify, with a statistical approach, the failed particle fraction and fission product release of a HTR fuel core under normal and accidental conditions (compact or pebble design). - to simulate irradiation tests or benchmark in order to compare measurements or others code results with ATLAS evaluation. These two objectives aim at qualifying the code in order to predict fuel behaviour and to design fuel according to core performance and safety requirements. A statistical calculation uses numerous deterministic calculations. The finite element method is used for these deterministic calculations, in order to be able to choose among three types of meshes, depending on what must be simulated: - One-dimensional calculation of one single particle, for intact particles or particles with fully debonded layers. - Two-dimensional calculations of one single particle, in the case of particles which are cracked, partially debonded or shaped in various ways. - Three-dimensional calculations of a whole compact slice, in order to simulate the interactions between the particles, the thermal gradient and the transport of fission products up to the coolant. - Some calculations of a whole pebble, using homogenization methods are being studied. The temperatures, displacements, stresses, strains and fission product concentrations are calculated on each mesh of the model. Statistical calculations are done using these results, taking into account ceramic failure mode, but also fabrication tolerances and material property uncertainties, variations of the loads (fluence, temperature, burn-up) and core data parameters. The statistical method used in ATLAS is the importance sampling. The model of migration of long-lived fission products in the coated particle and more generally in the fuel element (pebble or compact) aims at estimating the source term of the fission products release outside the fuel element in normal operation or accidental conditions. In ATLAS, the transport mechanisms are modelled in a single transport law using effective diffusion coefficients for the fission product species in the different constitutive materials. The Verification and Validation of ATLAS code rests on two main steps: - Testing plan on fuel particle thermal mechanical behaviour has been carried out regarding sensitivity on dimensional parameters and physical properties such as kernel diameter, density and layer thicknesses and pyrocarbon layer anisotropy. The obtained results allow justifying and specifying the design for the manufacture. - Regarding fission product release under core heat-up accident conditions, the IAEA Coordinated Research Project 6 on 'Advanced in HTGR Fuel Technology Development' benchmark is the basis of the ATLAS code verification step. The ATLAS results obtained on IAEA benchmark cases with analytical solutions demonstrate that the models used fit the physical, chemical and mathematical laws. Regarding past irradiation tests and heating tests, ATLAS results show good agreement with the experimental database measurements. Comparison between ATLAS code results with analytical and experimental data allows defining confidence zones where ATLAS code gives accurate results and critical limits. These limits show where R and D efforts on models and material properties are needed to refine laws and models. (author)