This article presents an overview of the graphical user interfaces (GUIs) developed at CEA/SERMA (Service d’Études des Réacteurs et de Mathématiques Appliquées) in Saclay, France, which have been used for over forty years by engineers and scientists to build geometries and meshes for general-purpose lattice transport calculations (neutrons and photons). Several applications make use of these calculations, from fuel assembly to full core design, criticality and safety, needing consistency check of the geometry and input properties before starting any lattice calculation. The software pattern design of the GUIs is briefly discussed, showing also the rationale behind the two interfaces for the construction of the geometries for simple fuel assemblies and complex motifs including the reflector (colorsets). The new GUI, ALAMOS, specifically developed for APOLLO3® with a Python Application Programming Interface (API), is here presented as the successor of Silène, which was the first GUI released in the 1990s to serve APOLLO2 calculations. The considerable experience gained by Silène over the years with plenty of various applications has provided a crucial support for the development of ALAMOS.
This work deals with the modeling of homogenized few-group cross sections by Artificial Neural Networks (ANN). A comprehensive sensitivity study on data normalization, network architectures and training hyper-parameters specifically for Deep and Shallow Feed Forward ANN is presented. The optimal models in terms of reduction in the library size and training time are compared to multi-linear interpolation on a Cartesian grid. The use case is provided by the OECD-NEA Burn-up Credit Criticality Benchmark [1]. The Pytorch [2] machine learning framework is used.
Since the 80’s, industrial core calculations employ the two-step scheme based on prior cross sections preparation with few energy groups and in homogenized reference geometries. Spatial homogenization in the fuel assembly quarters is the most frequent calculation option nowadays, relying on efficient nodal solvers using a coarse mesh. Pin-wise reaction rates are then reconstructed by dehomogenization techniques. The future trend of core calculations is moving however toward pin-by-pin explicit representations, where few-group cross sections are homogenized in the single pins at many physical conditions and many nuclides are selected for the simplified depletion chains. The resulting data model requires a considerable memory occupation on disk-files and the time needed to evaluate all data exceeds the limits for practical feasibility of multi-physics reactor calculations. In this work, we study the compression of pin-by-pin homogenized cross sections by the Hotelling transform in typical PWR fuel assemblies. The reconstruction of these quantities at different physical states of the assembly is then addressed by interpolation of only a few compressed coefficients, instead of interpolating separately each homogenized cross section. Savings in memory higher than 90% are observed, which result in important gains in runtime when interpolating the few-group data.
23 November 2020 in Journal of Computational and Theoretical Transport
In this work, we apply the Ronen method to obtain highly-accurate approximations to the solution of the neutron transport equation in simple homogeneous problems. Slab, cylindrical, and spherical geometries are studied. This method demands successive resolutions of the diffusion equation, where the local diffusion constants are modified in order to reproduce new estimates of the currents by a transport operator. The diffusion solver employs here finite differences and the transport-corrected currents are forced in the numerical scheme by means of drift terms, like in the CMFD scheme. Boundary conditions are discussed introducing proper approximations to save the particle balance in case of reflection in the slab. The solution from the Ronen iterations is compared against reference results provided by the collision probability method. More accurate estimates of the currents are provided by integral transport equations using first flight escape probabilities. Slow convergence on the scalar flux is analyzed, although the results match the reference solutions in the limit of fine meshes and far from the bare boundary.