Accurately modeling the coupled propagation of electron and photon in matter is crucial in several of CEA's flagship programs, notably radiation shielding and nuclear instrumentation. Such transport processes involve a very large number of interactions during the particle histories, which typically requires time-consuming Monte Carlo simulations. These features call for the use of High-Performance Computing (HPC), which allows to obtain statistically meaningful results in reasonable times even for such complex processes. The validation of simulation codes requires dedicated parametric studies and experimental measurements. Given the scarcity of publicly available experimental data, comparisons between simulation codes are a common  practice.

The TRIPOLI-4 Monte Carlo code, developed at CEA and largely used in reactor physics and radiation shielding, aims at extending its scope to nuclear instrumentation, and to improve its computational performance in an HPC context. Some research units of CEA also use the MCNP and GEANT4 codes for radiation shielding and nuclear instrumentation applications. The challenge in the coming years is to qualify these codes in a broad energy domain. As part of this effort, preliminary investigations carried out at CEA revealed unexpected discrepancies between their results, in particular in scenarios involving the coupled transport of neutrons, photons and electrons.

The VALERIAN project is designed to address the observed discrepancies by taking advantage of a unique data collection campaign planned at CEA between 2025 and 2027.

CEA has committed to perform the quality control of at least 700 pixel modules for the new inner tracker of the ATLAS experiment, which is part of the major upgrade on-going at the CERN Large Hadron Collider (LHC). Numerous measurements of crossing particle fluxes will be carried out with beta sources in 2025-2027 for the qualification of the modules. The tests will be carried out at Saclay on a bench developed at CEA, ensuring the full control of the experimental parameters. The small pixel pitch (50 μm) and the high performance of the ITkPix readout chips integrated into the modules will enable near state-of-the-art accuracy measurements.

The VALERIAN project aims at developing numerical twins of the IRFU test bench with TRIPOLI-4, MCNP and GEANT4, and to validate them using experimental data. The members of the project will also contribute to the daily operations and optimization tasks of the IRFU test bench.

PhD in one of the following fields: numerical simulation, nuclear physics or neutronics.