Over the last decade, there has been a rapidly expanding interest in applying Monte Carlo simulation to reactor physics problems at the scale of the whole core, encompassing stationary configurations, fuel depletion, and kinetics, most often including thermal-hydraulics and thermo-mechanics feedback. Such multi-physics simulations typically involve a large number of possibly time-dependent material compositions, as well as spatially continuous temperature, density, and nuclide concentration fields. These applications pose distinct challenges in terms of massive parallelism, memory footprint, and use of heterogeneous, quickly evolving computing architectures (including traditional processors, CPUs, as well as graphical processing units, GPUs). In order to cope with these challenges, several Monte Carlo codes have been either extensively modified or built from scratch.

​In this context, CEA, IRSN and EDF have joined forces in 2022 and started the development of TRIPOLI-5®, a new Monte Carlo code whose primary goal in the short term is to address reactor physics problems, integrating multi-physics feedback for stationary and non-stationary configurations. TRIPOLI-5 builds upon the experience accumulated with PATMOS [1], a Monte Carlo mini-app developed at CEA to explore the portability of particle-transport algorithms in High-Performance Computing (HPC) environments, including hybrid CPU/GPU architectures. A pilot version of TRIPOLI-5 has been released to beta-testers at the end of 2024, as a stepping stone towards the first official release covering stationary reactor problems [2-5].

For an overview of TRIPOLI-5, see: D. Mancusi et al., Overview of TRIPOLI-5, a Monte Carlo code for HPC, EPJ Nuclear Sci. Technol., 10 (2024) 26.

See also: TRIPOLI-5 website.

In the short term, the main focus of the development efforts for TRIPOLI-5 will be, on the one hand, on multi-physics feedback, depletion calculations, and reactor kinetics. The target architectures will be hybrid, massively parallel CPU/GPU machines, as well as office workstations for more modest workloads. In the long run, the development of TRIPOLI-5 will be eventually extended to cover a broader range of applications, encompassing nuclear instrumentation and radiation shielding with variance reduction. It is envisaged that TRIPOLI-5 will supersede the current-generation Monte Carlo codes TRIPOLI-4 and MORET6 at that stage.​

Testing PATMOS performances for the Hoogenboom-Martin benchmark at the TGCC supercomputer.

References

[1] E. Brun, S. Chauveau, F. Malvagi, PATMOS: A prototype Monte Carlo transport code to test high performance architectures, in Proc. M&C 2017, Jeju, Korea, April 16-20, 2017.

[2] D. Mancusi et al., Overview of TRIPOLI-5, a Monte Carlo code for HPC, EPJ Nucl. Sci. & Technol. 10, 26 (2024).​

[3] C. Larmier et al., Preliminary Investigation Of Nuclear Data Sampling For The New Monte Carlo Code Tripoli-5®, in Proc. M&C23, Niagara Falls, Canada, 13-17/8 (2023).​

[4] C. Montecchio, C. Larmier, D. Mancusi, A. Zoia, Benchmarking of probability tables with TRIPOLI-5, EPJ Web of Conferences 302, 07005 (2024).​

[5] C. Larmier, D. Mancusi, V. di Blasi, A. Zoia, Verification Methods for the AGORA Geometry Navigation Engine of the TRIPOLI-5 Monte Carlo Code, in Proc. M&C2025, Denver, USA (2025).​​