Speaker
Description
The LIBRTI program seeks to de-risk fuel-cycle technology through physical demonstrations of specific breeder concepts, and accompany this with a digital representation of the facility to enhance the understanding of any measurements obtained. Such endeavours provide a route to in-silico design and qualification of breeder blanket technologies, thereby accelerating the critical pathway to commercializable fusion energy. In this contribution we review the ongoing effort for LIBRTI Integrated Modelling work-stream in support of this mission.
The primary function is to provide the core multi-physics modelling capability, and integrate this within the broader digital ecosytem. Since underlying the activity is the intention to develop scientific insights and capture complex emergent phenomena, the approach presumes a high level of fidelity and hence scalability is necessary. As such, having already been proven on high-performance computing (HPC) systems, the Multi-physics Object-Oriented Simulation Environment (MOOSE) software has been selected to provide our base functionality. Supplementing existing physics domains (which spans neutronics, fluid dynamics, heat transfer and tritium transport), we have developed new interfaces to FISPACT-II and FLUKA which enable inclusion of activation and charged particle transport respectively. In parallel, we have modelled molten salt, solid and liquid metal experimental breeder concepts in order to demonstrate current capabilities.
Moving forwards, our focus will shift towards validation and uncertainty quantification (UQ), alongside reproducibility, automation, and connectivity with other digital systems. The current strategy to assess model confidence against experimental data, as well as the intention to leverage workflow orchestration software, is outlined. Finally, to provide an example of how such an integrated approach may in future be deployed to accelerate design, we describe a recent effort employing sequential learning to optimise the LIBRTI solid breeder experimental mock-up concept.
| Speaker affiliation | UKAEA |
|---|
