Keywords: fusion energy, tritium breeding, lithium blanket, hydrogen permeation, digital twin, artificial intelligence, experimental facility, low code architecture, simulation, hydrogen economy.

Significant international efforts are underway to advance fusion technology from scientific demonstration to a practical source of electricity. One of the key challenges is the fuel cycle, in particular the supply of deuterium–tritium fuel. While deuterium is readily obtained from seawater, tritium is scarce in nature and cannot sustain a power plant without on-site production. The accepted strategy is to breed tritium inside the reactor using lithium-based breeding blankets, which generate tritium when exposed to high-energy neutrons from the fusion reaction. Engineers are designing and testing mock-ups of these systems to demonstrate reliable tritium generation, storage, and handling. To support this effort, a new high-power neutron test facility, LIBRTI, is being established to evaluate candidate tritium breeding technologies. Our contribution is to develop a digital twin of LIBRTI, enabling experimental and simulation data to be integrated and used to train AI models that accelerate design and evaluation. We present preliminary work on a low-code digital twin architecture that leverages Nvidia Omniverse and open-source tools. As a proof of concept, we demonstrate a prototype digital twin of a gas-driven permeation system. This work will interest the fusion community and engineers engaged in hydrogen technologies.

download the full-text of this paper (PDF, 1 pages, 813 Kb)

go to the previous paper
go to the next paper
return to the table of contents
return to the volume description