In fusion reactors, a significant number of neutrons are generated, creating a harsh environment for reactor components. Testing sensitive devices, such as diagnostics and electronics is a key aspect to ensure proper and reliable operations in present and future tokamaks. To address this issue, the development of a dedicated facility is proposed: the GENeuSIS (General Experimental Neutron Systems Irradiation Station) project. GENeuSIS is a novel methodology designed to study and characterize the response of diagnostics, electronics, and other critical components of ITER, when exposed to the FNG (“Frascati Neutron Generator”) 14 MeV neutrons.
The GENeuSIS layout [1] consists of a layered structure made of moderating materials aimed at reproducing the expected neutron and gamma spectra in specific locations of the ITER machine under DT neutron irradiation.
Within this framework, a machine learning model helps automate the process of selection of the best materials and the configuration of assembly layouts to accurately reproduce the desired radiation environment. This work focuses on developing a supervised machine learning model (a neural network), that leverages a database generated from previous three-dimensional calculations of neutron and photon transport made using the Monte Carlo MCNP transport code. These simulations demonstrated the feasibility of GENeuSIS and its reliability in replicating the neutron spectrum in the ITER tokamak Port Interspace (GENeuSIS-I assembly) and the Port Cell (GENeuSIS-II assembly).
The machine learning model aims to streamline the pre-analysis phase and automatically determine the optimal combination of materials to replicate various neutron and gamma energy spectra.
This preliminary study presents the application of this new machine learning methodology to GENeuSIS, focusing first on reproducing fusion spectra given the different materials' configuration. The next step is to determine the best materials' configuration to replicate the ITER-relevant radiation field, given a chosen spectrum.
扫码关注我们
求助内容:
应助结果提醒方式:
