About this Abstract |
| Meeting |
Materials in Nuclear Energy Systems (MiNES) 2025
|
| Symposium
|
Materials in Nuclear Energy Systems (MiNES) 2025
|
| Presentation Title |
Use of Data Mining, Machine Learning and Computational Fluid Dynamics for the Fabrication of Coated Particle Fuel |
| Author(s) |
Eddie Lopez Honorato, Flavio Dal Forno Chuahy, Angel Diaz, Ryan Heldt, Bryan Conry |
| On-Site Speaker (Planned) |
Eddie Lopez Honorato |
| Abstract Scope |
Tristructural-isotropic (TRISO) and other coated particle fuels have been proposed for future microreactor designs for terrestrial use and space exploration. This expanded service envelope will require in some cases the use of coated particle fuels with different kernel sizes and compositions, as well as different coating layers, thicknesses and compositions. In this work we will discuss the use of data mining, machine learning and computational fluid dynamics to reduce the exploration space and accelerate the development of coated particle fuels. Data mining was able to generate pyrolytic carbon deposition maps to better understand the impact of deposition conditions such as particle loading, kernel size, kernel density, precursor concentration, surface area/volume ratio, etc. on the density and anisotropy of PyC. Furthermore, the database created was used to implement machine learning and generate a predictive model for the deposition of PyC. Furthermore, computational fluid dynamics was used to better understand fluidization behavior, coating rates and the impact of other variables such as kernel size and particle size distribution. |
| Proceedings Inclusion? |
Undecided |