TY - CHAP
T1 - Modeling radiation-induced segregation and precipitation
T2 - contributions and future perspectives from artificial neural networks
AU - Castin, Nicolas
AU - Malerba, Lorenzo
N1 - Score=10
PY - 2020/3
Y1 - 2020/3
N2 - Radiation-induced segregation and precipitation is one of the main responsibles for the changes in macroscopic properties taking place in steels under irradiation. For instance, reactor pressure vessel steels and ferritic-martensitic steels are known to harden and embrittle within the lifetime of the reactors. The development of quantitative predictive models for these changes is a very challenging task, because of the highly multiscale nature of the mechanisms taking place at the microstructural level and the chemical complexities involved. Fully physically based approaches such as object kinetic Monte Carlo models are promising tools to take this challenge, but their parametrization must be accurately and adequately elaborated. In this chapter, we revise how highly powerful and flexible numerical tools offered by machine learning systems, specifically artificial neural networks, have contributed to these models. Perspectives for future developments are also discussed,
AB - Radiation-induced segregation and precipitation is one of the main responsibles for the changes in macroscopic properties taking place in steels under irradiation. For instance, reactor pressure vessel steels and ferritic-martensitic steels are known to harden and embrittle within the lifetime of the reactors. The development of quantitative predictive models for these changes is a very challenging task, because of the highly multiscale nature of the mechanisms taking place at the microstructural level and the chemical complexities involved. Fully physically based approaches such as object kinetic Monte Carlo models are promising tools to take this challenge, but their parametrization must be accurately and adequately elaborated. In this chapter, we revise how highly powerful and flexible numerical tools offered by machine learning systems, specifically artificial neural networks, have contributed to these models. Perspectives for future developments are also discussed,
KW - Modelling
KW - Artificial Intelligence
KW - Kinetic Monte Carlo
KW - Atomic simulation
UR - http://ecm.sckcen.be/OTCS/llisapi.dll/open/31626701
U2 - 10.1007/978-3-319-44680-6_140
DO - 10.1007/978-3-319-44680-6_140
M3 - Chapter
SP - 2517
EP - 2538
BT - Handbook of Materials Modeling
PB - Springer
CY - Switzerland
ER -