| Abstract Scope |
In this presentation, I will discuss an integrated theoretical, computational, and experimental approach to understanding and tuning defect diffusion kinetics for the design of damage-resistant materials for nuclear energy applications. First, I introduce a neural network kinetics framework grounded on an efficient on-lattice structure and chemistry representation, enabling precise prediction and modeling of atomic diffusion and structural evolution in compositionally complex materials. Next, I will present how vacancy defects can be controlled and trapped by chemical compositions in tungsten alloys, with irradiation experiments proving the concept that vacancy cluster growth is effectively suppressed by this trapping effect. Finally, I will present new insights into collective dislocation motion and microscopic evolution in advanced structural materials under high stress, revealing fundamental behaviors that challenge classical models of material deformation. |