By Shokirbek Shermukhamedov
In nuclear fusion, the main challenge is holding extremely hot plasma in a confined space. So, the walls of a fusion device should withstand considerable particle and heat loads. Therefore, plasma-wall interaction is one of the main problems of fusion research. This article shows the possibilities of neural network potentials for the study of such interactions. The effect of the impact energy and the angle of deuterium atoms on the sputtering of beryllium has been studied. We also calculated the relationship between these quantities and the outgoing and reflection angles of particles.
For more information: https://iopscience.iop.org/article/10.1088/1741-4326/ac044e