(182d) Microsecond-Length Atomistic Simulations of Plasma-Facing Materials

Authors: 
Hammond, K. D., University of Missouri
Hu, L., University of Massachusetts Amherst
Maroudas, D., University of Massachusetts Amherst
Wirth, B. D., University of Tennessee, Knoxville

Helium from linear and tokamak-like plasma devices has been shown to cause "fuzz" to form on the surface of tungsten and other metallic surfaces after only a few hours of plasma exposure.  The mechanisms involved occur on multiple length and time scales, and understanding all of those details is the subject of significant multi-scale modeling.  This study examines the initial stages of fuzz formation by performing molecular dynamics simulations on both single-crystal and bicrystalline models of tungsten for times on the order of 0.5 to 2 microseconds.  These simulations employ state-of-the-art many-body potential energy models and implantation depth distributions constructed from implantation simulations of surfaces with different crystal orientations.  The large-scale atomistic simulations, which involve the simulation of up to 20 million atoms for up to 18 months or more, reveal the specifics of bubble formation and growth that drive surface evolution via the sequence of helium implantation, diffusion of helium to form bubbles, growth of bubbles and subsequent formation of tungsten self-interstitial atoms, organization of those self-interstitials into prismatic dislocation loops, annihilation of those prismatic loops at surfaces and grain boundaries, and finally, bubble rupture.  The orientation of the surface and the presence of grain boundaries and pre-existing dislocations have pronounced effects on the pattern and quantity of helium retained and the resulting surface features.  In particular, grain boundaries serve as sinks of helium, dramatically changing the distribution of bubbles in the vicinity of a grain boundary.  These results provide important information about the morphological evolution of plasma-facing surfaces and provide important benchmarks for models of nuclear fusion reactor materials.