Computational statistics of segregation and dislocation activities of hydrogen charged free surfaces and grain boundaries

Abstract

Revealing statistics of H-defect interactions provides insights into significant ductility loss due to the particular strain partitioning in H-charged structural alloys. Experimental investigation of these interactions is extremely difficult, labor-intensive, and costly. Here, we used MD and GCMC simulations and studied H-diffusion deformation at polycrystalline scale with atomic resolution efficiently. To study H-free surface interactions, large pillars including all possible angles and planes of free surfaces were modeled. To study H-grain boundary interactions, several polycrystalline models containing comparable statistics of low and high angle grain boundaries were examined. We studied the statistics of H-segregation tendencies based on free surface angles and grain boundary types. Dislocation activities were also classified for these various types and total density and strength were compared and analyzed compared to H-free samples. In the free surface model, it was observed that H was evenly distributed along the model's surface. Although the dislocation density was reduced compared to H-free samples, localized bands of dislocations were produced. Additionally in the polycrystalline samples, it was concluded that H tends to segregate along grain boundaries with misorientation angles less than or equal to 25 degrees. However, specific misorientation angle interactions -- namely $>45$ degrees -- led to increased dislocation density in H-charged samples compared to their H-free counterparts.