| About this Abstract |
| Meeting |
MS&T24: Materials Science & Technology
|
| Symposium
|
Advances in Materials and Systems for a Hydrogen Economy
|
| Presentation Title |
Fundamental Atomistic Study of H-Defect Interactions to Predict H Segregation Energy Spectra |
| Author(s) |
Matthew Melfi, S. Mohadeseh Taheri-Mousavi |
| On-Site Speaker (Planned) |
Matthew Melfi |
| Abstract Scope |
Hydrogen-embrittlement is the most critical life-limiting factor of structural alloys, causing the US to waste ~3% of the GDP per year. Understanding hydrogen-defect interactions offers insights into significant ductility loss in H-charged structural alloys due to strain partitioning. Experimental investigations of these interactions are challenging, prompting the development of simulation models to explore nanoscale interactions. Using hybridization and parallelization of molecular dynamics and grand canonical Monte Carlo, polycrystalline scale atomistic models of H-diffusion deformation can be simulated with high efficiency. Here we study the local atomic environment of H segregation by determining SOAP vectors and use feature extraction to determine the important features. Various machine learning techniques, such as linear and non-linear techniques will be applied to predict the H atom locations and segregation energies. We will discuss how these analyses will guide defect engineering, e.g. grain boundary engineering to inhibit H segregation in the microstructure. |