Woodruff School of Mechanical Engineering
COE/Structural Mechanics Seminar
On Kinetic Monte Carlo algorithms in materials science:theory and applications
Dr. Enrique Martinez
Los Alamos National Laboratory
Friday, February 27, 2015 at 11:00:00 AM
Love Building, Room 109
Dr. Laurent Capolungo
Multiscale Materials Modeling has become an essential approach in the quest to understanding and predicting materials properties in a wide range of applications. The ultimate goal focuses on designing materials with the desired properties for the application considered. One of the tools most use in this approach is the kinetic Monte Carlo (KMC) algorithm. The KMC algorithm has become an important tool in Materials Science to investigate the dynamic evolution of systems under a wide variety of conditions. Owing to its flexibility, the KMC algorithm has been applied at different length and time scales for remarkably different physical problems, from magnetic properties to dislocation mobilities. The algorithm was conceived as an efficient method to sample the system trajectory in phase space, avoiding the analytical solution of the Markovian Master Equation, which might be daunting. It relies on the knowledge of the probabilities per unit time for the system to go from state i to state j. I will show how, once these rates are known (or calculated on-the-fly), the algorithm accurately provides one realization of the Master Equation. I will review the different sampling methods and the most widely used searching algorithms. I will describe the existent parallelization schemes and explain why their development has long lagged behind. I will focus on the application of the algorithm to defect diffusion, out and under irradiation, for the FeCr system, showing how we can reproduce the right precipitation kinetics and the radiation-induced segregation profiles, and how we can use this methodology to predict the system behavior and therefore help design next generation materials for extreme environments.
My research interests lie on the study of the microstructural evolution of materials under extreme conditions of deformation, temperature and irradiation and its relation to the macroscopic system response. During my doctoral dissertation I developed, for the first time, a robust algorithm to incorporate the explicit treatment of partial dislocations accounting for the stacking fault energy independently of the dislocation character or the reactions it might undergo. I also developed a synchronous parallel kinetic Monte Carlo (KMC) algorithm to take advantage of parallel architectures with the possibility of reaching computational grid sizes and simulated times beyond what can be achieved currently. I have studied, using Density Functional Theory, the energetics of the ternary system Fe-Cr-He (of paramount importance in the nuclear industry) and developed an interaction model to study precipitation, radiation induced segregation and microstructure evolution. My current projects comprise: 1) Next Generation Quantum Molecular Dynamics (LDRD-DR) 2) Radiation-Induced Solid Redistribution (LDRD-DR) and 3) Exascale Co-Design Center for Materials in Extreme Environments (DOE Office of Science). Recently, I have signed a contract to write a book on the kinetic Monte Carlo algorithm with the tentative title: “The Kinetic Monte Carlo Algorithm: Principles and Applications in Materials Science”.
Refreshments will be served.