Artificial Intelligence For Use In Atomistic Kinetic Monte Carlo Simulations

Flyura Djurabekova; Roberto Domingos; Gennaro Cerchiara; Nicolas Castin; Lorenzo Malerba; Abderrahim Al Mazouzi

Artificial Intelligence For Use In Atomistic Kinetic Monte Carlo Simulations

Flyura Djurabekova, Roberto Domingos, Gennaro Cerchiara, Nicolas Castin, Lorenzo Malerba, Abderrahim Al Mazouzi

Research output › peer-review

Abstract

A new approach to atomistic kinetic Monte Carlo (AKMC) simulations based on the determination of vacancy migration barriers as functions of the local atomic environment has been developed, with a view to provide a better description of the kinetic path followed by the system through diffusion of solute atoms in the alloy via vacancy mechanism. Tabulated values of barriers versus local atomic configurations (LAC) including atoms up to 2nd nearest neighbour shell, obtained by molecular dynamics (MD) techniques, have been used to train an artificial intelligence (AI) system to recognize the LACs and predict the barriers accordingly. Here some details on the method and preliminary results are presented and briefly discussed.

Original language	English
Title of host publication	MMM*** Third International Conference Multiscale Materials Modeling
Place of Publication	Freiburg, Germany
Pages	721-723
Volume	1
State	Published - Sep 2006
Event	3rd Intl Conf on Multiscale Materials Modeling - Fraunhoffer Institute for Mechanics of Materials - University of Freiburg, Freiburg Duration: 18 Sep 2006 → 22 Sep 2006

Conference

Conference	3rd Intl Conf on Multiscale Materials Modeling
Country/Territory	Germany
City	Freiburg
Period	2006-09-18 → 2006-09-22

Cite this

@inproceedings{8e773b2438e44a68b65bcfcb0ecdbbf7,

title = "Artificial Intelligence For Use In Atomistic Kinetic Monte Carlo Simulations",

abstract = "A new approach to atomistic kinetic Monte Carlo (AKMC) simulations based on the determination of vacancy migration barriers as functions of the local atomic environment has been developed, with a view to provide a better description of the kinetic path followed by the system through diffusion of solute atoms in the alloy via vacancy mechanism. Tabulated values of barriers versus local atomic configurations (LAC) including atoms up to 2nd nearest neighbour shell, obtained by molecular dynamics (MD) techniques, have been used to train an artificial intelligence (AI) system to recognize the LACs and predict the barriers accordingly. Here some details on the method and preliminary results are presented and briefly discussed.",

keywords = "artificial intelligence, kinetic Monte Carlo, iron-copper",

author = "Flyura Djurabekova and Roberto Domingos and Gennaro Cerchiara and Nicolas Castin and Lorenzo Malerba and {Al Mazouzi}, Abderrahim",

note = "Score = 1; 3rd Intl Conf on Multiscale Materials Modeling ; Conference date: 18-09-2006 Through 22-09-2006",

year = "2006",

month = sep,

language = "English",

isbn = "978-3-8167-7206-4",

volume = "1",

pages = "721--723",

booktitle = "MMM*** Third International Conference Multiscale Materials Modeling",

}

TY - GEN

T1 - Artificial Intelligence For Use In Atomistic Kinetic Monte Carlo Simulations

AU - Djurabekova, Flyura

AU - Domingos, Roberto

AU - Cerchiara, Gennaro

AU - Castin, Nicolas

AU - Malerba, Lorenzo

A2 - Al Mazouzi, Abderrahim

N1 - Score = 1

PY - 2006/9

Y1 - 2006/9

N2 - A new approach to atomistic kinetic Monte Carlo (AKMC) simulations based on the determination of vacancy migration barriers as functions of the local atomic environment has been developed, with a view to provide a better description of the kinetic path followed by the system through diffusion of solute atoms in the alloy via vacancy mechanism. Tabulated values of barriers versus local atomic configurations (LAC) including atoms up to 2nd nearest neighbour shell, obtained by molecular dynamics (MD) techniques, have been used to train an artificial intelligence (AI) system to recognize the LACs and predict the barriers accordingly. Here some details on the method and preliminary results are presented and briefly discussed.

AB - A new approach to atomistic kinetic Monte Carlo (AKMC) simulations based on the determination of vacancy migration barriers as functions of the local atomic environment has been developed, with a view to provide a better description of the kinetic path followed by the system through diffusion of solute atoms in the alloy via vacancy mechanism. Tabulated values of barriers versus local atomic configurations (LAC) including atoms up to 2nd nearest neighbour shell, obtained by molecular dynamics (MD) techniques, have been used to train an artificial intelligence (AI) system to recognize the LACs and predict the barriers accordingly. Here some details on the method and preliminary results are presented and briefly discussed.

KW - artificial intelligence

KW - kinetic Monte Carlo

KW - iron-copper

UR - http://ecm.sckcen.be/OTCS/llisapi.dll/open/ezp_32928

UR - http://knowledgecentre.sckcen.be/so2/bibref/3668

M3 - In-proceedings paper

SN - 978-3-8167-7206-4

VL - 1

SP - 721

EP - 723

BT - MMM*** Third International Conference Multiscale Materials Modeling

CY - Freiburg, Germany

T2 - 3rd Intl Conf on Multiscale Materials Modeling

Y2 - 18 September 2006 through 22 September 2006

ER -