DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Accelerating discrete dislocation dynamics simulations with graph neural networks

Journal Article · · Journal of Computational Physics

Discrete dislocation dynamics (DDD) is a widely employed computational method to study plasticity at the mesoscale that connects the motion of dislocation lines to the macroscopic response of crystalline materials. However, the computational cost of DDD simulations remains a bottleneck that limits its range of applicability. Here, we introduce a new DDD-GNN framework in which the expensive time-integration of dislocation motion is entirely substituted by a graph neural network (GNN) model trained on DDD trajectories. As a first application, we demonstrate the feasibility and potential of our method on a simple yet relevant model of a dislocation line gliding through an array of obstacles. We show that the DDD-GNN model is stable and reproduces very well unseen ground-truth DDD simulation responses for a range of straining rates and obstacle densities, without the need to explicitly compute nodal forces or dislocation mobilities during time-integration. Our approach opens new promising avenues to accelerate DDD simulations and to incorporate more complex dislocation motion behaviors.

Research Organization:
Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)
Sponsoring Organization:
USDOE National Nuclear Security Administration (NNSA)
Grant/Contract Number:
AC52-07NA27344
OSTI ID:
2005099
Report Number(s):
LLNL-JRNL-838407; 1058699; TRN: US2405946
Journal Information:
Journal of Computational Physics, Vol. 487, Issue N/A; ISSN 0021-9991
Publisher:
ElsevierCopyright Statement
Country of Publication:
United States
Language:
English

References (38)

Scaling of Dislocation Strengthening by Multiple Obstacle Types journal June 2010
Advanced time integration algorithms for dislocation dynamics simulations of work hardening journal April 2016
Dislocation glide through non-randomly distributed point obstacles journal September 2013
Consistent formulation for the Discrete-Continuous Model: Improving complex dislocation dynamics simulations journal May 2016
Gaussian approximation potentials: A brief tutorial introduction journal April 2015
Chapter 88 Dislocation–Obstacle Interactions at the Atomic Level book January 2009
The ReaxFF reactive force-field: development, applications and future directions journal March 2016
An efficient implicit time integration method for discrete dislocation dynamics journal March 2020
Modelling crystal plasticity by 3D dislocation dynamics and the finite element method: The Discrete-Continuous Model revisited journal February 2014
Four Generations of High-Dimensional Neural Network Potentials journal March 2021
Machine Learning for Molecular Simulation journal April 2020
Efficient time integration in dislocation dynamics journal January 2014
Analysis of Obstacle Hardening Models Using Dislocation Dynamics: Application to Irradiation-Induced Defects journal May 2015
Multipole expansion of dislocation interactions: Application to discrete dislocations journal April 2002
Dislocation movement through random arrays of obstacles journal October 1966
Machine learning plastic deformation of crystals journal December 2018
A statistical model of irradiation hardening induced by non-periodic irradiation defects journal August 2021
A FFT-based formulation for discrete dislocation dynamics in heterogeneous media journal February 2018
A non-singular continuum theory of dislocations journal March 2006
The strengthening effect of voids journal August 1982
Connecting discrete and continuum dislocation mechanics: A non-singular spectral framework journal November 2019
Machine learning for interatomic potential models journal February 2020
Accurate and scalable graph neural network force field and molecular dynamics with direct force architecture journal May 2021
Machine Learning-Based Classification of Dislocation Microstructures journal June 2019
Machine Learning Force Fields journal March 2021
Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals journal April 2019
E(3)-equivariant graph neural networks for data-efficient and accurate interatomic potentials journal May 2022
Frontiers in the Simulation of Dislocations journal July 2020
Multiscale modelling of precipitation hardening in Al–Cu alloys: Dislocation dynamics simulations and experimental validation journal April 2020
Homogenization method for a discrete-continuum simulation of dislocation dynamics journal September 2001
Machine Learning of Dislocation-Induced Stress Fields and Interaction Forces journal October 2020
Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties journal April 2018
Effect of particle size distribution on strength of precipitation-hardened alloys journal September 2004
Enabling strain hardening simulations with dislocation dynamics journal July 2007
A FFT-based formulation for efficient mechanical fields computation in isotropic and anisotropic periodic discrete dislocation dynamics journal August 2015
Computer simulations of Orowan process controlled dislocation glide in particle arrangements of various randomness journal June 2002
Dislocation precipitate bypass through elastically mismatched precipitates journal February 2021
Implicit integration methods for dislocation dynamics journal January 2015