skip to main content
DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Accelerating high-strain continuum-scale brittle fracture simulations with machine learning

Abstract

Failure in brittle materials under dynamic loading conditions is a result of the propagation and coalescence of microcracks. Simulating this discrete crack evolution at the continuum level is computationally expensive or, in some cases, intractable, resulting in the need to make broad assumptions or neglect key physics. In this work, we have developed an approach using machine learning that overcomes the current inability to represent meso-scale physics at the macro-scale. Our approach leverages damage and stress data from a computationally expensive high-fidelity model that explicitly resolves microcrack behavior to build an inexpensive machine learning emulator. Once trained, the machine learning emulator is used to predict the evolution of crack length statistics, which then informs a continuum-scale constitutive model. This results in a significant speed-up of the workflow by four orders of magnitude. Both the machine learning emulator and the continuum-scale model are validated against the high-fidelity model and experimental data, respectively, showing excellent agreement. There are two key findings. The first is that we can reduce the dimensionality of the problem, establishing that the machine learning emulator only needs the length of the longest crack and one of the maximum stress components to capture the necessary physics. Another compelling findingmore » is that the emulator can be trained in one experimental setting and transferred successfully to predict behavior in a different setting.« less

Authors:
ORCiD logo [1]; ORCiD logo [1]; ORCiD logo [1];  [1]; ORCiD logo [1];  [2]; ORCiD logo [1]
  1. Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
  2. Univ. of Florida, Gainesville, FL (United States)
Publication Date:
Research Org.:
Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
Sponsoring Org.:
USDOE National Nuclear Security Administration (NNSA)
OSTI Identifier:
1774441
Alternate Identifier(s):
OSTI ID: 1778360
Report Number(s):
LA-UR-20-22890
Journal ID: ISSN 0927-0256
Grant/Contract Number:  
89233218CNA000001
Resource Type:
Accepted Manuscript
Journal Name:
Computational Materials Science
Additional Journal Information:
Journal Volume: 186; Journal ID: ISSN 0927-0256
Publisher:
Elsevier
Country of Publication:
United States
Language:
English
Subject:
36 MATERIALS SCIENCE; Machine learning; crack statistics; validation; cost reduction

Citation Formats

Fernandez, Maria Giselle, Panda, Nishant, O'Malley, Daniel, Larkin, Kevin, Hunter, Abigail, Haftka, Raphael T., and Srinivasan, Gowri. Accelerating high-strain continuum-scale brittle fracture simulations with machine learning. United States: N. p., 2020. Web. https://doi.org/10.1016/j.commatsci.2020.109959.
Fernandez, Maria Giselle, Panda, Nishant, O'Malley, Daniel, Larkin, Kevin, Hunter, Abigail, Haftka, Raphael T., & Srinivasan, Gowri. Accelerating high-strain continuum-scale brittle fracture simulations with machine learning. United States. https://doi.org/10.1016/j.commatsci.2020.109959
Fernandez, Maria Giselle, Panda, Nishant, O'Malley, Daniel, Larkin, Kevin, Hunter, Abigail, Haftka, Raphael T., and Srinivasan, Gowri. Tue . "Accelerating high-strain continuum-scale brittle fracture simulations with machine learning". United States. https://doi.org/10.1016/j.commatsci.2020.109959. https://www.osti.gov/servlets/purl/1774441.
@article{osti_1774441,
title = {Accelerating high-strain continuum-scale brittle fracture simulations with machine learning},
author = {Fernandez, Maria Giselle and Panda, Nishant and O'Malley, Daniel and Larkin, Kevin and Hunter, Abigail and Haftka, Raphael T. and Srinivasan, Gowri},
abstractNote = {Failure in brittle materials under dynamic loading conditions is a result of the propagation and coalescence of microcracks. Simulating this discrete crack evolution at the continuum level is computationally expensive or, in some cases, intractable, resulting in the need to make broad assumptions or neglect key physics. In this work, we have developed an approach using machine learning that overcomes the current inability to represent meso-scale physics at the macro-scale. Our approach leverages damage and stress data from a computationally expensive high-fidelity model that explicitly resolves microcrack behavior to build an inexpensive machine learning emulator. Once trained, the machine learning emulator is used to predict the evolution of crack length statistics, which then informs a continuum-scale constitutive model. This results in a significant speed-up of the workflow by four orders of magnitude. Both the machine learning emulator and the continuum-scale model are validated against the high-fidelity model and experimental data, respectively, showing excellent agreement. There are two key findings. The first is that we can reduce the dimensionality of the problem, establishing that the machine learning emulator only needs the length of the longest crack and one of the maximum stress components to capture the necessary physics. Another compelling finding is that the emulator can be trained in one experimental setting and transferred successfully to predict behavior in a different setting.},
doi = {10.1016/j.commatsci.2020.109959},
journal = {Computational Materials Science},
number = ,
volume = 186,
place = {United States},
year = {2020},
month = {9}
}

Works referenced in this record:

Ejecta source and transport modeling in the FLAG hydrocode
journal, August 2013


Evaporation Effects in Shock-Driven Multiphase Instabilities
journal, April 2017

  • Black, Wolfgang J.; Denissen, Nicholas A.; McFarland, Jacob A.
  • Journal of Fluids Engineering, Vol. 139, Issue 7
  • DOI: 10.1115/1.4036162

Earthquake Damage Patterns Resolve Complex Rupture Processes
journal, October 2018

  • Klinger, Yann; Okubo, Kurama; Vallage, Amaury
  • Geophysical Research Letters, Vol. 45, Issue 19
  • DOI: 10.1029/2018GL078842

Mesoscale informed parameter estimation through machine learning: A case-study in fracture modeling
journal, November 2020


Predictive modeling of dynamic fracture growth in brittle materials with machine learning
journal, June 2018


Surrogate Models for Estimating Failure in Brittle and Quasi-Brittle Materials
journal, July 2019

  • Mudunuru, Maruti Kumar; Panda, Nishant; Karra, Satish
  • Applied Sciences, Vol. 9, Issue 13
  • DOI: 10.3390/app9132706

Bridging machine learning and computer network research: a survey
journal, November 2018


Characterization of shocked beryllium
journal, January 2012


Fracture-permeability behavior of shale
journal, September 2015

  • Carey, J. William; Lei, Zhou; Rougier, Esteban
  • Journal of Unconventional Oil and Gas Resources, Vol. 11
  • DOI: 10.1016/j.juogr.2015.04.003

Big data classification: problems and challenges in network intrusion prediction with machine learning
journal, April 2014


The combined plastic and discrete fracture deformation framework for finite‐discrete element methods
journal, December 2019

  • Rougier, Esteban; Munjiza, Antonio; Lei, Zhou
  • International Journal for Numerical Methods in Engineering, Vol. 121, Issue 5
  • DOI: 10.1002/nme.6255

Validation of a three-dimensional Finite-Discrete Element Method using experimental results of the Split Hopkinson Pressure Bar test
journal, September 2014

  • Rougier, E.; Knight, E. E.; Broome, S. T.
  • International Journal of Rock Mechanics and Mining Sciences, Vol. 70
  • DOI: 10.1016/j.ijrmms.2014.03.011

Power Law of Crack Length Distribution in the Multiple Damage Process
journal, September 2019


Statistically informed upscaling of damage evolution in brittle materials
journal, August 2019


Some fundamental issues of damage mechanics
journal, October 1995


A combined finite‐discrete element method in transient dynamics of fracturing solids
journal, February 1995


The Evolution of Computing: AlphaGo
journal, July 2016


Reduced-order modeling through machine learning and graph-theoretic approaches for brittle fracture applications
journal, February 2019


Recent advances and applications of machine learning in solid-state materials science
journal, August 2019

  • Schmidt, Jonathan; Marques, Mário R. G.; Botti, Silvana
  • npj Computational Materials, Vol. 5, Issue 1
  • DOI: 10.1038/s41524-019-0221-0

A micromechanics based model to predict micro-crack coalescence in brittle materials under dynamic compression
journal, August 2019


Effective elastic moduli of two-phase composites containing randomly dispersed spherical inhomogeneities
journal, March 1994


An interacting micro-crack damage model for failure of brittle materials under compression
journal, March 2008


Numerical comparison of some explicit time integration schemes used in DEM, FEM/DEM and molecular dynamics
journal, September 2004

  • Rougier, E.; Munjiza, A.; John, N. W. M.
  • International Journal for Numerical Methods in Engineering, Vol. 61, Issue 6
  • DOI: 10.1002/nme.1092

A constitutive model for metals applicable at high‐strain rate
journal, March 1980

  • Steinberg, D. J.; Cochran, S. G.; Guinan, M. W.
  • Journal of Applied Physics, Vol. 51, Issue 3
  • DOI: 10.1063/1.327799

Scaling of fracture systems in geological media
journal, August 2001

  • Bonnet, E.; Bour, O.; Odling, N. E.
  • Reviews of Geophysics, Vol. 39, Issue 3
  • DOI: 10.1029/1999RG000074

Surrogate-based analysis and optimization
journal, January 2005


Deep multiscale model learning
journal, April 2020


Issues in Deciding Whether to Use Multifidelity Surrogates
journal, May 2019

  • Giselle Fernández-Godino, M.; Park, Chanyoung; Kim, Nam H.
  • AIAA Journal, Vol. 57, Issue 5
  • DOI: 10.2514/1.J057750

Digital selection and analogue amplification coexist in a cortex-inspired silicon circuit
journal, June 2000

  • Hahnloser, Richard H. R.; Sarpeshkar, Rahul; Mahowald, Misha A.
  • Nature, Vol. 405, Issue 6789
  • DOI: 10.1038/35016072

Simulation of Fracture Coalescence in Granite via the Combined Finite–Discrete Element Method
journal, March 2019


Quantifying Topological Uncertainty in Fractured Systems using Graph Theory and Machine Learning
journal, August 2018


Learning to fail: Predicting fracture evolution in brittle material models using recurrent graph convolutional neural networks
journal, May 2019