TD Swinburne


CNRS   CINaM

The design of more durable materials is key to reduce the carbon emissions of construction, transportation and energy production. This requires an understanding of mechanical strength and toughness at the level of jiggling atoms.

We design atomic simulation methods to discover how metals bend and break.

The range of atomic deformation mechanisms is vast- the challenge is to discover and characterize the most influential. A short summary is below; for more detail please see our publications and open source software.

  Main Research Themes

Autonomous exploration at scale: TAMMBER
In massively parallel computatation, decisions must be made every few minutes to optimise resources. TAMMBER uses novel Bayesian methods to autonomously allocate computational effort only in the most relevant regions of configuration space, building kMC/Markov models with robust UQ.
TDS and D Perez, NPJ Computational Materials (2020) TAMMBER
TDS and D Perez, PR Materials (2018)

ML-assisted DFT for dislocation-defect interations
(Postdoc of Petr Grigorev, from ANR JCJC MeMoPas)
Ab initio methods such as DFT are essential to train and validate atomic models, but extended defects are at or beyond current computational limits. We design efficient methods to embed DFT accuracy only where it is needed, using custom-built machine learning interatomic potentials. This builds on earlier developments that enabled energy calculations.
P Grigorev et al., TDS, arXiv (2022) LML-retrain
TDS and JR Kermode, PRB (2017)

Vibrational free energies beyond harmonicity: PAFI
PAFI is a path-constrained sampling routine, which returns exact vibrational free energy differences for linear-scaling effort. This enables evaluation of activation free energies for very large systems of e.g. dislocation or twin migration, which can not be treated by cubic-scaling HTST approximations.
TDS and M-C Marinica, PRL (2018) PAFI

Surrogate ML models for defect entropics
(PhD of Clovis Lapointe, with M-C Marinica)
A linear machine learning model typically used for energetics can predict the formation and migration entropy of defects. Generalised linear models allow a parameter-geometric interpretation of relations such as the Meyer-Neldel compensation law.
C Lapointe, TDS et al. Submitted (2022)
C Lapointe, TDS et al. PRMaterials (2020)

Analysis of Markov Models: PyGT
Much atomic data (from TAMMBER or other codes) is a connected network of metastable states, which can be mapped to a discrete state Markov model, or master equation. We have developed tools to automatically propagate atomic data, with UQ, to drift-diffusion equations or first passage distributions between distant regions.
TDS and D Perez, arXiv (2022)
TDS and DJ Wales, JCTC (2020) PyGT

Mean-field models for lattice vibrations: BLaSa
BLaSa is a lightening-fast mean field approach to evaluate anharmonic vibrational phase free energies for milliseconds of effort. For unary FCC we are within meV/atom of brute force thermodynamic integration, requiring hours of effort. Current developments are within the PhD of R Dsouza since 2020.
TDS et al., J Neugebauer PRB RC (2020) BLaSa

  Team

  • Tom Swinburne
  • me 10/18-:  CNRS researcher, CINaM
    04/17-06/18:  Postdoc, T-1, LANL
    03/15-03/17:  EuroFusion fellow, CCFE
    10/11-03/15:  MSc+PhD, Imperial College London

    Curriculum Vitae
  • Petr Grigorev
  • Petr Grigorev 20-23:  Postdoc, Machine learning assisted DFT methods
    Funded by ANR MEMOPAS (PI: TDS)

    Personal Website
  • Clovis Lapointe
  • 09/18-01/22:  PhD, Machine learning for defect entropics. Joint with Dr MC Marinica, CEA Saclay
  • Raynol Dsouza
  • 09/20-09/23:  PhD, mean field vibration models. External with Prof J Neugebauer, MPIE Dusseldorf

      Main Collaborators

  • Prof David Wales FRS, Univ. Cambridge, UK
  • Analysis and convergence of highly metastable systems; Markov models
    e.g.
    JCTC 2020 PyGT
  • Dr Mihai-Cosmin Marinica, SRMP, CEA Saclay, FR
  • Linear machine learning models, free energy calculation
    e.g.
    PRMat 2020 PRL 2018 PAFI
  • Dr Danny Perez, T-1, Los Alamos, USA
  • Massively parallel sampling of metastable systems
    e.g.
    NPJ Comp. Mat 2020 TAMMBER
  • Prof James Kermode, Univ. Warwick, UK
  • Hybrid DFT/empirical methods
    e.g.
    PRB 2017 arXiv 2021/2 LML-retrain
  • Prof Jörg Neugebauer, Max Planck für Eisenforschung, DE
  • Mean field models for anharmonic lattice vibrations
    PRB RC 2020 BLaSa

      Past Members

  • Mira Sharma
  • 21-21:  M2 project, Learning of transition states, Aix-Marseille University.
  • Deepti Kannan
  • Deepti Kannan 20-21:  MSc project, reducing Markov Chains. With Prof DJ Wales, Cambridge. Now PhD, MIT.

      News

      Recent Publications


    = corresponding author(s). Full list at Google Scholar.

      Funding


    anr

  • ANR MeMoPas (2020-2022)
  • PI on ANR JCJC project, Mesoscale models from massively parallel atomistic simulations: uncertainty driven, self-optimizing strategies for hard materials

    EuroFusion

  • EuroFusion / IREMEV program
  • Computational resources (2020-2021) for studies of timescale estimation of microstructural evolution

    genci

  • GENCI / IDRIS national supercomputing centers
  • Computational resources (2019-2020,2020-2021) for high-throughput studies of defect diffusion mechanisms

      Job Openings

    If you are interested in masters projects at AMU, or are considering an independent funding application, please email me.

    We currently have no open contracts, but are always open to applications from motivated postdocs or PhD candidates. Please email me to discuss funding possibilities.

      Some Presentations

  • Kink limited motion of line defects: multiscale simulation and analysis
  • Applied analysis seminar, Imperial College London, February 2019 (Invited)
    Slides   

  • Kink limited Orowan strengthening and the brittle to ductile transition of bcc metals
  • Oxford MFFP Meeting, September 2018 (Invited)
    Slides   

  • Uncertainty-driven construction of Markov models from accelerated molecular dynamics
  • Advances in Computational Statistical Physics, CIRM, France, September 2018 (Invited)
    Slides   

  • Fast, vacancy-free climb of dislocation loops in bcc metals
  • NUMAT conference, Montpellier, France, 2016
    Slides   

  • Multiscale analysis of nonlinear dislocation models
  • MMM conference, Berkeley, USA, 2014
    Slides   

      Contact

       swinburne cinam "." univ-mrs "." fr


       CINaM, Campus de Luminy, 13288 Marseille, FRANCE