LM3: Learning Multiscale Material Models

Department of Mechanical Engineering, University of Michigan, Ann Arbor
Thomas D Swinburne, PhD   tswin-at-umich-dot-edu

Metallic alloys uniquely combine strength, ductility and reusability due to an often mysterious hierarchy of plasticity mechanisms.
The LM3 lab develops new simulation methods connecting atomic dynamics, microstructure and mechanical properties. We collaborate widely, drawing from applied math, machine learning, physics and mechanics.

Current research topics

• Application focus:
  • Prediction of phase diagrams across the periodic table
  • Strain hardening and dislocation networks
  • Segregation and annealing in structural alloys for fusion
  • Error-aware experimental design
• Theory/method development focus:
  • Model-form uncertainty quantification (UQ) for ML surrogates
  • Data-driven forecasting of long-timescale behavior
  • End-to-end differentiable simulations for UQ and design
  • Inverse fine-tuning of atomic machine learning models
    See open positions or email tswin-at-umich-dot-edu. Past members have gone on to obtain competitive positions in academia and industry.
    
    Selected recent work
    See Google Scholar for an up-to-date list.
• Score matched free energies (NeurIPSAI4Mat 2025 / Nat. Comm. 2026
  MLIP phase diagrams as functions, for UQ and inverse problems !
• Entropy of dislocation glide (Nat. Comm. 25/PAFI/Rodney Group)
  Why ML potentials are required to accurately model bcc plastic flow.
• Implicit differentiation in MD (NPJ 2025 featured / code)
  MLIP energy landscapes as functions, for UQ and inverse problems !
• Misspecification uncertainty in MD (NPJ 2025 / code)
  Using POPS to explore model-form UQ of SNAP and MACE.
  
• POPS deterministic model-form UQ (ML:S&T 2025 / code)
  Solving Bayes’ ignorance of model-form UQ (misspecification).
• MACE foundation potentials (Csyani Group / JCP 2025)
  Section A.13. Dislocations are a tough extrapolation test for UMLIPs.
• Coarse-graining & forecasting atomic simulations ( PRL 2023)
  ML feature vectors used for MLIPs can also forecast complex futures
• Embedding ab initio in MLIPs for solute studies (Acta Mat. 2023 )
  QM/ML for dislocation-solute interactions. Applied to W alloys
  
• A15 defects in fcc (Nat. Comm. 2023 / TAMMBER / Marinica Group)
  Irradiation defects can grow as three-dimensional Laves phase clusters
  Some recent/upcoming presentations
  ☆ =invited, †=chair / symposium chair
  † MRS Spring 5/27, Seattle, USA
  ☆ TMS Spring 4/27, Orlando, USA
  ☆ $MMM12, Jeju, Korea, 11/26
  ☆ AI Methods for Discovery of Materials, Purdue, USA, 07/26
  ☆ EL2026, Telluride, USA, 07/26
  † DECLARE, IMSI, U Chicago, USA, 06/26
  ☆ SIAM UQ, Minneapolis, USA, 03/26
  ☆ Machine Learning for Chemical Modeling, Santa Fe, USA, 05/26
  ☆ USACM UQ, Webinar, USA, 01/26
  ☆ Dislocations 2025, Miami, USA, 11/25
  ☆ CECAM: UQ from DFT to ML, EPFL, Lausanne, 11/25
  ☆ CoMPASs workshop, ICMS, Edinburgh, 11/25 (unavailable)
  ☆ IPAM Electrochemistry, UCLA, 10/25
  ☆ USACM Nanomechanics 2025, Urbana-Champaign, USA, 09/25