Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Video

Updated: Nov 4, 2025

Analyzing Mitochondrial Morphology Through Simulation Supervised Learning
12:06

Analyzing Mitochondrial Morphology Through Simulation Supervised Learning

Published on: March 3, 2023

4.4K

Self-supervised learning and prediction of microstructure evolution with convolutional recurrent neural networks.

Kaiqi Yang1, Yifan Cao1, Youtian Zhang1

  • 1Department of Materials Science and NanoEngineering, Rice University, Houston, TX 77005, USA.

Patterns (New York, N.Y.)
|May 26, 2021
PubMed
Summary

Related Concept Videos

Microcracking in Concrete01:20

Microcracking in Concrete

255
Microcracking in concrete refers to the tiny cracks that can form within the material even before any external load is applied. These microcracks typically occur at the interface between the coarse aggregate and the hydrated cement paste, often as a result of differential volume changes prompted by variations in stress-strain behavior, as well as thermal and moisture movement. Initially, these microcracks remain stable and do not grow substantially until the concrete is stressed to about 30...
255

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Efficient method for calculation of low-temperature phase boundaries.

The Journal of chemical physics·2026
Same author

Monte Carlo Simulations of Crystal Defects in Open Ensembles.

Physical review letters·2026
Same author

Geo-mapping of caries and obesity in preschool children: a Swedish register-based study.

BMC oral health·2026
Same author

First principles free energy model with dynamic magnetism for<i>δ</i>-plutonium.

Reports on progress in physics. Physical Society (Great Britain)·2025
Same author

Serum Insulin-like Growth Factor-II Is Associated with Poor Poststroke Outcomes in Males: A Secondary Analysis.

International journal of molecular sciences·2025
Same author

Model-free estimation of completeness, uncertainties, and outliers in atomistic machine learning using information theory.

Nature communications·2025
This summary is machine-generated.

Convolutional recurrent neural networks can now predict material microstructure evolution, replacing complex partial differential equation (PDE) simulations. This data-driven approach offers faster, efficient material science predictions.

Area of Science:

  • Materials Science
  • Computational Materials Science
  • Artificial Intelligence in Materials

Background:

  • Understanding material microstructure evolution is crucial for predicting material properties.
  • Traditional modeling relies on computationally intensive coarse-grained simulations using partial differential equations (PDEs).

Purpose of the Study:

  • To demonstrate the capability of convolutional recurrent neural networks (CRNNs) to model and predict microstructure evolution.
  • To offer a data-driven alternative to traditional PDE-based simulations.

Main Methods:

  • Self-supervised learning using image sequences from various simulated microstructure processes.
  • Training CRNNs on data from plane-wave propagation, grain growth, spinodal decomposition, and dendritic crystal growth simulations.
Keywords:
convolution neural networkmachine learningmicrostructurepartial differential equationsphase field simulationsrecurrent neural networktime evolution

Related Experiment Videos

Last Updated: Nov 4, 2025

Analyzing Mitochondrial Morphology Through Simulation Supervised Learning
12:06

Analyzing Mitochondrial Morphology Through Simulation Supervised Learning

Published on: March 3, 2023

4.4K
  • Validating network predictions against established simulation results.

    • Main Results:

    • CRNNs accurately predict both short-term local dynamics and long-term statistical properties of microstructures.
    • The trained networks demonstrate extrapolation capabilities beyond training data in spatiotemporal and parameter spaces.
    • The data-driven approach significantly improves time-stepping efficiency compared to PDE simulations.

    Conclusions:

    • CRNNs offer a powerful and efficient alternative for modeling microstructure evolution.
    • This approach is particularly advantageous when material parameters or governing PDEs are uncertain.
    • The study highlights the potential of AI in accelerating materials discovery and design.