Dynamic heterogeneity at the experimental glass transition predicted by transferable machine learning - Laboratoire Charles Coulomb (L2C) Access content directly
Journal Articles Physical Review B Year : 2024

Dynamic heterogeneity at the experimental glass transition predicted by transferable machine learning

Abstract

We develop a machine learning model, which predicts structural relaxation from amorphous supercooled liquid structures. The trained networks are able to predict dynamic heterogeneity across a broad range of temperatures and time scales with excellent accuracy and transferability. We use the network transferability to predict dynamic heterogeneity down to the experimental glass transition temperature Tg, where structural relaxation cannot be analyzed using molecular dynamics simulations. The results indicate that the strength, the geometry, and the characteristic length scale of the dynamic heterogeneity evolve much more slowly near Tg compared to their evolution at higher temperatures. Our results show that machine learning techniques can provide physical insights on the nature of the glass transition that cannot be gained using conventional simulation techniques.
Fichier principal
Vignette du fichier
2310.20252.pdf (10.52 Mo) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-04514863 , version 1 (21-03-2024)

Identifiers

Cite

Gerhard Jung, Giulio Biroli, Ludovic Berthier. Dynamic heterogeneity at the experimental glass transition predicted by transferable machine learning. Physical Review B, 2024, 109 (6), pp.064205. ⟨10.1103/PhysRevB.109.064205⟩. ⟨hal-04514863⟩
0 View
0 Download

Altmetric

Share

Gmail Facebook X LinkedIn More