Predicting Bond Switching and Fracture in Simulated Al2O3 Glass Using Machine Learning

Tao Du, Han Liu, Longwen Tang, Søren Strandskov Sørensen, Mathieu Bauchy, Morten Mattrup Smedskjær

Publikation: Konferencebidrag uden forlag/tidsskriftKonferenceabstrakt til konferenceForskningpeer review

19 Downloads (Pure)


Flaw-free amorphous alumina (a-Al2O3) samples have recently been found to exhibit excellent nanoductility at room temperature and under high strain rate. A better understanding of the underlying ductile deformation mechanism could help to facilitate the design of damage-tolerant glassy materials. In this work, based on atomistic simulations and classification-based machine learning, we reveal that the propensity of simulated glassy Al2O3 to exhibit nanoscale ductility is encoded in its static (non-strained) structure. The machine learning based softness metric trained from the spontaneous dynamics of the system (i.e., under zero strain) is able to readily predict the fracture behavior of the glass (i.e., under strain). Specifically, lower softness facilitates Al bond switching and the local accumulation of high-softness regions leads to rapid crack propagation. These results are helpful for designing oxide glass formulations with improved resistance to fracture.
Publikationsdato6 jul. 2022
StatusUdgivet - 6 jul. 2022
Begivenhed26th International Congress on Glass - Marriott Hotel Berlin Central District, Berlin, Berlin, Tyskland
Varighed: 3 jul. 20228 jul. 2022
Konferencens nummer: 26


Konference26th International Congress on Glass
LokationMarriott Hotel Berlin Central District, Berlin


Dyk ned i forskningsemnerne om 'Predicting Bond Switching and Fracture in Simulated Al2O3 Glass Using Machine Learning'. Sammen danner de et unikt fingeraftryk.