SearcharxivSearch

arXiv subjects

Hidemasa Bessho

Publications and source records attributed to Hidemasa Bessho.

2 recordsLinked to original sources

Pre-yielding mechanical response near the jamming transition

The mechanical and rheological properties of jammed packings of frictionless particles under shear strain remain not fully understood, even when the strain amplitude is very small and well below the yielding threshold. Systems above the jamming transition point $\phi_J$ are known to display two anomalous mechanical behaviors with respect to the driving frequency $\omega$ (or time $t$) and the strain amplitude $\gamma$. In the linear-response regime ($\gamma\to 0$), the complex modulus exhibits an algebraic scaling, $G(\omega)\sim\omega^{1/2}$ (or $G(t)\sim t^{-1/2}$ in the time representation). In contrast, in the quasi-static limit ($\omega \to 0$), the modulus shows the nonlinear behavior, $G(\gamma)\sim\gamma^{-1/2}$, a phenomenon referred to as softening. The ranges of $\omega$ and $\gamma$ over which these algebraic scalings hold broaden as $\phi_J$ is approached from above, whereas both $G(\omega)$ and $G(\gamma)$ vanish for $\phi < \phi_J$. In this study, we investigate the mechanical response in the regime where these two anomalies coexist in the vicinity of $\phi_J$. To this end, we perform numerical analyses using two rheological protocols: oscillatory shear and transient stress relaxation. Our results demonstrate that the mechanical responses are not simply described as a superposition of the two algebraic relaxations and instead exhibit rich nonlinear viscoelastic behavior both above and even below $\phi_J$.

cond-mat.soft

Inferring the dynamics of glass-forming liquids from static structure across thermal states

In this study, we demonstrate the generalizability of graph neural networks in predicting the dynamic heterogeneity of model glass-forming liquids across different temperatures. While previous approaches have often been limited to making predictions at the specific temperatures used during training, we find that our proposed framework - T-BOTAN - enables interpolation to temperatures not included in the training set. We show that the dynamical behavior, the associated four-point correlations, and even the macroscopic temperature can be estimated with sufficient accuracy solely from static particle configurations at untrained temperatures. These results suggest that static configurations encode not only local structural features driving dynamic heterogeneity but also fundamental thermodynamic information.

cond-mat.soft