Analytic theory of shear localization in amorphous solids confined by Couette geometry

Yang Fu, Yuliang Jin, Itamar Procaccia · 2025

Read Analytic theory of shear localization in amorphous solids confined by Couette geometry by Yang Fu, Yuliang Jin, Itamar Procaccia (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-14522.txt.

Introduction

Yang Fu, Yuliang Jin, and Itamar Procaccia build an analytic theory for shear localization in amorphous solids. Their setting is Couette geometry, with material held between two rings that rotate at different rates. As stress builds, plastic avalanches can make nearby parts rotate in opposite senses. Shear then concentrates in a narrow zone between them.

Working in two dimensions with slow rotation of the inner ring, the authors derive the displacement field from theory alone. Their formulas predict the angle averaged response tied to each plastic drop. They also explain why localization sets in. The theory gives testable predictions without fitting to simulation after the fact.

Catalog record

CREATORYang Fu, Yuliang Jin, Itamar Procaccia
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE50 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-14522

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Analytic theory of shear localization in amorphous solids confined by Couette geometry Yang Fu, Yuliang Jin, Itamar Procaccia Abstract “Couette geometry” refers to two concentric rings in 2-dimensions (or cylinders in 3-dimensions with a medium in between. Typically the inner and outer rings (or cylinders) rotate at different rates and the response of the medium is studied. Here we study a medium which is a two-dimensional amorphous solid, and we rotate the inner ring quasi-statically. As stress accumulates, plastic avalanches can result in shear localization, characterized by adjacent parts of the system rotating in opposite directions, with the maximum shear localized between them. We derive an analytic theory that describes and explains the shear localization, providing a-priori predictions for the displacement field associated with the plastic drops and the shear localization. Introduction: Shear localization and shear banding are prevalent phenomena observed in shear strained amorphous solids Poirier 1980 ; Varnik et al. 2003 ; Dasgupta et al. 2012 ; Dasgupta et al. 2013a ; Dasgupta et al. 2013b . In metallic glasses, for example, shear banding is a ubiquitous response to accumulated shear stress, often leading to catastrophic material failure Li et al. 2002 ; Maass et al. 2014 . Typically shear localization results from plastic events (avalanches), and thus pure elasticity theory is unable to predict and describe the phenomenon. In spite of intense research and modeling, presently there is no theory that can provide a detailed displacement field (where particles move during the stress drop including a plastic avalanche). It is known from experimental results that shear localization can occur either near a moving boundary in a Couette geometry or in the bulk of the system Losert et al. 2000 ; Fenistein et al. 2004 . The aim of this Letter is to provide such a theory for amorphous solids that are loaded by quasi-static shear strain. We will see that localization can occur near a boundary or in the bulk and that the theory accounts for both possibilities. Note that we do not study here shear banding, a phenomenon that typically occurs at yielding, separating solid and plastic-flow states, and is associated with a large-scale, system-spanning plastic event. Shear localization occurs during smaller-scale plastic events that can occur both before and after yielding. In this study, we focus on shear localization and leave shear banding for future studies - the latter requires a theory incorporating strong nonlinear effects. The system under study is shown in Panel (a) of Fig. 1 (see Sec. S1 in the Supplementary Material (SM) for more details). Figure 1: Panel (a): The simulated system, here the area fraction is \phi=0.91 , see text for details. Panel (b): typical shear stress vs. accumulated rotation angle \Theta . The stress drop responsible for Panel (c) is indicated by an arrow. Panel (c): typical profile of the angle-averaged angular component of the displacement field d_{\theta}(r) normalized by inner rotation \Omega_{0} as a function of r , showing the reversal of particle displacement from anti-clockwise to clockwise. This leads to the stress localization shown in Fig. S1 . The blue curve represents the tangential displacement of the elastic solution shown in Eq. ( 4 ). The material is made of assemblies of N=3534 bidisperse 50:50 repulsive disks of radii 1 and 1.4 respectively. Below the smaller radius is used as the unit of length. The material is bounded between two concentric rings of radii R_{\rm in}=28 and R_{\rm out}=80.8 . The particles adjacent to the outer boundary are glued to the boundary, providing below a boundary condition {\bm{d}}(R_{\rm out})={\bm{0}} where {\bm{d}} is the displacement field. An initial random configuration of a chosen area fraction \phi is prepared, and subsequently brought to mechanical equilibrium by molecular dynamics with appropriate damping. For the present study we do not attempt to anneal further Gutiérrez et al. 2015 ; Singh et al. 2020 , we are interested in ductile response Bordignon et al. 2015 . The material is then loaded by a quasi-static protocol in which the inner ring is rotated anti-clockwise in steps of \delta\theta=0.024^{\circ} ; after each step the system is equilibrated by damped molecular dynamics. This imposes the inner boundary condition d_{\theta}(R_{\rm in})=\Omega_{0}=R_{\rm in}\delta\theta\pi/180\approx 0.012 and d_{r}(R_{\rm in})=0 for the tangential and radial components of the displacement field respectively. Initially the accumulated shear stress \Sigma^{r\theta} is increasing via mostly purely elastic steps. At higher values of stress, the response to a small additional rotation step can be a sharp stress drop due to plastic events. We are interested in the displacement field that is associated with such stress drops, {\bm{d}}=d_{r}(r,\theta)\hat{r}+d_{\theta}(r,\theta)\hat{\theta}. (1) In particular we would like a theory for the angle-averaged tangential and radial component, d_{\theta}(r)\equiv\frac{1}{2\pi}\!\int_{0}^{2\pi}d_{\theta}(r,\theta)d\theta,\ \quad d_{r}(r)\equiv\frac{1}{2\pi}\!\!\int_{0}^{2\pi}d_{r}(r,\theta)d\theta\ . (2) A typical plot of stress vs. accumulated angle is shown in Fig. 1 Panel (b). As explained, this plot is characteristic of ductile materials. The profile of d_{\theta}(r) which is associated with the stress drop that is marked in the figure of Panel (b) is shown in Panel (c). We note that because of the double zero boundary condition on the radial component of the displacement field, this component is negligible throughout this discussion. The data in the Panel (c) of Fig. 1 indicates that part of the system is displaced clockwise and a part anti-clockwise. This is in clear variance with the prediction of classical elasticity theory Landau and Lifshitz 1970 . If elasticity theory were appropriate for describing our simulations, the classical equation for the displacement field would have read \mu\Delta{\bm{d}}+\left(\lambda+\mu\right)\nabla\left(\nabla\cdot{\bm{d}}\right)=0\ ,\quad\text{purely elastic.} (3) Here \lambda and \mu are the classical Lamé coefficients. Solving Eq. ( 3 ) for our system with the appropriate boundary conditions, we would expect a tangential displacement field that reads d_{\theta}(r)=\Omega_{0}\frac{R_{\rm in}(R_{\rm out}^{2}-r^{2})}{r(R_{\rm out}^{2}-R_{\rm in}^{2})}. (4) This exact elastic solution is shown as the blue line in Fig. 1 Panel (c). Evidently, classical elasticity theory fails miserably in reproducing, let alone predicting, the actual displacement field. Below we present a theory that results in the black curve in Panel (c). We will show that the theory can predict the actual displacement fields that are found in the present simulation. We use this function (which is fully derived below) to compute the shear strain as u_{r,\theta}\equiv(\partial_{\theta}d_{r}+\partial_{r}d_{\theta})/2 . Realizing that the first term is negligible, we obtain the shear strain that is plotted in Fig. S1 . The realized displacement field and the shear localization that is seen as the minimum in this curve are the phenomena that we need to explain. Figure 2: The shear strain as a function of r , exhibiting the shear localization at r=51.4 where the displacement field in Panel (c) of Fig. 1 changes sign. The blue curve represents the shear strain of the elastic solution. In a recent series of publications, being theoretical Lemaître et al. 2021 ; Charan et al. 2023 ; Kumar and Procaccia 2024 ; Jin et al. 2024 , experimental Mondal et al. 2022 ; Cohen et al. 2025 and simulational Bhowmik et al. 2022 ; Kumar et al. 2022 ; Mondal et al. 2023 , it was shown already that to explain generic responses of amorphous solids to external loads one needs to take into account the plastic events that appear

Read the complete text