Table of Contents
Fetching ...

Compression-driven jamming in porous cohesive aggregates

Sota Arakawa

Abstract

I investigate the compression-driven jamming behavior of two-dimensional porous aggregates composed of cohesive, frictionless disks. Three types of initial aggregates are prepared using different aggregation procedures, namely, reaction-limited aggregation (RLA), ballistic particle-cluster aggregation (BPCA), and diffusion-limited aggregation (DLA), to elucidate the influence of aggregate morphology. Using distinct-element-method simulations with a shrinking circular boundary, I numerically obtain the pressure as a function of the packing fraction $φ$. For the densest RLA and the intermediate BPCA aggregates, a clear jamming transition is observed at a critical packing fraction $φ_{\rm J}$, below which the pressure vanishes and above which a finite pressure emerges; the transition is less distinct for the most porous DLA aggregates. The jamming threshold depends on the initial structure and, when extrapolated to infinite system size, approaches $φ_{\rm J} = 0.765 \pm 0.004$ for RLA, $0.727 \pm 0.004$ for BPCA, and $0.602 \pm 0.023$ for DLA, where the errors denote the standard error. Above $φ_{\rm J}$, the pressure follows $P \approx A {( φ- φ_{\rm J} )}^{2}$, which implies that the bulk modulus $K$ of jammed aggregates is proportional to $φ- φ_{\rm J}$. Rigid-cluster analysis of jammed aggregates shows that the average coordination number within the largest rigid cluster increases linearly with $φ- φ_{\rm J}$. Taken together, these relations suggest that the elastic response of compressed porous aggregates is analogous to that of random spring networks.

Compression-driven jamming in porous cohesive aggregates

Abstract

I investigate the compression-driven jamming behavior of two-dimensional porous aggregates composed of cohesive, frictionless disks. Three types of initial aggregates are prepared using different aggregation procedures, namely, reaction-limited aggregation (RLA), ballistic particle-cluster aggregation (BPCA), and diffusion-limited aggregation (DLA), to elucidate the influence of aggregate morphology. Using distinct-element-method simulations with a shrinking circular boundary, I numerically obtain the pressure as a function of the packing fraction . For the densest RLA and the intermediate BPCA aggregates, a clear jamming transition is observed at a critical packing fraction , below which the pressure vanishes and above which a finite pressure emerges; the transition is less distinct for the most porous DLA aggregates. The jamming threshold depends on the initial structure and, when extrapolated to infinite system size, approaches for RLA, for BPCA, and for DLA, where the errors denote the standard error. Above , the pressure follows , which implies that the bulk modulus of jammed aggregates is proportional to . Rigid-cluster analysis of jammed aggregates shows that the average coordination number within the largest rigid cluster increases linearly with . Taken together, these relations suggest that the elastic response of compressed porous aggregates is analogous to that of random spring networks.

Paper Structure

This paper contains 16 sections, 19 equations, 10 figures, 2 tables.

Figures (10)

  • Figure 1: Schematics of elastic interaction models. (a) Unbreakable contact model. $F_{\rm e} = 0$ when the two particles are separated (black line). A contact is formed when the particles collide (red filled circle). Once a contact has formed, $F_{\rm e}$ is given by Equation (\ref{['eq:F_e']}) (red line). A repulsive force acts when the compression length is positive, whereas an attractive force acts when it is negative. In this model, a contact never breaks once it has formed. (b) Breakable contact model. $F_{\rm e} = 0$ when the two particles are separated (black line). A contact is formed when the particles collide (red filled circle). Once a contact has formed, $F_{\rm e}$ is given by Equation (\ref{['eq:F_e_break']}) (red line). In this model, the contact breaks when $\delta$ reaches $- 2 \delta_{\rm c}$ (black open circle).
  • Figure 2: Three types of aggregates prepared using distinct aggregation procedures. Panels (a), (b), and (c) show the initial structures of DLA, BPCA, and RLA aggregates with $N = 16384$, respectively.
  • Figure 3: Snapshots of a compressed aggregate obtained using a two-step numerical procedure. (a) Initial configuration of BPCA-16384-1. (b), (c) Configurations at $\phi = 0.694$ and $0.717$ during the first compression stage. (d), (e) Configurations at $\phi = 0.694$ and $0.717$ after relaxation. The red particles in panels (d) and (e) form the largest rigid cluster (see Section \ref{['sec:rigid']}).
  • Figure 4: Normalized pressure $P/k$ after relaxation. The figure shows the relationship between ${\left( P / k \right)}^{1/2}$ and $\phi$ near the jamming transition. (a)--(c) Results for $N = 16384$. (d)--(f) Results for $N = 4096$. (g)--(i) Results for $N = 1024$. Data with positive $P$ are plotted using filled symbols, whereas data with negative $P$ are plotted as $- P$ using open symbols.
  • Figure 5: System-size dependence of the jamming point $\phi_{\rm J}$. Symbol shapes (circle, triangle, diamond, square, and pentagon) correspond to those in Figure \ref{['fig:4']}. The solid lines represent the best fits obtained by least-squares fitting to Equation (\ref{['eq:phi_J_fit']}).
  • ...and 5 more figures