Paper Number
GG58 

Session
Rheology of Gels, Glasses and Jammed Systems

Title
Network characteristics of heterogeneous reactive colloidal gels with varying interaction potentials

Presentation Date and Time
October 13, 2022 (Thursday) 11:15

Track / Room
Track 3 / Sheraton 5

Authors (Click on name to view author profile)

  1. Mangal, Deepak (Northeastern University, Mechanical and Industrial Engineering)
  2. Nabizadeh, Mohammad (Northeastern University)
  3. Goyal, Abhay (National Institute of Standards and Technology, Infrastructure Materials Group)
  4. Del Gado, Emanuela (Georgetown University, Department of Physics)
  5. Jamali, Safa (Northeastern University)

Author and Affiliation Lines (in printed abstract book)
Deepak Mangal1, Mohammad Nabizadeh2, Abhay Goyal3, Emanuela Del Gado4 and Safa Jamali2
1Mechanical and Industrial Engineering, Northeastern University, Boston, MA 02115; 2Northeastern University, Boston, MA; 3Infrastructure Materials Group, National Institute of Standards and Technology, Washington, DC; 4Department of Physics, Georgetown University, Washinton, DC 20057

Speaker / Presenter 
Mangal, Deepak

Keywords
computational methods; colloids; construction materials; gels

Text of Abstract
Colloidal gels that form from particle-level bonds between attractive colloids inherently result in multi-scale spatial inhomogeneities that are classically described through mesoscale clusters from a mean-field perspective. These mesoscale structural entities are believed to determine their mechanical properties under external deformations. Therefore, it is crucial to characterize the gel structure at different spatial scales to unravel new design possibilities. Particle-level simulations from one end, and continuum models from another have helped build a framework into rheology of these materials; However, these approaches generally do not account for cluster-scale features that are key to a mechanistic understanding of gels. Here, we apply network science tools to characterize the gel structure at the mesoscale and bridge it to the bulk material properties. To this end, we consider C-S-H (calcium-silicate-hydrate) gels which were generated using a combination of molecular dynamics and grand canonical Monte-Carlo simulations. Two DLVO interaction potentials with low and high barrier heights labeled as LS and ES, respectively, were used to generate C-S-H gels for different particle volume fractions φ in the range of 0.05-0.25. We use the Gaussian mixture clustering algorithm to identify sets of closely connected particle clusters and analyze the distribution of their sizes. We see that the ES clusters are rather heterogeneous compared to LS clusters. We also note that ES clusters have similar size distributions and same average cluster size across different φ, whereas average cluster size increases with increasing φ for LS. Our network analysis helps coarse-grain large scale simulations in an effort to bridge the particle-level information all the way to macroscopic measures of the system.