Dense suspensions, granular materials, non-Newtonian rheology, jamming and shear thickening, statistical physics, computational modeling and simulations.
Dense non-Brownian suspensions exhibit shear thickening, where viscosity increases with applied shear stress. This behavior arises from a stress-induced transition from lubrication-dominated interactions to frictional contacts. We investigate this shear-thickened regime using stress-controlled simulations via the Lubricated-Flow Discrete Element Method (LF-DEM), focusing on normalized shear stress σ/σ₀ = 100. Our simulations span volume fractions 0.72 ≤ φ ≤ 0.80 and bidispersity ratios δ = 1.0 to 4.0, with small particle volume fractions ζ = 0.25, 0.50, and 0.75. While prior studies suggest the relative viscosity ηᵣ follows the Maron-Pierce form ηᵣ = (1 - φ/φₘ)⁻², we find improved agreement with a generalized power law: ηᵣ = (1 - φ/φₘ)⁻ᵝ, where β ≈ 2.3. This scaling holds not only for ηᵣ but also for the normal stress components Σₓₓ, Σᵧᵧ, and the particle pressure Π, all of which collapse onto a unified form: y ≈ (1 - φ/φₘ)⁻ᵝ. Contact pair distributions reveal strong microstructural trends: at low bidispersity (δ = 1.4), contacts align with the compression axis, while at high bidispersity (δ = 4.0), increasing ζ causes large-large contacts to preferentially align along the flow direction. These trends suggest ζ-driven anisotropy in particle organization. We also explore shear jamming, where frictional contacts and rigid clusters arrest flow. Using the pebble game algorithm, we identify jammed states emerging when the instantaneous fraction of rigid particles exceeds m > 0.80. Finally, we examine angular velocity correlations between particles and their neighbors. Rigid particles show clear angular velocity correlations, while non-rigid ones exhibit anti-correlated behavior—highlighting rotational coupling as a structural signature of rigidity.