Reply to "Comment on 'Flow of wet granular materials: A numerical study' ".

In his Comment on our paper [Phys. Rev. E 92, 022201 (2015)10.1103/PhysRevE.92.022201], Chareyre criticizes, as inaccurate, the simple approach we adopted to explain the strong enhancement of the quasistatic shear strength of the material caused by capillary cohesion. He also observes that a similar form of the "effective stress" approach, accounting for the capillary shear stress, which we neglected, results in a quantitatively correct prediction of this yield stress. We agree with these remarks, which we deem quite relevant and valuable. We nevertheless point out that the initial approximation, despite ∼25% errors on shear strength in the worst cases, provides a convenient estimate of the Mohr-Coulomb cohesion of the material, which is directly related to the coordination number. We argue that the effective stress assumption, despite its surprising success in the range of states explored in Khamseh et al. [Phys. Rev. E 92, 022201 (2015)10.1103/PhysRevE.92.022201], is bound to fail in strongly cohesion-dominated material states.

Flow of wet granular materials: A numerical study.

We simulate dense assemblies of frictional spherical grains in steady shear flow under controlled normal stress P in the presence of a small amount of an interstitial liquid, which gives rise to capillary menisci, assumed isolated (pendular regime), and attractive forces, which are hysteretic: Menisci form at contact, but do not break until grains are separated by a finite rupture distance. The system behavior depends on two dimensionless control parameters, inertial number I and reduced pressure P*=aP/(πΓ), comparing confining forces ∼a2P to meniscus tensile strength F0=πΓa, for grains of diameter a joined by menisci with surface tension Γ. We pay special attention to the quasistatic limit of slow flow and observe systematic, enduring strain localization in some of the cohesion-dominated (P*∼0.1) systems. Homogeneous steady flows are characterized by the dependence of internal friction coefficient µ* and solid fraction Φ on I and P*. We also record normal stress differences, fairly small but not negligible and increasing for decreasing P*. The system rheology is moderately sensitive to saturation within the pendular regime, but would be different in the absence of capillary hysteresis. Capillary forces have a significant effect on the macroscopic behavior of the system, up to P* values of several units, especially for longer force ranges associated with larger menisci. The concept of effective pressure may be used to predict an order of magnitude for the strong increase of µ* as P* decreases but such a crude approach is unable to account for the complex structural changes induced by capillary cohesion, with a significant decrease of Φ and different agglomeration states and anisotropic fabric. Likewise, the Mohr-Coulomb criterion for pressure-dependent critical states is, at best, an approximation valid within a restricted range of pressures, with P*≥1. At small enough P*, large clusters of interacting grains form in slow flows, in which liquid bonds survive shear strains of several units. This affects the anisotropies associated with different interactions and the shape of function µ*(I), which departs more slowly from its quasistatic limit than in cohesionless systems (possibly explaining the shear banding tendency).