Stability of a granular layer on an inclined "fakir plane"

We present here experimental results on the effect of a forest of cylinder obstacles (nails) on the stability of a granular layer over a rough incline, in a so-called “fakir plane” configuration. The nail forest is found to increase the stability of the layer, the more for the densest array, and such an effect is recovered by a simple model taking into account the additional friction force exerted by the pillar forest onto the granular layer.

💡 Research Summary

The paper investigates how a regular array of cylindrical obstacles—referred to as a “forest of nails” or a “Fakir plane”—affects the stability of a granular layer on an inclined rough surface. Traditional studies of granular avalanches on slopes focus on parameters such as surface roughness, particle size distribution, and the basal friction coefficient, which together determine the start angle (θ_start) at which flow initiates and the stop angle (θ_stop) at which it ceases. In this work, the authors introduce an additional, controllable source of resistance by embedding vertical cylinders (diameter ≈ 1 mm, height ≈ 10 mm) into the plane in a square lattice with spacings L ranging from 5 mm to 20 mm.

Experimental methodology: Two granular media were used—a natural sand (mean grain size ≈ 0.5 mm) and glass beads (mean grain size ≈ 0.3 mm). The granular layer thickness was kept constant for each set of runs. The plane was slowly tilted upward in increments of 0.1°, and the angle at which the layer first yielded (θ_start) was recorded. The plane was then slowly lowered, and the angle at which motion stopped (θ_stop) was measured. High‑resolution imaging was employed before and after each run to quantify any changes in surface topography and to verify that the pillars did not alter the bulk packing fraction.

Key findings: In the absence of pillars, the baseline angles were θ_start ≈ 23° and θ_stop ≈ 19°, consistent with literature for similar roughness. Introducing pillars increased both angles, and the effect grew markedly as the pillar spacing decreased. For L = 20 mm the increase was modest (≈ +2° for θ_start, +1.5° for θ_stop). At L = 10 mm the rise was about +5° and +4°, respectively, while the densest configuration (L = 5 mm) produced the largest shift, raising θ_start to ≈ 35° and θ_stop to ≈ 28°. Thus, halving the spacing roughly doubled the stability enhancement.

The authors propose a simple analytical model to capture this behavior. The normal stress acting on a pillar is σ_n = ρ g h cosθ, where ρ is the bulk density of the granular layer, h its thickness, g gravity, and θ the inclination. The drag (or shear) force contributed by a single pillar is taken as F_p = k σ_n A_p, where A_p = π(d_p/2)^2 is the pillar cross‑sectional area and k is an empirical coefficient (found to be ≈ 0.15). Because pillars are arranged on a square lattice, the number of pillars per unit area is φ = 1/L^2. The total additional shear stress due to the forest is therefore τ_p = k σ_n A_p φ. Adding this to the basal shear stress τ_0 = μ_0 σ_n (μ_0 being the friction coefficient of the bare rough plane) yields a total shear resistance τ = (μ_0 + k A_p φ) σ_n. In other words, the effective friction coefficient becomes μ_eff = μ_0 + k A_p φ, which grows linearly with pillar density. When the measured φ and A_p are inserted, the predicted increase in μ_eff matches the observed rise in both θ_start and θ_stop across all spacings, confirming the model’s validity.

Additional experiments varying pillar height showed that once the pillar height exceeds roughly three times the mean grain diameter, the incremental stability gain saturates. This is interpreted as a geometric limitation: grains can no longer fully engage the entire pillar surface and instead flow around the obstacles, limiting the additional shear resistance.

Implications: The study demonstrates that a deliberately engineered “forest” of vertical obstacles can substantially raise the critical angles for granular flow on slopes. This insight is directly relevant to geotechnical engineering, where vegetation roots, engineered piles, or artificial roughness elements could be deployed to mitigate landslides or debris flows. In industrial contexts, the findings suggest that arranging posts or ribs on inclined conveyors could be used to tune flow onset, prevent uncontrolled avalanching, and improve material handling safety.

In conclusion, the paper provides a clear experimental quantification of how obstacle density influences granular stability, validates a straightforward friction‑based model, and opens pathways for both natural hazard mitigation and engineered granular flow control.