Lift force on an asymmetrical obstacle immersed in a dilute granular flow

Universality in quasi-two-dimensional granular shock fronts above an intruder

We experimentally study quasi-two-dimensional dilute granular flow around intruders whose shape, size, and relative impact speed are systematically varied. Direct measurement of the flow field reveals that three in-principle independent measurements of the nonuniformity of the flow field are in fact all linearly related: (1) granular temperature, (2) flow-field divergence, and (3) shear-strain rate. The shock front is defined as the local maxima in each of these measurements. The shape of the shock front is well described by an inverted catenary and is driven by the formation of a dynamic arch during steady flow. We find universality in the functional form of the shock front within the range of experimental values probed. Changing the intruder size, concavity, and impact speed only results in a scaling and shifting of the shock front. We independently measure the horizontal lift force on the intruder and find that it can be understood as a result of the interplay between the shock profile and the intruder shape.

Force and flow at the onset of drag in plowed granular media

We study the transient drag force FD on a localized intruder in a granular medium composed of spherical glass particles. A flat plate is translated horizontally from rest through the granular medium to observe how FD varies as a function of the medium's initial volume fraction, ϕ. The force response of the granular material differs above and below the granular critical state, ϕc, the volume fraction which corresponds to the onset of grain dilatancy. For ϕ<ϕc FD increases monotonically with displacement and is independent of drag velocity for the range of velocities examined (<10 cm/s). For ϕ>ϕc, FD rapidly rises to a maximum and then decreases over further displacement. The maximum force for ϕ>ϕc increases with increasing drag velocity. In quasi-two-dimensional drag experiments, we use granular particle image velocimetry (PIV) to measure time resolved strain fields associated with the horizontal motion of a plate started from rest. PIV experiments show that the maxima in FD for ϕ>ϕc are associated with maxima in the spatially averaged shear strain field. For ϕ>ϕc the shear strain occurs in a narrow region in front of the plate, a shear band. For ϕ<ϕc the shear strain is not localized, the shear band fluctuates in space and time, and the average shear increases monotonically with displacement. Laser speckle measurements made at the granular surface ahead of the plate reveal that for ϕ<ϕc particles are in motion far from the intruder and shearing region. For ϕ>ϕc, surface particles move only during the formation of the shear band, coincident with the maxima in FD, after which the particles remain immobile until the sheared region reaches the measurement region.

Lift and drag in intruders moving through hydrostatic granular media at high speeds

Recently, experiments showed that forces on intruders dragged horizontally through dense, hydrostatic granular packings mainly depend on the local surface orientation and can be seen as the sum of the forces exerted on small surface elements. In order to understand such forces more deeply, we perform a two-dimensional soft-sphere molecular dynamics simulation, on a similar setup, of an intruder dragged through a 50-50 bi-disperse granular packing, with diameters 0.30 and 0.34 cm. We measure, for both circular and half-circle shapes, the forces parallel (drag) and perpendicular (lift) to the drag direction as functions of the drag speed, with V=10.3-309 cm/s, and intruder depths, with D=3.75-37.5 cm. The drag forces on an intruder monotonically increase with V and D, and are larger for the circle. However, the lift force does not depend monotonically on V and D, and this relationship is affected by the shape of the intruder. The vertical force was negative for the half-circle, but for a small range of V and D, we measure positive lift. We find no sign change for the lift on the circle, which is always positive. The explanation for the nonmonotonic dependence is related to the decrease in contacts on the intruder as V increases. This is qualitatively similar to supersonic flow detachment from an obstacle. The detachment picture is supported by simulation measurements of the velocity field around the intruder and force profiles measured on its surface.