The effect of obstacles near a silo outlet on the discharge of soft, low-frictional grains

Soft, low-friction particles show peculiar characteristics in silo discharge, including, for example, non-permanent clogging and intermittent flow. This paper describes a study of hydrogel spheres in a quasi-2D silo. We enforce a more competitive behavior of these spheres during their discharge by placing an obstacle in front of the outlet. High-speed optical imaging is used to capture the discharge. All particles in the field of view are tracked by means of machine learning software using a mask region-based convolutional neural network algorithm. With particle tracking velocimetry, the fields of velocity, egress time, packing fraction, and kinetic stress are analyzed. While in pedestrian dynamics, placement of an obstacle in front of a narrow gate may reduce the stress near the exit and enable a more efficient egress, the effect is opposite for our soft grains. Placing an obstacle above the orifice always led to a reduction of the flow rates, in some cases even to increased clogging probabilities.