Predicting the drag coefficient of a granular flow using the discrete element method

Passive-protection structures, against snow avalanches, are designed with rough estimates of the drag coefficient, depending straightforwardly on the obstacle's geometry. In this paper, assuming an avalanche as a dry granular flow, a numerical model is presented for studying the influence on the drag coefficient of both the shape and size of an obstacle impacted by a granular flow. Small-scale laboratory experiments were conducted to validate the numerical model. During the experiments, velocity profiles were estimated using the particle image velocimetry method applied to the pictures recorded by a fast camera, which focused perpendicular to the lateral wall. Flow thickness variations along the slope were estimated by post-processing the images recorded by another fast camera, which films a laser line projected on the free surface flow. From the lateral motion of that line, the thickness could be determined. The granular impact force was measured through an instrumented obstacle positioned at the lower end of a canal. A 3D numerical model, based on the discrete element method using the YADE code, was set up to reproduce the experimental configuration. The law of contact between discrete elements involved elastic components (normal and tangential stiffnesses) and dissipative components (a normal restitution coefficient and a friction coefficient based on the Coulomb friction law). The model was validated by comparisons with both the experimental flow characteristics (velocity profiles and thicknesses) and the impact load history. Once validated, the numerical model was used to investigate the contribution of the height and shape of the obstacle to the drag coefficient. Finally, results are discussed and compared with ones from other studies.