Author(s): Vincenzo Sessa; Nils Rüther; Carlo Gualtieri
Linked Author(s): Nils Ruther
Keywords: River engineering; Sediment transport; Artificial gravel bed; Computation fluid dynamics; RANS equations
Abstract: The understanding of flow structures over static armor layers is essential for the development of modern sedim ent transport models. Over a rough boundary, such as in a gravel bed channel, friction created by individual gravel bed particles or cluster of particles retards the flow velocity creating near-bed turbulence, which plays an important role in the dynamics of a river. Very recently, the use of an artificial gravel bed in river engineering studies has proved to be highly advantageous for the investigation of steady flow and for a more easy comparison with the results obtained elsewhere. The present paper investigated the possibility to simulate the flow over an artificial gravel bed. The numerical results were compared with the experimental data collected in a physical model by Spiller et al. (2013) using Particle Image Velocimetry (PIV) .The commercial code STAR-CCM+was chosen to simulate the flow over the gravel bed. The flow was calculated by solving the Reynolds-Averaged Navier-Stokes equations in combination with the standard k-εmodel. The free surface was simulated by the Volume-of-fluid method. Five different discharges were chosen from PIV measurements and simulated to see whether the numerical model could reproduce the general and detailed flow characteristics. The numerical results for the streamwise component of the velocity were in good agreement with the experimental data. In particular, the match was very good for the velocities above the roughness height. Whereas within the roughness height the simulated results tended to deviate from the measured ones. The numerical model tended to overestimate the physical one indicating a percentage error between 4%and 7. 5%and lower accuracy near the bottom. In general it was observed that with an increasing simulated discharge, the deviation of the velocities increased due to grid resolution. The gravel bed created within STAR-CCM+lost the perfect roundness of those elements: the larger was the base size chosen for grid, the smoother was the bed and the effects were more visible with higher velocities.