Smoothed particle hydrodynamics (SPH) is a particle based Lagrangian method that can be adapted for simulating free surface flows. The method is particularly suitable for phenomena in which the flow changes rapidly. A weakness of the SPH method is the very long computational time. However, this can be significantly reduced using new techniques. In this research, the two-dimensional (2D) and three-dimensional (3D) models Tis Isat, developed at the University of Ljubljana, were used with an appropriate coupling procedure. The new 2D/3D coupled model was validated and calibrated against the results of laboratory experiments, the simulation results were also compared to the results of the fully 3D Tis Isat model and to the results of a one-dimensional (1D) finite difference (FD) model. The performed SPH simulations showed good agreement with measurements and the 1D model results in the symmetry axis of the channel. The two greatest advantages of the coupled model are a more realistic description of the water level below the expansion and the significantly shorter computational time as a result of the adopted coupling procedure.
COBISS.SI-ID: 6269025
The paper describes two different ways of defining the terrain roughness in smoothed particle hydrodynamics (SPH) simulations performed with the Tis Isat model, developed at the University of Ljubljana. The model introduces into the SPH method a non-discrete boundary condition with friction. Two basic definitions of terrain roughness are used: (a) as a hydraulically smooth wall, where roughness was controlled by the wall–particle eddy viscosity coefficient; and (b) as a hydraulically rough terrain by elevating the mesh-nodes. The undertaken SPH simulations relate to a dam break at the upper storage reservoir of the pumped-storage hydro power plant Kolarjev vrh in Slovenia. For the first time, such study was performed on a real topography. Water depths at the gauges along the valley were compared with measurements on a physical model and to results obtained using a finite volume (FV) model. The comparison showed satisfactory agreement with the measurements, which are comparable with the FV model simulations.