Abstract
Two cases are presented wherein the main flow pattern in sand traps changes considerably as a function of discretization scheme, grid resolution and turbulence model. Both cases involve channel flows directed into a desilting basin, where the main current changes from one part of the geometry to the other. The CFD computations are validated with field or laboratory measurements. The first case presented is one of the sand traps of Khimti hydro power plant in Nepal. According to the laboratory measurements, the recirculation zone for this case is close to the bed, with the main current following the water surface. This is reproduced by the numerical model when using a first-order upwind scheme. Using a second-order upwind scheme, the main current is close to the bed, and the recirculation is formed at the surface. The second case is one of the sand traps of Tonstad hydro power plant in Norway. CFD computations predict the main flow field to follow the right or the left sides or the centre of the expansion region, depending on the discretization scheme, grid resolution and turbulence model. Field measurements show that the main current follows the centre of the expansion zone.
Highlights
Since its early emergence in the field of aerodynamics (Hess & Smith, 1967), computational fluid dynamics (CFD) is currently employed to predict fluid flow characteristics in a diverse range of engineering applications
The current study introduces new examples where the solutions of the Navier-Stokes equations solved by CFD are not unique
The results should be validated by physical measurements, and sensitivity analyses for grid resolution, turbulence model and discretization scheme should be carried out
Summary
Since its early emergence in the field of aerodynamics (Hess & Smith, 1967), computational fluid dynamics (CFD) is currently employed to predict fluid flow characteristics in a diverse range of engineering applications. Real-life case studies using CFD have recently been performed in combustion engineering (Akbarian et al, 2018), thermal engineering (Ramezanizadeh, Nazari, Ahmadi, & wing, 2018) and within numerous applications of water engineering. CFD is considered an effective technique for computation of water and sediment flow in sand traps or desilting basins. The numerical model has to compute both the flow pattern and the concentration of the sediments. This has been carried out successfully both for suspended particle movements (Olsen & Skoglund, 1994; Ruether, Singh, Olsen, & Atkinson, 2005) and for computation of bed elevation changes (Esmaeili et al, 2017; Ruether & Olsen, 2006; Török, Baranya, & Rüther, 2017). Olsen and Kjellesvig (1999) computed bed elevation changes in a sand trap with satisfactory results. This article presents two case studies where multiple flow fields are identified for the flow field downstream the expansion zone
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
