Advances in Hydrodynamic and Morphodynamic Modelling

The present research focuses on the advanced modelling of complex phenomena of river hydraulics. Nowadays, two-dimensional (2D) depth-averaged hydrodynamic models are common tools in river hydraulics. Though outperforming classical one-dimensional models, 2D models still have limitations. The first theme of the present research is aimed at enhancing the predictive ability of 2D models by including effects induced by large-scale 3D flow structures. The focus is here on the secondary currents generated by the streamline curvature in river bends. A parametrization of curvature-induced helical flow is included in a 2D hydro- and morpho-dynamic model on cartesian unstructured grids by means of additional dispersive terms in the Shallow Water Equations. The non-linear saturation effect, which limits the growth of the helical-flow in case of relatively sharp bends, is modelled with a novel, purely 2D approach suitable for real-world applications. The model also accounts for the effects of helical flow on passive tracers mixing and on the bedload transport. Model applications to laboratory tests and to a real river, under fixed and mobile bed conditions, confirmed the validity of the proposed approach, the relevance of accounting for secondary flows in specific cases, and provided guidelines for a proper application of 2D hydrodynamic models to river flows in bends. Other kinds of geometrical irregularities of the solid boundary produce 3D turbulent structures that require the use of more sophisticated models. In the recent years, eddy-resolving Computational Fluid Dynamics (CFD) models has promoted to describe turbulent river flows. Then, the second theme of the present research concerns the application of these models to river flows over complex geometries. The Detached Eddy Simulation (DES) approach is used to simulate turbulent flows in natural beds in the presence of relatively large roughness elements (i.e., freshwater mussels) and structures (i.e., bridges). Studying the mutual interactions between turbulent flow and mussel shells has important ecological implications because mussels are among the most imperilled fauna, and understanding the flow at the organism scale may be of help for their conservation. Differently from previous works, which only considered smooth beds, the present research uses the representation of a real gravel-bed to simulate more realistic scenarios; indeed, mussels typically live in sand or gravel beds, where the bed roughness plays a role. The analysis moves from the case of flow around an isolated freshwater mussel and extends through considering large arrays of mussels (i.e., mussel beds), to understand the influence of different physical parameters (e.g., number of mussels per unit area, filtering activity, bed roughness, burrowing ratio) on flow, turbulent structures, drag forces on the mussel shells, bed shear stresses, and mixing of clean water exhaled from the mussels siphons. Finally, the DES approach is used to study the turbulent flow at a real bridge with multiple piers, which is of great importance because bridge failures are often caused by hydraulic-related reasons, such as bed erosion and trapping of floating debris. The DES simulations account for the detailed geometry of the bridge and for the real bathymetry of the riverbed; the Volume of Fluid (VoF) numerical technique is used to track the free-surface. Such simulations overcome the limitations of previous studies that either used turbulence-averaged models (e.g., RANS), or referred to idealized geometrical conditions. The goals include understanding how the flow and the shear stress at the bed are affected by the presence of the bridge, in both the free-surface conditions and in the pressure-flow with deck overtopping. The analysis focuses on the main turbulent coherent structures and their interaction with the riverbed, and allows investigating the accuracy of the RANS approach.