Numerical modeling of the seabed stresses and scouring at coastal structures by means of SWASH

This thesis was born from the partnership between UNIVPM (Università Politecnica delle Marche) and SAIPEM S.p.A. and the need for the company to evaluate the scouring phenomenon that occurred downstream of a temporary structure for the laying of the pipelines that approached the coast. To analyse the phenomenon of scouring, the study of the bottom shear stresses was carried out, since a variation of them is linked to an unbalance of solid transport. SWASH, a hydrodynamic wave propagation model, was used to analyse the stresses at the bottom. Through this model the necessary parameters for the calculation of the stresses produced by the currents and the oscillatory motion of the waves have been obtained. SWASH is not the only hydrodynamic model used; the SWAN model has also been used in this thesis. The use of the double model is motivated by the desire to reduce the calculation times, since the available data were located in deep waters (about 133 m deep, several km away from the coast) while the area where the phenomenon occurred it is in shallow water (about 3 m deep, about 700 m from the coast). The goal of the SWAN simulation is to find the sea state in shallow water in order to provide boundary conditions for SWASH analysis, with reasonable calculation times. The SWAN simulates the propagation of wave motion, starting from the known values on the boundaries of the domain and then propagates them using the information provided as inputs such as wind and bathymetry. The data used as boundary conditions are a series of storms derived from the DHI hindcast model. Once the state of the sea in shallow waters was defined, the SWASH model was used, which allows to know how the sea state interacts with an object with variable porosity (as in the case analysed, which is composed of two permeable layers and one in the impermeable layer in the middle) to produce an alteration of the currents that act and the oscillatory movement of the waves. In order to reach the definitive set-ups for the SWASH model, several sensitivities took place: two different measures of the computation domain grid (1.285 〖Km〗^2 and 3.209〖Km〗^2 ); Comparison between 2 of the 3 turbulent horizontal viscosity models usable by SWASH: Smagorinsky Prandtl Use and their effects on the results of a single or double layer of sponge; Modification of the mean wave direction (MWD) SWAN output (20 °N compared to 27.11 °N of SWAN simulation), so as not to have MWD too close to the angle of inclination of the grid (30 °N). The last step is to derive the stress down using the stress calculation formulas in the literature and the results obtained from the SWASH simulations. The simulation that best simulates the real case observed, is that with a single layer of sponge and an MWD that deviates from the angle of inclination of the grid. In configurations with a single sponge layer, the energy of the sea state in the area of interest is greater and for this reason it produces greater shear stresses, especially those due to the waves motion. From the simulations even a wave entry angle that is not perpendicular to the coastline has the effect of increasing shear stress.