Simulating 3D waves crashing against large 3D solid structures, such as ship hulls, piers, or large coastlines in the case of a tsunami, require a large number of cells even for a low-accuracy simulation. Traditional methods using the Navier-Stokes equations with a high number of cells require considerable amount of computational resources, and are, often, unfeasible. The 2D Shallow Water Equation (SWE) model delivers satisfactory simulations for the same scenarios, using a much lower number of cells, and therefore, at a lower cost. However, as a 2D model, it fails to resolve the 3D effects (e.g. near obstacles). This leads us to think of a solver in which we could combine the low cost of SWE with the capability of the Navier Stokes Equations to capture 3D effects. Previous research proposed a solver in which the domain is partitioned into 2D and 3D subdomains solved by OpenFOAM and coupled using the same framework to construct the global solution. In this thesis, we build upon this idea and develop a similar environment, moving the SWE computations outside of the OpenFOAM framework to a solver developed at the TUM SCCS, written in C++. On the 3D side, we continue to use interFoam as the driving solver. We implement the coupling of the solvers using the preCICE library by developing a new preCICE adapter for the SWE solver and by extending the preCICE adapter for OpenFOAM. This thesis builds the foundations of 2D-3D fluid-fluid simulations using preCICE, focusing on breaking-dam and open-channel-flow scenarios as validation cases. We compare partitioned to monolithic simulations, and observe qualitatively good results, with the error depending on the direction of the coupling, which in turn depends on the characterization of the flow (supercritical or subcritical).     «

Simulating 3D waves crashing against large 3D solid structures, such as ship hulls, piers, or large coastlines in the case of a tsunami, require a large number of cells even for a low-accuracy simulation. Traditional methods using the Navier-Stokes equations with a high number of cells require considerable amount of computational resources, and are, often, unfeasible. The 2D Shallow Water Equation (SWE) model delivers satisfactory simulations for the same scenarios, using a much lower number of...     »