Graduation of Jan van Gorsel

Lighthouses have guided seafarers for centuries and still play an essential role in navigating marine vessels and bringing them to safe harbours. With climate change, the already harsh ocean environment may become more severe. It is expected that future wave conditions can compromise the structural integrity of the lighthouses with their violent impact. The hydrodynamic processes of wave transition before, at, and after the point of impact are complex, highly non-linear, and turbulent. There is no exact description of these processes. Therefore, there is a need to further study them in order to better understand the broken wave load and its effect on the structural integrity of lighthouses.

 From the very beginning, wave breaking processes are complex. For instance, prior to impact with shoal-mounted cylindrical structures, waves undergo a rapid transition from deep to shallow water. During this transition the waves shoal, start to break, and once broken they hit the cylindrical structure of the lighthouse. It is therefore necessary to approximate the hydrodynamics. Since, it is difficult to measure the processes using only physical models. Consequently, this research focuses on broken wave load analysis using a process-based numerical model.

 An approach widely used for analysing wave forces on slender cylindrical structures is calibrated for breaking wave loads. Critical factors for determining these loads are the curling factor and slamming coefficient. However, there is still no scientific consensus regarding the exact values for these parameters. Further research, like this study, can help in determining these values.

The main goal of this thesis is to assess the pressure distribution on the structure and quantify the values for the curling factor and slamming coefficient for a shoal-mounted slender cylindrical structure. The STORMLAMP (STructural behaviour Of Rock Mounted Lighthouses At the Mercy of imPulsive waves) project's experimental data of Dassanayake et al. (2019) is used as a benchmark case to validate the proposed computational framework. In addition, the performance of two different Volume of Fluid methods, isoAdvector and MULES, are assessed.

 This study demonstrates that the curling factor and slamming coefficient are significantly different for the broken wave load compared to the breaking wave load. The curling factor is found to be twice as large in the case of a broken wave load. On the other hand, the peak slamming coefficient is five times smaller. Moreover, the isoAdvector solver in comparison to the MULES solver produces results that better resemble the benchmark case results. Although the two solvers do not produce the same peak impact forces, they do both produce similar values for the aforementioned critical parameters. Also, it was found that the isoAdvector solver is faster and more efficient than the MULES solver in the case of a 3D simulation.

 From the results it can be concluded that the chosen numerical framework represents the benchmark case well, and that the methodology used for breaking wave loads can be applied for broken wave loads with adjustment of the curling factor and slamming coefficient.