One of the emerging technologies in sustainable energy is Floating Offshore Wind. GustoMSC designs a structure for Floating Offshore Wind, the Tri-Floater. One of the problems that need to be investigated before building the unit is wave impact loading. Wave impact occurs when a large wave hits the structure, placing large forces on the structure.

In this research project, a model is developed to find the wave impact loads in 3-hour sea states in which the significant wave height and period can be set. The model will work using CFD solver ComFLOW and time-domain solver aNySIM.

In the first part of the thesis, a CFD model and an aNySIM model are created and verified with model test data using decay test analysis. Decent results were found for the period and damping of the motion by both the CFD and aNySIM model. When more time is spent on the CFD model, it could closely match the model test results.

In the second part of the thesis, waves are included in the simulation. A grid sensitivity study is done with 2 m, 1m and 0.5 m grid sizes. The difference between these grids is slight regarding wave height dissipation.
Irregular waves are used to do longer simulations. Two tests are done, a full 3D CFD simulation and a combination of a 2D CFD simulation with time-domain solver aNySIM. When comparing the computational time, the 2D CFD simulation proved to be the most practical solution, saving much computational time.

Finally, a methodology has been designed in which a long term wave impact study can be done.
A 2D CFD simulation is performed to simulate the wavefield in the domain. This 2D CFD simulation is used as input for aNySIM, in which the motion of the Tri-Floater is simulated. From this motion, the air gap is calculated, and events are selected that will be simulated in a full 3D CFD simulation. The 3D CFD simulation uses initial and boundary conditions from the 2D wave field. It also uses the motions that are calculated in aNySIM for stability.
In this 3D CFD simulation, the wave impact pressures can be obtained.

The methodology results are hard to judge for value, as there are still some problems that need to be solved. When these problems are solved and more time is spent on improving the individual parts of the simulation, it would be interesting to compare the results to model test data if this is to be done in the future.