Offshore wind energy is a fast-growing sector and it is quickly increasing its share in the European electricity mix. However, the costs are still high and the economic feasibility of offshore wind turbines is limited to shallow waters. This limitation plays a relevant role in the development of the sector and limits the offshore wind exploitation area to a small percentage of the potential one. In this scenario, floating support structures may represent a solution. Nevertheless, these type of support structures are not yet economically feasible as their design is challenging. In fact, with respect to bottom fixed structures, they present more complex dynamics and more degree of freedom. In order to cut the cost of floating support structures, it is necessary to be able to predict the wave- and wind-induced loads and motions of the floater and understand the coupling between them. The aim of this project is to develop a three-dimensional numerical model of a floating cylinder in order to study the Fluid-Structure Interaction (FSI) of a spar-buoy support structure for offshore wind applications. The model is based on the coupled use of the CFD solver fluidity, to resolve the fluid dynamics, and an in-house python-developed code, to numerically solve the equations of motion of a rigid body in three dimensions. To achieve this goal, firstly, the python code is developed. Secondly, a Numerical Wave Tank (NWT) containing both air and water is generated and validated with both linear and nonlinear waves. Here, both an unstructured adaptive mesh and a simplex structured mesh are used to describe the domain. Finally, a free heave decay test of a floating cylinder is performed to investigate the accuracy of the model in computing the hydrodynamic coefficients of the floater. Here, the immersed-body method is used to model the presence of the body in the fluid domain. The numerical wave tank developed in this work resulted to be quite accurate and capable of correctly describing both the linear and nonlinear waves propagation. The final model is developed with the use of a simplex structured mesh. In fact, mesh adaptivity resulted to be very challenging and the cost of its implementation exceeded the benefits. Also, a finite element based scheme including a Sweby slope limiter is used to limit advective fluxes in the setup of the NWT. Thanks to this, a less computationally demanding mesh could be used. Finally, the FSI analysis showed that, with the developed setup, the CFD solver is able to accurately predict the natural period of the floater but it underestimates the hydrodynamic damping. The cause of this was attributed to the use of the slope limiter aforementioned. In fact, it smooths the velocity field by means of numerical diffusion and this affects the resulting damping.