code to develop a fully integrated floating wind turbine model [96]. Among these simulation tools briefed above, it is obvious an advantage if the simulation tool could couple two existing independent computer programs [97].
Recently, the SIMO and RIFLEX codes have been contained in SIMA analysis tool, a collective program composed of several codes for modeling, simulation and post-processing, proving a ‘tightcoupling’ method to solve the Newton's equation [30]. The SIMA analysis tool has been enhanced to involve a model specifically for simulating the offshore wind turbine. More detailed information can be referred to Luxcey et al. [98] and MARINTEK [99].
Furthermore, a novel approach was proposed to couple a threedimensional flow solver EllipSys3D, with FLEX5, an aeroelastic code, to calculate the dynamics of floating wind turbines under the excitations of wind-wave combined loads [100]. While the Navier– Stokes equation governing the air flow is solved in the EllipSys3D code, the FLEX5 code calculates the aerodynamic forces acting on the blades based on the BEM theory. At each time step, the aerodynamic data output from the FLEX5 code is transferred into the EllipSys3D model. The wind velocities generated by solving the Navier–Stokes equation in EllipSys3D model, in return, are transferred back into the FLEX5 code for the further calculation of the aerodynamic loads. The main advantage of this EllipSys3D-FLEX5coupled model is that the unsteady simulation can be carried out without resolving the boundary layer problem occurred when trying to resolve the flow-flied near the rotor blades.
5.5. Key challenges and future researches
It is widely recognized in the numerical simulation of the dynamic behaviors of floating wind turbines that the currently available tools are generally in lack of algorithms simulating the aerodynamics of long soft leaves and tall towers [78]. In addition, Morison's equation [69] adopted in several programs to calculate the hydrodynamics of floating turbines has been found not fully reliable. Furthermore, the loads acting on the mooring system are usually calculated according to the user-defined relationship between the force and the displacement or quasi-static catenary equations, which is considered oversimplified. The more accurate and reliable dynamic cable method, however, demands vast computational power, which makes it infeasible to include in a comprehensive simulation model.
Future research should concentrate upon the improvements of the aero-hydrodynamic simulations of semi-submersible foundation [101]. Several challenges are foreseen as follows:
1. The simulation of ‘high Reynolds number flows’ around the blades is practically challenging due to the complex structures of wakes, the tip vortices and the high rotational speeds. Therefore, simulating the flowing fluid around the blades based on dynamic mesh demands refined algorithm and considerable computational resource.
2. Although the algorithm is already available to solve the FSIproblem, the complex geometry of the wind turbine and floating foundation requires unbearable computational power to conduct a FSI numerical simulation if no specific optimization is included. Currently, the conventional finite