Dear Referee RC1,

On behalf the authors of this work we appreciate your time and the comments about our work.

In this work we want to highlight two main outcomes. The first one is that we have tested a new feature in the SiL hybrid method: including the wind turbine moments (not only the thrust). Due to this improvement, the platform response in yaw, induced by the wind turbine out-of-plane moment around the vertical moment “Mz” was obtained. In the subsequent validation of the numeric model in OpenFAST of the SATH 10MW model in full scale it was observed a good agreement between numerical results and experimental measurements of the platform yaw response under the same wind and wave conditions. This shows the good performance of the new feature of the SiL method.

The second relevant outcome is related with a variation of the platform pitch natural frequency obtained in the experiments under high rotor thrust, compared to the natural frequency obtained in free decays. This variation is caused by a change of the hydrostatic stiffness when the platform tilts. During validation of numerical model it was observed that OpenFAST was not able to capture this variation in the natural frequency. The reason is that OpenFAST assumes a linear behavior of the hydrostatic stiffness matrix that is obtained at the un-displaced platform position. Nevertheless, the complex geometry of SATH platform produces a highly non-linear behavior of this stiffness. For this reason, once the stiffness matrix was recomputed for the tilted platform, the pitch natural frequency of the computation agrees with the experiments.

This limitation of the common linear potential hydrodynamic numerical model is an open issue. A better computation of the instantaneous buoyancy loads may improve the simulation of dynamic response of a FOWT with this type of floater geometry.

We have updated the document considering all your comments to highlight the scope of this article and we have listed some of the clarification below:

• For these experiments the tower for the scaled model was designed rigid, with a larger diameter to avoid any elastic response and reduce uncertainties.
• In this test campaign the SPM system was not implemented to ease the uncertainties on the initial validation of the numerical tools and floating design. The retention system constrains all the platform degree of freedom. In contrast, once the SPM is enabled, the platform will be able to freely rotate in yaw offering a different dynamics of the platform.
• The mooring system was modelled using a stiffness matrix (6x6) in Hydrodyn with coupled terms at the respective DoF. First, the matrix coefficients were defined mathematically and afterwards the coefficients were tuned according to the platform response in the free decays.
• A 3% of difference in platform pitch stiffness due to the cable bundle was estimated from experimental pitch free decays with and without cable bundle. The numerical model takes into account this effect by including an additional pitch stiffness coefficient.
• We have discarded that the difference in the PSD peaks are caused by low frequency 2^nd order effects because this difference also appears at the turbulent wind only cases. Thus, we believe that the difference in PSD peaks are related with uncertainties in the characterization of the couplings of the retention systems.

Dear Referee RC2,

On behalf the authors, we thank you for your time and dedication for reading our article.

We will update and prepare a new version of the documents considering your review.

The inclusion of a load cell, to measure force and moment, in the force actuator is an improvement that we are working on. For this tank test we could not fit the load cell in the force actuator. However, we run calibration tests on the actuator before send the device to the tank, to check the response of the motors and also we apply signal filtering to avoid interruptions casued by noise in the signal.

The equilibrium pitch for this floating platform was below -1deg, from the numeric estimation. The RNA mass of the scaled model was adjusted taking care of not exceeding this value. The COG location and the moments of inertia were calculated based on numerical mass distribution calculations. We added significant amounts of lead to the heave plate, the transition piece and the nacelle to achieve the correct distribution. The difference in the MOI is below 1% and the CoG position is below 5mm in any of the 3 directions.

The linear stiffness matrix that represent the mooring system was initially defined analytically according the mooring lines tension, fairlead positions and the scaled model geometry. The coefficient of the matrix were later adjusted to match the experimental results.

We have performed decay tests with wind during the setup of the SiL system to ensure the correct performance of the hybrid system. Nevertheless, for the paper we have focused on the turbulent wind only case that is another form of observe the aerodynamic damping from the wind turbine and the controller actions but under a more complex wind field. 

Comment on Figure 7. We do not consider that the pitch peak from experiments is related with a resonance because its amplitude would be much larger than the one obtained. Instead we see that the difference in PSD curves around pitch natural frequencies are related with changes in the water plane area of the scaled model that varies the pitch natural frequency. The Hydrodyn numeric model uses constant hydrostatic coefficients that are limiting the numeric response of the platform. 

Comment on Figure 8. The Fy from the rotor is included in the numeric model. The rotor is fully simulated in OpenFast producing the respective aerodynamic 3 forces and 3 moment. The multi-propeller actuator only can generate forces in the direction showed in Figure 3. It cannot produce forces in the perpendicular directions "Fy" and "Fz". The sway motion is relatively small compared with the surge motion and could be produced by the difference between experimental and numerical model. The yaw motion could be due to a coupling of the sway-surge degrees of freedom with yaw through the seakeeping system.  

The bundle cable was not constrained to move in any direction. We have accounted it effect during the decay test but for this cases is difficult to determine its influence. Certainly, it should not be discarded.

We are working in the reduction of number of cables to control the actuator and reduce any other bias in the experiment.

Comments related with Figure 9 and 10. After considering the differences between the simulations with second order effects and  just linear, we believe that the differences are small due to the low height of the wave. We have updated the text accordingly.

Comment on Figure 11. This is a very interesting comment and we will study it to get a better understanding of the experimental results for future analysis and publications. Nevertheless, the deadline for the revision of the paper are tight and we do not have time enough to include this analysis with the required background analysis in time.