We thank you for your review and helping us improve the manuscript. Your comments are addressed below:

• Abstract Line 18, 19 and conclusion line 352 onward: The text was changed like suggested. Currently, the GWP of the shaft makes up less than 1/3 of the overall main bearing unit. As the predesigns have heavy, over-dimensioned bearings, the shaft GWP share would be larger for smaller bearings.
• Line 37: Regarding the section modulus, the explanation assumes that the cross-section area of the shaft (=annulus) is fixed. In this case the outer diameter is a function of the inner diameter (D = sqrt(4*A/pi – d^2)) and similarly the section modulus (S = f(A, d)). In this case, a larger inner diameter results in a larger section modulus (see pdf figure), which enables the transfer of higher bending moments or reversely a cross-section reduction.
• Section 3.2 and Line 207: To improve readability, a table was inserted with the component masses as suggested. Similarly, the letter symbols (masses) were simplified/removed from the text. Likewise, the upwind/downwind main bearing have been marked in the figures and the bearing housing been differentiated form the machine frame (Figure 1).
• Line 85: The word surface surcharges mean the added material for forging/casting to ensure that the final geometry can be machined within the required tolerances from the forged/cast geometry. A more general term for this should be “allowance”.
• Line 199-203: The postprocessing (machining) cost are relatively small compared to the total shaft manufacturing costs. Given the KNIGHT references, they make up around 10% of the total costs. For solid forging, the required material for forging includes the material which is later drilled out. As the costs scale with the mass, the additional bore material therefore also increases the machining costs. How the machining costs split up (sawing off shaft ends, drilling, turning, …) is not considered given the low cost share.
• Line 350: The percentages are to be taken literally => The hollow forging surcharges are currently 160% of the final shaft mass and must be reduced to 50% of the final shaft mass (= -110%) to be evenly in costs compared to 32% for casting. The text was changed to avoid the confusion.
• Figure 6: The cost diagrams were updated. Instead of showing the lower and upper (0lower + spread) bounds of the shaft costs, they now include the mean costs and are scaled in size to reflect the overall cost ratio.

The minor grammatical comments have been corrected.

We thank you for your review and helping us improve the manuscript. Your comments are addressed below:

Clarification needed

• (1) Line 53: The capabilities of WISDEM / DrivetrainSE were indeed oversimplified. The tool uses a hollow shaft as a rotor shaft, which can be defined from the outside and iterated over multiple runs. What the tool lacks is – as stated by the reviewer – the ability to consider the manufacturing process constraints for the geometry and the corresponding surcharges in the strength assessment. As suggested, the addressed line was changed.
• (2) Sec. 3.2: In Line 155, the “both variants have similar bearing configurations” refers to the two predesign (hollow forged, cast) not a predesign and the original maxcap141 MBU. The two predesign use two SRBs each and therefore have a similar bearing configuration, which makes comparing the shafts easier. The comparison to the original MBU should only showcase, that the cast shaft mass is similar and that the predesigned cast shaft therefore is a reasonable comparison point.
  
  If the hollow shaft version were to be build, factors such as the change in bearing type would have to be examined beforehand and the design adapted if necessary. This also includes designing the bearing housings or checking if a ring creep lock is necessary.
  
  => A line was added that the bearing types chosen by the predesign tool differ from the original bearings.

Improvement Suggestions

• Material Property Uncertainty: It’s true that the FKM guideline doesn’t apply to AHD steel giving a lack of data and different material properties (e. g. airhardening). Lab and industry samples of AHD alloys show, that the tensile strength is relatively independent of the wall thickness compared to quench tempering (QT) steels like 42CrMo4. Similarly, the alloy hardens under cyclic loads (cyclic hardening), such that the fatigue strength increases unlike other steel types. Both effects are not reflected in the FKM guideline and the calculating of the shafts utilization as if the AHD steel were from QT steel therefore results in unused safety margins. It’s therefore assumed that the stress assessment is valid and oversizes the shaft.
• Lack of Reliability Analysis: Given that the predesign phase is prune to assumptions, the study intentionally only compares two predesigns, such that the load assumptions, drivetrain stiffness, …, are considered equally.
  
  Regarding the loads, the study uses an existing turbine as an example such that the uncertainty in the loads is comparably low. (The static and dynamic load cases were conservativly picked / combined from a set of load cases required in WT design.)
  
  The forging and casting surcharges are the result of industry feedback to ensure that these are realistic. Especially for hollow forging, the surcharges are taken from single-piece manufacturing such that the risk of forging errors (inability to machine the final shafts geometry from the forged shaft) is minimized. Larger surcharges in turn reduce the yield strength of the material and therefore increase the utilization rate, which increases robustness.
  
  Material properties see point above.
• Economic Viability Discussion: Given cartel/antitrust laws, it’s hard to get any kind of data for cost estimations. We therefore used available literature and tweak/verified the model with internal data and industry feedback. Given that the costs scale with the required material, the largest cost variable are the manufacturing surcharges – which is why the hollow forged shaft is so costly. Alloy/element prices and energy prices (e. g. gas) also effect the cost but over a longer period, such that the effect of, e. g. a 20% nickel price increase, cannot be translated to a day-to-day cost difference. The cost model was kept simple on purpose to get a general idea of how a hollow forged shaft compares to a cast shaft.
• Digital Twin or Monitoring Integration: As hollow forging a WT rotor shaft from AHD steel has a technical readiness level of 4 – feasibility proven with lab samples and simulations –, topics like digital twins and condition monitoring are outside the papers focus. Condition monitoring (and maintenance) focuses on components that are affected by wear and fatigue like bearings and gears. The rotor shaft is not considered critical in these cases, as it is designed such that no fatigue damage builds up. Using material testing and post forging quality assurance ensures that the shaft can take on the loads.
  
  A condition monitoring system would instead focus on the shaft interfaces, where wear can happen: e. g. ring creep damage between shaft and inner bearing ring. Especially ring creep is a topic for hollow forged shafts, as the thinner wall thicknesses compared to a cast shaft could increase the risk of structural induced ring creep.
  
  Digital twins similarly will most likely consider the rotor shaft as a structural component and focus on how the shaft deformation effect the bearings for example.
  
  What could be interesting is mirroring the hollow forging process with a digital twin, such that the twin can predict what local material strength the shaft will have (as the material strength depends on the amount of deformation during forging and how the material cools down). A digital twin might furthermore intervene with the forging process and adjust forging steps when the material flow gets undesirable, such that the amount of additional material to minimize forging errors can be reduced.