A Calibration Study with CFD Methodology for Self-Propulsion Simulations at Ship Scale

Round 1

Reviewer 1 Report

This paper shows the main findings from the CFD analyses conducted on a famous benchmark vessel by comparing different results data obtained in the JoRes project. The research is attractive for several engineering applications. The paper is quite well organized, and it has a good quality, however some requests can further improve its quality.

The introduction should be a bit simplified in the text, moreover I suggest to add some references on the UQ techniques that can be integrated in your analysis with also other recent applications.

In particular I suggest for example these references:

-  Xia, L., Zou, Z. J., Wang, Z. H., Zou, L., & Gao, H. Surrogate model based uncertainty quantification of CFD simulations of the viscous flow around a ship advancing in shallow water. Ocean Engineering 2021, 234, 109206

- Cravero, C.; De Domenico, D.; Marsano, D. The Use of Uncertainty Quantification and Numerical Optimization to Support the Design and Operation Management of Air-Staging Gas Recirculation Strategies in Glass Furnaces. Fluids 2023, 8, 76.

- Cravero, C; De Domenico, D; Marsano, D. “Uncertainty Quantification Analysis of Exhaust Gas Plume in a Crosswind”. Energies, 2023, Vol. 16, Issue 8, p. 3549.

The benchmark dataset of the vessel should be described in a different section from the introduction. The main geometrical data should be superimposed in a figure.

The mesh has been described and reported with some details, as the Y+ distribution obtained, however it is not clear if has been performed a mesh sensitivity.

Concerning the turbulent models adopted, some consideration must be reported on the better theoretically requested of a Y+< 5. The boundary conditions should be better resumed in some tables.

The time step has been described and the courant has been plotted, however also in this case it is not clear if has been performed a sensitivity.

The results section is complete and clear with many interesting diagrams, maybe you could enlarge some of these.

The conclusions well summarize the main target obtained.

Only minor correction.

Author Response

The authors would like to thank the Reviewer for their constructive feedback and suggestions for improvement. Our answer to the Reviewer's comments are listed below:

 

Comment: 

The introduction should be a bit simplified in the text, moreover I suggest to add some references on the UQ techniques that can be integrated in your analysis with also other recent applications. In particular I suggest for example these references: (list of three references)

Answer:

The authors would like to thank the Reviewer for the suggestion of additional references. These are included in the revised version of the paper. In addition, we provide quantification of uncertainties for CFD predictions using the tasks specific simulation templates employed in this work, which results from earlier verification and validation studies (see last section Introduction). Moreover, the entire text has undergone revision of the language.

 

Comment:

The benchmark dataset of the vessel should be described in a different section from the introduction. The main geometrical data should be superimposed in a figure.

Answer:

In the revised version of the paper, the benchmark vessel MV REGAL is described in a separate section (now section 2) with the ship's main particulars reduced in Table 1.

 

Comment:

The mesh has been described and reported with some details, as the Y+ distribution obtained, however it is not clear if has been performed a mesh sensitivity.

Answer:

Since the work reported in this paper was a "blind" calibration exercise, mesh sensitivity and time step sensitivity studies (which cannot be conducted independently) were not part of the scope. However, as explained in the section Introduction, the simulation templates and numerical settings employed in the present analyses have been subject to earlier verification and validation studies where these questions have been thoroughly addressed, resulting in recommended settings applied in this work. Quantification of said studies are added in the last paragraph of the Introduction.

 

Comment:

Concerning the turbulent models adopted, some consideration must be reported on the better theoretically requested of a Y+< 5. The boundary conditions should be better resumed in some tables.

Answer:

The authors agree with the Reviewer that the use of fine near-wall treatment with Y+<5 is advantageous for a more accurate prediction of the frictional component of forces and moments, as well as for the modelling of boundary layer separation/detachment. This is especially true in the case of Scale Resolving Simulations. In fact, according to the authors' experience using Y+<5 (and preferably <1÷2) is required on propeller and hull appendages in model-scale simulations. However, in full-scale simulations such as the case studied in this paper, fine near-wall treatment renders the setup impractical due to exceedingly large mesh size and a small time step required to keep the desired Courant number level in the areas of mesh refinement. The other reason for choosing high Y+ near-wall treatment is the inclusion of surface roughness in full-scale simulations which relies on the use of roughness modified wall functions. These considerations are added in the revised version of the paper (see section Research Methodology). While the authors agree with the Reviewer that listing the applied BCs in a table would be beneficial for reproduction, the amount of different simulations presented in the paper renders it difficult to create such a table in a good manner.

 

Comment:

The time step has been described and the courant has been plotted, however also in this case it is not clear if has been performed a sensitivity.

Answer:

Please see our reply to the comment regarding mesh sensitivity

 

Comment:

The results section is complete and clear with many interesting diagrams, maybe you could enlarge some of these.

Several figures have been improved with both larger legends and size. Some figures (15, 16 and 18) could not be enlarged without compromising the flow of the manuscript.

 

Reviewer 2 Report

A calibration study with CFD methodology for self-propulsion simulations at ship scale

1.      Include more explanations regarding the MV REGAL.

2.      In abstract there are several abbreviations used which should be discussed properly. For instance e CFD,

3.      Figure 1 is copied which is not acceptable. Justify via your own approach. Or copy right needed.

4.      Include some mathematical steps for CFD simulations. Which software is used for simulations.

5.      Authors claim that “One can hypothesize that…” . There may be some assumptions here.

6.      Include some recent research on current topic.

7.      Discuss some fluctuation for CFD computations for velocity profile and Blasius profile in inlet and vicinity of the plate leading edge. If any.

8.      Source of table 1 needed.

9.      Why r_g=150  um is assumed on line 337?

1  Check English language again and correct some minor typo errors.

 

1Try to improve the quality of figures, some legends are very small.

1Its better to write conclusion in built form.

No serious issues are detected. 

Author Response

The authors would like to thank the Reviewer for their constructive feedback and suggestions for improvement. Our answer to the Reviewer's comments are listed below:

 

Comment 1:

Include more explanations regarding the MV REGAL.

Answer 1:

In the revised version of the paper, the benchmark vessel MV REGAL is described in a separate section (now section 2) with the ship's main particulars reduced in Table 1.

 

Comment 2:

In abstract there are several abbreviations used which should be discussed properly. For instance e CFD.

Answer 2:

The explanation to the abbreviation Computational Fluid Dynamics (CFD) is now added to the Abstract in the revised version. 

 

Comment 3:

Figure 1 is copied which is not acceptable. Justify via your own approach. Or copy right needed.

Answer 3:

The following acknowledgement is added to the title of Figure 1: "(Reproduced with permission from reference [9])". The authors would like to point out that, as mentioned in the section Acknowledgements, the present manuscript has been reviewed by Dr. Ponkratov, manager of JoRes. Thus, all necessary permissions regarding the use of images and data originating from JoRes project have been received. 

 

Comment 4:

Include some mathematical steps for CFD simulations. Which software is used for simulations.

Answer 4:

The CFD software and its version are mentioned in the first sentence of section 3 Research Methodology : " For the numerical simulations conducted in the present study the commercial CFD software STAR-CCM+ (version 15.06.007-R8) was employed."

The numerical approach is described in detail for open water, resistance and self-propulsion simulations, providing the information about the solvers, physics models and numerical schemes in use. Otherwise, the authors do not find it necessary to reproduce the well-known equations governing the solution, since such equations can be found in the references.

 

Comment 5:

Authors claim that “One can hypothesize that…” . There may be some assumptions here.

Answer 5:

An alternative formulation is now used in the revised version: " Such a high reported value of average hull roughness may be caused by the following main factors: vessel age, quality of underwater hull cleaning, and method used for conversion of locally measured values." These were indeed the assumptions made regarding possible reasons for having such a high measured value at the time when the "blind" simulation exercise was performed. As mentioned further in the text, at a later stage of JoRes project, a corrected, lower value for average hull roughness was suggested. However, the latter suggestion was made already after the present calibration exercise had been completed.

 

Comment 6:

Include some recent research on current topic.

Answer 6:

As suggested, we have added in the section Introduction some references regarding the recent research in the field of CFD simulations of ship performance in full scale.

 

Comment 7:

Discuss some fluctuation for CFD computations for velocity profile and Blasius profile in inlet and vicinity of the plate leading edge. If any.

Answer 7:

We provided more observations regarding the solution behaviour in the flat plat case. Beginning from line 314:

" This was done following the recommendation from JoRes case description [11], in the attempt to mitigate a slight acceleration in the boundary layer that may occur locally when using a constant velocity profile. The Blasius profile applied at the inlet was calculated at the distance of x/L = 0.005 from the plate leading edge. A comparison between the two methods to set up the inlet velocity, has indeed revealed that local flow acceleration downstream of the inlet is reduced, and the displacement velocity decreases somewhat faster along the plate, when the Blasius profile is applied. However, only very minor differences between the two solutions were found in the computed distributions of the local friction coefficient, and in its integral value."

 

Comment 8:

Source of table 1 needed.

Answer 8:

As requested, we have provided additional references to the ITTC57 friction line and roughness correction by Townsin used in the 1978 ITTC Performance Prediction Method, which are used in the comparisons presented in Table 1 (now 2). The reference to the Österlund's work was already given in the paper.

 

Comment 9:

Why r_g=150  um is assumed on line 337?

Answer 9:

As mentioned in the same sentence, the main idea was to assess the influence of hull surface roughness variation on calculation results, also considering that higher than usual values of averaged hull roughness were reported in JoRes for this vessel. At the same time, applying too high values of sand-grain roughness height may compromise the performance of roughness modified wall functions. Therefore, at the time of the simulation exercise, it was decided to check the impact of roughness by applying r_g=150 um, which is about twice the value estimated from the flat plate calculations, and equals to the standard value applied in 1978 ITTC performance prediction procedure (even though, the latter value does not refer to equivalent sand-grain roughness used in CFD).

 

Comment 10:

Check English language again and correct some minor typo errors.

Answer 10:

The entire text has undergone revision of the language.

 

Comment 11:

Try to improve the quality of figures, some legends are very small.

Answer 11:

Several figures have been improved with both larger legends and size. Some figures (15, 16 and 18) could not be enlarged without compromising the flow of the manuscript.

 

Comment 12:

Its better to write conclusion in built form.

Answer 12:

This suggestion from the Reviewer was not entirely clear. Given the fact that the other Reviewer commented that "The conclusions well summarize the main target obtained.", the authors decided not to modify the section Conclusion.

Round 2

Reviewer 1 Report

All my comments have been correctly addressed and added in the revised work, now the paper has increased its quality and it is ready for the publcation in this form.

Author Response

The authors thank the Reviewer for their consideration and acceptance of our manuscript.

The 2nd revision is uploaded at the other Reviewer's request for rewriting the conclusion and correcting a few typos still present in the manuscript.

Reviewer 2 Report

Authors claimed that conclusion section is already good which is not correct. Conclusion is so long. The observations are not summarized properly. Therefore, I repeat my suggestion again. 

Re write conclusion section. 

Express main findings properly in built form. 

There are some minor typo errors which need to be correct. 

Author Response

The authors thank the Reviewer for their thorough review of our manuscript.

 

Several typos were still present and have been corrected.

The Conclusion section has been rewritten at the Reviewer's request, shortening it down and making it more structured. The authors have structured the Conclusion as following: Objective -> Main findings -> Supporting figures - > Implications/recommendations.

If the Conclusion section is still not found acceptable, the authors would like an example of the Reviewer's preference of "built form".