Verification of the LOGOS Software Package for Tsunami Simulations

Round 1

Reviewer 1 Report

Please see the attached file.

Comments for author File: Comments.pdf

Author Response

Authors response to the remarks of the referee 1 in the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

This paper attempts to solve several benchmark problems to use the LOGOS software developed by the authors in tsunami simulation. The validity of the proposed software is verified by comparing the numerical results obtained with the experimental results.

In the paper, comparisons with a number of benchmarks are made, and I think that the numerical calculation method and the proposed model are valid. However, the paper needs to be revised as the description of the calculation conditions and the method for evaluating the errors are not fully discussed.

I have the following concerns.

1. Overall: LOGOS software the finite volume method (FVM) to solve the 3-dimensional NS equations. On the other hand, in the area of CFD research, a software called OpenFOAM, which is based on the same FVM method, is available as free software and is already being used in various fields including tsunami. What are the advantages of the proposed LOGOS software compared to OpenFOAM? Additional references need to be added and clearly stated.

2. In this paper, there seems to be a lack of uniformity in the evaluation of the errors compared to the experimental results of each benchmark. In some cases, it is sufficient to compare the error of the maximum value of the waveform, while in other cases, it is necessary to compare the consistency of the entire waveform. At the very least, please include a detailed description of how the error calculated by each benchmark is evaluated, including its definition. For example, we think it will be easier for the reader to understand if we describe the errors of all the benchmarks together and discuss the implications of these values.

3. Overall: The performance and computation time of the computers used to run the LOGOS software for this benchmark will be of practical use to the readers.

4. 8, 3.9: Since these benchmarks are experiments in a 3D water tank, I think that the method of setting up the computational domain is of great interest to the readers. In the paper, however, only the number of cells is described. Therefore, you should add more detailed information about the cell structure in the calculation conditions. At least the number of cell divisions in the horizontal and vertical directions should be described.

5. 23, 24: The caption explains the gray notation in the caption. On the other hand, the colors in the diagram are shown in red and blue.

6. 27, 29, 32: Analytical solution in the caption should be written as "experiment".

Author Response

We very much appreciate the overall positive attitude of the referee to our manuscript and thank him for particularly useful comments. The comments and suggestions of the referee are presented in italics.

1. Overall: LOGOS software the finite volume method (FVM) to solve the 3-dimensional NS equations. On the other hand, in the area of CFD research, a software called OpenFOAM, which is based on the same FVM method, is available as free software and is already being used in various fields including tsunami. What are the advantages of the proposed LOGOS software compared to OpenFOAM? Additional references need to be added and clearly stated.

LOGOS and OpenFOAM have a lot in common in terms of mathematical model and numerical method. However, there are differences that can give our software an advantage when simulating free surface flows, for example, the original algorithm for accounting for gravity forces to prevent velocity oscillations in free surface problems, which are typical when using FVM and pressure-based solver (Efremov V.R.; Kozelkov A.S.; Kurkin A.A.; Kornev A.V.; Strelets D.Y.; Kurulin V.V.; Tarasova N.V. Method for taking into account gravity in free-surface flow simulation. Computational Mathematics and Mathematical Physics. 2017. 57(10), pp. 1720-1733.) This information has been entered into the article.

2. In this paper, there seems to be a lack of uniformity in the evaluation of the errors compared to the experimental results of each benchmark. In some cases, it is sufficient to compare the error of the maximum value of the waveform, while in other cases, it is necessary to compare the consistency of the entire waveform. At the very least, please include a detailed description of how the error calculated by each benchmark is evaluated, including its definition. For example, we think it will be easier for the reader to understand if we describe the errors of all the benchmarks together and discuss the implications of these values.

We agree with the remark, and the following changes are made for the manuscript: for each problem, a method for assessing the accuracy in commonly used terms (standard deviation, maximum deviation, and so on) is indicated.

3. Overall: The performance and computation time of the computers used to run the LOGOS software for this benchmark will be of practical use to the readers.

For each problem a description of the resources used and the time of the problem calculation are added.

4. 8, 3.9: Since these benchmarks are experiments in a 3D water tank, I think that the method of setting up the computational domain is of great interest to the readers. In the paper, however, only the number of cells is described. Therefore, you should add more detailed information about the cell structure in the calculation conditions. At least the number of cell divisions in the horizontal and vertical directions should be described.

Corrected.  Additional information about computational grids is added.

5. 23, 24: The caption explains the gray notation in the caption. On the other hand, the colors in the diagram are shown in red and blue.

Corrected.

6. 27, 29, 32: Analytical solution in the caption should be written as "experiment".

Corrected.

Reviewer 3 Report

The present article deals with the comparison of results obtained from LOGOS software package for Tsunami Simulations. Both free-surface flow simulations on the test cases of a collapsing water column, gravity water sloshing in a tank and propagation and run-up of a single wave onto a flat slope and a vertical wall, sliding of a wedge-shape body down a slope, flow around an island and wave run-up over an obstacle were performed by the authors.

The authors have reported the 10 to 15 % error for the results predicted by LOGOS software package, which is quite satisfactory results for such large scale fluid dynamic problem. The quality and soundness of the article is quite interested. In my opinion, the article can be considered, after improvements to the figures in the quality and visibility of text in legends for Figures 19, figure 31,figures 34 and figure 37. 

Author Response

We very much appreciate the overall positive attitude of the referee to our manuscript and thank him for particularly useful comments.

Figures 19, figure 31,figures 34 and figure 37 (and its legends) are corrected.

Round 2

Reviewer 1 Report

I hope you make clear the disadvantages and limitations of the application by using your simulation model. Because usually, the simulation model has both the advantage and disadvantage.

It is useful information for other researchers to reveal them.

Some questions were not responded to in your revised manuscript though you answered “Corrected” in the reply report. You should respond to them appropriately and sincerely.

1. I understood “LOGOS” is not an abbreviation. So, what is “LOGOS”? I want to know the meaning of this name.

2. You did not respond to my comment below though you write “Corrected”.

“L.247-250: I think that the water surface profile in pictures (Fig.4 (c)) and results (Fig.4 (a) and (b)) do not match. Fig.4 (a) and (b) at t=0.6s are similar to Fig.4 (c) at t=1.0s. (1) is it appropriate to compare the results (a), (b), and (c) by time? (2) dimensionless time used in Fig.5 and 6 should also be used to compare the results in Fig.4.”

3. You did not respond to my comment below though you write “Corrected.”

“What is L?. L is not used in Eq.(17).”

(I’m sorry that I might delete the first “L“ by my mistake in my review report.)

4. In L.479, it is mentioned that two cases of H=0.016m and H=0.0032m were conducted. But Fig.31 is the result of H=0.064m. What is right?

Author Response

We very much appreciate the overall positive attitude of the referee to our manuscript and thank him for particularly useful comments. The comments and suggestions of the referee are presented in italics.

1. I understood “LOGOS” is not an abbreviation. So, what is “LOGOS”? I want to know the meaning of this name.

LOGOS is a name that we associate with ancient Greek philosophy, in which this word means a universal rational principle.

2. You did not respond to my comment below though you write “Corrected”.

“L.247-250: I think that the water surface profile in pictures (Fig.4 (c)) and results (Fig.4 (a) and (b)) do not match. Fig.4 (a) and (b) at t=0.6s are similar to Fig.4 (c) at t=1.0s. (1) is it appropriate to compare the results (a), (b), and (c) by time? (2) dimensionless time used in Fig.5 and 6 should also be used to compare the results in Fig.4.”.

It was a mistake with the drawing at 1.0 seconds. We fixed it. The time in the figure is indicated in dimensional terms, and on the graph in dimensionless form. Here we follow the format of the original article (Ubbink [30]) and hope that this will be more convenient for the readers.

3. You did not respond to my comment below though you write “Corrected.”

“What is L?. L is not used in Eq.(17).”

(I’m sorry that I might delete the first “L“ by my mistake in my review report.).

Thanks for reading carefully, this fix was unfortunately not included in the original article. We removed the L.

4. In L.479, it is mentioned that two cases of H=0.016m and H=0.0032m were conducted. But Fig.31 is the result of H=0.064m. What is right?

H=0.032 m is correct.