Modelling Time-Dependent Flow through Railway Ballast

Round 1

Reviewer 1 Report

The present study proposed a numerical methodology to predict the water flow through railway ballast under consideration of different permeabilities, geometries (rectangular and trapezoid) and fouling. The information from the numerical methodology, when it is used to predict the water flow in ballast, can be interesting for the railway ballast management. The methodology seems reasonable to consider the complex fluid flow in the porous structure on a macro-scale basis and was also successfully used in the past by the authors (see [9] and [16]). However, the major concern in this study is the novelty. So, nearly one half of the presented results were published before in a slightly refined version. In the present study the main focus was the transient characteristic of the fluid flow through ballast based on a transient boundary condition at the inflow. For this purpose, steady-state simulations were carried out in section 3 and 4.1 for validation of the numerical model. Especially this steady-state validation process is redundant since it was extensively shown in [9] and [16] that the model works for railway ballast with different geometries, fouling etc. Even the parameters are the same (see Table 1 and figure 16). Only the cases with the transient boundary at the inflow seems new in the paper, which is very short here. Even the transient simulation in section 3.1 seems a validation process, because it was compared to data from literature. Only the simulation in section 4.2 can be seen as new to the scientific community. Unfortunately, the discussion and presentation of the “new” stuff is very short and conclusions about the effect of fouling etc. on the water flow are hard to do based on the limited results. Therefore, I cannot recommend the paper for publication in the present form.

--> Since the main focus of the study is the transient behaviour of the free surface in realistic railway ballast for a fast increase of the inflow height, a detailed discussion and presentation of the results is missing (only limited to section 4.2). In section 4.2 a detailed discussion about Model 2 and Model 3 is missing.

--> Figure 18: The Model 1 represents a lower conductivity then the clean ballast. As expected due to the lower conductivity between 0.1 and 1 second the difference on the free surfaces increase between the models. However, between 1 and 5 seconds the free surfaces are getting closer to each other (clean ballast and Model 1). How is this possible? Because the Model 1 has a lower conductivity and the difference between the free surfaces should be steadily increasing.

--> Which software was used for the study? Was this an in-house code?

--> Table 1: Units are missing.

--> Line 93: “Figure ??”

--> A better explanation of the hydraulic conductivity and the specific yield should be implemented.

--> Figure 9 and Figure 13: The reference Zuyang et al. is wrong. It should be Ye et al.

--> In some figures, arrows and the time captions are shifted and cannot be identified sometimes. The same for the zoom in Figure A1, A2 and A3

--> I would change “nonhomogeneous” to “inhomogeneous”.

--> Figure 16: It should be Ks not k.

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The work is very interesting, the authors have chosen a very complex object for modeling.

a) General remarks.

1) The article contains a lot of references to other articles that contain significant points for understanding, in order to fully understand this work, it is necessary to find and work through a significant part of the cited works, so this article is not self-sufficient. In particular, it is not clear what calculation model is used and where equation 1 comes from, what approximations are used, how the properties of the material and liquid affect, etc. Perhaps this is all in other articles, but it is not clear from this article.

2) The conclusions are too short, and they talk about what was done, this is already clear from the abstract and the text of the article. In the conclusions, it is necessary to show what kind of scientific results are obtained. The conclusion that the time step must be chosen with care can hardly be considered a full-fledged scientific result.

b) Technical notes

1) The authors consider a 2D problem, for non-specialists in railway construction, from Figure (1) it is not clear how the computational area is oriented relative to the railway tracks. It is advisable to provide an explanatory picture.

2) Figure 1 shows the calculation area with vertical outer boundaries. Do such boundaries occur in practice? Or is it just a test task?
3) Decide if you are using the /phi or /varphi symbol.
4) Line 93 contains "??" instead of picture numbers.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors have addressed some reviewer comments, however, only some slight changes in the manuscript were done. Furthermore, the authors responded that some other comments were revised, but the reviewer was not able to find the changes in the revised manuscript. This is weird. Since the authors did not address or revise all the reviewer comments I cannot recommend publication.

à Ad comment 1: The authors mentioned they added more material but the reviewer cannot find new data in the revised manuscript.

à Ad comment 2: accepted

à Ad comment 3: The authors responded that details in section 4.1 were expanded. However, no new information can be found in the revised manuscript.

à Ad comment 4: accepted

à Ad comment 5: I cannot find a Figure 18b.

à Ad comment 6: Why is this information not included in the revised manuscript?

à Ad comment 7: Not revised

à Ad comment 8: The error is still there. Now in line 94.

à Ad comment 9: This was not done.

à Ad comment 10: Accepted

à Ad comment 11: Accepted

à Ad comment 12: This was not done.

à Ad comment 13: Accepted

 

Author Response

Hi,

I am very sorry - but we uploaded the wrong file by mistake.  The correct file is now uploaded (I sincerely hope).  Changes should be clearly marked in blue. 

Cheers Mike

Round 3

Reviewer 1 Report

The authors have addressed the reviewer comments. The paper can be published. However, some slight changes should be done before:

à Ad comment 1: accepted

à Ad comment 2: accepted

à Ad comment 3: accepted

à Ad comment 4: accepted

à Ad comment 5: accepted

à Ad comment 6: Please include the information that MATLB was used in the manuscript.

à Ad comment 7: accepted

à Ad comment 8: accepted

à Ad comment 9: Coarse description of the specific yield should be implemented.

à Ad comment 10: accepted

à Ad comment 11: accepted

à Ad comment 12: accepted

à Ad comment 13: accepted

 

Author Response

1) We have added that the finite element code is built from the code in matlab

2) We have highlighted the description of the specific yield.