Matching Analysis of Carbon-Ceramic Brake Discs for High-Speed Trains

Round 1

Reviewer 1 Report

·         The abstract must be improved to include the brief research background/gap, objectives, methods, and findings

·         In the abstract please also include your recommendations

·         The introduction section was poorly written, no extensive review of previous studies. How does the authors come up with the research gap and problem? I recommend extensive review to be done

·         Even the background review was not done properly. It needs to be redone

·         At the end of section 1 or section 2, I could not find the summary of research gap, research problem and objectives. This will confuse readers to understand the paper

·         The results were not properly discussed. I suggest you to rediscuss your results and compare with previous studies, and back up with scientific reasons

·         Please simplify your conclusions, if possible in numbering or font format

·          

Author Response

We have substantially revised the article in the light of the comments made by the reviewers, please see the updated manuscript in the attachment.

Reply to Q1 and Q2: the abstract has been redone, and it is divided according to your comments as follows:

research background: "Matching analysis is a key-step in the progress of verifying the adaptation of carbon-ceramic brake discs to hign-speed train braking systems"

gap: "relevant research on matching analysis tends to be carried out only on a single parameter of the brake disc, lack of comprehensive analysis"

objective: "a method of parametric and data-driven is proposed to address the problem."

methods: "The method based on the feasibility of modelling the parameters, complete the analysis of non-modelled parameters through comparative conformity check, and modelled parameters through statistical analysis of experimental data."

findings: "Conformity comparison results show that: the example carbon-ceramic brake disc is well suited to high-speed trains, and better matching than example cast-steel brake discs in terms of mass and average frictional coefficient. Analysis of simulated experimental shows that: under high-speed braking conditions, the maximum disc surface temperature and wear of example carbon-ceramic disc is higher than that of cast-steel disc, trains equipped with carbon ceramic discs have shorter emergency braking distances and higher average braking deceleration, exhibits better matching performance."

Reply qu Q3,Q4,Q5: The introduce part has been redone, and mainly expand the section on relevant research and derived research gaps.

in the third paragraph mainly talk about researchers study matching performance by analysing the frictional coefficient, and fourth paragraph mainly talk about research on temperature, wear of brake disc, to analyse matching. The fifth paragraph identifies a research gap: the lack of a comprehensive approach to match performance analysis.

Reply to Q6 and Q7:  the conclusion part was divided into three sections by serial number. The first section summarises the results of the fit analysis for the non-modelled parameters.The second part summarises the modelling of the modelling parameters, and the method of selecting the braking conditions.The third section summarises the results of the computational analysis of the modelled parameters，due to the large number of parameters involved in the conclusions, the third part is longer.

Author Response File: Author Response.docx

Reviewer 2 Report

 

This paper presents a parametric and data-driven approach to study the matching of carbon ceramic brake discs in comparison to cast steel brake disc. However, a couple of key issues need to be carefully addressed by the authors .

（1）The author should give a focused attention in the writing of the article. There are some expressions which confuse the reader. For example, in section 4.2 the expression “Under  the  typical  braking  conditions,  the  parameters  include  brake  deceleration, brake distance, average brake deceleration, maximum temperature and wear were calculated by using a friction pair formed by a carbon ceramic brake disc with a copper -based brake pad and assembled in a CR400AF train as the simulated test group (carbon ceramic group) and a friction pair formed by a cast steel brake disc with a copper-based brake pad and assembled in a CR400AF train as the control group (cast steel group).” was presented as a one sentence. This and other writing formats should be well revised.

Section 3.2.1, this sentence “Among the parameters of the physical properties dimension, the mechanical effects and friction coefficients are also inherent properties of the carbon ceramic dynamic discs, which are tabulated using the same method as in 3.1, as shown in Table 2.” should be rewritten.

The format of the subtitle is inconsistent in section 3.2.2. “a) Maximum temperature of brake disc surface” is in bold text, and “b) The wear of brake disc” is not bold.

(2) The author should re-print all of the figures in the manuscript to make it clear and keep uniform format. The problems mainly include: In figure 2, the two left side figures (one emergency mode figure and another maximum common braking mode) show an overlapping. In figure 3 the legends are completely invisible to the reader.

(3) In section 3.2.1. it was stated “Non-model simulation class parameters”, what does non-model mean and how does it represent the content which is the mechanical property of the two disc.

In this section, it is also stated that the average friction coefficient of the disc. How to obtained these values?

(4) Literature shows there are a couple of methods to test normality of a data. Which one did the author used and why?

(5) In section 4, it is shown that there are 6, 7 data’s for each braking mode. Studies show that the normality test for a small value of data will lead to a higher p-value (>0.05). How can the presented analysis be accepted with the given data?

(6) How to determine the parameters of K, H when calculate the wear brake disc? The details to calculate the wear should be given.

 

Author Response

We have made extensive changes to the article in response to the reviewers' comments. please see the attachment.

Reply to Q1: This text is intended to illustrate the grouping of the virtual experiment, which shall have an experimental group (carbon ceramic) and a control group (cast steel). A significant amount of passive voice has been added in this revision and the meaning of the individual sentences will be made clearer. Please refer to the first paragraph of section 4.2 for the revised text.

Reply to Q2：all of the picures and tables have been redone during this revision.

Reply to Q3: non-modelled and modelled parameters were divided baesd on:

there whether be a good Mathematical models to simulate the value of this parameter. For example, the density of carbon-ceramic brake discs it a static parameter, and frictional coefficient could only be obtained on  test stands. While parameters like brake distance,  although it has more accurate data of field measurement, it is difficult to implement, and the data obtained from calculations using a mathematical model is simple in its approach and has a low margin of error.

In the section 2.2 of revised article, I also explain the reasons for this classification.

The average coefficient of friction data provided in this article were obtained from field tests based on the C6 procedure in the standard “Provisional technical conditions for locomotive gates”. And in section 3.1, paragraph e of the revised article, I have also cited a literature, no. [18], where the method for the average friction coefficient of brake discs is also described in detail.

Reply to Q4: The calculated braking distance and maximum temperature data were imported into “origin” software and proved to be normally distributed using the software's own normality analysis tool.

The two mean-value hypothesis tests done in the article are both problems that can be categorised as multiple sample function type data mean-value tests. And the for the specific process of setting up a rejection domain , reference is made to document number 24.

Reply to Q5: The analysis of braking distances as well as the maximum temperature in the article，the main aim was to draw conclusions from a comparative analysis of the data between the different materials, and therefore not specifically for the different materials and braking modes. We generalised the analysis by fitting all braking distances under one material to the normal fitting, which we acknowledge as a shortcoming of the study.

In the proof process of statistics, if the rejection domain of a hypothesis test is not accepted, it means that the hypothesis cannot be denied, i.e. it is proved.

The updated statistical analysis process will be clearer, please refer to section 4.2.2.

Reply to Q6: the data of brake discs' hardness is given in Table 3. And the K is obtained by generalization from previous experiments, it's intended to fit the model realistically, with different values for different experimental results. For confidentiality reasons, its value will not be made public.

Author Response File: Author Response.docx

Reviewer 3 Report

1. In Section 3.2.1, it is mentioned that "the friction coefficient data of the friction pair under each

set of working conditions were measured on the test stand, and the average friction coefficient values were obtained by numerical fitting method, which were used as the data for the brake simulation calculation. Table 4 shows the average friction coefficient values for the brake discs paired with copper-based brake pads to form a friction pair, with 0.36 for the carbon ceramic friction pair and 0.338 for the cast steel friction pair.". Please explain in detail which numerical fitting method was used to obtain the average friction coefficient? And I suspect “Table 4” here should be “Table 2”.

 

2. In Table 3, what is meaning of the values of train load AW_0、AW_1、AW_2、AW_3? What are the specific values of these values? Please note the specific meaning of the abbreviation that appears for the first time.

 

3. In the Figures of this study (Figure 3, Figure 4), the mean value should be biased, so the data in the figure should have a deviation line.

 

4. There is no field or test data analysis in the paper. How to check the accuracy of simulation results?

 

5. There are “disc” and also “disk” shown in the study. Can you explain the difference between these two words?

 

6. The braking is based on the contact friction between the brake disc and the brake pad. The friction pair of brake disc, namely brake pad, is not introduced in this paper. It is suggested to add a brief introduction of the structure, material and other characteristics of the brake plate.

Author Response

We have revised the article extensively in the light of the reviewers' comments, please see the attachment.

Reply to Q1: We tested the friction coefficient of the brake disc according to the C6 procedure in the standard document "Provisional technical conditions for locomotive gates".

the specific numerical fitting process is: one C6 procedure test will get about 50 friction coefficient data under three braking conditions, after removing the abnormal data to take the arithmetic average as the average friction coefficient obtained from this procedure test.Ten more C6 program tests are performed and the arithmetic mean of the s is taken as the average coefficient of friction obtained from the test.

In paragraph e of section 3.1 of the revised article, we also cite a paper [18] in which a test procedure for a C6 program is documented in detail.

Reply to Q2: The meaning and specific values of these parameters are given in the list in section 0. Other abbreviations are explained as we cover them.

Reply to Q3: For the normal analysis of braking distances in the article, we focused on the differences in data between the different materials and did not analyse the data for the different braking modes. On reflection we felt that the inclusion of deviation lines was unnecessary and therefore no changes were made to Figures 3 and 4.

Perhaps you could express this change in more detail to facilitate our understanding, thank you.

Reply to Q4:  The virtual tests in this paper do not aim to approximate reality at all, as the data we obtained for the brake discs came from prior studies or were provided by third parties. It is intended to verify the feasibility of the matching analysis method proposed in this paper.

We have also explained this in the last paragraph of the introduction.

Reply to Q5: We are very sorry for the misunderstanding caused by the translation. In the article we only use 'disc'.

Reply to Q6: The material of the gate used for the friction sub is a copper-based alloy and the material parameters of the gate are also given in Table 4 of the revised article.

 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

All comments raised by me has been successfully addressed 

Reviewer 2 Report

Most of the issues have been modified. However, some figures are still unclear, figure 3 for example.