Development of a New Test-Rig for Wheel–Rail Contact Experiments under Large Slip Conditions
Round 1
Reviewer 1 Report
This article concerns the development of a new experimental wheel-rail contact simulator. The design of the device has already been presented in ref. 32 (doi: 10.1088/1742-6596/2355/1/012038) and much of the information contained in the current article is merely an extension of an earlier publication. New experimental results are presented but not adequately discussed. The manuscript might be suitable for publication, but the following major objections need to be addressed:
1. There is no discussion of the results presented in chapter 4, especially 4.2, 4.3 and 4.4. For example, the following questions should be answered: How much does the data correspond to other papers? What is the mechanism of the “transient” behaviour (during the continuous water application?)?
2. What is the water application process during the experiments in Chapter 4.2? It appears that there is a continuous application through the sprinkler system. However, the “cleaning effect” is usually associated with an immediate drop in adhesion due to a single water application.
3. Adhesion coefficient usually increases over time due to the “wear-in” process in test rigs with repeated contacts (under “dry” and water-lubricated conditions). That could also be a possible explanation for the behaviour in Fig. 13 and 14. Time test with constant slip could confirm or deny it.
4. What is the explanation for the opposite trend in the case of leaves in Fig. 15(a)? On my opinion, it's simply that the adhesion coefficient decreases in time during the run-in of the leaf layer.
5. Chapter 5. “Conclusions” is only a summary of “what has been done” in the study and does not bring any real conclusions. This section must be modified.
6. Check the equations for Hertz theory in Chapter 2.2. A and B are designated as “equivalent radii of wheel and rail”, but both the radii for wheel and rail appear in equations (4) and (5). Why there are negative signs in equation (5)?
7. What is the Hertzian contact pressure corresponding to axle loads used in chapter 4.3? From my perspective, Hertzian pressure is more indicative than the axle load in the scaled test rig.
8. The authors mention that the design of the new test rig with slip control (instead of traction control) is required to allow the study of the contact under large slip conditions. Most of the situations where the “cleaning effect” have been presented were with traction/braking control instead of slip control. The fact that the creep is coupled with adhesion is usually perceived positively in an experimental setup, as it corresponds to a real contact and allows studying transient phenomena in a much more natural form.
Author Response
Point 1: There is no discussion of the results presented in chapter 4, especially 4.2, 4.3 and 4.4. For example, the following questions should be answered: How much does the data correspond to other papers? What is the mechanism of the “transient” behaviour (during the continuous water application?)?
Response 1: Thank you for your comment. We are very sorry for not presenting discussion of the results clearly. Considering your suggestion, We have added discussion in chapter 4.2 (line 354-358), 4.3 (line 371-372) and 4.4 (line 393-395). As shown in other experimental investigations about adhesion recovery for wet conditions [18,21], with the increase of running speed, the adhesion coefficient of the second maximum increases. The mechanism of this “transient” behaviour is believed to be related to the changes of wheel rail interface. Large creepage in the wheel–rail contact may be used to alter the surface conditions, removing surface layers, affecting roughness, increasing the temperature, and causing the evaporation of water, all contributing to restoring adhesion [3]. The phenomenon that the adhesion recovery decreases as the axle load increases is in accordance with experimental results [18]. However, the rolling contact theory of large creepage, especially containing surface contaminants, involves multiple factors, and it is very complex, still in development. So this phenomenon of adhesion recovery is currently lack of accepted explanation.
In oil and leaves contaminated surface conditions, low adhesion without adhesion recovery during full creepage process can be simply explained by full hydrodynamic lubrication theory: The greater the creepage, the greater the relative velocity, the greater the film thickness, and the smaller the friction coefficient. The measured adhesion coefficient in chapter 4.4 fits well with the theory.
Point 2: What is the water application process during the experiments in Chapter 4.2? It appears that there is a continuous application through the sprinkler system. However, the “cleaning effect” is usually associated with an immediate drop in adhesion due to a single water application.
Response 2: Thank you for your comment. During the experiments in Chapter 4.2, the sprinkler flow of water application is constant at 4L/min. According to lubrication theory, this continuous watering process builds a fixed initial water film thickness, simulating natural rainfall. However, during the transient process of large slip, the water film thickness between wheel and rail decreases due to removal of surface layer, water evaporation caused by temperature rise, changing the lubrication state, leading to an increase in adhesion coefficient. Existing theories cannot fully explain this phenomenon, and the microscopic mechanism of it is still under research [15,16].
Point 3: Adhesion coefficient usually increases over time due to the “wear-in” process in test rigs with repeated contacts (under “dry” and water-lubricated conditions). That could also be a possible explanation for the behaviour in Fig. 13 and 14. Time test with constant slip could confirm or deny it.
Response 3: Thank you for your comment. The wear-in phenomenon does exist, but it is not a transient process. In fact, it has been wear-in for minutes under small creepage with the continuous sprinkler flow before each test to ensure the stability of initial adhesion coefficient. The adhesion coefficient in time test with constant 5% creepage is as following:
As we can see, after several seconds, adhesion coefficient remains stable and is much lower than the second maximum under large slip conditions in Fig. 13 and 14. The adhesion recovery caused by large creepage is observed and repeated by many researchers [18-21]. The mechanism of this phenomenon remains to be studied.
Point 4: What is the explanation for the opposite trend in the case of leaves in Fig. 15(a)? On my opinion, it's simply that the adhesion coefficient decreases in time during the run-in of the leaf layer.
Response 4: Thank you for your comment. In my opinion, the reason why the opposite trend of adhesion coefficient in the case of leaves is that the leaves were not completely crushed at first, with the increasing creepage, leaves were turned into some kind of thin film between rail and wheel, which was colloidal and caused low adhesion like oil. The essence of this phenomenon is that the material parameters at the wheel rail interface was changed after large creepage. Photos of the track in the test-rig are as following:
Point 5: Chapter 5. “Conclusions” is only a summary of “what has been done” in the study and does not bring any real conclusions. This section must be modified.
Response 5: Thank you for your comment. According to your suggestion, We have modified Chapter 5 carefully, highlighting the test results in accordance with other papers and the possible mechanism of the experimental observation (line 425-431). May I point it out, with all my respect, the title of this paper is “Development of a New Test-Rig for Wheel-Rail Contact Experiments under Large Slip Conditions”. As far as I'm concerned, the main conclusions of this paper are the reasonable form of test-rig based on the requirements for adhesion recovery research (line 406), key parameters of test-rig calculated from the equivalence principle (line 410), the accuracy, capability and potential of the test-rig performing adhesion experiments under large slip conditions (line 425). On the contrary, the mechanical explanation of the test results is not the focus of this article. It should be fully discussed in further experimental investigations following this work.
Point 6: Check the equations for Hertz theory in Chapter 2.2. A and B are designated as “equivalent radii of wheel and rail”, but both the radii for wheel and rail appear in equations (4) and (5). Why there are negative signs in equation (5)?
Response 6: Thank you for your comment. In Hertz theory, considering the two elastic bodies in contact, they will meet at a single point, where the normal distance between them is minimal. Near this contact point, without load, the surface shapes of the bodies are represented by second-order polynomials. The vertical relative distance d(x, y) between the two bodies can be written as [33]:
In rail-wheel contact situation, ψ=0, then:
Point 7: What is the Hertzian contact pressure corresponding to axle loads used in chapter 4.3? From my perspective, Hertzian pressure is more indicative than the axle load in the scaled test rig.
Response 7: Thank you for your comment. The Hertz contact pressure is easily calculated according to equation (1). When applying the axle loads used in chapter 4.3 to the equation,we can get 818.4MPa for 14t, 836.7MPa for 15t, 855.0MPa for 16t and 871.8MPa for 17t. We have added the corresponding Hertz contact pressure in the paper (line 362-364). In the rail vehicle field, axle load or vertical force is more widely used because it is more direct and easier to measure.
Point 8: The authors mention that the design of the new test rig with slip control (instead of traction control) is required to allow the study of the contact under large slip conditions. Most of the situations where the “cleaning effect” have been presented were with traction/braking control instead of slip control. The fact that the creep is coupled with adhesion is usually perceived positively in an experimental setup, as it corresponds to a real contact and allows studying transient phenomena in a much more natural form.
Response 8: Thank you for your comment. As you pointed out that creep is coupled with adhesion and force control is a much more natural form of vehicle traction/braking control. Therefore, in our previous studies [18], adhesion recovery was investigated using test-rig based on actual vehicle braking devices. However, the change in interface state caused by slip between wheel and rail is the mechanism of adhesion recovery, so studying creepage is more essential than adhesion, meanwhile the latter is often interfered by braking devices. In the future, the research on applications of adhesive recovery will raise concern for traction/braking control, especially in combination with anti-skid control. Then, the traction/braking control of adhesion becomes more meaningful.
Author Response File: 
Author Response.pdf
Reviewer 2 Report
The article highlights peculiarities of an innovative new test-rig for wheel-rail contact experiments under large slip conditions in order to further theoretical and experimental research on adhesion recovery. The authors showed potentials of a new test-rig for exploring adhesion recovery under various surface conditions. They performed adhesion experiments under large slip conditions with different speeds, axle loads and surface contaminants.
The article is interesting, but a number of shortcomings need to be corrected:
1. The text in Figure 9 cannot be recognized.
2. The font size in Figure 11 should be increased.
3. In Fig. 13, the authors should set the same maximum value of the adhesion coefficient on the ordinate axis for better visualization, as shown, for example, in Fig. 14.
4. The authors should explain in more detail the differences between the trends of change in the adhesion coefficient, presented in Fig. 14a, b and 14c, d, in the process of the creepage falling.
Author Response
Point 1: The text in Figure 9 cannot be recognized.
Response 1: Thank you for your comment. We are very sorry for not drawing the figure clearly. We have reprocessed Figure 9 and have increased the font size in the pictures. Now Figure 9 in line 283 is readable. Figure 9 is as following:
Point 2: The font size in Figure 11 should be increased.
Response 2: Thank you for your comment. Considering your suggestion, the font size has been increased in Figure 11 in line 309 to make it readable. Figure 11 is as following:
Point 3: In Fig. 13, the authors should set the same maximum value of the adhesion coefficient on the ordinate axis for better visualization, as shown, for example, in Fig. 14.
Response 3: Thank you for your comment. We have set the same maximum value of the adhesion coefficient on the ordinate axis in Figure 13 according to your suggestion. Now Figure 13 (line 343) becomes more visualized. Figure 13 is as following:
Point 4: The authors should explain in more detail the differences between the trends of change in the adhesion coefficient, presented in Fig. 14a, b and 14c, d, in the process of the creepage falling.
Response 4: Thank you for your comment. According to your suggestion, We have added the explanation of the differences between the trends of change in the adhesion coefficient, presented in Fig 14 (line 372-377):
“Another interesting phenomenon is that the trends of adhesion coefficient in the process of the creepage falling appear to be different. In Figure 13(a)and(b), slip-adhesion curves are above those in the whole process of creepage rising. In Figure 13(c)and(d), adhesion coefficient drops below the first maximum. The same phenomenon was reported in former experimental investigations [18,19], but the mechanism of it remains to be studied.”
In the cited literatures, the same phenomenon is shown below, but lack of mechanism explanation currently.
Author Response File: Author Response.pdf
Reviewer 3 Report
This could be modified. VAS already has a new type.
In general the paper is ok. With this test rig concept, there are always transverse forces. These test rig concepts have not actually become established in Europe. Only the TU Deft has this.
Author Response
Point 1: In general the paper is ok. With this test rig concept, there are always transverse forces. These test rig concepts have not actually become established in Europe. Only the TU Deft has this.
Response 1: Thank you for your comment. Exactly as you said, transverse forces occur to the wheel-on-ring test-rig because of the spin creepage. We have studied the effect of lateral forces and found that the ratio of lateral force to longitudinal force on the test-rig is very small when the circling diameter of the test-rig is greater than 2 m and the creepage is greater than 0.1, as shown in the following figure. Due to the focus of this paper on longitudinal forces and length constraints of this paper, this part of the research has not been included. Similar test-rigs have been introduced in Table 1 (line 101). Special thanks to you for your good comment.
Reviewer 4 Report
+ General view;
This paper is a valuable paper concerning test-rigs for wheel-rail contact under large longitudinal slip, but some revisions are necessary for sufficient considerations.
+ Individual comments;
(1) The title of this paper is concerned on “test-rig for wheel-rail contact”, and not only on “longitudinal creepage or slip”. The researches concerning “creepage or slip” extends into many different fields, not only longitudinal slip, i.e traction or/and braking. You had better consider and describe various fields of wheel-rail contact researches and test-rigs, if you use the present title of this paper. The “wheel-on-ring” type test rigs are suitable for longitudinal large slip, but rather not suitable for the other kinds of creepage. The additional descriptions should be desirable in chapter 1 and 2 , especially in section 2.1.
The reviewer recommends to consider the following papers concerning the test-rig and experiments for general purpose of wheel-rail contact, as examples of previous papers;
1) Matsumoto A.; Sato Y.; Ohno H.; Mizuma T.; Suda Y.; Tanimoto M. & Oka Y. Study on curving performance of railway bogies by using full-scale stand test. Vehicle System Dynamics. 2006, Vol.44 sup 1, 862-873
2) Matsumoto A.; Sato Y.; Tanimoto M. & Oka Y. Wheel-rail contact mechanics at full scale on the test stand. Wear. 1996, Vol.191, 101-106
(2) The “axle load” and “test velocity” should be described in the caption of figures;
“load=14t” in Figure 13, “velocity=60km/h” in Figure 14, “load=14t, velocity=40km/h” in Figure 15, etc. in order to grasp at a glance.
Author Response
Point 1: The title of this paper is concerned on “test-rig for wheel-rail contact”, and not only on “longitudinal creepage or slip”. The researches concerning “creepage or slip” extends into many different fields, not only longitudinal slip, i.e traction or/and braking. You had better consider and describe various fields of wheel-rail contact researches and test-rigs, if you use the present title of this paper. The “wheel-on-ring” type test rigs are suitable for longitudinal large slip, but rather not suitable for the other kinds of creepage. The additional descriptions should be desirable in chapter 1 and 2 , especially in section 2.1.
Response 1: Thank you for your comment. We have described various fields of wheel-rail contact researches and test-rigs in section 2.1 and in Table 1 (line 101). For example, bogie/vehicle roller is suitable for railway vehicle dynamics investigation (line 107), wheel-on-rail test-rig is suitable for investigating rolling contact fatigue (line 113) and wheel-on-ring test-rig is used to the study of noise and vibration (line 115). Then, the applicability of wheel-on-ring test-rigs for adhesion research under wet and contaminated surface conditions is highlighted (line 116-120, 135-136). The wheel-on-ring test-rigs are suitable for longitudinal creepage research only when the distribution of longitudinal creepage on the contact patch similar to the actual vehicle. This has been fully discussed in section 2.3. We hope our response will solve your doubts.
Point 2: The reviewer recommends to consider the following papers concerning the test-rig and experiments for general purpose of wheel-rail contact, as examples of previous papers;
1) Matsumoto A.; Sato Y.; Ohno H.; Mizuma T.; Suda Y.; Tanimoto M. & Oka Y. Study on curving performance of railway bogies by using full-scale stand test. Vehicle System Dynamics. 2006, Vol.44 sup 1, 862-873
2) Matsumoto A.; Sato Y.; Tanimoto M. & Oka Y. Wheel-rail contact mechanics at full scale on the test stand. Wear. 1996, Vol.191, 101-106.
Response 2: Thank you for your comment. Considering your suggestion, more papers are cited when discussing the category and usability of test-rigs in line 105-107:
“Bogie/vehicle roller like Full Scale Roller Rig of State Key Laboratory of Traction Power in SWJTU and Full-scale Stand Test in NTSEL [29,30], is capable for railway vehicle dynamics investigation on account of suspension system.”
Point 3: The “axle load” and “test velocity” should be described in the caption of figures;“load=14t” in Figure 13, “velocity=60km/h” in Figure 14, “load=14t, velocity=40km/h” in Figure 15, etc. in order to grasp at a glance.
Response 3: Thank you for your comment. We are very sorry for not presenting figures clearly. We have reprocessed Figure 13-15 and some comments have been added. Now Figure 13 in line 343, Figure 14 in line 366 and Figure 15 in line 389 are as following:
Figure 13. Measured slip-adhesion curves in wheel-rail contact experiments under large slip conditions with running speed of (a) 20 km/h; (b) 40 km/h; (c) 60 km/h; and (d) 80 km/h.
Figure 14. Measured slip-adhesion curves in wheel-rail contact experiments under large slip conditions with axle load of (a) 14 t; (b) 15 t; (c) 16 t; and (d) 17 t.
Figure 15. Wheel-rail contact experiments under large slip conditions with various surface contaminants: (a) measured slip-adhesion curves; (b) contaminants on rail surface of the test-rig.
Author Response File: Author Response.pdf
