Wear Mechanism of Current Collecting Materials due to Frictional Heat

Abstract

To reduce wear of current collecting materials such as contact wires and contact strips on electric railways, it is necessary to clarify the wear mechanism, especially mechanical wear due to frictional heat. In this study, the authors carried out wear tests using a new rotary wear tester capable of measuring the contact temperature and analyzed the temperature rise using a new contact model that takes into account the number of contacts. From the test and analysis results, the authors classified wear modes into four types based on the softening of materials and clarified that the seizure wear mode occurs when the contact pressure exceeds the hardness of the material.

1. Introduction

In electric railways, current collecting materials such as the contact wire and the contact strip are subject to wear due to friction under current flow conditions and the timing of their replacement depends on the rate of wear. Therefore, in order to reduce the maintenance costs, it is necessary to clarify the wear mechanism and take measures for reducing the wear.

The wear modes of current collecting materials have been conventionally classified into two types: electrical wear caused by an arc discharge at contact loss point, and mechanical wear caused by contact adhesion [1, 2, 3, 4]. However, in reality, wear phenomena are diverse and cannot be explained by just these conventional classifications.

Recently, the authors carried out wear tests with a linear wear tester capable of measuring the contact temperature caused by Joule heat and analyzed the temperature distribution around contact spots [5, 6]. From test and analyses results, the authors classified electrical wear modes into three types and clarified that these mode transitions occur due to a molten bridge occurring prior to arc discharge. However, mechanical wear mechanisms have not yet been clarified.

In this paper, the authors therefore develop a rotary tester capable of measuring the contact temperature caused by frictional heat and carry out wear tests in order to classify mechanical wear modes. The authors then clarify the mechanism of occurrence of these modes based on the contact temperature and the number of contacts.

2. Wear tests and test results

2.1 Wear tester and test conditions

The rotary wear tester developed to monitor the contact temperature during the test is shown in Fig. 1. The contact strip specimen is pressed against the rotating copper plate which imitates the contact wire. The contact force N [N] and the friction force F [N] are measured with bi-axial load cells set between the contact strip and the linear actuator. The contact temperature T [K] is calculated from thermoelectromotive force V [µV] using (1) [7].

  

$$T=\frac{V}{S}+T_0$$(1)

Here, S is the relative Seebeck coefficient [µV/K], T₀ is the temperature at the electric potential measurement point [K].

Fig. 1　Diagram of rotary wear tester

The material properties of each specimen are shown in Table 1. The copper plate and contact strip should be of different materials because otherwise the relative Seebeck coefficient is 0. As test conditions, the apparent contact area between the contact strip and the copper plate is 10×10⁻⁴ m², the contact force is set to 30 N, 60 N, 80 N, and the sliding speed is set to 0.56-27.8 m/s. The sliding time is set to within 60 s so that the temperature at the electric potential measurement point does not rise more than 10 K.

Table 1　Material properties

Specimen	Copper plate	Contact strip
Material	Tough pitch copper	Iron-based sintered alloy
Principal component	Cu	Fe
Additive	-	Cr, MoS2, Bi, BN
Softening point	Approximately 473 K	Approximately 673 K
Relative Seebeck coefficient	9.6 µV/K

2.2 Test results

Figure 2 shows the classification of four types of wear mode for each test condition. The wear modes are classified based on the contact temperature shown in Fig. 3, the friction coefficient shown in Fig. 4 and the observation results of typical wear surface and wear particles shown in Fig. 5. In Fig. 3, the horizontal axis is the average temperature, and the vertical axis is the statistical maximum temperature calculated by adding 3σ (σ: standard deviation) to the average temperature. The softening points of both materials are shown as dashed lines in Fig. 3. In Fig. 5, wear surfaces are observed using an optical microscope and the wear particles are observed using scanning electron microscope (SEM). The four types of wear modes are specifically classified as follows:

Fig. 2　Occurrence condition of each wear mode (test condition)

Fig. 3　Occurrence condition of each wear mode (contact temperature)

Fig. 4　Relationship between sliding speed and friction coefficient for each mode

Fig. 5　Optical microscope images of wear surface and SEM images of wear particles (sliding directions in wear surface are from left to right)

(i) Wear mode I

This wear mode is observed under conditions of relatively low sliding speed and low contact force as shown in Fig. 2. Note that neither material will soften, since even the maximum temperature is below the softening point of the copper plate as shown in Fig. 3.

The features of this mode are that the friction coefficient is in the range of 0.2-0.3 regardless of the sliding speed, as shown in Fig. 4, and the size of wear particles is in the order of a few microns, shown in Fig. 5. Coulomb's law [8] states that the friction coefficient does not depend on the sliding speed and that the wear particles become finer in the adhesive wear mode.

Based on these features, wear mode I is classified as the adhesive wear mode. This classification indicates that the other wear modes are not adhesive wear modes, and is an important classification.

(ii) Wear mode II

This wear mode is observed under conditions of relatively low sliding speed and high contact force as shown in Fig. 2. In Fig. 3, the average temperature is below the softening point of the copper plate, but the maximum temperature is around the softening point. This means that wear mode II occurs under the conditions where the copper plate temporarily softens but the contact strip never softens.

The features of this mode are that the friction coefficient is greater than for wear mode I and tends to decrease as the sliding speed increases, as shown in Fig. 4, and the contact force fluctuates significantly due to stick-slip. In addition, the copper plate surface becomes significantly rougher as shown in Fig. 5.

These features suggest that significant sticking, like a seizure, occurred at the contact spot, causing stick-slip and roughening of the wear surface. It should be noted that although the wear mode II is clearly different from the wear mode I, both modes can occur under the same conditions as shown in Fig. 2. The mechanism of the wear mode II is clarified by considering the difference from wear mode I and the speed dependence of the friction coefficient is considered in Chapter 3.

(iii) Wear mode III

This wear mode is observed under conditions of moderate sliding speed as shown in Fig. 2. In Fig. 3, the average temperature is above the softening point of the copper plate, but the maximum temperature is below the softening point of the contact strip. This means that the wear mode III occurs under the conditions where the copper plate softens on average, but the contact strip never softens.

The features of this mode are that the friction coefficient increases as the contact force increases as shown in Fig. 4, and the copper plate surface shows wear tracks along the sliding direction as shown in Fig. 5. Wear tracks on the copper plate are also observed in wear mode I. The differences between wear mode III and wear mode I are that many copper pieces are transferred to the contact strip surface and the wear particles of the wear mode III are long and thin.

These features of the wear mode III are consistent with those of abrasive wear mode [9]. The abrasive wear occurs when there is a large difference in hardness between two materials, and the harder material digs into the softer material, producing long and thin wear particles. The friction coefficient in the abrasive wear mode increases as the amount of indentation of an asperity increases [10]. Therefore, it is sufficiently considered that the indentation of contact strip asperities increases as the contact force increases, resulting in an increase in the friction coefficient as shown in Fig. 4.

(iv) Wear mode IV

This wear mode is observed under conditions of high sliding speed and high contact force as shown in Fig. 2. As shown in Fig. 3, the average temperature is above the softening point of the copper plate, and the maximum temperature is above the softening point of the contact strip. This means that the wear mode IV occurs under the conditions where both materials soften.

The features of this mode are that the friction coefficient tends to decrease as the sliding speed increases, as shown in Fig. 4. From the observation result as shown in Fig. 5, the copper plate surface is smooth and an iron, which is the main component of the contact strip, is transferred to the surface. There are some cracks and peeling marks on the copper plate surface. On the other hand, some plastic flow marks in the sliding direction are observed on the contact strip surface. Wear particles are wavy flakes that seem to have peeled off from the copper plate surface.

Based on these features, the authors consider the mechanism of the wear mode IV. In this mode, both materials soften and undergo plastic flow, resulting in iron transfer to the contact wire surface. The shearing force due to friction force accumulates in the subsurface of the copper plate because the wear amount due to the plastic flow is small, resulting in cracks on the surface. These cracks lead to delamination wear [11] and peeling wear [12], resulting in flake wear particles. It is assumed that the sliding contacts between the copper plate and the contact strip are terminated by softening of the contact strip, regardless of the sliding speed. Since the softening point of the contact strip is constant and the softening occurs due to temperature increase caused by the friction force and sliding speed, the friction force decreases as the sliding speed increases in this mode. This is the reason why the friction coefficient in wear mode IV decreases as the sliding speed increases in Fig. 4.

3. Analysis of wear mode II mechanism

In this chapter, the authors analyze the wear mode II mechanism in terms of differences with wear mode I. It has been previously reported that a reduction in the number of contacts is one of the factors leading to sticking and seizure at the contact spot as observed in the wear mode II [13]. Therefore, the authors estimate the number of contacts in wear mode I and II using a heat conduction analysis.

3.1 Analysis model

The authors use the analysis software FEMTET, which is capable of unsteady heat conduction analysis. The analysis model is a static contact between two cylinders, as shown in Fig. 6. There is the boundary layer between the copper plate and the contact strip which is a mixture of both materials. By inputting frictional heat into this layer, it is not necessary to consider the heat distribution rate which has not been determined. The heat transfer properties of the boundary layer are assumed to be the average of the transfer properties of the copper plate and the contact strip as follows:

  

$${\rho }_{\mathrm{B}}=\frac{{\rho }_{\mathrm{C}}+{\rho }_{\mathrm{S}}}{2}$$(2)

  

$${\lambda }_{\mathrm{B}}=\frac{{2\lambda }_{\mathrm{C}}{\lambda }_{\mathrm{S}}}{{\lambda }_{\mathrm{C}}+{\lambda }_{\mathrm{S}}}$$(3)

  

$$c_{\mathrm{B}}=\frac{c_{\mathrm{C}}+c_{\mathrm{S}}}{2}$$(4)

Where, ρ is the weight density [kg/m³], λ is the thermal conductivity [W/(mK)], c is the specific heat [J/(kgK)], and the subscripts B, C and S represent the boundary layer, the copper plate and the contact strip, respectively.

Fig. 6　Analysis model

The frictional heat Q [W] input into the boundary layer is calculated by (5). It is assumed that the work done by friction is not 100% converted into frictional heat but also converted into friction noise and contact erosion. Therefore, the work done by the friction is converted into frictional heat by multiplying it by the heat conversion efficiency η.

  

$$Q=\eta \frac{\mu Nv}{n}$$(5)

Where, μ is the friction coefficient, N is the contact force [N], v is the sliding speed [m/s] and n is the number of contacts calculated using (6) proposed by Lim and Ashby [12].

  

$$n=\left(\frac{r_0}{r_{\mathrm{a}}}\right)\left(\frac{N}{A_0{H}_0}\right)\left(1-\frac{N}{A_0{H}_0}\right)+1$$(6)

Where, r_a is the radius of real contact area [m], r₀ is the radius of apparent contact area [m], A₀ is the apparent contact area [m²] and H₀ is the hardness of the copper plate [Pa]. The analysis time t [s] is calculated by (7) as relative sliding time.

  

$$t\text{=}\frac{2r_{\mathrm{a}}}{v}$$(7)

The boundary conditions are that the temperatures at the end of both cylinders are set to 298 K and that the circumferences of both cylinders are insulated. Since both cylinders are assumed to be in perfect contact, the heat transfer resistance between them is set to zero. The analysis parameters and material properties are shown in Tables 2 and 3.

Table 2　Analysis parameters

Radius of real contact area ra, m	64x10−6
Apparent contact area A0, m2	100x10−6
Radius of apparent contact area r0, m	5.6x10−3
Vickers hardness of copper plate H0, Pa	980x106
Heat conversion efficiency η	0.4-0.8

Table 3　Material properties

	Copper plate	Contact strip	Boundary layer
Weight density ρ, kg/m3	8,910	6,910	7,910
Heat conductivity λ, W/(mK)	361 (293 K) 337 (573 K) 312 (783 K) 286 (1,173 K)	20.0 (293 K) 19.3 (573 K) 18.5 (783 K) 17.7 (1,173 K)	37.9 (293 K) 36.5 (573 K) 34.9 (783 K) 33.3 (1,173 K)
Specific heat c, J/(kgK)	394 (293 K) 416 (573 K) 440 (783 K) 464 (1,173 K)	491 (293 K) 576 (573 K) 666 (783 K) 756 (1,173 K)	443 (293 K) 496 (573 K) 553 (783 K) 610 (1,173 K)

The output of analysis is the maximum temperature rise in the boundary layer from room temperature, since the thermoelectromotive force in Fig. 1 is theoretically dependent on the temperature difference between contact surface and the electric potential measurement point (= room temperature). According to previous study between the analytical accuracy and the model size [14], the bulk height is set to 400 μm, the thickness of the boundary layer is set to 0.2 μm, and the mesh size of the boundary layer is set to 0.1 μm.

3.2 Analysis results

As an example of the analysis results, the estimated contact temperature rise under the condition of η 0.6 is shown in Fig. 7 (a), and the average temperature rise measured for comparison is shown in Fig. 7 (b). The analysis error δ for each wear mode is calculated by (8).

  

$$\delta \text{=}\sum \left(\left|∆T_{\mathrm{A}}-∆T_{\mathrm{T}}\right|/T_{\mathrm{T}}\right)/D$$(8)

Where, ∆T_A is the analytical temperature rise [K], ∆T_T is the measured average temperature rise [K], T_T is the measured average temperature [K], and D is the number of data.

Fig. 7　Relationship between sliding speed and contact temperature rise (analysis and experimental result)

Figure 8 shows the analysis errors for the four types of wear mode and the average error for all the types of wear mode when η is varied from 0.4 to 0.8. This figure shows that the average error is smallest when η is 0.6, and the value of η at which the analysis error is smallest varies depending on the wear mode. The reason for this variation is that unconsidered factors not considered in the analysis model, such as softening of material and formation of an alloy layer at the contact boundary, may affect the temperature rise.

Fig. 8　Analysis error

3.3 Occurrence mechanism of wear mode II

To estimate the number of contacts from experimental and analysis results, it is necessary to apply a relationship between friction, taking into account the number of contacts, and the contact temperature rise. The authors assumed the contacts to be a semi-infinite solid as shown in Fig. 9 and supplied the heat flux q₀ [W/m²] to the contact surface (x = 0). In the thermal conduction theory [15], the temperature T_x,t [K] at the distance from contact surface x [m] after the contact time t [s] is calculated by (9).

  

$$\frac{\lambda \left(T_{x,t}-T_{\mathrm{i}}\right)}{q_0\sqrt{\alpha t}}=\frac{2}{\sqrt{\pi }}\mathrm{e}\mathrm{x}\mathrm{p}\left(-\frac{x^2}{4\alpha t}\right)-\frac{x}{\sqrt{\alpha t}}\mathrm{e}\mathrm{r}\mathrm{f}\mathrm{c}\left(\frac{x}{2\sqrt{\alpha t}}\right)$$(9)

Where, T_i is the initial temperature [K], α is the thermal diffusivity [m²/s]. The heat flux supplied to one contact spot is calculated by (10).

  

$$q_0=\beta \eta \frac{\mu Nv}{n\pi r_{\mathrm{a}}^2}$$(10)

Where, β is the heat distribution rate. Since the analysis result shown in Fig. 7 (a) is the temperature in the boundary layer, which means contact surface, the surface temperature T_0,t [K] and the temperature rise ΔT_0,t [K] of the solid are calculated by (11) and (12) as x =0.

  

$$T_{0,t}=T_{\mathrm{i}}+\frac{2\sqrt{\alpha t}}{\lambda \sqrt{\pi }}\beta \eta \frac{\mu Nv}{n\pi r_{\mathrm{a}}^2}$$(11)

  

$$\Delta T_{0,t}=T_{0,t}-T_{\mathrm{i}}=\frac{2\eta \beta \sqrt{\alpha }}{{\pi r}_{\mathrm{a}}^2\lambda \sqrt{\pi }}\frac{\mu Nv}{n}\sqrt{t}=C\frac{\mu Nv}{n}=\mathit{CE}$$(12)

Since the material properties and the radius of real contact area are constant, the temperature rise is proportional to the work done by friction and square root of the contact time per contact spot. Therefore, the authors define the variable in (12) as E [W/s^−0.5].

Fig. 9　Semi-infinite solid model for heat conduction calculation

Figure 10 shows the analyzed temperature rise obtained by converting the horizontal axis of Fig. 7 (a) into E. From Fig. 10, it is confirmed that the coefficient of determination is 0.9999, so that the proposed analysis model shown in Fig. 6 satisfies (12). Figure 11 shows the measured temperature rise in wear mode I and II obtained by converting the horizontal axis of Fig. 7 (b) into E and approximate curve of Fig. 10. From Fig. 11, it is found that measured temperature rise in the wear mode I is generally proportional to E, but in the wear mode II it is not proportional to E and some data exceed the approximate curve.

Fig. 10　Analytical relationship between temperature rise and friction, taking into account number of contacts and contact time (Analysis result: η = 0.6)

Fig. 11　Experimental relationship between temperature rise and friction, taking into account number of contacts and contact time (Experimental result: wear mode I and II)

The reason why the temperature rise in the wear mode II exceeds the approximate curve is thought to be because C and E in the experiment are greater than those in the analysis. The factors that increase C and E are as follows:

(a)　The number of contacts n in the experiment was lower than that estimated by (6) in the analysis model, so that E was greater than it was in the analysis model. Previously, it has been proposed that the transition to the seizure wear mode under unlubricated conditions is caused by a reduction in the number of contacts due to wear particles and uneven contact [13].

(b)　The real contact area ${\pi r}_{\mathrm{a}}^2$ in the experiment was smaller than in the analysis model, so that C was greater than it was in the analysis model. On the other hand, it has been reported that the real contact area increases in the seizure wear mode [13], so it is unlikely that the real contact area decreased in this experiment.

(c)　The heat conversion rate η in the experiment was greater than the 0.6 in the analysis model, so that C was greater than it was in analysis model. As mentioned in section 2.2, stick-slip occurred during the wear mode II test. Since friction noise and vibration due to stick-slip were significantly higher, the heat conversion rate might lower than in other wear modes. Additionally, even if the heat conversion rate differs according to wear mode, the temperature rise must be proportional to another approximate curve, but the measured temperature rise in wear mode I is not proportional to E. Therefore, it is unlikely that the heat conversion rate increased in this experiment.

(d)　The contact time t in the experiment was greater than that in the analysis model, so that E was greater than it was in the analysis model. According to (7), the factor that increases the contact time is the increase in the radius of the contact spot r_a. However, as the radius of the contact spot increases, the heat flux supplied to the contact surface decreases, so the temperature rise must decrease. It is therefore unlikely the contact time increased in this experiment.

From the above, it can be assumed that the reduction in the number of contacts is the most influential factor.

The number of contacts in the experiment can be calculated using the approximated relationship between the temperature rise and E. From the approximation and the temperature rise curve ΔT_T in Fig. 11, the number of contacts in the experiment n_T is calculated by (13).

  

$$n_{\mathrm{T}}=1206.7\frac{\mu Nv\sqrt{t}}{\Delta T_{\mathrm{T}}}$$(13)

Figure 12 shows the number of contacts calculated using (13), and Fig. 13 shows the contact pressure per contact. These figures indicate that the number of contacts in wear mode II is less than that in wear mode I, as expected above, and that the contact pressure boundary between the wear mode I and II is approximately 900 MPa, which is similar to the hardness of the copper plate, 980 MPa. According to the Goddard's report [10], the condition for the occurrence of the seizure wear mode is expressed by (14).

  

$$\frac{N}{n_{\mathrm{T}}\pi r_{\mathrm{a}}^2}=H$$(14)

Here, H is a hardness [Pa]. This equation shows that if the contact pressure exceeds the hardness of material, plastic flow occurs rapidly around the contact spot, causing the seizure.

Fig. 12　Estimated number of contacts

Fig. 13　Estimated contact pressure

Based on the above considerations, the wear mode II is classified as the seizure wear mode. It is assumed that if the seizure occurs between the copper plate and the contact strip, crack will occur inside the contact point of the copper plate. It is also thought that seizure at the contact point is released by the softening of the copper plate regardless of the sliding speed. Since the softening occurs due to the friction force and sliding speed, the friction force decreases as the sliding speed increases in this mode. This is the reason why the friction coefficient in wear mode II decreases as the sliding speed increases in Fig. 4.

In this study, the authors classified the mechanical wear mode, which had previously been considered to be solely the adhesive wear, into four types based on the material softening. The transition mechanism of wear modes due to material softening is universal and can be applied to different equipment, so that the wear phenomena caused in an actual railway field can be explained using this study. In actual railway field, it is thought that differences in the wear rate and the wear surface of the contact wire lead to the occurrence of different wear modes. Since the wear mechanism differs according to wear mode, this knowledge will be useful for considering fundamental measures for wear reduction.

4. Conclusions

In this study, the authors carried out wear tests with a new rotary wear tester, which can measure the contact temperature, and analyzed the contact temperature with the new contact model, taking into account the number of contacts. From the test and analysis results, the authors classified wear modes into four types and clarified the conditions under which each mode occurs. The results obtained in this study are as follows:

(1)　The wear modes under the material combinations of the hard-drawn copper plate and the iron-sintered alloy contact strip are classified into four types as follows:

Wear mode I: Adhesive wear mode

Wear mode II: Seizure wear mode

Wear mode III: Abrasive wear mode

Wear mode IV: Soften and flow wear mode

(2)　The conditions of occurrence of each wear mode are clarified based on the softening of materials due to frictional heat as follows:

Wear mode I, II: Neither material is softened.

Wear mode III: Only copper plate is softened.

Wear mode IV: Both materials are softened.

(3)　Heat transfer analysis taking into account the number of contacts revealed a mechanism of occurrence of wear mode II, which is caused by the surface pressure at the real contact point exceeding the hardness of the copper.

(This paper is the updated version of the reference [16] [17])

References

• [1]  Iwase, M., “Current Collecting by the Pantograph and its Wear (III),” Railway Technical Research Report, Vol. 217, 1961 (in Japanese).
• [2]  Kohno, A., Ohyabu, H. and Soda N., “Effect of Discontact Arc on Wear of Materials for Current Collection (Part 1),” Journal of Japan Society of Lubrication Engineers, Vol. 27, No. 4, pp. 283-287, 1982 (in Japanese).
• [3]  Kohno, A., Ohyabu, H. and Soda N., “Effect of Discontact Arc on Wear of Materials for Current Collection (Part 2),” Journal of Japan Society of Lubrication Engineers, Vol. 27, No. 7, pp. 527-532, 1982 (in Japanese).
• [4]  Jia, S. G., Liu, P., Ren, F. Z., Tian, B. H., Zheng, M. S., Zhou, G. S., “Sliding wear behavior of copper alloy contact wire against copper-based strip for high-speed electrified railways,” Wear, Vol. 262, No. 7-8, pp. 772-777, 2007.
• [5]  Yamashita, C., Adachi, K., “Influence of Current on Wear Modes and Transition Condition of Current Collecting Materials,” Journal of Japanese Society of Tribologist, Vol. 58, No. 7, pp. 496-503, 2013 (in Japanese).
• [6]  Yamashita, C., Adachi, K., “Wear Mode Map of Current Collecting Materials,” Journal of Japanese Society of Tribologist, Vol. 62, No. 2, pp. 129-136, 2017 (in Japanese).
• [7]  Hirata, T., Tanaka, M., Kumano, H., Hada, Y., Illustrated Energy Engineering, Morikita Publishing Inc., pp. 173-174, 2011 (in Japanese).
• [8]  Sasaki, S., Elucidation and Control of Tribology by numerical Simulation and Surface Analysis, Technosystem, pp. 47-56, 2018 (in Japanese).
• [9]  Mizumoto, M., Usami, K., “Why does wear occur,” Turbomachinery, Vol. 24, No. 5, pp. 267-273, 1996 (in Japanese).
• [10]  Goddard, J., Wilman, H., “A theory of friction and wear during the abrasion of metals,” Wear, Vol. 5, No. 2, pp. 114-135, 1962.
• [11]  Suh, N. P., “The delamination theory of wear,” Wear, Vol. 25, pp. 111-124, 1973.
• [12]  Ashby, M. F., Lim, S. C., “Wear-mechanism Maps,” Acta metal, Vol. 35, No. 1, pp. 1-24, 1987.
• [13]  Sasada, T., Wear, Yokendo, 2008 (in Japanese).
• [14]  Nemoto, K., Yamashita, C., “Elucidation of Transition Mechanism of Adhesive Wear Mode and Seizure Wear Mode between Copper Plate and Iron Based Sintered Alloy Contact Strip by Analysis of Contact Temperature,” Journal of Japanese Society of Tribologist, Vol. 69, No. 4, pp. 301-311, 2024 (in Japanese).
• [15]  The Japan Society of Mechanical Engineers, Heat Transfer Engineering, Maruzen Publishing Inc., 2005 (in Japanese).
• [16]  Yamashita, C., Nemoto, K., “Classification of Mechanical Wear Mode of Contact Wire and Contact Strip caused by Frictional Heat,” RTRI Report, Vol. 35, No. 12, pp. 11-16, 2021 (in Japanese).
• [17]  Nemoto, K., Yamashita, C., “Transition Mechanism between Adhesive Wear Mode and Seizure Wear Mode of Current Collecting Materials,” RTRI Report, Vol. 38, No. 10, pp. 23-30, 2024 (in Japanese).

Authors

	Chikara YAMASHITA, Ph.D. Senior Chief Researcher, Head of Current Collection Maintenance Laboratory, Power Supply Technology Division Research Areas: Overhead Contact Line, Tribology, Electric Contact, Fatigue
	Koki NEMOTO Researcher, Current Collection Maintenance Laboratory, Power Supply Technology Division Research Areas: Overhead Contact Line, Tribology