Properties of Cold Spray Coatings for Restoration of Worn-Out Contact Wires

The influence of Cu-Al2O3-Zn powder mixtures for cold spraying on the properties of the coatings is studied. The coatings were deposited to restore worn-out copper contact wires, which were used as a substrate. The main parameters, such as adhesion, specific electrical conductivity, hardness, and wear resistance, were examined. The content of Al2O3 in the mixture varied from 30% to 60%, and the content of Zn, from 0% to 20%. The results obtained indicate that the 60Cu-40Al2O3 coating was the most adhesive and showed the best conductivity, while 40% Cu-50% Al2O3-10% Zn was found to be the most wear-resistant. The optimum spraying temperature was 500 °C.


Introduction
Increased wear of the contact wire during operation occurs in locations where its displacement is limited. This type of contact wire wear is called local wear. The local wear limit is achieved several times faster than that which is achieved over the rest of the length of the contact wire. The length of the local wear section is up to 0.5 m. The wear surface width is usually about 10 mm [1][2][3][4][5][6]. At present, on conventional railway lines, a worn-out section of a contact wire is cut out, and a piece of a new, not worn-out contact wire is inserted [7]. On high-speed lines, such a replacement is not possible since the tensile stress of the contact wire on high-speed lines is 2-3 times higher than on conventional lines [8]. Therefore, on high-speed lines, due to local wear with a length of 0.5-1 m, 1.2-1.8 km (about 1500 kg) of the contact wire are changed [9,10]. The cold spray technology was chosen to restore the local wear of contact wires [7]. Relatively simple and light technological equipment allows using the technical capabilities of the railcar to restore local wear of contact wires without dismantling them. The following basic requirements are imposed on the sprayed layer. The adhesion strength of the sprayed layer to the substrate must be at least 30 MPa, the specific electrical conductivity must be at least 20 MS/m, the wear rate of the sprayed layer must be no more than the wear rate of the copper contact wire, the spraying speed must be at least 0.5 m/h. Cold spray technology allows the deposition of thick coatings up to 2 mm without annealing in a short time [11]. In order to avoid extra gas supply units, compressors, and gas vessels, air was used as carrier gas with a working pressure of 6 atm.
A vast number of researches, for example [12][13][14][15][16][17][18][19][20][21][22][23][24][25], have been devoted to the application and study of the gas dynamic cold spraying technology. The research works cover numerous theoretical and experimental problems associated with gas-dynamic cold spraying. There are a significant number of studies on the deposition of copper particles, for example [26][27][28][29]. A large portion of research works is associated with spraying on steel, titanium, aluminum, and other metallic substrates. Of these, only a very small part is devoted to deposition on a copper substrate. Reviews [12,[30][31][32][33] give only a few examples of deposition of a coating on a copper substrate.
In study [34], the corrosion behavior of low-pressure cold sprayed coatings Al-Al 2 O 3 on copper and Cu-Al 2 O 3 on aluminum is studied. The thickness of the coatings was just 300 µm, and no adhesion was examined. In study [35], the authors applied cold spraying to protect copper bars from corrosion to extend the life cycle of the electrical contact. Cold spraying provides thin and accurate deposition of tin coating (45-50 µm) on the surface of the copper rod. However, the rest of the properties were ignored. In studies [36,37], the effect of preheating the copper substrate on the bonding of nickel particles during deposition was considered. Sufficient adhesion to the copper substrate was achieved only after heat treatment of the sprayed layer, which is proscribed for contact wire due to recrystallization issues [11]. In study [38], a well-known arc-resistant coating of a mixture of Ag-SnO 2 powders was deposited on a copper substrate. It is not clear why the authors of study [38] protected the copper with an arc-resistant but expensive, Ag-SnO 2 coating and not a relatively cheap arc-resistant Cu-W or Cu-Cr coating [39]. In studies [40,41], the influence of the critical velocity and annealing on the deposition of copper powder on a copper surface and its microstructure and properties were studied. In studies [38,40,41], nitrogen was used as a carrier gas.
Until now, far too little attention has been paid to choosing the proper powder mixture composition for cold spraying to provide high adhesion with the given substrate material for a particular carrier gas. Furthermore, among tribologists, much uncertainty still exists about material couples prone to seizure in a greater or lesser manner. Therefore, the selection of powder mixtures to restore local wear of contact wires is based on the functional properties of the layer.
High-conductive metal, for instance, aluminum or copper, must be the base material of the mixture. In the former case, an aluminum-copper galvanic couple is formed in the wire which leads to intensive corrosion at atmospheric conditions. Copper was chosen as the base material for the powder mixture since it is more electroconductive, corrosion-resistant, and harder to eliminate the drawback.
In study [42], the authors found that the introduction of the second component to the copper powder increases the coating rate from 0% to 65%. This is due to the second component bonding more easily to steel while copper bonds to the additive. Applying non-equilibrium thermodynamics and self-organization theory to the process of seizure during friction showed that in order to increase the seizure area, relatively light elements should be transferred to the friction zone [7]. Thus, tin, aluminum, and zinc are the most rational available materials to meet the criteria. Tin is restricted due to its low melting point. Previous practice has shown that tin particles melt during spraying and adhere to the nozzle, limiting the outcome of the copper powder. Although aluminum exhibits the lowest flow stress, it forms the hardest oxides, which complicate obtaining juvenile surface on them. Therefore, within this work, zinc was chosen as a second component to copper powder to restore the contact wires.
In studies [33,43], a natural wave-like shape characteristic of explosion welding was noted, which the substrate surface acquired after cold spraying. Alumina coarse powder particles up to 100 µm will be used in the preprocessing stage to obtain a natural wavelike shaping of the surface. However, a copper-zinc powder mixture cannot provide the necessary hardness of the deposited layer (at least 100 HB), which can lead to increased wear. Alumina particles up to 50 µm will be added to the mixture to provide dispersive hardening to improve the mechanical properties.
The given examples of gas-dynamic cold spraying on a copper substrate do not contain data on the adhesion of the coating to the substrate and other properties. In previous studies, the seizure process for the gas-dynamic spraying of metal powders was reviewed from a non-equilibrium thermodynamics standpoint [7], and the problem of copper softening during cold gas-dynamic spraying was investigated [11]. In this work, the necessary properties of coatings obtained by the method of cold gas-dynamic spraying for the restoration of local wear of contact wires are investigated.

Materials and Methods
The following properties of the sprayed layers were investigated: adhesive strength, electrical conductivity, hardness, wear.
Copper contact wire produced according to [44] with a cross-sectional area of 120 mm 2 was used as a substrate for the deposited layer to study its wear resistance. The hardness of the wires was 115-125 HB. Adhesion and conductivity tests were performed on copper specimens of the same hardness according to the scheme shown in Figure 1. Copper, zinc, and alumina powders with a particle size distribution from 30 to 50 µm were used as initial materials for powder mixture preparation. Alumina powder with particles from 100 µm to 150 µm was used for preliminary substrate blasting. To deposit a coating Dimet 405 cold spraying system (Joint-Stock Company "DIMET", Obninsk, Russia) with air as a carrier gas was utilized.
Within the experiments, the gas temperature varied from 200 • C to 600 • C in increments of 100 • C. The following compositions of powder mixtures were investigated: Cu, 60%Cu- Adhesion tests were performed on the LFM-L10 tensile machine (Walter + Bai AG, Löhningen, Switzerland) with a maximum load of 10 kN. Specific electroconductivity was measured by the eddy-current measuring device Konstanta K6 (Russia), with measuring limits from 0.05 to 59 MS/m.
Wear tests of the restored contact wires were carried out on the upgraded friction machine UMT1 (Russia). It allows performing friction tests with the current collector at linear velocities from 2 km/h to 200 km/h with a current density up to 30 A/mm. A segmented octagon was chosen as an element imitating a contact wire. There were 8 pieces of contact wire 120 mm long each on the aluminum base (Figure 2b,d). This approach provided wear tests of 8 specimens of the contact wire simultaneously. Moreover, the zigzag of the contact wire was also reproduced. Each segment of the contact wire was attached to the holder using 2 screw connections located on the back of the aluminum disc. The aluminum disc was isolated from the friction machine with an electro-insulating gasket. Local wear was simulated on segments 1, 3, 5, 7 by milling ( Figure 2). Each worn area was restored with a powder mixture according to Table 1. Cold-sprayed areas were machined after to provide the necessary friction surface without significant asperities. Figure 3 shows an image of the octagon of the contact wire after restoration and milling. Segments 2, 4, 6, 8 represented a non-worn contact wire.
The rotation of the aluminum disk with the sections of the contact wire took place at a speed of 500 rpm. Samples of SK01Cu current-collecting materials were pressed against the contact wire at diametrically opposite points. The hardness of the current-collector was 83 HS, electrical resistivity, 6.7 µΩ·m. The length of the contact was 15 mm, pressing force, 30 N. Density of electric current was set at 7 A/mm. Wear tests continued until 100,000 rotations were reached. Linear wear of contact wires was measured after the tests.
To determine the effect of preliminary treatment of the surface of the copper substrate with an airflow with an abrasive, large Al 2 O 3 particles, the size of which exceeded 100 µm, 3 mixtures were sprayed at a temperature of 500 • C. The adhesive strength of the sprayed layer is shown in Table 2.

Results and Discussion
The measured adhesion values of the cold-sprayed layer are presented in Table 2. Table 2 shows that substrate preprocessing with coarse alumina increased adhesion by 3-5 times. Further experiments were always preceded by this preprocessing technique.
To reveal the optimum spraying temperature, such parameters as adhesion and electroconductivity of the 40%Cu-60%Al 2 O 3 cold sprayed coating were plotted versus temperature (Figures 4 and 5).  The dependence of electroconductivity from spraying temperature of 40%Cu-60%Al 2 O 3 coating.
It can be concluded from Figure 4 that when the airflow temperature was raised from 200 • C to 500 • C, the adhesion to the substrate also increased. Moreover, the most intensive adhesion growth was observed in the interval from 400 • C to 500 • C. However, further heating of carrier gas to 600 • C adversely affected adhesion. Thus, a temperature of 500 • C was found to be optimal for cold spraying 40%Cu-60%Al 2 O 3 coating on preprocessed copper substrate.
A relatively good correlation of adhesion and spraying temperature may be attributed to both the increased airflow velocity and decreasing yield strength of the metal particles. As the copper temperature rose from 20 • C to 600 • C, the yield stress of copper decreased by a factor of 5 [45]. Increased kinetic energy and plasticity contributed to the intensification of local bonds to the substrate, hence, better adhesion. On the other hand, higher spraying temperature led to more intense oxidation of powder particles and substrate surface. This factor explains lower adhesion values of the deposited material after 500 • C. At temperatures below 600 • C, a fragile porous layer formed on the copper surface, consisting of a mixture of CuO and Cu 2 O oxides, which was relatively easily destroyed by impact and plastic deformation of the particles. At a temperature of 600 • C and above, a dense and strong layer of Cu 2 O oxide formed, which remained on the surface of particles and substrate after impact and plastic deformation [45,46]. Moreover, heating the substrate up to 600 • C may cause critical degradation of the material, which must be prevented due to the possible recrystallization of the contact wire [11,47]. At a temperature of 600 • C, dynamic recrystallization of copper particles in the coating can occur [48], which can lead to an intensive decrease in the hardness of the coating.
Although the specific electroconductivity of the layer sprayed with 40%Cu-60%Al 2 O 3 powder weakly depended on the airflow temperature, the maximum value was achieved at a temperature of 500 • C being similar to that of adhesion ( Figure 4). The correlation observed may be explained in the same way. An increase in the seizure intensity with increasing temperature led to an increase in the seizure area of the particles. Therefore, the maximum electroconductivity was registered at the spraying temperature of 500 • C. As was described earlier, after this critical temperature, intensive copper oxidation led to decreasing bonding area hence, decreasing electroconductivity.
The properties of coatings sprayed at a temperature of 500 • C with mixtures of different compositions after preliminary treatment of the substrate surface are shown in Table 3. Compared to the 50%Cu-50%Al 2 O 3 mixture, the 45%Cu-50%Al 2 O 3 -5%Zn mixture contained 5% zinc and 5% less copper, which resulted in significant hardness and adhesion increase of the deposited material while electroconductivity lowered. Replacing 20% of copper in the 60%Cu-40%Al 2 O 3 mixture by 20% of Zn led to minor adhesion improvement but halved electroconductivity value of the coating, while hardness remained unaffected. When the zinc content increased by 10%, specific conductivity and hardness degraded strongly, but it did not affect adhesion.
Varying Al 2 O 3 content from 30% to 60% in the two-component system with copper revealed its influence on the properties. Increasing alumina particles in the mixture from 30 to 50% had little effect on electroconductivity and adhesion. Further increasing up to 60% doubled adhesion strength, while electroconductivity and hardness rose only by 10%. The correlations between adhesion and Al 2 O 3 content in the two-component mixture are represented in Figure 6. Adding 10% Zn to all the mixture powder compositions increased hardness and adhesion and reduced electrical conductivity.
Cold spray coating obtained from 40% Cu-60%Al 2 O 3 mixture outperformed the rest of the compositions involved in this study. The adhesion behavior of the deposited coating, depending on alumina and zinc content in Cu-Al 2 O 3 -Zn powder system, is shown in Figure 7.
Two peaks can be observed on the adhesion plot in Figure 7. The absolute adhesion maximum was achieved when spraying the 40% Cu-60% Al 2 O 3 powder. The second-best result was performed by the 40% Cu-50%Al 2 O 3 -10%Zn powder The 40% Cu-60% Al 2 O 3 powder mixture was considered to be the best composition according to its properties. However, the spraying of a mixture of this composition was characterized by high consumption of powders. Therefore, the duration of spraying increased significantly. This problem was tackled by adding 5%-10% of zinc which saved material and cut operation time by four. In addition to 40%Cu-60%Al 2 O 3 material, zinc added to the powder increased adhesion to the substrate. The results of wear tests of the restored contact wires were compared to that of regular contact wires (Table 4).  It can be concluded from Table 4 that the wear results of both restored and regular contact wires were comparable and did not differ much. The lowest wear was possessed by a layer sprayed with the mixture 40Cu-10Zn-50 Al 2 O 3.

Conclusions
The optimal airflow temperature for cold spray restoration of worn-out contact wires was found to be 500 • C providing the maximum values of adhesion and specific electric conductivity.
Pre-treatment of the surface of the copper substrate with an airflow with large Al 2 O 3 particles increased the adhesion of the sprayed coating by 3-5 times.
The adhesion strength increased with the Al 2 O 3 content in the sprayed Cu-Al 2 O 3 mixture. Cold spray coating obtained from the 40%Cu-60%Al 2 O 3 powder composition demonstrated the highest adhesion and specific electric conductivity, while the Cu-50%Al 2 O 3 -(5-10)% Zn coatings were the most wear-resistant. The latter powder may be recommended for cold spray restoration of worn-out contact wires to balance wear resistance, feeding rate, and operation time.

Conflicts of Interest:
The authors declare no conflict of interest.