Abstract：The copper-steel composite was prepared by laser cladding technology, the alloy layer, interface and matrix microstructure of the copper-steel composite were characterised, the mechanical properties and friction properties of the copper alloy layer and interface were detected. The results show that the fine crystal reinforcement caused by laser cladding and nickel diffusion to copper alloy layer make copper alloy layer show excellent mechanical properties, which is attributed to the presence of nickel in copper tin alloy generating complex ternary phase, the grain refinement weakens the tin segregation. Meanwhile, the increase of the grain boundaries separates the δ phases, and the disappearance of the continuous coarse δphases is transformed into the scattered precipitation of intragranular and grain boundaries. The average tensile strength of the
prepared copper-steel composite alloy layer is 502.3 MPa, the average yield strength is 356.6 MPa, the average elongation is 11.05%, the average section shrinkage is 8.55%, the average elastic modulus is 106.863 GPa, the average shear strength is 450.9 MPa, the interfacial bonding strength of the copper alloy is 549.7 MPa, and the friction coefficient is between 0.786-0.846.
Key words：copper/steel composite; laser cladding; δ phase

Tin bronze is a copper alloy with tin as the main alloying element. The tin element can be dissolved in the copper matrix to form α-solid solution and δ phase (Cu41Sn11). α-solid solution has a face-centered cubic structure and is the main component phase in tin bronze. δ phase is a complex cubic structure, hard and brittle. The existence of these two phases makes tin bronze have high strength and hardness, good corrosion resistance and excellent wear resistance. Studies have found that the addition of nickel elements can improve the strength, wear resistance and plasticity of the alloy, and can also reduce the stress corrosion sensitivity of the alloy. Tin bronze is usually compounded with steel matrix to form tin bronze/steel composite materials. Copper/steel composite materials combine the advantages of copper and steel, overcome their respective shortcomings, and obtain better comprehensive performance. Therefore, copper/steel composite materials have been widely used in different fields such as electronic information, communication engineering and marine engineering.

At present, researchers have conducted many studies, such as using laser powder bed and electron beam powder bed melting technology, using laser or electron beam as heat source, and preparing copper-steel composite materials based on powder bed powder additive manufacturing. It has the characteristics of metallurgical bonding, complex structure integrated molding, high molding accuracy and good mechanical properties, which improves manufacturing efficiency. However, the copper-steel composite materials prepared by this technology are relatively expensive, and the cladding layer still has defects such as pores, cracks, and inclusions, which will affect the performance of the composite materials. At present, the interface bonding mechanism theory of copper-steel composite materials is not yet mature, and the strength and stability of the interface bonding are crucial to the performance of the composite materials. Therefore, further optimization is needed.

In this study, tin bronze/steel layered composite materials were prepared by laser cladding technology, and the characteristics of the interface bonding mechanism, microstructure and mechanical properties of copper-steel composites were explored. The research results can open up new technical paths for the preparation of key parts such as sliding bearings, reducer turbines, and oil distribution plates, which has a clear engineering significance, and at the same time provides a theoretical basis for the preparation of high-strength copper/steel composite materials.

1. Test materials and methods
The test substrate is a quenched and tempered 42CrMoA steel forging rod with a size of Φ 150mm×600 mm. The copper-steel transition layer welding wire is Inconel718 alloy with a diameter of Φ1.2 mm. The copper alloy welding wire is CuSn12 with a diameter of Φ1.2 mm. The chemical compositions of the two welding wires are shown in Tables 1 and 2. The oxide layer of the 42CrMoA steel forging rod needs to be polished before cladding. After polishing, the substrate is wrapped with a heater for preheating to ensure the fluidity of the liquid metal during the laser cladding process. When the surface temperature of the substrate rises to 200℃, the heating is stopped. All welding wires used need to be preheated in a high and low temperature box before laser cladding to remove the influence of oil, moisture, etc., and kept at 80℃ for 2 hours. After cladding, the sample is subjected to stress relief annealing treatment, and the treatment process is 300℃/2h, air cooling.

The microstructure was characterized by ultra-high resolution field emission scanning electron microscope (SEM, SEMEV018) and optical metallographic microscope (OM, Olympus-GX71). Before the test, the samples were ground and polished with 600#, 800#, 1000# and 2000# sandpaper in order to remove the oxide layer on the surface of the sample, so that the surface of the sample was bright and without obvious scratches. Then the corresponding corrosive agent was used for corrosion, and the corrosion time was about 15 s. Then the residual corrosive liquid was quickly rinsed with alcohol, and the sample surface was blown dry with air. The ratio of the corrosive agent was 10 g FeCl3 solution + 10 mL HCl solution + 120 mL H2O.

The microhardness test used a KB3000 BVR-Video type Brove automatic hardness tester with a loading force of 31.25 kgf (1 kgf = 9.8 N) and a dwell time of 15 s. The hardness test of each sample was repeated at least 5 times. The tensile, shear and bonding strength tests were conducted using a UTM5504HA
double-column floor-standing electronic universal testing machine. Copper alloy tensile, shear and bonding strength test specimens were taken along the length of the steel bar. Two tensile specimens (No. 1 and 2) were taken, the specimen width bo=15 mm, the thickness ao=2.40 mm, the specified plastic extension εp=0.2%, the extensometer gauge length Le=25mm, and the strain rate was 0.00025 s-1. The tensile rate of the shear and bonding strength tests was 1.0mm/min. Three shear specimens (No. 1, 2, 3) were taken, and four bonding strength specimens (No. 1, 2, 3, 4) were taken.

The friction and wear performance of the copper alloy layer was tested using a friction and wear testing machine. The copper alloy layer sample was gradually polished with sandpaper to 1200 mesh, and then the sample surface was polished with a polishing cloth to a mirror surface. Friction and wear test The grinding ball radius is 5 mm, the material is Al2O3, the friction speed is 6 mm/s, and the load is set to 3 and 5 N, respectively, and a 30-min reciprocating friction and wear test is carried out. The 3D morphology of the copper alloy layer after friction and wear is observed using a laser confocal microscope, and the wear volume is calculated.

2 Results and discussion

2.1 Microstructure analysis
2.1.1 Analysis of copper alloy layer structure Figure 1 is a 500-fold metallographic microstructure of the copper alloy layer. The microstructure shows the typical phase distribution of the copper alloy layer in the copper-steel composite structure. At room temperature, the microstructure consists of an α-Cu matrix and intergranular and intragranular δ phases. Irregular spherical δ (Cu41Sn11) phases are gray and intermittently distributed along the grain boundaries. Pentagonal δ phases are scattered in the grains, and α-Cu is bright white. The grain size of the copper alloy layer is 8 to 9. No obvious pores are found in the 500-fold microstructure photo. The density of the copper alloy layer is high. The δ phase is a hard and brittle phase. This type of grain boundary phase transformation can strengthen the various properties of the material. According to the Cu-Sn phase diagram, the δ phase is a non-steady-state phase and will decompose at 350°C.

2.1.2 Interface structure analysis
Figure 2 (a) shows the interface between the copper alloy layer and the transition layer. As shown in Figure 2 (a), the transition layer and the copper alloy layer have achieved good metallurgical bonding. During the laser cladding of the first layer of copper alloy, the high-temperature metal liquid in the molten pool flows through the transition layer, and the nickel element diffuses into the copper alloy layer, resulting in the miscibility of nickel and copper elements. From the crystal structure, it can be seen that Cu and Ni have the same structure (fcc) and similar lattice constants (aCu=0.3615 nm, aNi=0.3524 nm). In liquid conditions, nickel and copper can be infinitely soluble in each other, which is also an important reason for choosing nickel-based materials as transition layers. When copper and steel are directly composited, liquid copper will have a peeling effect on the steel grain boundary and produce penetration cracks, while nickel-based alloys can achieve metallurgical bonding with steel and copper. The first layer of laser cladding copper alloy showed non-uniform dendrite segregation, with gray structure sandwiched between white dendrites, while the second layer of copper alloy had an equiaxed crystal microstructure. During the solidification process, the distribution of Ni, Sn/Cu elements in the dendrite structure formed a segregation phase. The nickel content decreased during the laser cladding of the second layer of copper alloy, and the segregation phase could not be formed. The microstructure of the copper alloy layer was equiaxed.

Figure 2 (b) shows the interface between the steel substrate and the nickel-based transition layer. From the metallographic structure photo of the interface in Figure 2 (b), it can be seen that the average thickness of the heat-affected zone is 1.7 mm. There are no microcracks or defects near the fusion line or near the base material, and there is no deterioration of the structure, indicating that the quality of the heat-affected zone is good.

From Figure 3, it can be seen that the structure of the heat-affected zone is mainly composed of ferrite and martensite, and the ferrite is distributed in a needle shape. As can be seen from Figure 3 (a), the heat-affected zone near the nickel-based transition layer is a coarse needle-shaped structure, most of which is a coarse martensite structure. Compared with the structure near the base metal area (Figure 3 (b)), the size of the structure near the nickel-based transition layer is smaller, the martensite content is reduced, and the size is smaller. The remaining small block structures are ferrite. Since 42CrMo steel is a medium-carbon low-alloy steel, the addition of alloying elements such as Cr and Mo significantly improves the hardenability of the substrate. The heat-affected zone near the nickel-based transition layer belongs to the fully quenched area. The temperature of this area during laser cladding is above Ac3, and coarse martensite is obtained. The heat affected zone near the base material has a relatively low temperature during laser cladding, which is between Ac1 and Ac3. Under the condition of rapid heating of laser energy, a small amount of ferrite dissolves into austenite, while pearlite, bainite and troostite transform into austenite. During cooling, austenite transforms into martensite, but ferrite remains in austenite and gradually grows during the transformation. Subsequently, a mixed structure of martensite and ferrite is formed, which is also called incomplete quenching zone.

2.2 Mechanical properties analysis
Figure 4 is the stress-strain curve of the copper alloy layer of copper/steel composite material. The test results show that the tensile strength Rm of sample No. 1 is 503.8MPa, the specified plastic extension strength Rp is 359.2MPa, and the elongation is 9.4%. The elastic modulus is calculated to be 107.650 GPa, and the cross-sectional shrinkage is 7.7%. The tensile strength Rm of sample No. 2 is 500.8MPa, the specified plastic extension strength Rp is 354.0MPa, and the elongation is 12.7%. The calculated elastic modulus is 106.076GPa, and the cross-sectional shrinkage is 9.4%. The average tensile strength of the copper alloy layer is 502.3MPa, the average yield strength is 356.6MPa, the average elongation is 11.05%, the average cross-sectional shrinkage is 8.55%, and the average elastic modulus is 106.863 GPa. After measurement and calculation, the average hardness of the alloy layer is 191HB.

BS-EN-1982 standard usually requires that the Rp of the centrifugal casting CuSn12 binary alloy is 150 MPa, Rm is 280 MPa, and the elongation is 5%. In this study, nickel-based alloy was used as the transition layer. Under the dissolution and diffusion of laser energy, the nickel element in the transition layer diffused into the copper alloy layer of the copper-steel composite structure. The test results showed that the Ni-containing copper-tin alloy had higher yield strength, tensile strength and elongation, indicating that the presence of nickel is one of the key factors in strengthening the copper-tin alloy. Some studies have also pointed out that the microstructure and phase formation of nickel-tin bronze are significantly different from binary bronze, which is particularly obvious in alloys with high nickel content, forming a complex ternary phase. However, alloys in the high Cu region of the ternary system are usually analyzed as pseudo-binary alloys with a fixed Ni component. In these ternary phase diagrams, it is found that the phase boundary of the CuSn12Ni region has not yet been determined. As is known to all, the morphology, distribution and precipitation position of the alloy strengthening phase have a great influence on the comprehensive mechanical properties of the alloy. In the microstructure of the centrifugal casting CuSn12 binary alloy, the continuous coarse δ phase is distributed at the grain boundary, while the microstructure δ phase obtained in this study is scattered in the grain and grain boundary in the form of a five-pointed star or irregular sphere. The laser cladding technology used in this study is a rapid solidification process. Under such rapid thermal cycle conditions, the grains will not have enough time to grow, so the grains of the obtained copper alloy layer are finer than the casting structure. Since the grain refinement weakens the segregation of tin, and the increase in grain boundaries separates the δ phase, the continuous coarse δ phase disappears. Therefore, it can be concluded that the five-pointed star morphology or irregular spherical δ phase scattered in the grain and grain boundary has a good strengthening effect on the copper alloy layer, while the continuous clusters of δ phase distributed at the grain boundary have a negative effect. Lin et al. reported the absence of ternary phases when studying the Sn-Cu-Ni solder system, while Schmetterer et al. only identified two ternary compounds with low Cu content. Lin et al. also proposed a continuous solid solution between the Cu and Ni ternary angles. The addition of nickel (Ni) to the copper-tin (Cu-Sn) system has a significant effect on the microstructure, mechanical properties and processing parameters of the alloy. However, the interaction between copper, tin and nickel atoms in this alloy is still unclear.

Figure 5 is a shear force and displacement relationship diagram of the copper alloy layer of the copper/steel composite structure. The shear specimen is shown in the upper right corner of Figure 5. The size of the small cuboid above the specimen is 9.7 mm long, 2.5 mm wide and 2.5 mm high. The small cuboid and the interface with the large cuboid below are all copper alloy layer organizations. The test results show that the shear strength of sample No. 1 is 453.1MPa, the shear strength of sample No. 2 is 465.7MPa, the shear strength of sample No. 3 is 433.9MPa, and the average shear strength of the copper alloy layer is 450.9MPa. From the above data, it can be seen that the copper alloy layer exhibits good shear failure resistance. The copper-steel composite structure can be applied to engineering structures and components that need to withstand large shear forces, and can provide better safety. Shear strength is related to factors such as the organizational structure, grain size, strength, and defects of the material. Therefore, the copper-steel composite structure prepared by laser cladding technology in this study has excellent comprehensive performance.

The shape test sample and the tensile test fixture shown in Figure 6 are processed on the prepared copper-steel composite material using a CNC machine tool. The rightmost is the assembled bonding test sample. The bonding interface between the bonding test sample and the copper alloy is a nickel-based transition layer. The assembled bonding test sample is clamped at both ends of the tensile machine for bonding test. Before the bonding test, the mobility of the bonding test sample in the test fixture was manually tested to prevent the upper and lower stuck from causing large errors in the bonding test value. Figure 6 is a graph of the interface bonding strength and displacement of the copper alloy layer. In this test, a total of 4 bonding test samples were prepared, and the diameter of the bonding surface with the copper alloy was 3.6 mm. As shown in Table 3, the interface bonding strength of sample No. 1 is 616.5 MPa, the interface bonding strength of sample No. 2 is 604.1 MPa, the interface bonding strength of sample No. 3 is 549.7 MPa, and the interface bonding strength of sample No. 4 is 627.1 MPa. The average bonding strength is 599.35 MPa. Samples No. 1, 2, and 4 show obvious necking, and the interface bonding strength is high. It can be seen from the EDS analysis of the fracture (Figure 7 (e)) that the chemical composition of the fracture is mainly nickel and a very small amount of titanium, which proves that the fracture occurs in the nickel-based transition layer and there are a large number of dimples on the fracture. The fracture occurred in the transition layer, and the specimens all showed obvious necking. During the fracture process, not only a large number of dimples appeared, but also river and tongue-shaped patterns appeared, forming a quasi-cleavage surface (as shown by the white arrows in Figure 7 (b)). In addition, fine dimples were observed in the fracture morphology, as shown by the solid line in Figure 7 (b). The formation of these fine dimples is related to the nano-scale precipitates in the nickel-based transition layer.

The fracture photo in Figure 8 shows that the bonding strength test specimen did not undergo obvious necking. As shown in Figure 8 (f), the fracture is mainly composed of Cu, Sn and a small amount of Ni elements. Therefore, it can be concluded that the fracture of the specimen occurred at the interface between the nickel-based transition layer and the copper alloy layer. The interface bonding strength is 549.7MPa, which is lower than the interface bonding strength of the fracture in the transition layer. As can be seen from Figure 8 (a), the fracture is relatively flat and perpendicular to the direction of the applied tensile stress. Obvious steps and cleavage surfaces can be seen in the microstructure (as shown by the black arrows, Figure 8 (b)). The fracture mechanism is a ductile-brittle combination fracture. From the fracture, we can see the existence of microcracks (indicated by red arrows, see the electronic version for color pictures, the same below) and hole defects (indicated by yellow arrows). The brittle second phase that precipitates discontinuously along the grain boundary provides the possibility for the crack to expand at the grain boundary, which improves the strength of the copper alloy layer but reduces the plasticity.

Figure 9 is the SEM images of the sample surface at different magnifications after the friction and wear test of the copper alloy layer under different load conditions with a friction time of 30 minutes. As can be seen from Figure 9, with the increase of load, the width and depth of the wear scar of the alloy increase. During the reciprocating friction process, the accumulation of the test friction surface causes the edge of the wear scar to bulge, indicating that the Cu alloy has undergone obvious plastic deformation under the action of friction shear force; and delamination occurs during the friction process, which leads to the unevenness inside the wear scar. When the alloy is subjected to a friction and wear test under a load of 5 N, the friction-induced plastic deformation of the surface layer will penetrate into an area far away from the wear surface, resulting in more serious wear. This is because the load increases, the heat generated during the friction process increases, some debris softens and adheres to the friction surface, and after reciprocating friction, the alloy surface further peels off and the wear intensifies.

Figure 10 is a graph of the friction coefficient of Cu alloy changing with time under the same frequency conditions when the load is 3 N and 5 N for 30 min. As shown in Figure 10, during the running-in stage, due to the uneven surface roughness of the alloy samples, the friction coefficient of all samples is relatively low and has certain fluctuations. When the load is 3 N and 5 N, the friction coefficient of the alloy is in a stable trend and the friction process is stable. The main reason is that under lower loads, friction may be mainly controlled by adhesion and micro deformation, which dominates the two load ranges. The friction mechanism does not change significantly, so the friction coefficient remains relatively stable. When the load increases from 3N to 5N, the average friction coefficient of the copper alloy layer increases from 0.786 and 0.807 to 0.832 and 0.846, ranging from 0.786 to 0.846. This is mainly because when the friction time is constant, the greater the load, the friction heat generated during the friction process will change the state of the contact layer on the surface of the material, causing a large area of ​​the surface to be molten, and the wear mechanism has changed. The wear mode of the copper alloy layer gradually changes from abrasive wear to adhesive wear. This change is usually accompanied by an increase in the friction coefficient.

Figures 11 and 12 are the three-dimensional wear morphology and contour diagrams of the copper alloy under different loads (3 and 5 N), respectively. It can be seen from Figures 11 (a) and 12 (a) that the width and depth of the wear scar of the copper alloy increase with the increase of the load. During the reciprocating friction process, the accumulation of the test friction surface causes the edge of the wear scar to bulge, indicating that the alloy has undergone obvious plastic deformation under the action of friction shear force. When the same alloy is subjected to different loads, the greater the load, the more serious the bulge on the edge of the alloy. This is because as the load increases, the heat generated during the friction process increases, and some debris softens and adheres to the friction surface. After reciprocating friction, the alloy surface further peels off, and the wear increases. When the load increases, the large load squeezes the contact surface into the softer grinding surface, the actual contact area increases, the rough peak is quickly flattened under the lateral shear force, the wear depth increases, and the wear amount increases; as the load increases, friction heat is generated, the contact surface temperature rises, the friction surface softens and the hardness decreases, resulting in an increase in wear. When the friction load is 3N and 5N, the wear rate of sample No. 1 is 2.4×10-3 mm3/ (N·m) and 2.6×10-3 mm3/ (N·m), respectively, and the wear rate of sample No. 2 alloy is 1.9×10-3mm3/ (N·m) and 2.5×10-3mm3/ (N·m), respectively. From the results, we can see that the reason why the wear rate of copper alloy at room temperature is relatively high is that the hardness of the grinding ball is higher than that of copper alloy. During the friction process, copper alloy is more susceptible to wear. The formation and transfer of wear products affect the friction behavior, which ultimately leads to a high wear rate.

3 Conclusions

In this paper, copper/steel composites were prepared by laser additive manufacturing technology. The microstructures of the copper alloy layer, interface, and matrix of the copper/steel composites were characterized by OM and SEM. The mechanical properties and friction properties of the copper alloy layer and interface were tested by tensile and friction and wear tests. The effects of transition layer elements and copper alloy layer precipitation on the alloy structure and mechanical properties were revealed. The fracture morphology and fracture mechanism were studied and analyzed.

(1) The microstructure of the copper/steel composite showed irregular spherical δ phases in gray and discontinuously distributed along the grain boundaries, pentagonal δ phases scattered in the grains, and α-Cu in bright white. No obvious defects were found in the copper alloy layer, and the grain size was 8 to 9. The average thickness of the interface heat affected zone was 1.7 mm. There were no micro cracks or defects near the fusion line and near the base material, and there was no structural deterioration.

(2) The copper alloy layer exhibits excellent mechanical properties, which is attributed to the diffusion of nickel elements from the transition layer into the copper alloy layer. The nickel element causes the copper-tin alloy to form a complex ternary phase. The alloy structure prepared by laser additive technology has fine grains. The refined grains weaken the segregation of tin. The increase in grain boundaries causes the separation of the δ phase. The disappearance of the continuous coarse δ phase is transformed into scattered precipitation within the grain and at the grain boundary.

(3) The average tensile strength of the copper/steel composite prepared by laser additive manufacturing is 502.3MPa, the average yield strength is 356.6MPa, the average elongation is 11.05%, the average cross-sectional shrinkage is 8.55%, the average elastic modulus is 106.863 GPa, the average shear strength is
450.9MPa, the interface bonding force between the copper alloy and the transition layer is 549.7MPa, the friction coefficient is between 0.786 and 0.846, and the copper alloy layer exhibits good mechanical and wear resistance.

(4) The fracture of the copper alloy layer is relatively flat and perpendicular to the direction of the applied tensile stress. Obvious steps and cleavage planes can be seen. The fracture mechanism is a combination of ductile and brittle fracture. The brittle phase that precipitates discontinuously along the grain boundary is the crack source. The expansion at the grain boundary increases the strength of the copper alloy layer but reduces the plasticity.