Development of a Multimaterial Structure Based on CuAl₉Mn₂ Bronze and Inconel 625 Alloy by Double-Wire-Feed Additive Manufacturing

Abstract

This work studied the possibility of producing multimaterials consisting of aluminum bronze CuAl₉Mn₂ and nickel-based superalloy Inconel 625 by double-wire electron beam additive manufacturing. Samples with 5%, 15%, 25%, and 50% of the nickel-based alloy in aluminum bronze were produced for the research. The structural features of these multimaterials were analyzed, and tensile properties, microhardness, and dry sliding friction properties were measured. The results showed that 50% of the nickel-based alloy in the multimaterial composition provides the formation of a dendritic structure. Such a material shows worse values of ductility and wear resistance. Samples containing 5%, 15%, and 25% of Inconel 625 provide similar friction coefficient values, whereas, with increasing concentration of the nickel-based alloy, the material’s ultimate tensile strength and microhardness increase significantly.

1. Introduction

Copper and copper alloys have good ductility, thermal and electrical conductivity, high corrosion resistance, and good formability and machinability because they are applied in instrumentation, vehicle, aerospace, electric-power, defense, and other industries [1]. However, the issue with homogeneous materials is that most do not meet the properties required for specific industries, such as energy or marine. While having good corrosion resistance, copper alloys do not have sufficient strength [2]. Multimaterials are actively used in extreme environmental conditions, for example, in the aerospace or nuclear power industries. At the same time, the use of multimaterial structures in the fabrication of components can reduce the volumetric weight of the final product [3,4,5].

It is known that copper–nickel alloys have high mechanical strength and ductility, corrosion resistance, and electrical and thermal conductivity, much higher than similar characteristics of copper- or nickel-based alloys [6,7]. For example, the corrosion resistance of copper–nickel alloys is significantly higher than that of copper alloys due to the embedding of nickel in the protective oxide layer of copper and the reduction in the electrical conductivity of this oxide layer. However, as the nickel content in the alloy is above a specific value, corrosion resistance begins to decrease again. It is associated with forming new vacancies at the interface between the oxide layer and the electrolyte [8,9]. Inconel 625 is a nickel-based alloy. The molybdenum and niobium in the nickel–chromium alloy matrix give the alloy high mechanical stiffness and resistance to high-temperature effects, such as oxidation and carburization [10,11]. Due to these properties, copper–nickel alloys are used in the aircraft, marine, instrumentation and vehicle engineering, and energy industries, for example, for heat exchangers, heating cables, thermocouples, resistors, condenser pipes of power plants, and pipelines on offshore drilling platforms and ships [5,6,12].

The difficulty in producing copper–nickel alloys is that the high melting temperature of copper and nickel (1085 °C and 1455 °C, respectively) makes the production process of these alloys expensive, and the homogeneity of the alloy structure is violated when the alloy solidifies [13]. The authors of the paper [14] used electrodeposition to lower the process temperature (900 °C), but it becomes difficult to control the final composition of the deposited layer. The development of additive technologies makes it possible to manufacture single- and multicomponent products of different shapes, sizes, and purposes, the properties of which are difficult or even impossible to achieve by conventional manufacturing technologies [15,16,17,18]. Additive-manufactured multimaterial structures have demonstrated improved values of such indicators as hardness, ductility, and wear resistance compared to components made of mono-materials [19,20]. Wire-feed electron beam additive manufacturing is effective in producing complex-shaped metal components, and the operation in a vacuum allows working with easily oxidized materials [21,22]. At the same time, wire-based additive manufacturing is a much less expensive method than a powder-based one [23]. In addition, it is possible to feed two or more materials into the melting pool, between which layers a gradient transition can be achieved [15,20,24]. In this regard, the fabrication of various multimaterials with unique physical and mechanical properties is an urgent task, the solution of which will allow obtaining a new generation of functional materials for application in various industries. Thus, this work will investigate the fabrication of multimaterials based on aluminum bronze and nickel-based superalloy by the double-wire additive manufacturing method and study its structural features and characteristics.

2. Materials and Methods

The samples of multimaterials shaped as vertical walls were produced for the studies. The samples were produced using a self-developed electron beam additive manufacturing machine (ISPMS SB RAS, Tomsk, Russia). The printing process occurred in a vacuum chamber. The 3D printing was performed using a double-wire-feed system. A 1.2-mm-diameter CuAl₉Mn₂ aluminum bronze wire and a 1.2-mm-diameter Inconel 625 nickel-based superalloy wire were used for the 3D printing. Feeding of the feedstock material was carried out simultaneously in one melting pool. Aluminum–manganese bronze CuAl₉Mn₂ was chosen as a matrix material, while Inconel 625 alloy was used as a hardening additive. As a result, samples with 5, 15, 25, and 50 vol.% of the Inconel 625 alloy were produced.

The printing was carried out according to the scheme shown in Figure 1. Samples (1) were printed on a steel substrate (2) by feeding filaments of the matrix and hardening alloys (4) through nozzles (3). Melting of filaments was performed by an electron beam (5) from a gun (6), directed through a magnetic focusing system (7) to the printing zone, and forming a melting pool (8). Additive content was adjusted by decreasing the feeding rate of the bronze wire and increasing the feeding rate of nickel-based alloy wire from corresponding feeding devices (9) to achieve the required ratios. Parameters of the electron beam additive manufacturing were accelerating voltage 20 kV, linear printing speed −400 mm/min, and beam current −72.5–77.5 mA at the bottom of the sample and 45–46 mA at the top of the sample.

Through electrical discharge machining, specimens for metallographic studies, static tensile tests, and dry sliding friction tests were cut from the samples produced (10 in Figure 1). EDM-machine DK7750 (Suzhou Simos CNC Technology Co., Ltd. Suzhou, China) was used for cutting. The samples were ground on abrasive paper with different grit sizes for metallographic studies and polished with a suspension. Then, the prepared surface was chemically etched to reveal the material microstructure.

The multimaterial formation process is presented in more detail on the scheme in Figure 1b. In the printing zone, the electron beam is fed through the magnetic focusing system with the formation of a sweep as a circle, diameter 5 mm (11), dividing the melting pool into two parts. The first part is represented by the melted part (12) of the previously deposited layer (13). The second part is represented by a melting pool of filaments (14,15) mixed in a common melting pool. The use of beam sweep is caused by the need to melt both wires and the substrate and the difficulty in ensuring high accuracy of wire entry into the printing zone due to the presence of kinks and internal stresses. The beam cross section was circular. The beam was focused to a depth of 1 mm below the print surface. The filament arrangement is set in such a way that the main component of the polymetallic material is fed over the hardening component, which prevents the wire of the second component from dripping when its feed rate is low. If the wires are perfectly spaced, the second component, fed at a slower rate during melting, may not have time to enter the melting pool by the flow and form a drop at the end of the wire because of the surface tension of the molten metal. As the drop increases, it falls into the melting pool due to its weight, which exceeds the resistance force of the surface tension. In this case, there is a transition to the droplet mode, characterized by significant heterogeneity of component distribution in the composite after printing. Using the scheme shown in Figure 1 allows for eliminating such a phenomenon.

The structure of the material was investigated using an Olympus LEXT OLS4100 confocal laser scanning microscope (Olympus Corporation, Tokyo, Japan) and an Apreo 2 S LoVac scanning electron microscope (Thermo Fisher Scientific Inc., Waltham, MA, USA) with the EDAX Octane Elect EDS System (AMETEK, Inc., Berwyn, PE, USA). The phase composition was studied by X-ray diffraction analysis using an XRD instrument XRD-7000S, CoKα (Bourevestnik, JSC, Saint-Petersburg, Russia).

Static tensile tests were performed using dog-bone flat specimens with the gauge length of 12.0 mm on a universal testing machine UTS 110M-100 (Testsystems, Ivanovo, Russia). Dry sliding friction tests according to the pin-disk scheme were performed on the TRIBOtechnic tribometer (Tribotechnic, Clichy, France). For this purpose, pins of 5 × 5 × 10 mm were prepared in the shape of parallelepipeds. The pins were made from multimaterial samples according to the layer deposition direction (DD) and the wall growth direction (GD). AISI 420 steel was used as a counterbody. The friction surfaces of the samples and counterbodies were grinded to 2000 grit on sandpaper. The following parameters were chosen for the test: load is 15 N, friction path radius is 10 mm, speed is 400 rpm, and time is 3 h. The linear dimensions of the pins and their weight before and after the test were measured to determine the wear. The material microhardness was measured with 0.25 mm increments and a load of 200 g using a Duramin 5 microhardness tester (Struers A/S, Ballerup, Denmark).

3. Results

3.1. X-ray Diffraction Analysis

To describe the overall phase composition of the samples obtained, X-ray diffraction patterns were taken of all the multimaterial samples (Figure 2). The characteristic changes are revealed immediately, even with the addition of 5% nickel alloy to the bronze. As shown in Figure 2, the main phase of the resulting multimaterial is the fcc-Cu phase, or more exactly the solid solution of nickel in copper. According to X-ray diffraction analysis, the volume fraction of the copper matrix here is 91.1 vol.%, while the detected NiAl and Cu3Al phases occupy 8.2 and 0.7 vol.%, respectively. In the other two cases, 15%Ni and 25%Ni, the NiAl phase occupies 17.5 vol.% in both cases, while the Cu3Al phase is undetectable. The peculiarity of this study is that the β-phase (NiAl), whose volume increased with the addition of nickel alloy, was present in the lowest concentration when the nickel alloy was added at 50%. This is indicated only by the reflection (111) of the NiAl phase, which overlaps with the reflection of phase γ′ (Ni₃Al), which was formed in a high volume (7.5 vol.%) in the 50%Ni sample. In addition to phase γ′, a brittle σ-phase (Cr3Ni2) is formed in the 50%Ni sample in a volume of ~4.5 vol.%.

3.2. Microstructure Analysis

Figure 3 shows optical microscopy images of the characteristic areas for all the samples obtained. All samples were homogeneous at the macro level except for the implicit interlayer boundaries, which can be observed in Figure 3e.

The multimaterial samples produced using aluminum bronze CuAl₉Mn₂ show significant differences depending on the content of the nickel-based alloy in the composition. With the addition of 5% nickel-based alloy, the typical structure of aluminum bronze with the formation of small grains of intermetallic phases β-Cu3Al during solidification of α-Cu matrix grains, which are defined as dark areas in the metallographic image (Figure 3a,b). At the same time, small inclusions of fine nickel-based alloy particles in both α-Cu matrix and β-Cu3Al are observed (Figure 3b).

When 15%Ni is added, the β-phase fraction significantly decreases (Figure 3c,d). However, the amount of nickel-based alloy particles significantly grows, and particles are formed, which appear to be the result of alloying elements of the nickel-based alloy interacting with the aluminum bronze (dark gray inclusions in Figure 3d). With an increase in the Ni-alloy fraction up to 25%, such particles begin to predominate in the material structure, and β-Cu3Al is practically not detected (Figure 3f). A further increase of up to 50%Ni leads to the fact that the structure of the multimaterial acquires the features of a Ni-alloy rather than bronze (Figure 3g,h). This manifests in forming a dendritic structure characteristic of nickel-based alloys produced by additive methods [25].

It should be noted that the sizes of α-Cu grains are clearly distinguished only in the case of 5%Ni. Further increasing the content of the Ni-alloy makes it impossible to distinguish the boundaries of such grains in metallographic images.

Scanning electron microscopy and elemental composition analysis using the energy dispersive method were carried out to determine the nature of the formed structures. For each multimaterial composition, chemical element mapping and elemental analysis of local areas were performed to determine the types of formed particles in the structure (Figure 4).

At 5%Ni, three types of formed compounds are distinguished. The first type is the compound Ni–Cu–Al (1 in Figure 4a). The local elemental analysis made it possible to determine that the dark inclusions of type 1 represent a compound based on 47 wt% Cu-33 wt% Ni-17 wt% Al. The second type is fine particles containing Ni and Cr (2 in Figure 4a), with Cr observed only in particles of this type. Moreover, the third type of particle is a compound of Mo and Nb, which stand out in the image as bright light particles (3 in Figure 4a). Moreover, the point analysis shows that this is a solid Mo(Nb) solution in the 1/1 component ratio.

When 15%Ni is added to the aluminum bronze in the copper matrix, a Ni content of not more than 9 wt% is observed. The SEM image clearly shows small platelike inclusions of dark color, which are NiAl compounds (4 in Figure 4c). It is also possible to distinguish light gray areas enriched with Mo and Cr (5 in Figure 4c). The brightest glow in the SEM image also shows particles containing Nb and Mo, but in this case, these particles no longer represent a 1/1 ratio. Local chemical analysis shows about 65 wt% Mo, 10 wt% Nb, and 15 wt% Cr in these particles (6 in Figure 3c).

At 25% of Ni-alloy, the formed particle pattern is practically unchanged. Ni₃Al particles and other types of particles are observed, but Mo(Nb) particles are represented in smaller amounts, and Mo and Nb compounds with Cr and Ni prevail. A more detailed map of element distribution can be seen in Figure 5.

Figure 4g,h shows that, with the addition of 50% nickel-based superalloy to aluminum bronze, the matrix of the material undergoes a serious change, which now contains alloying elements of both CuAl₉Mn₂ bronze and Inconel 625 alloy. Thus, at point 7 the composition of the matrix is as follows: Cu-42.1 wt%, Ni-40.3 wt%, Cr-9.4 wt%, Al-4.2 wt%, Mn-1.5 wt%, Nb-1.4 wt%, and Mo-1.0 wt%. At the same time, the particles formed during the solidification of the sample do not contain MoNb compounds.

Because the aluminum bronze matrix is enriched with nickel and other elements of the Inconel 625 alloy when 50% of it is added, the structure undergoes significant changes and goes from a grained to a dendritic one. Now, it is necessary to understand how structural changes in bronze/Inconel multimaterials affect the improvement of mechanical properties.

3.3. Mechanical Properties

Multimaterial microhardness measurements were made at three locations in the lower (10 mm), middle (20 mm), and upper (30 mm from the substrate) parts of the vertical walls with different Inconel 625 alloy content. The obtained average values demonstrate a significant increase in the material microhardness with increasing Ni-alloy content in aluminum bronze (Figure 6). Therefore, at 5% Ni, the microhardness of the multimaterial (134–166 MPa) is 16–44% higher than the values for the pure aluminum bronze (115 MPa). When the nickel alloy content is increased to 15%, the increase is at least 69% (193 MPa), and at most, the microhardness doubles (230 MPa). A further increase to 25%Ni results in an increase in microhardness values of 2.8 times at the upper part of the wall (319 MPa), and at 50%Ni, the maximum microhardness value of 414 MPa is observed, which is 3.6 times higher than the microhardness of aluminum bronze.

However, it is worth noting that samples with 5%, 15%, and 25% of Ni-alloy show a tendency for the material microhardness to increase with the wall growth. This is probably due to the cooling rates in the printing process. Since heat removal efficiency through the substrate decreases with moving away from the substrate, solidification is slower on the top layers, so the probability of reacting nickel-based alloy and bronze elements increases. This leads to the fact that with the growth in the wall, the proportion of secondary intermetallic phases in the matrix of the multimaterial increases, which leads to an increase in microhardness. At 50%Ni, such a situation is not observed. This is because, with such a fraction of nickel-based alloy, a dendritic structure is formed from the beginning of 3D printing to the top of the wall.

Static tensile tests also show an increase in strength characteristics with an increase in the content of Inconel 625 alloy in the aluminum–manganese bronze matrix. Nevertheless, the results are not so unambiguous.

Table 1. Tensile test results for multimaterials and initial materials.

Property	UTS, MPa	Ductility, %	UTS, MPa	Ductility, %
Direction	Wall Growth Direction		Layer Deposition Direction
CuAl9Mn2	468 ± 33	50 ± 0.8	430 ± 5	63 ± 1.2
Inconel 625	815± 19	30± 3.0	868± 28	21± 4.2
5%	525 ± 13	36 ± 1.4	519 ± 31	35 ± 3.5
15%	560 ± 14	22 ± 2.1	584 ± 17	20 ± 4.1
25%	794 ± 55	11 ± 1.8	824 ± 88	13 ± 1.8
50%	796 ± 55	8 ± 1.2	530 ± 167	6.4 ± 1.5

shows that the tensile strength of material with 5% nickel alloy is higher than that of pure CuAl₉Mn₂ alloy (519–525 MPa against 440 MPa), and the ductility also increases. It is known that the elongation of CuAl₉Mn₂ alloy material under tensile test can reach 20%, while at 5%Ni, the multimaterial demonstrates values of about 35%.

A further increase in the Ni-alloy content leads to a nearly twice increase in the ultimate tensile strength—up to 824 MPa at 25%Ni, although a decrease in the material’s ductility is observed. The sample with 25%Ni appears to be the most optimal in terms of strengthening the material since it provides an average increase in the ultimate tensile strength of 1.8 times, while the ductility of the material obtained remains at the acceptable level for the CuAl₉Mn₂ alloy.

The sample with 50%Ni showed the most unstable result, as its tests showed the lowest ductility of the material and too low values of the ultimate tensile strength when tested in the layer deposition direction. Apparently, this is due to the structural heterogeneities of the produced material.

The identified changes in the mechanical properties of the multimaterial should also affect its operational properties, in particular, its tribological characteristics.

3.4. Tribological Characteristics

The dry sliding friction tests of the multimaterials showed that adding 50% of Ni-alloy to the sample composition significantly increases the friction coefficient (Figure 7). Consequently, this sample demonstrates the worst tribological properties among all tested samples, which is also evidenced by the data on the degree of friction wear of the samples (Figure 8).

At the same time, the results show relative friction coefficient values for samples with 5, 15, and 25% of nickel-based alloy in the layer deposition direction and almost identical values in the wall growth direction in the steady-state friction mode.

The sample with 15%Ni shows the lowest wear values, while the sample with 5% shows the highest (Figure 8). The sample with 25% nickel alloy shows the average values for both friction coefficient and material wear (Figure 7 and Figure 8).

4. Discussion

The results demonstrate that the fabrication of multimaterials based on the aluminum bronze CuAl₉Mn₂ and the superalloy Inconel 625 makes it possible to achieve material characteristics that cannot be obtained for each component separately. This becomes possible primarily due to structural transformations during the solidification of the material during electron beam additive manufacturing. The analysis of the formation mechanisms of the phases shown in Figure 4 is determined by the thermodynamics of formation in a multicomponent system and will be studied in detail in the future work. The formation of Ni_xAl_y phases, when adding nickel alloy to aluminum bronze is thermodynamically justified [26]. The addition of nickel to aluminum bronze leads primarily to the formation of Ni_xAl_y phases rather than the dissolution of nickel in the copper matrix. At low concentrations (Figure 4a,b) of Ni in aluminum bronze, a Cu₃Al eutectic is formed with dissolved Ni in the copper matrix, as the amount of Ni does not allow the Ni_xAl_y phases to form. When a definite concentration of nickel is reached in the bronze, NiAl is formed, and when the concentration of Ni is further increased, the Ni₃Al phase is formed and the concentration of NiAl phase decreases, as shown in Figure 4c–h and confirmed by X-ray analysis (Figure 2). However, the formation of particles of the Cr–Mo–Nb system is because these elements do not form mutual phases with copper. In this case, from the nickel alloy melt introduced into the bronze, the nickel forms Ni_xAl_y phases, the remaining material based on the alloying elements Inconel 625 agglomerate forms into large particles, and with increasing concentration of the nickel alloy introduced into the bronze, the size and number of particles increases. Thus, the organization of the material structure during melt solidification acquires a dendritic character, a feature of heat-resistant nickel-based alloys [25]. At the same time, despite the structure’s proximity to a pure nickel alloy, not even close values of strength and ductility could be achieved in this multimaterial.

Moreover, the ductility of the sample with 50% nickel alloy is almost ten times lower than that of the Inconel 625 alloy at room temperature. It is also worth noting that the strength characteristics of this sample are volatile, exhibiting a wide variation in the ultimate tensile strength and inferior to those of the sample with 25% nickel alloy. In terms of friction characteristics, this sample also shows the worst result of all tested samples. This work suggests that the degradation of tribological and mechanical properties is due to the presence of brittle incoherent sigma phase particles, detected by XRD analysis or NiAl to Ni₃Al phase transition, when 50% nickel alloy is added to the bronze. Obviously, sigma-phase presence has a negative effect on the tribological properties of the material. The negative influence of the second component, in the form of gamma’ phase, is not so obvious and will be investigated further using the method of transmission electron microscopy. Therefore, it can be concluded that the high content of nickel alloy in aluminum bronze-based multimaterials is not necessary to achieve improved performance.

All samples with a nickel-based alloy content of 5 to 25% show a gradual increase in microhardness and tensile strength with satisfactory ductility values. The samples exhibit a similar friction coefficient and wear values in the dry sliding friction test. This allows the nickel-based alloy content to be varied in different areas of the multimaterial product to achieve different hardness, tensile strength, or ductility values without significantly affecting the tribological characteristics.

Comparing the results obtained with other multimaterials from aluminum bronze- and nickel-based alloys, the multimaterials produced in this work with 5, 15, and 25% of the Inconel 625 alloy show lower friction coefficient values compared to the composite based on CuAl₉Mn₂/Udimet 500 [26]. At a load of about 15 N, the specimens with Udimet 500 additions show friction coefficient values of about 0.45 at 5% Udimet, which is the best result among all concentrations of this composite, while higher concentrations show friction coefficient values of 0.55–0.6. At the same time, using 5% Inconel 625 alloy, the coefficient of friction is at 0.425, and at higher values does not exceed 0.475. Consequently, the multimaterial produced in this work provides better dry friction wear resistance. Similarly, using Inconel 625 instead of Udimet 500 results in a higher strength product: the ultimate tensile strength of the 25% Inconel 625 is, on average, 15% higher than Udimet 500 at the same concentration of the alloy.

Thus, in this work, it has been possible to produce multimaterials based on CuAl₉Mn₂ aluminum bronze and Inconel 625 superalloy using double-wire-feed electron beam additive technology, which exhibits improved strength characteristics compared to their analogs and provides high resistance to dry friction wear.

5. Conclusions

Based on the structural studies, and mechanical and tribological tests, the following conclusions have been drawn:

• In the manufacture of multimaterials based on aluminum–manganese bronze CuAl₉Mn₂ and nickel-based superalloy Inconel 625 by double-wire-feed electron beam additive manufacturing, increasing the content of Nu-alloy up to 25% leads to an increase in the number of particles of phases based on the Ni-Al, Mo-Nb, Ni-Cr, and Mo-Cr systems in the bronze matrix. At the same time, the character of structure formation does not undergo significant changes. A further increase in the content of the nickel-based alloy up to 50% leads to the formation of a dendritic structure characteristic of Inconel 625.
• Samples with 50% of nickel-based alloy in the composition of the bronze-based multimaterial show worse ductility and unstable ultimate tensile strength values when tested along the layer deposition direction and the wall growth direction. The wear resistance of the samples of this composition is also poor compared with 5, 15, and 25% of Ni-alloy.
• Tribological characteristics of multimaterials with 5, 15, and 25 % nickel alloy demonstrate close values of friction coefficient in the range from 0.425 to 0.475; herewith, strength characteristics of the materials differ considerably: 519–525 MPa for 5 %, 560–584 MPa for 15 %, and 794–824 MPa for 25 %. This makes it possible to vary the ratio of nickel-based alloy in the multimaterial composition in different parts of the product with the preservation of wear resistance.
• Produced compositions of multimaterials using Inconel 625 as a modifying material demonstrate superiority compared with the similar composite material based on Udimet 500 in terms of wear resistance and strength characteristics.