On the Peculiarities of Wire-Feed Electron Beam Additive Manufacturing (WEBAM) of Nickel Alloy–Copper Bimetal Nozzle Samples

Abstract

In order to gain insight into the unique characteristics of manufacturing large-scale products with intricate geometries, experimental nozzle-shaped samples were created using wire-feed electron beam additive technology. Bimetal samples were fabricated from nickel-based alloy and copper. Two distinct approaches were employed, utilizing varying substrate thicknesses and differing fabrication parameters. The two approaches were the subject of analysis and comparison through the examination of the surface morphology of the samples using optical microscopy, scanning electron microscopy, and X-ray diffraction analysis. It has been demonstrated that the variation in heat flux distributions resulting from varying the substrate thicknesses gives rise to the development of disparate angles of grain boundary orientation relative to the substrate. Furthermore, it is demonstrated that suboptimal choice of the fabrication parameters results in large disparities in the crystallization times, both at the level of sample as a whole and within the same material volume. For example, for the sample manufacturing by Mode I, the macrostructure of the layers is distinguished by the presence of non-uniformity in their geometric dimensions and the presence of unmelted wire fragments. In order to characterize the experimental nozzle-shaped samples, microhardness was measured, uniaxial tensile tests were performed, and thermal diffusivity was determined. The microhardness profiles and the mechanical properties exhibit a higher degree of strength than those observed in pure copper samples and a lower degree of strength than those observed in Inconel 625 samples obtained through the same methodology. The thermal diffusivity values of the samples are sufficiently close to one another and align with the properties of the corresponding materials in their state after casting or rolling. The data discussed above indicate that Mode II yields the optimal mechanical properties of the sample due to the high cooling rate, which influences the structural and phase state of the resulting products. It was thus concluded that the experimental samples grown by Mode II on a thinner substrate exhibited the best formability.

1. Introduction

Additive manufacturing represents a sophisticated technology for the design and fabrication of finished products for a variety of applications [1,2,3]. In contrast to conventional technologies, additive technologies allow the stepwise manufacturing of three-dimensional components through the layer-by-layer deposition of molten feedstock materials, such as powder or wire. These technologies are distinguished by their capacity to fabricate intricate and continuous components with minimal additional processing and material expenditure. Among the various additive technologies, those that utilize wire as a raw material are becoming increasingly popular [4,5,6,7,8]. The utilization of wire instead of powder results in a reduction in product cost and the time required for the preparation of the raw material. Furthermore, these methods are distinguished by their high production speed, reduced possibility of defect formation, and the ability to create complex geometries in the final product. Therefore, due to the high productivity and versatility of the process, as well as the possibility of fabricating thin metallic products, the fabrication of nozzles is proposed as a means of achieving innovative applications [9]. The fabrication of a nozzle with a complex structure often presents significant challenges when employing conventional methods. Nozzles are often manufactured from copper or its alloys, which exhibit high thermal diffusivity and melting temperature but are also characterized by low laser absorption. Copper exhibits a strong reflection of the laser beam, particularly at 1064 nm, which is a common wavelength used in laser powder bed fabrication (LPBF), with a reflectivity value as high as 90–95% reported [10,11]. Jadhav et al. [12] demonstrated that it is challenging to achieve the complete melting of copper alloy powder even at elevated laser power levels, which often results in an increased prevalence of defects due to spattering and melt instability. To address the challenge of manufacturing copper-based products, it is essential to consider electron beam additive manufacturing. The electron beam can be effectively combined with any electrically conductive material, including highly reflective alloys such as aluminum and copper. Furthermore, the utilization of a vacuum chamber eliminates the possibility of oxide and impurity formation, thus enabling the formation of impurity-free interface patterns with specific elements in the material employed. Furthermore, the integration of wire technology and a vacuum chamber with electron beam printing markedly diminishes the financial burden of the manufacturing process, enhances its efficiency, enables the attainment of superior physical and mechanical characteristics, and guarantees the superior mechanical strength of the final product, minimizing the likelihood of internal defects.

It is well established within the field of materials research that multi-component products may offer a superior balance between strength and ductility in comparison to homogeneous single-phase products. Copper-based alloys are employed extensively in a variety of sectors, including aerospace, nuclear, and automotive, due to their high thermal and electrical conductivity. In particular, copper alloys are utilized in the production of heat exchangers [13,14,15,16]. However, in comparison to other metals, such as nickel, titanium, and iron alloys, copper alloys exhibit relatively low levels of strength, hardness, and wear resistance. One avenue of promising research is the development of bimetallic components. Bimetallic materials comprising a Cu/Ni combination exhibit excellent thermal diffusivity, characteristic of copper, and high strength, typical to Ni-based alloys [17,18,19,20]. Copper-based alloys are frequently employed in the fabrication of a range of industrial components, including liquid fuel rocket engine chambers, continuous casting crystallizers, etc. Nickel-based alloys exhibit high mechanical strength and wear resistance at elevated temperatures and loads [21]. Consequently, the utilization of bimetallic materials considerably broadens the scope of applications for copper alloys while also enhancing the practical deployment of nickel alloys. The fabrication of bimetallic materials has been demonstrated to result in the improvement of thermomechanical properties, as evidenced by the combination of two aerospace alloys, e.g., the copper-based alloy GRCop-84 and the nickel-based alloy Inconel 718, described in [22]. An additively formed alloy, GRCop-84, has been developed for use as a lining material for the main combustion chamber and nozzle in rocket engines with regenerative cooling. Furthermore, it shows great potential for use in specific applications involving high heat fluxes at temperatures up to 700 °C [23]. The typical production method for GRCop-84 is rapid solidification and powder metallurgy. The most crucial aspect of joining dissimilar materials is the quality of the interfacial bonding region, which determines the metallurgical performance of the bond and the reliability of multi-material components. As reported in [24], experimental samples were produced using LDED. However, the high reflectivity of the fiber laser, which is widely used in LDED, and the high thermal diffusivity of the In718 layer on the surface of the CuCr0.8 substrate, which is deposited by LDED, proved a significant challenge [25,26,27]. Given the complex thermal history typical for additive manufacturing, it is possible for a range of defects to occur if the process parameters, such as the power and scanning speed, are not chosen optimally [28]. Thus, the significant potential of additive manufacturing for controlling the properties of 3D-printed metallic materials must be realized through the judicious selection of the methods and conditions of fabrication. The processing parameters exert an influence on the flow dynamics, heat transfer, and solidification characteristics of the molten metal in the melt pool, which in turn gives rise to microstructural variations in the grain size, morphology, and texture [29]. The fabrication of materials by additive manufacturing methods is influenced by a number of factors, including energy absorption that affects temperature gradients, the crystallization process, the microstructure formed, and the mechanical properties that ensue. Energy absorption is contingent upon the characteristics of the heat source and the material reflectivity vs. absorptivity. The advancement of additive manufacturing technologies has resulted in a direct correlation between the structure of a product and its intended use. The aim of the present work is to study the features of forming experimental nozzle-shaped samples based on a nickel alloy and copper bimetal combination by wire-feed electron beam additive manufacturing (WEBAM). In this context, the structural design of the material used for parts and structures becomes a crucial factor in determining the product’s suitability for specific operational tasks. The mechanical, thermal, and tribological properties of the material are essential considerations in this regard.

2. Materials and Methods

The nozzle-shaped samples under investigation were produced using an experimental setup for wire-feed electron beam additive manufacturing (WEBAM) of metal products from wire. The wire, with a diameter of 1.2 mm, was copper C11000 (with a reported weight percentage of 99.9% Cu, less than 0.005% Fe and 0.002% Ni) and of nickel alloy Inconel 625 cladding material with a reported weight percentage of 58% Ni, 20–23% Cr, 8–10% Mo, 3.15–4.15% Nb, less than 5% Fe, 0.05% Mn, and 0.1% C. The wire supplier is Metal Expedition. A rectangular plate was constructed from AISI 321 stainless steel with a reported weight percentage of 67% Fe, 17–19% Cr, 8–9.5% Ni, less than 2% Mn, and 0.8% C, with thicknesses of 0.5 cm and 2 cm in the case of 3D printing in Modes I and II, respectively.

The principal parameters for large-scale fabrication that control heat deposition and dissipation are the source power and scanning speed. Based on this methodology, samples of complex geometry were obtained using two different modes [30]. In accordance with the first mode (I), a thick substrate of 20.0 mm thickness was selected, whereas in the second mode (II), a thin substrate of 5.0 mm thickness was employed. The thickness of the substrate is a significant factor influencing heat dissipation in 3D printing. The choice of substrate thickness may allow selection of the optimal approach for fabricating large-scale products with complex geometries. In both experimental sample printing modes, two vertical copper walls were grown in parallel in the initial stage (Figure 1a). In this instance, the electron beam was swept in a tapering spiral motion with a diameter of 5 mm. The dimensions of the melt pool during copper wire printing were a diameter of 8 mm, and the overlap between the walls was 2 mm. Consequently, two rows of pure copper C11000 layers were deposited on the substrate to form a cone. Thereafter, the direction of 3D printing was altered in relation to the substrate at an angle of 90° (Figure 1b). In this configuration, the nickel-based alloy layers were deposited in a sequential manner. In this instance, the sweep of the electron beam was elliptical in shape, with a larger diameter of 5 mm. This is comparable to the dimensions of the melt pool utilized during the printing of nickel wire, which measured 7 mm. The nickel layers exhibited a thickness of approximately 2–3 mm, with an overlap of 1 mm (Figure 1b). The difference between modes was that in Mode I, the 3D printing process was terminated to permit the cooling of the entire workpiece, after which the wire deposition was resumed until the finished product was formed. In contrast, for Mode II, the 3D printing process was continuous.

The operating current of the electron beam and the 3D printing method were found to have a significant influence on defectivity and macrostructure, along with the thickness of the substrate used and the size of the experimental sample formed [31]. The manufacturing process parameters, including accelerating voltage, beam current, focal length, beam sweep, wire feed rate, and linear printing speed, were maintained at a constant level for both processes and Modes (

Table 1. 3D printing parameters for manufacturing.

Parameters		Beam Sweep	Liner Speed, mm/min	I, A (1st Layer)	I, A (last Layer)
Mode I	C11000	Spiral	400	80	60
	Inconel 625	Ellipse	300	70	50
Mode II	C11000	Spiral	400	68	67
	Inconel 625	Ellipse	300	70	50

). The accelerating voltage was set at 40 kV, while the sweep frequency was fixed at 1 kHz. In the case of heterogeneous materials with disparate thermophysical properties being employed as the initial material for printing, it is essential to vary the low and high values of the linear energy in accordance with the layer in question. The principal parameters governing the additive manufacturing process are the electron beam current, the linear rate of formation of each layer, and the wire feed rate. It was determined that the optimal value for the electron beam accelerating voltage was 40 kV. In the case of printing with nickel and copper alloy wires, the selection of parameters was based on previously acquired data [30,31,32,33]. Another crucial element was considered when optimizing the parameters for wire printing on previously deposited layers. In particular, the objective was to minimize the values of the linear energy. Since an insufficient energy supply does not allow the initial feed material to be remelted in full, and an excessive value of the linear energy leads to melting of the product, an incorrect selection of basic parameters will result in overheating of the material. This will increase the thickness of the product, which is an undesirable phenomenon.

The utilization of different wires gives rise to disparate thermal conditions during the printing process, attributable to the dissimilar thermophysical properties of the wires. In light of the aforementioned considerations, it is imperative to modify the printing parameters, specifically the beam current and linear printing speed. Consequently, the value of heat input calculated on the basis of technological parameters undergoes either a sudden change [32] or a smooth alteration [33]. Once a height order of 20–30 mm has been reached, the value of the heat input becomes constant for subsequent layers. It is noteworthy that the ranges of the optimal process parameters were similar for both regimes. However, in this paper, only the results for those parameters that demonstrated the most favorable outcomes for each regime are presented. Two nozzle-shaped experimental samples, differing in size, were thus produced: the smaller sample, grown in Mode I, had a diameter of 250 mm and a height of 60 mm, while the larger sample, grown in Mode II, had a diameter of 135 mm and a height of 90 mm, with wall thicknesses of 15 mm.

In order to facilitate a comparative analysis of the structural and mechanical characteristics of the experimental nozzle-shaped samples obtained via different regimes, samples were cut for structural studies, mechanical testing, and thermal diffusivity assessment. The methodology employed for the excision of experimental coupons utilizing electric discharge machining (EDM) is illustrated in Figure 2a.

The samples were prepared by grinding with abrasive paper and polishing with diamond paste and then were subjected to etching for 15 min in a solution of 8 g of CuSO₄, 40 mL of ethanol, and 40 mL of HCl at 80 °C. Macroscopical images of the specimens were taken using the Pentax K-3 camera using a lens with a focal length of 100 mm. Optical studies were performed using a confocal microscope OLYMPUSLEXT (Olympus NDT, Inc., Waltham, MA, USA). In order to determine the elemental composition of the samples, the scanning electron microscope SEMTRACK (Nikisso, Tokyo, Japan/Microtrac, Montgomeryville, PA, USA) mini equipped with an energy-dispersive elemental microanalysis attachment was used. Transmission electron microscope JEOL JEM-2100 (JEOL, Tokyo, Japan) was also employed. In the case of electron microscopic studies of the microstructure and phase composition, the transition zone was the principal subject of investigation. Samples with dimensions 4 × 2 × 1 mm³ were EDM cut off the gradient zone (transition area between nickel- and copper-based alloys), and then ∅3 mm disks were drilled out using a Dimpling Grinder Model 200 (Fischione Instruments, Export, PA, USA), followed by dimpling at the center down to 10–20 μm. Next step was argon ion beam milling at 7 kV for 9 h in a TEM Mill Model 1051 dual-beam ion polishing system (Fischione Instruments, USA) until perforating a hole with surrounding 100–200 nm thickness edges.

X-ray diffraction analysis was conducted on a DRON-7 diffractometer (Bourevestnik, Saint-Petersbourg, Russia) utilizing cobalt Kα radiation. The thermal diffusivity was determined on a Netzsch LFA 457 (Selb, Germany) microflash device on samples of square cross-section with a side of 10.0 ± 0.02 mm and a thickness of 3.5 ± 0.5 mm by the laser flash method, in which the lower part of the sample was heated by an unfocused pulsed laser beam. Concurrently, the temperature increase over time was gauged on the upper surface of the specimen with the aid of an infrared detector (Figure 2b). Subsequently, additional analysis of the temperature evolution over time was conducted, employing the mathematical models offered by the software provided with the instrument.

3. Results

3.1. Macrostructure

One of the defining characteristics of copper–nickel-type compounds is the unlimited mutual solubility of the major components. The structural-phase state of the gradient zone material becomes more complex as a result of the interaction between the alloying elements present in the alloys and the formation of secondary phase particles. Furthermore, the parameters employed in 3D printing play a pivotal role in the manufacturing of such compounds. The high speed of the local melting and crystallization of the material, coupled with a high degree of heterogeneity in chemical and phase composition, contributes to the emergence of a structure formed due to multiple thermal effects resulting from the application of subsequent layers. Another crucial element in the electron beam printing process is the discrepancy in the thermal expansion coefficients of the polymetallic components, which can give rise to thermal stresses during the printing process. Furthermore, the inhomogeneity of stress distribution in complex-shaped products introduces an additional complication, whereby the product may undergo significant distortion or fracture upon cooling. The characteristics of the resulting products, obtained by wire-feed electron beam additive manufacturing (WEBAM), depend on a number of factors, including the power of the heat source, the speed of the scanning, the spacing of the scanning lines, the thickness of the clad layer, the scanning trajectory, and the presence of residual stresses. The aforementioned additive manufacturing process parameters exert a considerable influence on the solidification of metals, the metallurgical processes occurring within the melt pool, and the formation of microstructures subsequent to solidification.

In Mode I, copper layers were deposited in two rows, resulting in the formation of a truncated cone with a diameter of 135 mm and a height of 90 mm. At this step, it is possible to discern the presence of defects on both the internal and external surfaces of the sample. In consideration of the physical and thermal properties of copper (high thermal diffusivity), a reduction in heat input was applied. Consequently, the heat input, specifically the beam current responsible for the changes in the heat input, exhibited similar values for both the first and second copper vertical walls (Table 1).

On the inner side of the specimen, areas of unmelted copper wire 5–10 layers thick are clearly discernible. During the layer-by-layer deposition of the material by WEBAM, a melt pool is formed, in which the feed wire is melted. At this juncture, the operational temperature must exceed the melting point of the material in use. The relatively low melting temperature of copper (Cu 1358 K) and a high value of thermal diffusivity (Cu 401 W/mK) present a significant challenge for local unsteady metallurgy. The thermal diffusivity of copper tends to decrease when it is heated, but at temperatures above 1400 K, the value of thermal diffusivity begins to increase once more. The density of copper decreases when it is heated due to its expansion. The density of liquid copper is approximately 8000 kg/m³ at temperatures up to 1573 K [34], compared to the 8960 kg/m³ density of solid copper at room temperature. The combination of regions of different densities leads to a thermally unstable state. Temperature inhomogeneity therefore may lead to local structural instability, resulting in the presence of unmelted wire and an excessive penetration of the previously deposited layer. An increase in the volume of material fed into the melt pool results in an uneven distribution of material, which in turn gives rise to the formation of a “swell” of material on the previously deposited layer, as well as on the outer surface of the sample produced in accordance with Mode I (Figure 3). In this instance, the molten starting material assumes the form of a roll. When the total surface area of the melt pool exceeds that of a sphere with an equivalent volume, the free surface energy decreases, leading to the formation of “swells” [35]. The insufficient melting of the feed wire or the formation of “swells” results in the formation of non-uniform layers during 3D printing (Figure 3). In this region, wire deposition occurs on the inner side of the wall.

The macrostructure of the layers is distinguished by the presence of non-uniformity of geometric features and the presence of unmelted wire fragments. The anomaly in the deposition process manifests from the outset of the curvilinear section of the layer. Upon the melting of the feed wire by the electron beam, a localized increase in the substrate temperature is observed in the vicinity of the melt pool. Consequently, steep temperature gradients emerge, giving rise to internal stresses. Subsequent to a series of thermal cycles, the deformation of the layers occurs as a consequence of temperature-induced stresses. This results in the formation of a curved interface between the first layer that has already undergone the hardening process and the layer that is being deposited. During the deposition of the upper layer, it is constrained by the significantly cooler lower layer, resulting in elastic compressive deformation. However, at elevated temperatures, the yield strength of the top layer is reduced, thereby allowing for plastic compressive deformation. The cooling of the top layer, which has undergone plastic compression, causes it to shrink and induce bending at an angle with respect to the direction of the layer deposition. This consequently gives rise to tensile stress in the direction of growth. The characteristics observed are typical of those seen in products manufactured using additive manufacturing, in which overlapping melt pools of a half-ellipse shape are formed. The dimensions of the melt pools are statistically 1.1–1.3 mm in width and 0.5–0.7 mm in depth. The boundaries of the melt pool in the sample are observed to exhibit a “fish-scale” pattern [36,37], marked by a discernible grain structure. During the solidification process, the elongated grains undergo growth from the edge to the center of the melt pool, exhibiting alignment along the direction of the growth, which coincides with the direction of the largest temperature gradient.

The physical parameters and mechanisms that have a significant impact on the quality of a sample produced by additive manufacturing are residual stresses and cracking. The formation of cracks during the fabrication process is predominantly attributable to thermal residual stresses that emerge as a consequence of substantial temperature gradients and the rapid solidification of the material, which gives rise to melt pool shrinkage [38]. The application of relatively low scanning speeds with high heat input, resulting from an increased electron beam current, can lead to the development of significant thermal residual stresses. The aforementioned stresses are of a compressive nature at the center of the parts and tensile at the edges, with a high concentration of tensile stress occurring in the vicinity of the substrate. As the level of applied heat input increases, the molten metal is subjected to greater compression, which can result in the onset of cracking [39]. Two common types of cracking are layer solidification cracking and grain boundary cracking. Solidification cracking occurs when the energy content of the melt pool is excessive, resulting in the emergence of stresses between the previously applied and currently applied layers. The nucleation of grain boundary cracking is attributed to the presence of impurity elements at the grain boundaries and/or the growth of grain boundary precipitations [39,40]. Figure 3 depicts the presence of extensive hot cracks extending to the surface. The formation of these cracks is contingent upon the heating of the metal to temperatures in the range of 1200 to 1600 K. At this point, the metal has already initiated the process of solidification. In contrast, other parts of the material remain in a plastic state, with ruptured intercrystalline bonds. The grains that form at the edges are relatively small in size. Subsequently, columnar grains emerge, exhibiting growth perpendicular to the axis of the junction between the smaller grains. Furthermore, the angle between adjacent grains comprising the polycrystalline structure also has an influence on the rupture between the grains.

Following the establishment of two vertical copper walls, a rotation was initiated to facilitate the continuation of printing with the nickel wire positioned perpendicular to the substrate. Accordingly, an experimental nozzle-shaped sample was grown additively in accordance with the initial methodology, as illustrated in Figure 4. Defects of the cracking type were observed in the vicinity of the substrate. In this instance, the occurrence of cracking can also be attributed to the existence of a temperature gradient, which is the consequence of the different melting temperatures of the two materials, and the heat dissipation characteristics of the water-cooled table.

The impact of the differing melting temperatures of copper and nickel-based alloys is evident upon the deposition of the first layer. In this area, a band emerges that mirrors the outline of the melt pool. It is important to note that prior to the deposition of nickel alloy layers, the copper billet underwent an additional intermediate stage of growth cessation. During this stage, the billet was cooled, thereby dissipating heat through the substrate into the cooling table. The additional stopping and cooling of the copper billet had a beneficial effect on the formation of a sharp interface between the materials. The greater thickness of the substrate results in a greater amount of heat transfer occurring during the printing process when Mode I is employed. In such a case, it is necessary to increase the heat input value during fabrication by increasing the current. Consequently, the greater thickness of the substrate employed in Mode I serves to offset the excessive radiation heat transfer during the fabrication of large-scale products.

Mode II was employed for the growth of an experimental nozzle-shaped sample with reduced geometric dimensions (Figure 5). In order to circumvent the formation of the aforementioned defects, it is imperative to attain a state as close as possible to a thermal steady state. Insufficient heat input precludes the remelting of the filament material in its entirety, whereas excessive heat input results in the melting of the product.

Presently, the resolution of these issues is predominantly reliant on empirical data, largely due to the dearth of models that facilitate the immediate establishment of the interrelationship between process parameters, material microstructure, and its physical and mechanical properties [41]. Whilst the required models should describe the structure of the product in its entirety, this is currently a technically challenging proposition for large-scale products [41]. Inconel 625 and copper alloys do not exhibit lack-of-fusion defects in the joint zone. However, the formation of macro-defects is observed when the cone is printed using Mode I. This is a consequence of the combination of the thicker steel substrate and the high thermal diffusivity of copper, which results in altered heat transfer conditions. The formation of porosity is a consequence of the combination of high temperature gradients and the rapid solidification that occurs during the fabrication Mode I as a result of the use of non-optimized process parameters. In the initial method of 3D printing, the elevated power of the electron beam amplifies the heat input and, consequently, the temperature at which pore-type defects are prone to form. Thus, the formation of porosity is evident when two rows of copper layers are deposited, with an average pore size of 75 µm. Furthermore, non-melting and non-flattening defects with an average size of 2.75 mm are observed. In contrast to the experimental nozzle-shaped samples produced by Mode I, the samples obtained by Mode II are free of defects.

Figure 6 illustrates the microstructure of the experimental nozzle-shaped samples produced using Mode II. It is important to note that the samples exhibit variation in their microstructural characteristics.

Consequently, the initial layers of the sample produced by Mode I exhibit a fine-grained structure (with an average grain size of 20 µm) as a result of primary recrystallization. Given that the WEBAM method is distinguished by its high fabrication speed and the limited mobility of small-angle grain boundaries, any alteration in the volume fraction of these boundaries may only occur at elevated temperatures. The considerable disparity in the range of disorientation angles and axes between recrystallized grains results in enhanced grain boundary mobility. The additively grown copper in the first row exhibits a fine-grained structure, with an average grain size of 52 microns. The structure of the second row of additively grown copper is comprised of large grains of columnar form, with an average size of 3.5 mm. The formation of large grains can occur in two ways: firstly, as a result of the neighboring nuclei coalescing, and secondly, due to gradual growth. SEM-EDS analysis of the boundary region structure reveals that the transition zone can be formed of two distinct types: one with the presence of mechanical mixing of components and another practically without it. In the initial scenario, the boundary transition region thickness is observed to be approximately 15–20 microns, while in the subsequent case, it is noted to be in the range of 3–5 microns. The first type is characteristic of samples obtained at high beam currents (Mode I), while the second is characteristic of samples obtained at lower currents (Mode II) during the manufacturing of experimental nozzle-shaped samples.

3.2. Microstructure

In order to gain insight into the structural and phase formation patterns of Cu–Ni alloy bimetal samples, a comprehensive examination by scanning electron microscopy (SEM) and transmission electron microscopy (TEM) was conducted. This analysis aimed to identify and elucidate the patterns of inhomogeneity formation, phase distribution, and properties of interphase boundaries at the microscopic level. The analysis of the sample structure is entirely consistent with the observed change in the chemical composition (Figure 7).

A gradient zone, up to 20 microns thick, can be discerned on the graph of the dependence of elemental composition on the distance to the substrate. This zone exhibits a distinct composition of nickel alloy, differing from the nominal composition, and a notable concentration of copper. When two wires, comprising nickel and copper, are fed simultaneously, a gradual transition in chemical composition is observed, followed by a sudden change, which forms a gradient zone. Some observed variation in the elemental composition is attributable to the fact that the phases within the matrix are partially trapped during the analytical process. The initial area is linked to the emergence of primary copper droplet fractions within the nickel matrix. The second region pertains to the formation of nickel alloy inclusions within the copper matrix. Both areas are distinguished by the inhomogeneity of microstructure and phase distribution (Figure 8).

Upon solidification, the primary droplet fraction assumes a spherical shape. Furthermore, the interfacial boundaries of inclusions are sharp, exhibiting no transition regions. Additionally, structural inhomogeneities are observed in the vicinity of the boundary (Figure 9).

Concurrently, the X-ray diffraction analysis data revealed the absence of peaks associated with intermetallic and carbide phases (Figure 10). The samples are therefore found to be in the form of solid solutions.

To examine the distinctive characteristics of the inclusions and the formation of interfacial boundaries in the samples, transmission electron microscopy results were obtained for foils cut from the gradient zone (Figure 11).

The presence of a Ni solid solution, a Cu solid solution, and the M₆C carbide phase was confirmed. M₆C carbides are infrequently identified by scanning electron microscopy. It is possible for carbides to provide limited hardening in two ways: directly, through dispersion hardening, or indirectly, through grain boundary stabilization, as is often the case. M₆C-type carbides, which contain nickel, molybdenum, chromium, and niobium, are formed in alloys and result in the hardening of the γ-phase. During the crystallization process, carbides are formed at the initial stages due to their higher melting point relative to other phase components. Carbides of this type are enriched in chromium, while the surrounding melt is depleted of chromium. In the present case, the depletion of chromium is insignificant. The shape of these carbides is arbitrary and takes the form of irregular morphologies.

The identified features of microstructure exert an influence on the mechanical properties. As previously stated, the printing process resulted in the formation of a homogeneous and defect-free structure for both the main components and the boundaries between them. A comprehensive strength analysis of additive manufacturing products necessitates the consideration of the physical and mechanical characteristics that are contingent upon the specific additive manufacturing technology and process parameters employed. The aforementioned approach enables the identification of critical values for process parameters, thus facilitating the choice of appropriate manufacturing configuration for specific products. The microstructure that is formed subsequently to solidification is the determining factor for the mechanical characteristics of the clad metal. Thermal diffusivity measurements were conducted on experimental nozzle-shaped samples during isothermal holding at the designated temperatures. The total duration of measurements at each temperature was limited to 30 min. The heating process between isothermal holding temperatures was conducted at a rate of 3 °C/min in a static argon atmosphere. In calculating the thermal diffusivity, the change in the thickness of the sample under study with temperature changes was not considered (Figure 12).

The thermal diffusivity of the specimens demonstrates a direct proportionality with the test temperature. The thermal diffusivity of the specimens exhibits a rapid growth in the temperature range from 273 K to 373 K. Furthermore, the rate of thermal diffusivity growth demonstrates a decline in the temperature range from 423 K to 673 K. In the temperature range from 723 K to 823 K, the growth rate of thermal diffusivity is observed to be relatively slow. However, when the temperature reaches 823 K and above, the growth of thermal diffusivity accelerates. The thermal diffusivity values of the samples are sufficiently close to one another and align with the properties of the corresponding materials after casting or rolling.

3.3. Mechanical Properties

In the experimental samples, mechanical properties were determined from the stress-strain curves (Figure 13).

The mechanical properties of the obtained bimetal samples are summarized as follows: the 0.2% proof strength σ_0.2 = 325 MPa, σ_T = 450 MPa, and ε = 6.5%. Apparently, their mechanical properties exhibit higher strength than those of pure copper samples and lower strength than those of Inconel 625 samples obtained by the same WEBAM route. It is important to note that the observed reduction in ductility can be attributed to the rapid crystallization of the material that occurs when nickel is deposited upon pure copper. This phenomenon, coupled with the rapid cooling that ensues, contributes to the diminished ductility of the resulting material.

The microhardness profiles exhibit variation with the build height (Figure 14). The analysis of microhardness distribution in the copper region yielded an average value of 0.57 GPa, while the nickel alloy region exhibited an average value of 2.45 GPa. Upon crossing the boundary, a notable shift in the mechanical properties is evident, with the microhardness value undergoing a significant change from 1.54 GPa to 3.31 GPa. The sudden shift in microhardness at the boundary can be attributed to the partial integration of copper within the nickel alloy region. During the microhardness measurement, the Vickers pyramid partially falls on the “soft” copper inclusions in the transition zone. These changes are essentially associated with the formation of mechanical mixtures of system components at varying structural and scale levels. The observed increase in microhardness values above those of the strongest material in this pair (nickel alloy) can be attributed to the formation of new phases.

The observations discussed above indicate that Mode II yields the optimal mechanical properties of the sample due to the high cooling rate, which influences the structural and phase state of the resulting products. Nevertheless, despite the optimal mechanical properties, an extensive interface is present between disparate materials. One potential solution to this issue is the implementation of a forced stop during the 3D printing process to facilitate the cooling of the entire workpiece. Nevertheless, this approach may encounter challenges in disrupting the thermal gradient and decelerating the manufacturing process at the factory level. Therefore, the formation of the transition boundary between dissimilar materials is contingent upon the selection of optimal process parameters that ensure the production of a high-quality sample with the requisite geometry. The impact of wire feed geometry is mitigated in this instance due to the fixed positioning of the wire feed nozzle in relation to the product trajectory. The sole challenge encountered pertains to the fabrication of large-scale products with intricate geometries. In this instance, it is essential to regulate the rotation of the product throughout the 3D printing process. This will prevent overheating and the leakage of wall material, thereby extending the range of optimal printing modes for the creation of parts with the required shapes. The primary challenge in this case is to enhance the cooling conditions, which will facilitate the formation of a more optimal structure and, consequently, the acquisition of superior mechanical properties.

4. Conclusions

The present paper contains proof that during the 3D printing of nozzle-shaped bimetal samples by wire-feed electron beam additive manufacturing (WEBAM), the most important parameters that affect the outcome are the thickness and material of the substrate, the wire feed geometry, and the printing path, along with other parameters such as the beam power and the wire feed rate. It has been found that when a thick substrate is used, a defective structure is likely to be formed. The most undesirable defect among those described is the presence of unfused feed wires stacked in layers. The presence of such major defects makes it impossible to use the part. This phenomenon can be attributed to the suboptimal temperature regime, which results in widely differing crystallization times during the manufacturing process. Consequently, the thickness of the material varies across different parts of the experimental sample. It can be posited that higher values of the heat input provide the most favorable structure and enable the attainment of the highest degree of precision in the product shape and surface quality. Further optimization of the parameters employed in the manufacturing of the component will result in an improvement of the thermal conditions of the process and the reduction or complete elimination of defects. It was demonstrated that both geometric parameters and electron beam parameters exert a considerable influence on the accuracy and quality of the products obtained through the wire-feed electron beam additive manufacturing process.