Inconel 718-CoCrMo bimetallic structures through directed energy deposition-based additive manufacturing

Bimetallic structures and coatings through additive manufacturing (AM) have demonstrated a high degree of freedom for tailoring properties depending on the application. In this study, Inconel 718 and CoCrMo were used as both are common alloys and exhibit unique properties, such as high-temperature oxidation, wear, and fatigue resistance. Using directed energy deposition-based metal AM, bimetallic structures containing these two alloys were manufactured, and the resulting structures exhibited no intermetallic phase formation, cracking, or porosity. Scanning electron microscopy and energy dispersive spectroscopy revealed a smooth elemental transition between the two compositions. Hardness testing showed a linear transition in the interfacial zone, validating no brittle intermetallic phase formation. Compression testing and fracture surface analysis revealed that the failures were not dependent on the interface properties. High-temperature oxidation showed no distinct effect on the interface, a firmly attached chromium oxide layer on the Inconel 718 side and a loosely attached chromium oxide layer on the CoCrMo side. There was also evidence of pit formation on the Inconel 718 surface, but not on the CoCrMo. These findings confirm a stable bimetallic system in which one of the two alloys can be used on the other material to improve the structure’s high-temperature oxidation or wear/corrosion resistance.


Introduction
Advances in metal additive manufacturing (AM) technology have allowed designing multi-material structures to impart unique and site-specific properties for various applications. For example, AM-processed FeCrAl coating on a Zr alloy increased the oxidation resistance by a factor of 50 [1] . It has been reported that a 30Cr15MoY alloy steel on a C45 substrate increased the corrosion resistance of neat C45 steel [2] . For strength and thermal conductivity, adding a tungsten alloy coating to Inconel 718 increased the strength and thermal conductivity compared to pure Inconel 718 [3] . A GRCop84 coating on Inconel 718 increased thermal conductivity by 300% more than pure Inconel 718 [4] . Volume  Finally, for increased bioactivity, AM-processed titanium (Ti)-tantalum (Ta) bimetallic structure has shown a significant improvement in biocompatibility similar to pure Ta [5] . A recent review documents many such performance enhancements through AM-processed bimetallic and multi-material structures that highlight the innovation opportunities using AM with the next-generation metallic materials [6] . The interface between the coating and the bulk material can form various intermetallic compounds, which have been shown to increase the strength of some bimetallic systems, such as Ti6Al4V/Al12Si and Ti/Ni [7,8] . However, this interfacial behavior can be detrimental in some cases, which has prompted studies to understand better and model the behavior of these bimetallic systems [9] . In addition, research has been conducted to manufacture multi-material and bimetallic systems more efficiently to be better suited for industry [10] .
Inconel 718 is a nickel-based superalloy consisting of about 55% nickel (Ni), 18% chromium (Cr), 18% iron (Fe), 5% niobium (Nb), and 3% molybdenum (Mo). This alloy is used extensively in aerospace applications due to its excellent high-temperature strength, along with corrosion and oxidation resistance [11] . Inconel 718 is a standard alloy used in AM due to its excellent weldability [12] . This alloy is particularly strong for the formation of the γ' (Ni 3 Ti and Ni 3 Al), the γ" (cubic Ni 3 Nb), and the δ (orthorhombic Ni 3 Nb) phases, all of which have been observed in heattreated Inconel 718 samples that were printed by selective laser melting (SLM) [13] . Different heat treatments can control the amount of these phases formed in Inconel 718 [14][15][16] . This allows manufacturers to tailor the properties of Inconel 718 depending on the application and further demonstrates the versatility of this alloy. The corrosion resistance of Inconel 718 comes from the Cr and Mo, which forms a surface passivation layer limiting oxygen flow into the bulk structure. Because of microstructural and phase evolutions, corrosion resistance for Inconel 718 also depends on the manufacturing and post-processing methods used. It has been reported that excessive carbide formation during solution treatment caused Cr and Mo depletion throughout the bulk Inconel 718 [17] . Oxidation resistance is typically attributed to the Cr 2 O 3 film, which prevents oxygen diffusion into the bulk structure; however, a NbO film also forms at temperatures higher than 800°C to further aid oxidation resistance [18] .
Cobalt-chromium molybdenum (CoCrMo) alloy consists of around 60% cobalt, 30% Cr, and 5% Mo, with 5% being other elements such as Ti. Due to its excellent wear and corrosion resistance, this alloy is used in cutting tools and other applications requiring better wear resistance, such as articulating surfaces of biomedical implants and dental crowns [19][20][21] . However, this alloy's high hardness and wear resistance make it very difficult to machine [22] . As a result, many studies look at the cutting force and anisotropic effects to explain how to properly machine these alloys without compromising the alloy surface or the milling tools [23,24] . The strengthening mechanism in CoCrMo alloy comes from the various carbides, such as Co 23 C 6 and Cr 6 C [25] . Therefore, the properties of CoCrMo alloys can be altered using different heat treatments, which results in various carbide formations [26,27] . The oxidation resistance of CoCrMo at high temperatures is not well reported; however, the alloy's carbide forming abilities and the chromium oxide layer offer good oxidation resistance.
Since CoCrMo is wear-and corrosion-resistant and Inconel 718 is high-temperature oxidation-resistant, a coating of one of these materials may prove effective in increasing the bulk and surface properties of the bimetallic structures. In cases where a strong bulk structure is needed with high oxidation resistance, Inconel 718 coating on CoCrMo would be preferred. A CoCrMo coating on Inconel 718 could be preferred in cases where superior corrosion resistance is needed. A recent study using laser powder bed fusion (L-PBF) of Inconel 718 and CoCrMo functionally graded structure exhibited low porosity and smooth composition transitions [28] ; however, there is still a knowledge gap on processing-structure-properties relationships of DED-manufactured bimetallic structures of Inconel 718 and CoCrMo. Addressing this gap is the primary focus of this research. Inconel 718 and CoCrMo bimetallic structures were manufactured by DED-based AM and were subjected to various tests for microstructure and phase analysis, hardness, compression testing, and measuring high-temperature oxidation resistance. Figure 1 outlines the processing strategies for the Inconel 718 and CoCrMo bimetallic structures using DED.

Directed energy deposition (DED) of bimetallic structures
Inconel 718 and CoCrMo bimetallic structures were manufactured using a powder-based DED printer (FormAlloy, CA). The computer uses a G-code input file derived from a computer-aided design (CAD) file containing the shape information. This G-code input file dictates the machine parameters while printing, including speed, layer thickness, laser power, powder flow rate, and shield gas flow rate, which prevents powders from melting onto the nozzle and clogging it. These alloys were printed onto 316L stainless steel substrate. Inconel 718 powder (Powder Alloy Corporation, Ohio) was used with a particle size of 50-150 μm. CoCrMo powder (Stellite Coatings, Goshen, IN) was used with a particle size range of 50-100 μm. The bimetallic structures were optimized until the part height was close to the theoretical height. The final parameters for each sample are listed in Table 1.

Characterization of bimetallic samples
Bimetallic samples were cut in half and underwent a typical grinding or polishing procedure. The samples were then imaged using a field emission scanning electron microscope (FESEM, FEI-SIRION, Portland, OR) to determine if any defects existed at the interface. Using the energy dispersive spectroscopy (EDS; EDAX), elemental maps were obtained to examine elemental distributions at the interface. Vicker's hardness testing (Penn Tool Co, NJ) was performed along the interface and compared to the base materials' hardness. At least 15 measurements were taken for each point. For the compression tests, an Instron servohydraulic compression tester was used to determine the stress-strain curves for the longitudinal and transverse bimetallic samples and the base materials using a strain rate of 1.3 mm/min. For the oxidation tests, a muffle furnace with no other additions was used to oxidize samples for 96 h at 800°C and then underwent the same characterization process mentioned above. Figure 2 shows a stereoscope image of the bimetallic sample on a 316L substrate and the SEM/EDS analysis of the interface. A small but distinct interface is formed between the CoCrMo and substrate. The same can be observed between the Inconel 718 and CoCrMo junction,  with the bimetallic interface appearing smaller than the substrate interface. The illuminated circles are tiny gas pores, which are more frequent and extensive in the Inconel 718 region than in the CoCrMo region. For the SEM image, a seemingly blank photo is observed with no qualitative distinction between the two materials and the interface. However, the EDS mapping indicates a relatively smooth transition between the elements, meaning there is little to no elemental segregation. No cracking or other defects are observed at the interface, which confirms that these materials are compatible [28] .

Hardness
Vicker's hardness tests were conducted on polished samples and the results are shown in Figure 3. The tests show that the base hardnesses in the Inconel 718 and CoCrMo zones were about 260 HV and 430 HV, respectively. A hardness profile was established, measuring at every 50 μm distance within the interface zone until the pure alloys were reached. The profile shows a linear variation within the interface region. The linear profile confirms no brittle intermetallic phase formations and reveals that the interface is about 150 µm wide. Figure 4 shows stress-strain curves for the four types of nonheat treated compression samples. The test's primary purpose was to understand the role of the interface in compressive deformation behavior. It was found that the interface properties did not limit the compressive deformation in bimetallic structures but were instead controlled by the bulk materials. The transverse bimetallic samples had the worst yield strength, with the base CoCrMo sample performing the best. The Inconel 718 samples did not fail, so the plot only displays up to 0.4 strain for that sample. The failure behavior of the two bimetallic structures is illustrated in Figure 5.

Oxidation studies of bimetallic samples
For the oxidation and followed by hardness measurements, samples were tested at 800°C for 96 h. No significant variations in hardness were observed. Only a slight decrease in hardness (430 HV to 400 HV) was observed in the CoCrMo part, which occurred as a byproduct of the oxidation test since recrystallization would also occur. Figure 6 shows the SEM/EDS analysis of the oxide layer formed in each constituent. For Inconel 718, a strongly adherent chromium oxide layer was formed along with some pits. This pitting resulted from the depletion of Cr near the interface, which made the material under the oxide layer prone to the effects of oxidation. For the CoCrMo side, spalling was observed, meaning that the oxide layer was not strongly attached to the base material. Furthermore, a Cr depletion zone was not observed in the CoCrMo part.

Discussion
The laser DED process is used to form bimetallic structures of Inconel 718 to CoCrMo. The results show that the Inconel 718 and CoCrMo form a stable interface. This indicates that using a coating of one or the other could tailor a structure with specific properties depending on the need, with the CoCrMo coating working better for wear and corrosion resistance applications with static conditions and with the Inconel 718 coating working better for high-temperature oxidation resistance applications with dynamic conditions. Furthermore, the microstructural findings of this study are in line with published data where functionally graded structures were printed through powder bed fusion (PBF) [28] .
It is essential to discuss what coating should be used for what application. Inconel 718 coating on CoCrMo would work better in environments with dynamic conditions in high-temperature environments. This is due to the oxide layer in Inconel 718 being firmly attached, unlike  Figure 4. Stress-strain curves for each compression sample type. Note that only one of each sample type is shown and that the Inconel 718 base sample did not fail and therefore is only plotted until 0.4 strain.

Sample
Inconel 718  the CoCrMo oxide layer. Although it was found that the CoCrMo showed less oxidation damage compared to Inconel 718, the Inconel 718 coating oxide layer would be able to resist further oxidation damage, but the coating would also be able to keep the CoCrMo oxide layer from being washed away, preventing degradation of the bulk structure. In the opposite scenario with a CoCrMo coating on Inconel 718, this structure would be best suited for corrosive environments and situations with high wear, as the corrosion and wear resistance of CoCrMo are unmatched compared to most alloys.
Wen et al. recently published functionally graded structures of Inconel 718 and CoCrMo printed using laser PBF [28] . In this study, we produced defect-free and low-porosity structures without any concerns related to the compatibility of the base materials. Our results also show a compatible interface between Inconel 718 and CoCrMo, even with the case of direct bimetallic transition, which is much more prone to issues arising from material incompatibility due to a sharp change in composition. The hardness of the CoCrMo part in our study compared to the FGM study differed by about 30 HV (420 HV compared to 400 HV, respectively), while the same for Inconel 718 part differed by 90 HV (360 HV compared to 270 HV). However, this difference is expected as the printing techniques are different and it is evident that different amounts of carbides formed in the FGM parts compared to the bimetallic parts due to faster cooling rates in DED. Finally, the stress-strain plots in each study showed considerable failure strain, reinforcing the notion that unwanted intermetallic phases are not a concern for these bimetallic structures.
Although oxidation kinetics are well understood for single material systems, more studies are needed to understand the oxidation resistance of coated or bimetallic systems. For single materials, the oxidation resistance can significantly vary depending on factors, such as oxide formation rate, defects, and mean free path of oxygen. [29] . The oxide layer's composition also matters significantly due to the defect density in the said oxide layer. For example, chromium oxide layers typically exhibit a very low defect density, preventing oxygen diffusion [30][31][32] . Material constants are typically determined experimentally by fitting a curve to the data. In the case of parabolic-type oxidation, the rate equation is given by: Where Kp is the rate constant, and x is the scale thickness [29] . Most metallic alloys follow this parabolic behavior due to the continual formation of the oxide layer, while ceramic materials follow a more linear relationship. In the case of an alloy's thin coating, oxygen's mean free path will be significantly higher to diffuse to the bulk material under the coating. However, the constants could potentially be determined by first determining the oxidation constants for the coating, then using it as a correction curve for the coated structure. In this work, Inconel 718 coating on CoCrMo proved to be a better combination to enhance oxidation resistance of the bimetallic structure due to the strongly adherent oxide layer on Inconel 718 even after 96 h at 800°C.

Conclusions
Inconel 718-CoCrMo was successfully manufactured using the laser DED-based AM. No difficulty was observed in printing these structures, even during optimization, and the finished parts had low porosity without any cracking. The interface could not be distinguished using SEM imaging but was revealed in EDS mapping of elemental transitions. Hardness testing revealed a smooth yet brief transition between the two alloys. Compression testing showed that the CoCrMo had the highest yield strength of the four structures tested and determined that the compression behavior was not dependent on the bimetallic interface. Oxidation tests showed that the bimetallic interface was not affected by the high temperatures and revealed that the Inconel 718 formed a firmly attached chromium oxide layer, while the CoCrMo showed spalling of the same oxide layer.