The Impact of Plastic Deformation on the Microstructure and Tensile Strength of Haynes 282 Nickel Superalloy Produced by DMLS and Casting

The article presents the results of research on the influence of plastic deformation on the microstructure and tensile strength of Haynes 282 nickel superalloy produced by direct metal laser sintering (DMLS) and a conventional technique (casting). Samples were tested for dimensional accuracy using a 3D scanner. Then, the samples were subjected to plastic deformation by rolling. The microstructures of the DMLS and the as-cast samples were analysed using a scanning electron microscope. The strength properties of the samples were determined in a static tensile test. Microhardness measurements of the samples were also performed. Based on the analysis of the dimensional accuracy, it was found that the surface quality of the components produced by DMLS is dependent on the input parameters of the 3D printing process. Using the DMLS method, it is possible to produce Haynes 282 with a fine-crystalline microstructure containing dendrites. The fine-crystalline dendritic microstructure and low porosity showed very good tensile strength compared to the as-cast material. It was also found that the increase in the degree of plastic deformation of the as-cast Haynes 282 and the samples produced by the DMLS technique resulted in an increase in the strength of the tested samples, with reduced ductility.


Introduction
The precipitation-hardened alloy Haynes 282 has excellent properties, such as high strength and creep resistance, due to its γ hardening. Additionally, Haynes 282 is highly susceptible to forming, more so than nickel alloys with similar creep strength [1]. Due to this combination of properties, this nickel superalloy is used in the aviation industry for compressors and turbine housings, as well as for exhaust nozzles and diffusers, due to its good weldability [2].
As is well known, the aviation industry is characterised by short-series production. A common method of producing superalloys in aviation is investment casting, which, though widespread, is ineffective in terms of both cost and time. Limitations, such as the development of new wax models and tools for producing them, and the low strength of ceramic molds and core materials [2,3], has led to a search for replacement technologies that would enable more profitable and time-efficient production. Additive manufacturing (AM) may become such a technology. This technique can be defined as a process of combining materials in layers to create physical objects based on data from a 3D model [4]. AM is already used in the aviation industry [5] because of its suitability for short production series, but it can also be used to manufacture a wide range of products. conventional methods, e.g., casting. This is prompting researchers to experiment with the plastic deformation of materials produced using AM methods.
Although both cast and DMLS Haynes 282 nickel superalloy are studied extensively, most published studies focus on conventional heat treatment and post-processing as described above. Moreover, no study has yet compared the microstructure and mechanical properties of the as-cast and deformed Haynes 282 superalloy. Apart from the obvious uses of the Haynes 282 alloy, the search for new applications through plastic deformation leading to increased strength and deliberate anisotropy can be a very interesting area of research.
In this study, samples made using the DMLS method and by casting were tested. Threedimensional scanning of the products was performed to compare how their geometries deviated from the nominal values. The quality of the products was assessed, and the accuracy of the two methods compared. In the next stage of the research, the samples were subjected to plastic deformation by rolling with different degrees of deformation: 20%, 40% and 60%. The microstructures of the DMLS and as-cast samples were then analysed, and static tensile tests and microhardness measurements were performed. The effect of deformation on the structure and mechanical properties of the DMLS and as-cast samples was examined.

Materials and Methods
In the first step, two samples were produced from the Haynes 282 nickel superalloy. The chemical composition of the alloy is shown in Table 1. The thickness of the samples was 5 mm. The DMLS sample was produced on an EOS M 100 device using an EOS M100 3D printer (EOS GmbH Electro Optical Systems, Krailling, Germany) operating in DMLS technology. A test specimen in the shape of a cuboid with dimensions of 50 mm × 70 mm × 5 mm was produced using the following parameters: layer thickness = 20 µm, laser power = 90 W, scanning speed = 800 mm/s, volumetric energy density = 94 J/mm 3 . A scanning strategy was used in which the scanning lines relative to the previous layer were rotated through an angle of 67 • to reduce anisotropy. The DirectBase S15 platform was heated to a temperature of 80 • C during the process. The oxygen content in the working chamber was <0.1% due to the use of argon as a protective gas. The as-cast sample was tested in the annealed condition. The test samples are shown in Figure 1.
plastic deformation of materials produced using AM methods.
Although both cast and DMLS Haynes 282 nickel superalloy are studied exten most published studies focus on conventional heat treatment and post-processing scribed above. Moreover, no study has yet compared the microstructure and mec properties of the as-cast and deformed Haynes 282 superalloy. Apart from the o uses of the Haynes 282 alloy, the search for new applications through plastic defor leading to increased strength and deliberate anisotropy can be a very interesting research.
In this study, samples made using the DMLS method and by casting were Three-dimensional scanning of the products was performed to compare how their etries deviated from the nominal values. The quality of the products was assesse the accuracy of the two methods compared. In the next stage of the research, the s were subjected to plastic deformation by rolling with different degrees of deform 20%, 40% and 60%. The microstructures of the DMLS and as-cast samples were th lysed, and static tensile tests and microhardness measurements were performed. fect of deformation on the structure and mechanical properties of the DMLS and samples was examined.

Materials and Methods
In the first step, two samples were produced from the Haynes 282 nickel sup The chemical composition of the alloy is shown in Table 1. The thickness of the samples was 5 mm. The DMLS sample was produced on M 100 device using an EOS M100 3D printer (EOS GmbH Electro Optical Systems, ing, Germany) operating in DMLS technology. A test specimen in the shape of a with dimensions of 50 mm × 70 mm × 5 mm was produced using the following para layer thickness = 20 μm, laser power = 90 W, scanning speed = 800 mm/s, volume ergy density = 94 J/mm 3 . A scanning strategy was used in which the scanning lines r to the previous layer were rotated through an angle of 67° to reduce anisotropy. T rectBase S15 platform was heated to a temperature of 80 °C during the process. T gen content in the working chamber was <0.1% due to the use of argon as a protect The as-cast sample was tested in the annealed condition. The test samples are sh Figure 1.  In the first stage of the research, 3D scanning of the produced samples was performed. A Gom Atos Core 200 3D scanner (Carl Zeiss GOM Metrology GmbH, Braunschweig, Germany) was used for this purpose. The accuracy of the device was 0.018 mm. No scanning spray was used for the scanning. Gom Inspect 2018 software was used to analyse the scans; they were compared with the CAD files in order to determine the geometric deviation. Surface topography was determined using the appropriate functions of the Gom Inspect software.
In the next stage, samples for plastic deformation were prepared. The samples were cut from the base material with an EDM machine, taking into account the building direction of the DMLS sample. The plastic deformation was carried out using a quarto reversible rolling mill with a roll diameter of 60 mm and a speed of 20 rpm. The degrees of deformation were 20%, 40% and 60%. The plan of the deformation experiment is presented in Table 2. Table 2. Deformation plan of the DMLS and as-cast sample.

No
Manufacturing Method Orientation Cast -0 10 Cast -20 11 Cast -40 12 Cast -60 Afterwards, samples for the microstructure studies were prepared from the base materials and after plastic deformation. The research was carried out on a Hitachi SU-70 (Tokyo, Japan) scanning electron microscope. The observations were made at a voltage of 5 kV with a magnification of 1000×. A point analysis of the chemical composition was carried out.
Microhardness tests HV 0.05 were performed on the samples prepared for the microstructure studies. The research was carried out using a Shimadzu HMV-2T E (Kyoto, Japan) hardness tester. Six measurements were made from each sample.
The as-cast and additively manufactured samples following plastic deformation were also characterized in terms of their mechanical properties as measured in uniaxial tensile tests. The miniaturized tensile specimens with a gauge length of 5 mm and a cross-section of 0.8 mm × 0.6 mm (as presented in Figure 2 and described in more detail in [22,23]) were cut using an electrical discharge machine (EDM) in the rolling direction. The tensile experiments were performed at an initial strain rate of 10 −3 s −1 using a Zwick/Roell Z005 (Ulm, Germany) static testing machine with a loading capacity of 5 kN and a digital image correlation (DIC) system for the strain measurements. Digital images of the sample surface ( Figure 2b) were registered at a frequency of 4 Hz using an AVT Pike F-505B ASG (Allied Vision Technologies GmbH, Stadtroda, Germany) CCD camera and processed using VIC-2D Correlated Solutions (Correlated Solutions Inc, Irmo, SC, USA) software. Each material was represented by at least three test specimens. The 0.2% offset yield strength (YS), ultimate tensile strength (UTS), and elongation to failure (A) were calculated from the stress-strain curves obtained, following the ISO 6892-1 standard procedure.  Figure 3 shows the results of a comparison of the 3D scans of the DMLS and as-cast samples with the nominal CAD models. The statistics generated in the Gom Inspect software are shown in Figure 4. Figure 5 shows the surface topography analysis of the DMLS and as-cast samples.

Results and Discussion
Based on these results (Figures 3 and 4), it can be concluded that the surface analysis results for the DMLS sample and for the as-cast sample do not differ significantly. The scatter of the results for both samples was similar (for the DMLS sample = 5.377 and for the as-cast sample = 5.655). The standard deviation of the results shows that the scatter of the results around the mean was larger for the DMLS sample (0.205) than for the as-cast sample (0.078). As indicated in the literature [6], the surface quality of 3D-printed metallic components depends on many factors, including the input parameters related to the raw material, the design, the process parameters, and finally post-processing. A final surface treatment such as grinding, post HIP or chemical polishing is recommended for 3Dprinted metallic elements [24].  Figure 3 shows the results of a comparison of the 3D scans of the DMLS and as-cast samples with the nominal CAD models. The statistics generated in the Gom Inspect software are shown in Figure 4. Figure 5 shows the surface topography analysis of the DMLS and as-cast samples.  Figure 3 shows the results of a comparison of the 3D scans of the DMLS and as-cast samples with the nominal CAD models. The statistics generated in the Gom Inspect software are shown in Figure 4. Figure 5 shows the surface topography analysis of the DMLS and as-cast samples.

Results and Discussion
Based on these results (Figures 3 and 4), it can be concluded that the surface analysis results for the DMLS sample and for the as-cast sample do not differ significantly. The scatter of the results for both samples was similar (for the DMLS sample = 5.377 and for the as-cast sample = 5.655). The standard deviation of the results shows that the scatter of the results around the mean was larger for the DMLS sample (0.205) than for the as-cast sample (0.078). As indicated in the literature [6], the surface quality of 3D-printed metallic components depends on many factors, including the input parameters related to the raw material, the design, the process parameters, and finally post-processing. A final surface treatment such as grinding, post HIP or chemical polishing is recommended for 3Dprinted metallic elements [24].     Many parameters of the 3D printing process determine the AM product topography. These parameters are related to the quality of the batch powder, the spatial design of the part, the process parameters and the final treatment [6]. Of the process parameters, the most important are the laser beam guidance strategy, the laser spot diameter, the laser power, the laser feed, and the layer thickness. The surface roughness of AM objects is a critical element in process design. High quality is required for components used in aviation, so it is extremely important to plan appropriate post-treatment for products.
Elements manufactured with the AM for high-quality applications require an average surface roughness below 1 mm [25]. For the samples analysed during our work (Figure 5), the Ra1C parameter for the DMLS sample was 4.73 μm, whereas the Ra2C was 1.87 μm. The Ra1DMLS parameter for the as-cast sample was 1.97 μm, whereas the Ra2DMLS was Based on these results (Figures 3 and 4), it can be concluded that the surface analysis results for the DMLS sample and for the as-cast sample do not differ significantly. The scatter of the results for both samples was similar (for the DMLS sample = 5.377 and for the as-cast sample = 5.655). The standard deviation of the results shows that the scatter of the results around the mean was larger for the DMLS sample (0.205) than for the as-cast sample (0.078). As indicated in the literature [6], the surface quality of 3D-printed metallic components depends on many factors, including the input parameters related to the raw material, the design, the process parameters, and finally post-processing. A final surface treatment such as grinding, post HIP or chemical polishing is recommended for 3D-printed metallic elements [24].
Many parameters of the 3D printing process determine the AM product topography. These parameters are related to the quality of the batch powder, the spatial design of the part, the process parameters and the final treatment [6]. Of the process parameters, the most important are the laser beam guidance strategy, the laser spot diameter, the laser power, the laser feed, and the layer thickness. The surface roughness of AM objects is a critical element in process design. High quality is required for components used in aviation, so it is extremely important to plan appropriate post-treatment for products.
Elements manufactured with the AM for high-quality applications require an average surface roughness below 1 mm [25]. For the samples analysed during our work (Figure 5), the Ra 1C parameter for the DMLS sample was 4.73 µm, whereas the Ra 2C was 1.87 µm. The Ra 1DMLS parameter for the as-cast sample was 1.97 µm, whereas the Ra 2DMLS was 2.46 µm. The results demonstrate the high surface quality of the manufactured parts. The Ra 1DMLS parameter was more than 2.5 times higher than the Ra 2DMLS parameter, which was the correct tendency given that the Ra 1DMLS was analysed perpendicular to the building direction. The authors of [26] measured roughness with standard devices for this purpose, i.e., with profilometers or using SEM. In this study, an unconventional method of measuring roughness is proposed.
2.46 μm. The results demonstrate the high surface quality of the manufactured pa Ra1DMLS parameter was more than 2.5 times higher than the Ra2DMLS parameter, wh the correct tendency given that the Ra1DMLS was analysed perpendicular to the b direction. The authors of [26] measured roughness with standard devices for this p i.e., with profilometers or using SEM. In this study, an unconventional method o uring roughness is proposed. Based on the microstructure analysis, no cracks or other material defects w served. Porosity (Figure 6h) with a maximum diameter of approx. 2 μm was ob The point analysis of the chemical composition (Figure 6i) showed that the alloy i Based on the microstructure analysis, no cracks or other material defects were observed. Porosity (Figure 6h) with a maximum diameter of approx. 2 µm was observed. The point analysis of the chemical composition (Figure 6i) showed that the alloy is rich in nickel and chromium. Figure 6g-h shows the dendritic structure in the initial state. The microstructure was columnar-grained. It can be observed that the dendrites have a similar growth direction to the sample building direction and grow along it (Figure 6h), which is a characteristic phenomenon for nickel superalloys produced by AM techniques [9,12,27]. Compositional contrast from the back-scattered electron (BSE) detector shows bright spots marked with white circles (Figure 6c,e,g). This represents a segregation of heavy element compounds in the interdendritic areas. As deformation increases, the grains elongate due to a deformation of the dendritic structure. In the strongly deformed structure (Figure 6c) the matrix grain boundaries are not visible; however, the bright zones in the structure are still clearly noticeable. This proves that the grains deformed more easily than the precipitation. Similar observations have been made by other authors [12], using a laser power of 400 W and an EOS 3D printer. The EDX analysis maps presented in that work highlighted the segregation of C, Ti, and Mo to the interdendritic regions. The dendrite cores were conversely rich in Cr, Ni, and Co, and Al was found to be evenly distributed throughout the solidification microstructure.
nickel and chromium. Figure 6g-h shows the dendritic structure in the initial state. The microstructure was columnar-grained. It can be observed that the dendrites have a similar growth direction to the sample building direction and grow along it (Figure 6h), which is a characteristic phenomenon for nickel superalloys produced by AM techniques [9,12,27]. Compositional contrast from the back-scattered electron (BSE) detector shows bright spots marked with white circles (Figure 6c,e,g). This represents a segregation of heavy element compounds in the interdendritic areas. As deformation increases, the grains elongate due to a deformation of the dendritic structure. In the strongly deformed structure (Figure 6c) the matrix grain boundaries are not visible; however, the bright zones in the structure are still clearly noticeable. This proves that the grains deformed more easily than the precipitation. Similar observations have been made by other authors [12], using a laser power of 400 W and an EOS 3D printer. The EDX analysis maps presented in that work highlighted the segregation of C, Ti, and Mo to the interdendritic regions. The dendrite cores were conversely rich in Cr, Ni, and Co, and Al was found to be evenly distributed throughout the solidification microstructure. The microstructure of the base as-cast sample is characterized by large grains (Figure  7a) with clearly visible grain boundaries. After deformation, the grains are elongated, and their boundaries are no longer clear (Figure 7b-d). In the microstructure there is also randomly distributed MC, such as (Ti, Mo)C precipitates, preferentially in the interdendritic The microstructure of the base as-cast sample is characterized by large grains (Figure 7a) with clearly visible grain boundaries. After deformation, the grains are elongated, and their boundaries are no longer clear (Figure 7b-d). In the microstructure there is also randomly distributed MC, such as (Ti, Mo)C precipitates, preferentially in the interdendritic regions and grain boundaries (Figure 7b-d). A point analysis of the precipitations and mapping (Figure 7e-f) showed that the carbides are rich in titanium, molybdenum (Ti, Mo-MC), and chromium (M 23 C 6 ). Based on the microstructure photography, it can be concluded that the precipitations are brittle following intensive plastic deformation. Vacuum casting and deoxidation processes are advisable to improve the mechanical properties of Haynes 282 alloy. These treatments can help inhibit the formation of large Ti-rich MX. It is important to reduce the amount of Ti-rich CX type carbides because Ti is the component of the main strengthening phase γ of the alloy [28].
regions and grain boundaries (Figure 7b-d). A point analysis of the precipitations and mapping (Figure 7e-f) showed that the carbides are rich in titanium, molybdenum (Ti, Mo-MC), and chromium (M23C6). Based on the microstructure photography, it can be concluded that the precipitations are brittle following intensive plastic deformation. Vacuum casting and deoxidation processes are advisable to improve the mechanical properties of Haynes 282 alloy. These treatments can help inhibit the formation of large Ti-rich MX. It is important to reduce the amount of Ti-rich CX type carbides because Ti is the component of the main strengthening phase γ′ of the alloy [28]. Representative stress-strain curves for the DMLS and as-cast Haynes 282 after further plastic deformation are presented in Figure 8. The mechanical properties calculated are summarized in Table 3. The as-cast Haynes 282 exhibited a YS of 421 ± 8 MPa, UTS = Representative stress-strain curves for the DMLS and as-cast Haynes 282 after further plastic deformation are presented in Figure 8. The mechanical properties calculated are summarized in Table 3. The as-cast Haynes 282 exhibited a YS of 421 ± 8 MPa, UTS = 742 ± 28 MPa, and A = 35.0 ± 1.7%, whereas the additively manufactured samples showed a much higher YS (594 ± 8 and 656 ± 18 MPa for X-Z and Y-Z orientation, respectively) and UTS (835 ± 7 and 879 ± 26 MPa), with a similar elongation to failure (A = 36.9 ± 3.4 and 31.7 ± 3.1%). The results obtained for the DMLS Haynes 282 are comparable with other literature data (YS = 633 MPa, A = 31.5% [9]). Further plastic deformation of both the as-cast and the DMLS Haynes 282 resulted in a gradual improvement in strength, with an increasing deformation degree at the expense of gradually reduced ductility (Table 3). It should also be noted that the Y-Z DMLS specimens exhibited higher strength and slightly lower ductility than their counterparts with an X-Y orientation. These differences resulted from the microstructure anisotropy of the DMLS Haynes 282; that is, the columnar grains growing in the building direction (Z-axis) provided enhanced strength in the Y-Z orientation, whereas the more homogeneous microstructure in the X-Y plane ensured slightly higher ductility of the X-Y specimens. Finally, a comparison of the mechanical properties of the as-cast and DMLS samples showed that the as-cast Haynes 282 was more prone to strengthening by plastic deformation than the DMLS samples. Both the YS and UTS improved much more in the case of as-cast Haynes 282 (by approximately ∆YS = 840 MPa and ∆UTS = 670 MPa for the highest deformation degree). The plastic deformation of the DMLS samples was not so effective (∆YS = 629 MPa and ∆UTS = 510 MPa for X-Y and ∆YS = 609 MPa and ∆UTS = 539 MPa for Y-Z). This resulted from the significantly improved mechanical properties of the DMLS samples directly after the DMLS processing and because of the much higher strain hardening capability shown by the as-cast Haynes 282, i.e., the difference between the YS and UTS (UTS-YS = 321 MPa) was higher than in the DMLS samples (241 and 223 MPa for X-Y and Y-Z orientation, respectively). Nevertheless, the combination of the highest strength and ductility was still obtained for the DMLS sample with a Y-Z building orientation. Based on the results of the microhardness tests (Figure 9), it was found that, in the case of DMLS samples in the initial state, the direction of building did not affect the hardness (HV 0.05 = 355 for X-Y and HV 0.05 = 354 for Y-Z). The 20% deformed DMLS sample built in the X-Y direction had an average HV value higher (HV 0.05 = 435) than the DMLS sample built in the Y-Z direction (HV 0.05 = 390). Then, as the deformation degree increased (to 40% and 60%), the hardness of the sample built in the Y-Z direction proved to be higher than that of the sample built in the X-Y direction (by 21 units at 40% strain and 55 units at 60% strain). These results are consistent with the tensile test results. The samples built in the Y-Z direction were characterized by higher strength, whereas the samples built in the X-Y direction were characterized by higher ductility. The hardness of the cast material was lower for the base sample and for the deformed samples than for the DMLS samples. The results of the microhardness tests are consistent with the results of the mechanical properties tests. The Haynes 282 cast material was more susceptible to deformation strengthening than the DMLS material. This is proved by the intense increase in the microhardness of the cast sample, along which the sample deformation was increased. mechanical properties of the DMLS samples directly after the DMLS processing and cause of the much higher strain hardening capability shown by the as-cast Haynes i.e., the difference between the YS and UTS (UTS-YS = 321 MPa) was higher than in DMLS samples (241 and 223 MPa for X-Y and Y-Z orientation, respectively). Neverthe the combination of the highest strength and ductility was still obtained for the DMLS s ple with a Y-Z building orientation.

Conclusions
-Based on the results of the 3D scanning, it can be concluded that the surface quality of DMLS and as-cast samples in the base state are at a similar level. - The quality of the surfaces of 3D printed materials depends on factors such as the parameters of the raw material (the metallic powder), the design process, the 3D printing parameters, and the post-processing. -During the microscopic examinations of the DMLS samples, no cracks were found. There was slight porosity in the structure. In the structure of the cast samples, precipitations of MX-type carbides were observed.