Characterization of the microstructural, magnetic, and thermal properties of Fe–45Ni fabricated by laser powder bed fusion
Nuri Sim, Hyo Yun Jung, Kee-Ahn Lee
TL;DR
This paper studies the properties of an Fe–45Ni alloy made using laser powder bed fusion, showing it has strong magnetic and thermal characteristics suitable for industrial applications.
Contribution
The study provides a detailed characterization of Fe–45Ni alloy fabricated via L-PBF, highlighting its magnetic and thermal performance.
Findings
Fe–45Ni alloy achieved 99.28% relative density with optimal laser parameters.
Magnetic permeability and coercivity varied between Z- and Y-axes.
Thermal expansion coefficient was 6.0834 × 10−6 and Curie temperature was 414 °C.
Abstract
Fe–Ni alloys are binary alloys composed of iron and nickel in various proportions and are widely used in industry owing to their excellent magnetic properties, corrosion resistance, and thermal stability. Among the various Fe–Ni alloys, the Fe–45Ni alloy is suitable for applications requiring both thermal stability and strong magnetic properties. Therefore, this study investigates the effects on metallurgical, magnetic, and thermal properties of an Fe–45Ni alloy manufactured using laser power bed fusion (L-PBF). The microstructure of the sample manufactured under the process conditions with the highest relative density (99.28%) (laser power of 85 W and scan speed of 300 mm/s) was analyzed using electron backscatter diffraction, focusing on both the ZY and XY planes. For the magnetic properties, the permeability and coercivity along the Z- and Y-axes differed, with values of 60.77 × 10−…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 10
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8
Figure 9- —https://doi.org/10.13039/501100003626Defense Acquisition Program Administration
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAdditive Manufacturing Materials and Processes · High Entropy Alloys Studies · Laser Material Processing Techniques
Introduction
Fe–Ni alloys are binary alloys composed of iron and nickel in various ratios. These alloys are widely used in industry because of their excellent magnetic properties, corrosion resistance, and thermal stability^1–3^. The magnetic, thermal, and mechanical properties of Fe–Ni alloys vary significantly depending on the Ni content. Low-Ni alloys with a Ni content of 50% or less are known to exhibit high strength, strong magnetism, and thermal expansion properties close to zero^4^.
Previously, Fe–Ni-based alloys were primarily manufactured through casting, rolling, and powder metallurgy methods^5^. These conventional manufacturing methods limit the shape and, owing to exposure to high-temperature processes, cause excessive grain growth, leading to a reduction in soft magnetic performance or decline in mechanical properties^6^. Furthermore, because secondary processes such as fine machining and grinding are generally required after manufacturing, these methods become time-consuming and expensive and produce material waste^7^. Recently, the aerospace industry required complex components to be manufactured in shapes not previously used, necessitating manufacturing processes with secondary processing and welding^8^. However, Fe–Ni alloys have a narrow solidification temperature range and low ductile-to-brittle transition temperature, resulting in hot cracking issues during welding^9^. Accordingly, the application of new manufacturing methods for the diverse utilization of Fe–Ni alloys is necessary^10^.
Additive manufacturing (AM) is a process whereby parts are manufactured with geometrically complex shapes based on 3D modeling, which has the advantage of significantly reducing production costs, time, and material consumption, even when the production shape is complex or requires the use of expensive raw materials^11,12^. Initially, plastic-based AM was primarily used for industrial applications^13^; however, over the past decade, the utilization and validation of metal AM have gradually progressed, and its reliability for industrial use has increased^14^. Consequently, metal AM has been actively used in recent years to manufacture advanced components in various industrial sectors such as defense, aerospace, and medical fields^15^.
Among AM processes, laser powder bed fusion (L-PBF) is the most actively utilized metal AM process. L-PBF involves spreading a layer of powder and then irradiating it with a concentrated laser beam, fusing one layer at a time to build up the metal layer-by-layer to create the desired shape^16^. By fabricating layers with a thickness of 100 μm or less, high-performance metal parts with fine and complex shapes can be produced^17^. In particular, the L-PBF process has been widely used for producing precision parts for aerospace, defense, and medical applications, owing to its high precision and topological advantages for weight reduction^15,18^.
Recent growth in the private sector-led aerospace industry^19–22^ has increased the industrial demand for Fe–Ni-based alloys with high thermal stability^5,23^. In particular, lightweight design and high-strength characteristics are being highlighted as key factors in the aerospace industry^24^. However, owing to geometric complexities, production using existing manufacturing processes has been difficult^25^. Recently, research has begun to apply the L-PBF process to manufacture components with high dimensional safety and complex shapes, such as satellite frames, orbital satellites, and space telescope brackets, by utilizing the advantages of 3D design and AM^11,26,27^. When appropriate parameters are applied, the L-PBF process can produce high quality products and allow microstructure control, enabling the effective realization of material properties. Consequently, research on microstructure control through process optimization has been actively conducted^28^.
Alloys manufactured using L-PBF have a microstructure that differs from those produced by conventional methods owing to the rapid cooling rate that is forcibly applied^29^. According to Hooper^30^, the cooling rate applied by L-PBF is in the range of 10^3^–10^5^ K/s, resulting in the appearance of predominantly fine dendritic or cellular microstructures. Cloots et al.^31^ attributed the solidification rate to the formation of these microstructures, and Krakhmalev et al.^32^ confirmed that the dendritic microstructure formed according to the dominant heat flow. In addition, the frequently observed microstructures with crystallographic anisotropy were attributed to local preheating and rapid solidification. When the thermal gradient direction was parallel to the build direction, dendritic grains formed, whereas when it was perpendicular, equiaxed grains formed. This microstructure formation was attributed to grain growth along the thermal gradient direction^33^.
Moreover, microstructure formation characteristics resulting from the application of L-PBF directly affected the thermal and magnetic properties of the material. An important research area when applying L-PBF to high-performance materials is the correlation between process capabilities and microstructural characteristics without defects, as well as their mechanical and functional properties. Fe–Ni alloys are not widely used in various fields owing to their high cost and processing limitations but have been employed as an essential industrial material to leverage their excellent thermal and magnetic properties. Therefore, research on the correlation between the AM conditions and magnetic properties of Fe–Ni alloys manufactured using the L-PBF process has recently begun. Harrison et al. ^34^ investigated whether the unique low thermal expansion properties of Invar were maintained after manufacturing the Fe–36Ni alloy using the L-PBF process. They reported that the Fe–36Ni alloy produced by the L-PBF process had a lower thermal expansion coefficient than the conventionally manufactured Fe–36Ni alloy. Zhang et al.^35^ manufactured Fe–80Ni alloys using the L-PBF process at various scan speeds and compared their magnetic properties under different manufacturing conditions. They reported that the magnetic properties of the Fe–80Ni alloy could be influenced by the parameters.
To date, research on the thermal and magnetic properties of Fe–Ni alloys exhibiting Invar characteristics is still in its early stages and limited to the representative Invar alloy, Fe–36Ni. Although several studies have investigated Fe–Ni alloys fabricated by the L-PBF process, these efforts have mainly focused on either low-Ni Invar compositions or Ni-rich alloys, leaving a clear knowledge gap regarding Fe–45Ni alloys that simultaneously exhibit strong ferromagnetism and low thermal expansion. Fe–Ni alloys are known to exhibit low thermal expansion properties close to zero, that is, Invar properties, when the Ni content is less than 50%^36^. Among Fe–Ni alloys, the Fe–45Ni alloy has a slightly larger thermal expansion coefficient than the 36Ni alloy but still maintains a low thermal expansion coefficient close to zero while exhibiting extremely strong magnetism. The Fe–45Ni alloy is suitable for applications requiring both thermal stability and strong magnetic properties^37^. However, despite its technological relevance, systematic studies on the microstructure evolution and the combined magnetic and thermal properties of Fe–45Ni alloys manufactured via the L-PBF process remain scarce.
Therefore, this study investigated the effects of additive manufacturing parameters on the microstructure of an Fe–45Ni alloy fabricated using the L-PBF process under various conditions, including laser power and scan speed. In particular, this work aims to clarify the build-direction-dependent magnetic behavior and thermal stability of L-PBF-fabricated Fe–45Ni alloys, thereby assessing the feasibility of this alloy system for additively manufactured functional components.
Method
Fe–45Ni powders produced by the gas atomization method were used. The chemical composition is presented in Table 1. The chemical composition and impurity concentration of the powders were analyzed using energy dispersive spectroscopy (JEOL, JSM-7100 F, Japan). The morphology of the powder was examined using scanning electron microscopy (SEM, JEOL, JSM-5500, Japan), as shown in Fig. 1a, b. The particle size distribution of the Fe–45Ni powder was analyzed using a laser particle size analyzer, as shown in Fig. 1c. The D10, D50, and D90 of the Fe–45Ni powder were 19.16, 30.89, and 50.68 μm, respectively. The Hall flow rate value of the powder measured using a Hall flowmeter (ACuPowder International LLC) was 14.96 s/50 g. The Hausner ratio, which is often used to evaluate the flowability of powders for AM, was calculated to be 1.16, confirming excellent flowability suitable for L-PBF^38^.
An L-PBF machine (GE additive, Concept Laser Mlab, USA) was used to manufacture the Fe–45Ni specimens under the process conditions presented in Table 2. Among the L-PBF process conditions, the layer thickness and hatch spacing were fixed at 0.025 and 0.105 mm, respectively, in all experiments. The cubic Fe–45Ni specimens manufactured in a chess pattern during the L-PBF process are schematically depicted in Fig. 2. In this study, the specimens fabricated using the L-PBF process were layered into cubes measuring 7 × 7 × 7 mm, with the build direction defined as the Z-axis and the directions perpendicular to the build direction defined as the X- and Y-axes.
The central part of the fabricated cubic specimen was cut and polished in the build direction. After polishing the surface of the cut specimen, the relative density was measured using Image J software (Table 3) and the polished surface was etched for 1 min with a mixed solution of 10 ml HNO_3_ and 90 ml C_2_H_5_OH. The microstructure of the cross-section was observed using optical microscopy (OM, Nikon, ECLIPSE MA200, Japan) and field emission scanning electron microscopy (FE-SEM, JEOL, JSM-7100 F, Japan). The crystallographic texture was analyzed using FE-SEM (Hitachi, SU5000, Japan) equipped with electron backscatter diffraction (EBSD, EDAX Ametek, Velocity Ultra, USA) camera. A step size of 0.38 μm was used for each scan. The scanned data were analyzed using TSL-orientation imaging microscopy (OIM) software.
For the crystal structure analysis, an X-ray diffractometer (XRD, D8 DISCOVER, BRUKER, Germany) with a scan range of 30–100°, step size of 0.02°, scan speed of 2°/min, and Cu-Kα radiation with a 1.5402 Å wavelength was utilized. The magnetic properties were analyzed by cutting the fabricated specimen into 3 × 3 × 1 mm pieces and applying 1.2 Tesla to the specimen using a vibrating sample magnetometer (Lake Shore Cryotronics 7407, USA). Thermal measurements (NETZSCH, Germany) were conducted to observe thermal expansion behavior in the temperature range of 25–500 ℃ with a heating rate of 5 ℃/min and a load of 5 g under N_2_ atmosphere.
Table 1. Chemical composition of gas-atomized Fe–45Ni powders.ElementFeNiO(wt%)Balance45.140.14
Table 2. Processing parameters used to fabricate Fe–45Ni samples via the L-PBF process.Processing parametersApplied valuesLaser power, W75, 85, 95Scan speed, mm/s100, 300, 500Hatching space, mm0.105Layer thickness, mm0.025Volumetric energy, J/mm^3^57.1–361.9Laser rotation angle90°
Table 3. Energy density calculation and relative density for L-PBF process parameters.Laser power (W)Scan speed (mm/s)Energy density (J/mm^3^)Relative density (%)7550057.195.6730095.298.85100285.794.378550064.896.38300107.999.28100323.897.399550072.498.66300120.699.08100361.997.30
Fig. 1. Gas atomized Fe–45Ni powders: SEM images of the (a) particle morphology and (b) cross-section. (c) Particle size distribution.
Fig. 2. Schematic of the (a) cube-shaped sample showing the build and scanning direction, and (b) chessboard scanning pattern.
Results and discussion
Figure 3 shows the cross-sectional OM images of the Fe–45Ni alloy deposited by the L-PBF process under different processing parameters. First, no noticeable cracks are observed in any of the cases. Moreover, regardless of the laser power, when the scan speed is 300 mm/s, the porosity is lower than that under other process conditions. When the laser power is 75 W and the scan speed is 100 mm/s, irregular small pores and irregularly sized lack of fusion can be observed, and the porosity decreases as the laser power increases. At a scan speed of 500 mm/s, gas porosity and lack of fusion can be observed, regardless of the laser power, resulting in a high porosity value. Overall, even in this case of a scan speed of 300 mm/s with the lowest porosity value, extremely fine pores are present in an irregular distribution. However, the differences in porosities owing to changes in laser power are not clear in this data. Therefore, further examination through high-magnification analysis was conducted, as shown in Fig. 4.
Fig. 3OM images of the polished surface corresponding to different laser parameters.
Figure 4 shows the SEM images of the porous regions of the specimens fabricated at a scan speed of 300 mm/s and laser power of 75–95 W. The pore defects observed in Fig. 4a are lack-of-fusion defects that are considered to have occurred because the newly formed melt pool did not completely bond with the underlying layer owing to low laser power^39^. Therefore, increasing laser power was expected to enhance remelting between subsequent layers, eliminating lack-of-fusion defects. By comparing Fig. 4a,b, irregular pores and lack-of-fusion defects are observed in Fig. 4a, whereas only fine spherical pores are present in Fig. 4b. In the PBF process, gas pores form when inert gas, which is insoluble in liquid metals, becomes trapped in the molten pool^40,41^. Hojjatzadeh et al.^42^ reported that at high energy densities, pores can be formed because of the evaporation of volatile impurities or the expansion of small gases trapped within the material. In addition, gas-atomized powders often contain entrapped gas bubbles—such as He and Ar—that reside between powder particles or within the powder itself during the gas atomization process^43^. Wu et al.^44^ reported that when the spherical pore size is 5–10 μm, porosity within the powder cannot be eliminated by a laser and can form pores. Therefore, the spherical pore defects observed in Fig. 4 appear to have been caused by inert gases entrapped within the powder produced via the gas atomization process (Fig. 1b). To reduce porosity, a practical approach involves replacing He and Ar with N_2_, a non-inert shielding gas^45^. A hot crack can be observed in Fig. 4c. Hot cracks occur during material solidification at high temperatures, often propagating along grain boundaries and frequently resulting in intergranular cracks^46^. Rapid cooling rate during the L-PBF process promotes the formation of elongated grains, which provide pathways for crack propagation.
Fig. 4SEM images showing defects in the Fe–45Ni alloy fabricated via the L-PBF process at a constant laser power of 75–95 W and scan speed of 300 mm/s: (a) lack of fusion and gas pore, (b) gas pore, (c) gas pore and hot crack.
The correlation between relative density and laser power and scan speed is shown in Fig. 5 through individual bar charts and a contour map. The bar chart (Fig. 5a) confirms that as the laser power increases, the relative density tends to increase, and the maximum relative density appears at a scan speed of 300 mm/s under each condition. In particular, the maximum relative density is observed at a laser power of 85 W and scan speed of 300 mm/s. The contour map, constructed according to the cubic interpolation method, was used to determine the optimal process parameter window. The contour plot reveals that the optimal process variable range is a laser power of 80–90 W and scan speed of 250–300 mm/s.
Fig. 5(a) Bar chart showing the relative density at different laser powers and scan speeds. (b) Contour map highlighting the optimal process area with high relative densities.
\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathrm{2}}{{\mathrm{d}}_{{\text{hkl }}}}{\mathrm{sin}}\uptheta ={\text{ n}}\uplambda ,$$\end{document}Figure 6a shows the XRD analysis results for Fe–45Ni samples with high relative density manufactured using the L-PBF process. Under all conditions, a diffraction peak for the face-centered cubic (FCC) FeNi phase can be observed. However, as the laser power increases from 75 to 95 W, the position of the (111) peak shifts to a lower angle (Fig. 6b). According to Bragg’s Law, the peak shift is closely related to the lattice constant.
where d_hkl_ denotes interplanar spacing, λ is the wavelength of the incident wave, θ is the diffraction angle, and n is a constant. During L-PBF processing of the Fe–Ni alloy, laser power increased and the position of the (111) peak shifted to a lower angle. This phenomenon suggests that as laser power increased, the alloy was exposed to higher temperatures and then rapidly solidified, resulting in an increase in the lattice constant. Martucci et al.^47^ reported that the lattice constant increases under high-temperature conditions (800 ℃), which is consistent with the results of this study. Figure 6c shows a phase diagram of the Fe–Ni alloy; the Fe–45Ni alloy has a dual-phase microstructure of FCC+ body-centered cubic (BCC) at room temperature^48^. However, the dual-phase microstructure is not present owing to the high temperature of the L-PBF process and rapid cooling that occurs simultaneously with powder melting. As the powder is melted by the laser and forms a liquid phase, it begins to solidify, creating the primary phase of the FCC FeNi. Owing to the rapid cooling in the L-PBF process, the FCC FeNi cannot undergo phase transformation to the equilibrium phase of BCC + FCC and thus solidifies, producing a final microstructure similar to that of FCC FeNi with contributions from FCC FeNi and the ordered FeNi phase. Miller et al.^49^ demonstrated that the FCC γ-phase of the Fe–27Ni alloy can be stabilized by rapid cooling processes, highlighting its potential in magnetic applications. This indicates that the FCC FeNi structure can influence magnetic properties.
Fig. 6(a) XRD patterns of the Fe–45Ni alloys fabricated with different laser powers, (b) magnified view of the (111) peak, (c) Fe–Ni phase diagram.
Figure 7 shows the SEM image of the microstructure under the following processing conditions: laser power of 75–95 W and a scan speed of 300 mm/s. Regardless of the laser power, particles composed of a cellular structure can be observed in the melt pool. The analysis confirmed that the particles generally grow parallel to the build direction. This cellular structure grows rapidly in the direction consistent with the thermal gradient when the molten powder solidifies, resulting in grains with a crystallographic texture generally parallel to the build direction. The growth direction of the cellular structure is indicated by the yellow arrows. The melt pool boundary shows a microstructure with columnar and equiaxed grains, depending on the temperature gradient and solidification rate (Fig. 7b). Equiaxed grains are present in cases with low temperature gradients and low solidification rates (Fig. 7a). In contrast, in cases with a high-temperature gradient and high solidification rate, larger columnar grains are formed (Fig. 7c). According to Kim et al. ^50^, as the laser power increases, a higher energy density is induced in the melt pool, causing the metal to remain in a molten state at higher temperatures for a longer period; this increases the grain growth time, resulting in larger grains. Therefore, in this study, experiments were conducted wherein the laser power increased from 75 to 95 W while maintaining the scan speed at 300 mm/s. As laser power increased, the columnar grains increased because of the high-temperature gradient and rapid solidification rate. Conversely, at a lower laser power, the temperature gradient decreased and solidification rate slowed, promoting the formation of equiaxed grains. In other words, as the laser power changed, the temperature gradient and solidification rate varied, which in turn altered the direction of the solidification process. As a result, the liquid-solid interface could move toward the center of the melt pool. When the temperature gradient decreased, the formation of equiaxed grains was promoted, and the columnar grain growth induced by equiaxed grains was halted^51^. That is, the microstructure formed in the melt pool gradually changed depending on the temperature gradient and solidification rate.
Fig. 7. Solidification microstructure of the processed Fe–45Ni alloy via the L-PBF process at a scan speed of 300 mm/s with different laser powers: (a) 75, (b) 85, and (c) 95 W.
The inverse pole figure (IPF) maps show the microstructure fabricated at a laser power of 85 W and scan speed of 300 mm/s both parallel (Fig. 8a) and perpendicular (Fig. 8b) to the build direction. The ZY plane primarily contains elongated grains aligned parallel to the building direction, with fine equiaxed and columnar grains coexisting with an average grain size of 43.18 μm. According to the grain size distribution in Fig. 8e, grains are mainly located below 60 μm, with distinct peaks also appearing in the 65–70 μm and 90–100 μm ranges. The formation of the microstructure is influenced by the thermal gradient and rapid cooling rate in the L-PBF process, following the solidification direction. The melt pools form randomly in different directions with distributed orientations and no preferred micro-texture (Fig. 8a). A larger grain size and a suitable texture are known to reduce the resistance to magnetic domain walls movement, thereby increasing magnetic permeability^52^. In contrast, the XY plane exhibits a chessboard-like grain pattern, arising from the scanning strategy utilizing a 90° rotation (Fig. 2). The change in the thermal gradient direction caused by this rotation allows a few grains with favorable orientations to grow toward the next layer during remelting, resulting in the grain pattern from the previous layer being partially inherited^53^. Grains in the < 101 > direction are mainly observed with a grain size of 31.94 μm and a square or columnar shape (Fig. 8b). The grain size distribution on the XY plane (Fig. 8f) exhibits right-skewed characteristics, indicating a higher of fine grains and a reduced fraction of large grains.
The kernel average misorientation (KAM) maps of the ZY and XY planes are shown in Fig. 8c, d. KAM is defined as a measure of local grain misorientation, with KAM maps showing the average misorientation of each pixel relative to its neighbors. Generally, higher KAM values indicate greater plastic strain and increased dislocation density^54^. The KAM maps use a rainbow scale, with blue indicating the minimum misorientation and red representing the maximum misorientation (0–5°)^55^. The blue areas (0°) exhibited low levels of misorientation, indicating minimal residual stress. Regions with high lattice rotation, as indicated by yellowish-green areas on the KAM maps, result from localized thermal stress and plastic deformation, leading to the accumulation of dislocation density. In particular, fine grain regions tend to exhibit greater dislocation concentrations owing to the influence of grain boundaries. The orientation deviation near the grain boundary in both the ZY and XY planes was approximately 2°, which reflects the characteristics of a grain boundary with high dislocation density. These results are consistent with the behavior of residual stress accumulation during rapid heating and solidification, as well as the repeated re-heating and re-cooling processes in L-PBF^56^.
Fig. 8EBSD results of the ZY and XY planes of the sample with a laser power of 85 W and scan speed of 300 mm/s: (a,b) IFP map, (c,d) KAM map, and (e,f) grain size distribution histograms.
To investigate the magnetic properties of the Fe–45Ni alloy fabricated by the L-PBF process, the hysteresis loop was measured along the Z-axis parallel to the build direction and Y-axis perpendicular to the build direction, as shown in Fig. 9. The magnetic property values for each axis are summarized in Table 4. The Fe–45Ni alloy exhibited favorable soft magnetic properties owing to the narrow hysteresis loops observed along both the Z- and Y-axes (Fig. 9b). According to Meng et al.^52^, a large grain size, appropriate microstructure, and low residual stress reduce the resistance to domain wall motion, leading to higher permeability, which is consistent with the results of this study. The saturation magnetization values measured in the vertical and parallel directions were similar; however, the sample parallel to the build direction had a permeability and coercivity of 60.77 × 10^− 3^ emu/(g Oe) and 20.93 Oe, respectively, whereas the sample perpendicular to the build direction had a permeability and coercivity of 28.20 × 10^− 3^ emu/(g Oe) and 34.78 Oe, thus exhibiting anisotropic magnetic properties. This difference is considered to originate from process-induced magnetic anisotropy associated with microstructural directionality during the L-PBF process^57^. Samples parallel to the build direction had high permeability and became easily magnetized, whereas those in the vertical direction exhibited less sensitive magnetization responses, resulting in differences in magnetization values as the external magnetic field strengthened. In addition, manufacturing defects, such as pores and lack of fusion generated during the L-PBF process, hindered the movement of domain walls, suppressing rapid magnetization, thereby increasing coercivity^57^. According to Gao et al., Fe–50Ni can exhibit the lowest coercivity of about 10 Oe under optimized processing conditions. However, in the present work, Fe–45Ni exhibits a coercivity of 20.93 Oe, which is significantly higher than the reported Fe–50Ni value^58^. Similarly, Zou et al.^59^ reported that the higher coercivity observed in a Ni-Fe based soft magnet fabricated using the L-PBF process could be attributed to microstructural defects such as fine porosity, grain boundaries, and cracks. In soft magnetic materials, the high density of grain boundaries is known to provide pinning sites for magnetic domain walls, thus suppressing their movement^60^. Zhang et al. ^3^ reported that the increased coercivity in L-PBF processed soft magnets can be attributed to porosity and microstructure defects. Therefore, the defects shown in Fig. 4 could degrade the soft magnetic properties of the Fe–45Ni alloy produced by the L-PBF process. Consequently, to improve the soft magnetic properties of the Fe–Ni alloy, a process must be developed that can minimize defects such as lack of fusion and pores that hinder the movement of the magnetic domain walls.
On the other hand, for the aspect on saturation magnetization, Kim et al. investigated the magnetic properties of Fe–50Ni fabricated by DED under different scan speeds and reported a saturation magnetization of 151.7 emu/g^50^. In the present study, the Fe–45Ni alloy shows a higher saturation magnetization of 162.62 emu/g. This increase in saturation magnetization can be primarily attributed to the higher Fe content in Fe–45Ni compared to Fe–50Ni, since Fe possesses a significantly larger magnetic moment than Ni. Additional contributions from specimen quality and processing-induced microstructural differences may also influence the measured Ms^61^.
Table 4. Summary of magnetic properties of L–PBF-fabricated Fe–45Ni alloy measured along the Z- and Y-axes.Laser power [W]Scan speed [mm/s]Axis \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:Ms$$\end{document} (emu/g) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:Mr\:$$\end{document} (emu/g)Initial slope x 10^− 3^ (emu/(g*Oe)) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:Hc\:\:$$\end{document} (Oe)85 W300Z162.621.5660.7720.93Y161.411.1729.2334.91
Fig. 9. Hysteresis loop of the Fe–45Ni alloy fabricated at a laser power of 85 W and scan speed of 300 mm/s: (a) hysteresis loop in the Z- and Y-axes, and (b) overview.
Fe–Ni alloys have a low coefficient of thermal expansion^62^, and maintaining this low thermal expansion coefficient is important, regardless of the manufacturing method or post-treatment^63^. The L-PBF process forms a large thermal gradient because of localized heating and rapid cooling, inducing slight volume changes that create residual stress^64^. Figure 6b shows that residual stress induces lattice distortion. Wegener et al.^65^ reported that residual stress significantly affects the thermal expansion coefficient. Figure 10 shows the thermal expansion behavior of the Fe–45Ni alloy manufactured by the L-PBF process (laser power of 85 W and scan speed of 300 mm/s), measured along the Z- and Y-axes. Thermal expansion measurements were performed once for each condition, and the results are discussed in terms of comparative trends. Accordingly, in the temperature range of 25–400 ℃, the thermal expansion coefficients were 6.0834 × 10^− 6^/℃ and 6.0321 × 10^− 6^/℃ along the Z- and Y-axes (Table 5), respectively. The linear coefficients were determined to be 1.28 × 10^− 3^/℃ and 1.26 × 10^− 3^/℃ (Fig. 10b), respectively, within the same temperature range. The Curie temperature of the manufactured Fe–45Ni alloy was measured to be 414.8 ℃ (Z axis) and 414.5 ℃ (Y axis). Furthermore, the length of the Fe–45Ni alloy exhibited a linear increase with temperature. This indicates that the low thermal expansion coefficient of the manufacturing specimen is exceptionally affected by residual stress^34^. However, the expansion is slow before 414 ℃ (Curie temperature), and a rapid expansion occurs above 414 ℃. This is because, owing to the strong magnetic attraction between the atoms of the Fe–Ni alloy, the alloy has a low thermal expansion rate, even with an increase in temperature, but above the Curie temperature, which exhibits ferromagnetic–antiferromagnetic properties, normal thermal expansion occurs with an increase in temperature^62,66^.
The low thermal expansion of Fe–Ni invar alloys is considered to be related to the magnetic characteristics of the material^67^. Huang et al. investigated the correlation between magnetic properties and the coefficient of thermal expansion in Fe–Ni invar alloys and reported that thermal expansion increased with increasing saturation magnetization. Rao et al.^68^. validated the Masumoto rule and proposed a parameter, M, to measure the magnitude of thermal expansion coefficients, as described by Eq. (2).
\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:M=\:Ms/Tc,$$\end{document}where Ms represents the saturation magnetization, and Tc denotes the Curie temperature. The measured Ms and Tc values were analyzed using Eq, (11), while the corresponding M values for the Z- and Y-axes were 0.392 and 0.389, respectively. The M values showed negligible variation, indicating that the thermal expansion coefficients of the samples oriented in the Z- and Y-axis directions via the L-PBF process were isotropic. Thermal expansion coefficients of L–PBF-fabricated Fe–45Ni alloy measured along the Z- and Y-axes.
Table 5. Thermal expansion coefficients of L–PBF-fabricated Fe–45Ni alloy measured along the Z- and Y-axes.AxisTemperature (℃)25–100 ℃25–200 ℃25–300 ℃25–400 ℃25–500 ℃Z6.636.556.286.087.62Y6.546.406.156.037.56
Fig. 10. Thermal expansion behavior of the Fe–45Ni alloy fabricated with a laser power of 85 W and scan speed of 300 mm/s: (a) thermal expansion coefficient and (b) thermal linear expansion.
The results of this study suggest the possibility of manufacturing Fe–45Ni alloys with high magnetic properties and a low thermal expansion coefficient using various parameters. Optimizing the microstructure (e.g., grain size and crystallographic texture) and minimizing defects (e.g., pores, residual stresses, and cracks) are further required to achieve superior soft magnetic properties. Future work could focus on further optimizing the microstructure through scan patterns to induce appropriate textures, thereby promoting the development of a straightforward magnetization axis.
Conclusion
In this study, an Fe–45Ni alloy was fabricated using the L-PBF process, and its microstructural, magnetic, and thermal properties were investigated under various processing conditions. A high relative density was obtained under optimized parameters, although a small amount of porosity originating from the gas-atomized powder was still observed.
The Fe–45Ni alloy produced by the L-PBF process exhibited anisotropic magnetic behavior depending on the build orientation. In particular, specimens parallel to the build direction showed more favorable soft magnetic properties. In addition, the alloy exhibited a relatively high saturation magnetization and maintained a low and stable thermal expansion coefficient over a wide temperature range.
These results indicate that Fe–45Ni alloys fabricated by the L-PBF process may be suitable for applications requiring both thermal stability and soft magnetic performance. Further improvements in magnetic properties and build quality are expected to be achieved through optimization of L-PBF process parameters.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Huang, G., He, G., Liu, Y. & Huang, K. Anisotropy of microstructure, mechanical properties and thermal expansion in invar 36 alloy fabricated via laser powder bed fusion. Additive Manuf.104025. 10.1016/j.addma.2024.104025 (2024).
- 2Kumar, M. S. et al. Impact of print orientation on morphological and mechanical properties of L-PBF based Al Si 7Mg parts for aerospace applications. Silicon 1–15. 10.1007/s 12633-021-01474-w (2021).
- 3Cornet, B., Chavy-Macdonald, M. A., Kneib, J. & Foray, D. EU-SPRI Conference, Utrecht.
- 4Katz-Demyanetz, A., Popov Jr, V. V., Kovalevsky, A., Safranchik, D. & Koptioug, A. Powder-bed additive manufacturing for aerospace application: Techniques, metallic and metal/ceramic composite materials and trends. Manuf. Rev.610.1051/mfreview/2019003 (2019).
- 5B 213, A. Standard Test Methods for Flow Rate of Metal Powders Using the Hall Flowmeter Funnel.
- 6Elmer, J., Vaja, J., Pong, R., Gooch, T. & Barth, H. The effect of Ar and N 2 shielding gas on laser weld porosity in steel, stainless steel, and nickel. Weld. J.2015. 10.2139/ssrn.4094533 (2015).
- 7Lee, J. D. Hot cracking behavior of Fe–Ni alloys via direct observation. ISIJ Int.43, 54–62. 10.2355/isijinternational.43.54 (2003).
- 8Miller, K. J., Sofman, M., Mc Nerny, K. & Mc Henry, M. E. Metastable γ-Fe Ni nanostructures with tunable curie temperature. J. Appl. Phys.107. 10.1063/1.3334198 (2010).