您当前的位置: 首页 > 网页快照
Aluminum with dispersed nanoparticles by laser additive manufacturing
Results.AMNC powders.We overcame significant challenges to design and fabricate aluminum powders that contain surface-coated and embedded nanoparticles suitable for LAM experiments. Here we demonstrate that laser-deposited aluminum that contains a high density of dispersed titanium carbide (TiC) nanoparticles (up to 35 vol.%) can be achieved. TiC was selected due to its chemical stability above 780 °C in aluminum melt (with a surprising twist that TiC is not chemically stable under 780 °C17. Aluminum matrix nanocomposite (AMNC) powders with dense TiC nanoparticles were systematically fabricated (see Methods). By simply tuning x, the volume ratio between TiC nanoparticles and liquid aluminum, AMNC powders with different TiC loadings (e.g., x = 0.25; and x = 1) can be fabricated (Also see chemical composition in Supplementary Fig. 1). As shown in Figs.1a and b, the AMNC powders (x = 0.25 and x = 1) are spherical with an average size of 11.3 ± 7.2 µm and 5.9 ± 4.6 µm, respectively (Supplementary Fig. 2). Most of the TiC nanoparticles would first assemble at the surface of the Al droplets, as shown in Fig. 1c (also see Supplementary Fig. 2b), to achieve a favorable energy state. As x increases, the favorable energy state is not available for the additional TiC nanoparticles; hence the nanoparticles are forced to enter the Al droplets. We conducted the experiment at 820 °C, at which the TiC is chemical stable with Al17 and the wetting angle of TiC/Al is less than 70°18, using ultrasonic processing (See Fabrication of AMNC powders in Methods). The cross-section images in Fig. 1d and Fig. 1e indicate that the TiC nanoparticles (shown as gray spots) were also effectively pushed into and distributed inside the core of the Al powders. Spherical particles enable a better flowability and a higher packing density than irregularly shaped particles for LAM19. The surface-coated and uniformly dispersed TiC nanoparticles inside the AMNC powders can enhance laser beam absorption because of the high absorptivity of non-oxide ceramic nanoparticles20,21. The reflectivity measurements, as shown in Fig. 1f and Fig. 1g, illustrate that the reflectivity of AMNC powders with x = 0.25 and x = 1 are 14.58 ± 0.46% and 7.46 ± 0.47%, respectively, at the laser wavelength of 1070 nm for the LAM process in this study. The reflectivity of AMNC powders is thus significantly lower than that of pure aluminum powders (58.12 ± 0.81%), which enhances the laser absorption by almost one order of magnitude. A comparison between theoretical values and experimentally measured reflectivity (Supplementary Fig. 3) shows that the actual reflectivity (x = 0.25; x = 1) are much lower, i.e., 70% and 80%, than the predictions by the equation. It could be attributed to the multiple reflection and absorption that is inherent to the porous structure of the powder bed and the fact that the concentration of TiC nanoparticles on the nanocomposite powder surface is significantly higher than overall concentration.Fig. 1Morphology, microstructure and reflectivity of aluminum matrix nanocomposite (AMNC) powders. a, b SEM images of AMNC powders with x = 0.25. c Magnified image of b, showing TiC nanoparticles coated on the surface of an Al micro particle. d, e SEM images of the cross sections of specimens with x = 0.25 and x = 1, respectively. The AMNC powders (x = 1) have a higher loading of embedded TiC nanoparticles than the AMNC powders (x = 0.25). f Reflectivity of aluminum powders without and with nanoparticles. The dashed vertical line indicates the laser wavelength at 1070 nm. g Comparison of the reflectivity at the wavelength of 1070 nm for aluminum powder specimens with and without nanoparticles. Scale bar, 40 µm in a, 1 µm in b, c, e, and 4 µm in d. The error bars in the measured reflectivity are the standard deviations (s.d.) calculated over three measurementsFull size image .Structures of laser-deposited AMNCs.To obtain structures of laser-deposited AMNCs, extensive experiments were conducted on LAM of AMNC powders (see Methods). Then we characterized the micro/nanostructure and mechanical properties of the laser-deposited AMNC. Figure 2a shows the AMNC with 35 vol.% TiC with a thickness of 309 ± 16 µm was layer by layer deposited by laser melting of the Al powder (x = 1) bed, which was preheated to 300 °C. A SEM image was captured from the top of the deposited specimen (8 mm × 18.5 mm), showing the AMNC has a good uniformity and is well bonded to the previous layers. To reveal the interior microstructure, the polished AMNC specimen was tilted 52° to acquire cross-sectional images (See Microstructure characterization in Methods), as shown in Fig. 2b and Fig. 2c, indicating that a high volume fraction of TiC nanoparticles was dispersed and distributed homogeneously throughout the Al matrix. The uniform dispersion and distribution of TiC nanoparticles in Al matrix can be attributed to the unique nature of laser processing and the good wetting of TiC in molten Al. The laser-induced rapid cooling rate can reach up to 106–7 K s−1 22, and therefore the movement of atoms and particles freezes within milliseconds. The initial AMNC powders were fully melted and then solidified rapidly. During the non-equilibrium laser-induced rapid melting and solidification, both the TiC nanoparticles at the surface and inside of AMNC powders did not agglomerate to form clusters. Despite the formation of some larger particles with an average size of 159 nm (as shown by TEM in Supplementary Fig. 4) after solidification, TiC nanoparticles were still uniformly dispersed and distributed in AMNC specimens, as shown in the cross-sectional image of Fig. 2b, c. Moreover, TiC nanoparticles bond with Al matrix extremely well as confirmed in the FFT filtered high resolution TEM image (Fig. 2d). The TiC nanoparticles on the powder surface can absorb the laser beam more effectively to achieve a much higher temperature relative to the melting point of aluminum, enabling a rapid dispersion and diffusion of surface TiC nanoparticles into the core of the molten aluminum powders to expose liquid aluminum for bonding into dense layers. The rapid heating and cooling during the LAM process also limited the chemical reaction of TiC below 780 °C in aluminum melt.Fig. 2Surface and micro/nanostructure of laser-deposited aluminum matrix nanocomposites (AMNC). a Top view of laser-deposited AMNC (35 vol.% TiC) specimens with dimensions of 3 mm × 3 mm, 3 mm × 18.5 mm, and 8 mm × 18.5 mm. The insert SEM image shows an area of the laser-deposited specimen. b, c 52° tilted cross-sectional SEM images of laser-processed AMNC (35 vol.% TiC) were captured under different magnifications, showing that TiC nanoparticles are uniformly dispersed and distributed in aluminum. d FFT filtered high resolution TEM image shows good bonding between TiC nanoparticle and aluminum. Inserts are the fast Fourier transforms corresponding to the planes of (2 2 0) aluminum matrix and of (2 0 0) TiC nanoparticles. e, The grain maps of laser-deposited AMNC (35 vol.% TiC). Scale bar, 1 mm in a (right), 5 µm in b, 1 µm in c, 2 nm in d, and 500 nm in eFull size image It is argued that the high density of uniformly distributed TiC nanoparticles plays a critical role of refining the Al grains because the nanoparticles can act as nucleation sites and also restrict the growth of the Al grains during solidification. The EBSD mapping results revealed the grain size and crystallographic texture difference from the laser-deposited specimens of pure Al and AMNC (35 vol.% TiC), as shown in Supplementary Fig. 5a, 5b and Fig. 2e. Clearly, TiC particles, i.e., the black spots shown in Fig. 2e, were uniformly distributed in the Al matrix grain, indicating that TiC nanoparticles were well dispersed and distributed throughout the Al matrix. For the high (i.e., 35 vol.%) volume fraction of reinforcing TiC nanoparticles and the good dispersion of these nanoparticles, it is necessary to remove the TiC phase from Fig. 2e to better reveal the grain size of the refined Al matrix (see further details in Supplementary Fig. 5). Whereas the average grain size for the pure aluminum is approximately 2.7 ± 1.4 µm, the average grain size for the AMNC (35 vol.% TiC) was refined to 331 ± 95 nm (see further details in Supplementary Fig. 5). Recent studies also observed that the grain size of the laser additive Al specimen can be reduced after the incorporation of ceramic nanoparticles20,23. The TiC nanoparticles promote grain refinement via two mechanisms. First, they provide a high density of nucleation sites, leading to a finer grain size once the liquid solidifies. Second, the TiC nanoparticles impede migration of the newly formed grains, thereby stabilizing the grain size. The TiC nanoparticles can be used as grain growth inhibitors in Al as they can be treated as pinning points, inhibiting the grain growth during solidification, recrystallization, and recovery. The phenomenon has been reported in other Al-TiC nanocomposites24, showing that a high volume fraction of fine particles is very effective for grain growth retardation. In addition, the rapid cooling rate from laser processing can further contribute to the small grain size. Thus, the decrease of grain size can be attributed to a combined effect of augmented nucleation sites, restricted grain growth and high cooling rate during LAM.Mechanical properties of AMNCs.To evaluate the enhancement of mechanical properties due to such a high density of TiC nanoparticles, we first conducted microcompression tests at room temperature. Micropillars with a diameter of 4.0 ± 0.1 µm and a height of 10.0 ± 0.5 µm were carefully machined by FIB from the laser-deposited specimens with and without nanoparticle reinforcements. It should be noted that the locations of the micropillars were chosen randomly, and all testing data shown in this study were conducted more than 3 times. As shown in Fig. 3a, the pure aluminum specimen has a yield strength of only about 92 ± 16 MPa (Fig. 3a, the curve in black), while the AMNC specimens (with 17 vol.% TiC, processed at 25 °C, i.e., no preheating) offer a yield strength of up to 300 ± 52 MPa (Fig. 3a, the curve in blue). With a higher TiC loading (35 vol.%), the yield strength of the as-deposited AMNC reaches 868 ± 104 MPa with a plasticity greater than 10%, as shown in the curve in purple in Fig. 3a. Data for each curve was obtained by at least three sets of experiments. To improve the layer uniformity during laser melting, the powder layers were pre-heated at 300 °C. The result shows that the yield strength of AMNC (35 vol.% TiC, pre-heated) is about 906 ± 105 MPa (Fig. 3a the curve in red) with a plasticity greater than 10%, slightly higher than that of result without preheating the powder bed. The compression performance is expected to improve because of the different thermal gradients. Specifically, the preheating AMNC powder bed can avoid solidification cracking since the cooling rate is affected25. This can result in the improvement of residual stresses and distortion of the counterpart during the layer-by-layer process26. We then characterized the micropillar deformation after the compression test, as shown in Fig. 3b. Multiple slip bands appeared in the laser-deposited pure aluminum specimens, which are common for microcompression tests of face-centered cubic micropillars27. In contrast, the AMNC specimens have significantly fewer slip bands as compared to those in the pure Al specimens. It is highly likely that TiC nanoparticles in the Al specimens can sustain higher compression loads, resulting in significant higher yield strength. This hypothesis can also be validated by the compression test at the elevated temperature of 400 °C (See Supplementary Movie 1 and Supplementary Fig. 6).Fig. 3Room-temperature mechanical behavior of laser-deposited aluminum with and without nanoparticles. a Typical engineering stress-strain curves of laser-deposited Al specimens with and without nanoparticles. b SEM images of micropillars after microcompression tests. c Young’s modulus of laser-deposited Al and AMNC specimens. Error bars represent SD for at least twenty data sets. d Specific Young’s modulus and specific yield strength of AMNC and other materials (all data from microcompression tests without size effect). Scale bar, 3 µm (left), 2 µm(middle) and 1 µm(right) in bFull size image To understand the strength obtained in as-deposited AMNCs, the strengthening mechanisms for the AMNC (35 vol.% TiC) can be possibly attributed to Orowan strengthening28, Hall-Petch effect29, and load-bearing transfer, which are estimated to be approximately 294, 104, and 525 MPa, respectively (see Mechanical Strengthening Mechanisms in Methods). The strong interfacial bonding as shown in Fig. 2d suggests that its theoretical value would be approximately 1000 MPa with a load-bearing transfer strengthening of 525 MPa. However, it should be noted that there is no direct evidence indicating interfacial bonding strength.It is postulated that the good interfacial bonding between the nanoparticles and Al matrix results in the superior elastic modulus in the AMNC specimens. Figure 3c shows that the Young’s modulus of the laser-deposited AMNC specimens is significantly enhanced when compared to that of pure aluminum. While the pure aluminum specimen has a Young’s modulus of 68 ± 4 GPa in our tests, the AMNC (17 vol.% TiC) and AMNC (35 vol.% TiC) specimens offer a Young’s modulus of 108 ± 10 GPa and 197 ± 27 GPa, respectively. Figure 3d shows the specific Young’s modulus and specific yield strength of the AMNC specimens and other engineering alloys, indicating that the AMNC (35 vol.% TiC) exhibits the highest specific Young’s modulus and one of the best specific yield strengths among all structural metals (see Comparison of specific mechanical properties with other materials in Methods). An extensive review of conventional aluminum matrix composites (AMCs), i.e. especially aluminum reinforced with TiC micro particles, is included in Supplementary Table 1. All AMC-TiC composites offer much lower mechanical properties and Young’s modulus than our laser printed AMNCs (See Supplementary Table 1). The comparison of the specific modulus and yield strength between the AMNCs and any other aluminum-based materials (aluminum alloys and composites) is shown in Supplementary Fig. 7. While the microcompression tests conducted using micropillars without size effect provide scientifically-meaningful yield strength values, Young’s modulus, and uniform plasticity to characterize the laser printed AMNCs, tensile testing would pose a more serious challenge for ductility in the laser-printed specimens, which heavily depends on engineering optimization. It should be noted, however, that uniform plasticity of about 10% in the laser-printed AMNCs is a good indication that the material can withstand some plastic deformation. It is worth noting, however, that in the case of high-temperature applications, ductility may not be an issue.High temperature stability.It is well known that most aluminum alloys lose their strength at elevated temperatures due to the rapid coarsening of grain size and loss of strengthening precipitates30. To evaluate the high-temperature stability of the laser-deposited AMNC (35 vol.% TiC) specimens, microcompression tests were conducted at 200, 300, and 400 °C (See in-situ microcompression test at elevated temperatures in Methods). The results from the in-situ microcompression (See Supplementary Movie 1) after testing reveal that the AMNC specimens can still reach a yield strength of 200 ± 43 MPa with a plasticity greater than 15% at 400 °C, as shown in Fig. 4a. To understand the strength contribution, the microstructure of 400 °C tested specimen was studied in detail (See Supplementary Fig. 8a), revealing that the fine particles are still well dispersed and distributed. Since the average particle diameter and the grain size are expected to change under high-temperature testing condition, the estimated strengthening from Hall-Petch effect and Orowan strengthening are ~77 and 25 MPa, respectively (See detailed in Supplementary Fig. 7). The strong interfacial bonding is still very likely the main strengthening contribution, which is a promising direction for further study to validate the hypothesis. Moreover, to further determine the thermal stability after a heating period of 1.0 h at 400 °C, the AMNC specimens were cooled down to room temperature and again tested using microcompression testing at 25 °C (See High-temperature stability measurement in Methods). The results show that the AMNC (35 vol.% TiC) specimens still exhibit a yield strength greater than 800 MPa with a plasticity greater than 10%. Figure 4b also shows the AMNC (35 vol.% TiC) specimens offer exceptional strength when compared with other engineering materials at different elevated temperatures. At 400 °C, the AMNC specimens offer a higher strength than any other aluminum materials and even greater than stainless steel SS304. These results clearly suggest that AMNC with dispersed nanoparticles deposited via laser additive manufacturing, not only exhibits yield strength and plasticity that are superior to those of previously reported Al-based materials but also provides exceptional high-temperature stability (see Comparison of yield stress at elevated temperatures in Methods).Fig. 4Mechanical behavior of laser-deposited AMNC at elevated temperature. a Typical engineering stress–strain curves for microcompression of laser-deposited AMNC (35 vol.% TiC) at 400 °C. b Yield strength of AMNC (35 vol.% TiC) at test temperatures of 25, 200, and 400 °C in comparison with other materials. Error bars show s.d. of three tested samplesFull size image In summary, aluminum with dense dispersed nanoparticles was layer-deposited via laser additive manufacturing of AMNC powders, delivering the highest specific Young’s modulus and one of the best specific yield strengths among all structural metals, as well as an thermal stability at 400 °C amongst all aluminum-based materials. The AMNC powders allow a higher laser absorption by almost one order of magnitude than the pure aluminum powders. The pathway for laser 3D printing of nanoparticles reinforced metals can be readily extended to other materials for widespread applications.
From: