Review of Selective Laser Melting Materials and Applications
Sci Technol Adv Mater. 2018; 19(one): 370–380.
Selective laser melting of high-performance pure tungsten: parameter design, densification beliefs and mechanical properties
Chaolin Tan
a School of Materials Science and Engineering science, South China University of Technology, Guangzhou, China,
b National Engineering Laboratory for Modern Materials Surface Engineering Technology, Guangdong Institute of New Materials, Guangzhou, China,
c School of Metallurgy & Materials, University of Birmingham, Birmingham, United kingdom,
Kesong Zhou
a School of Materials Scientific discipline and Engineering science, South China Academy of Technology, Guangzhou, Prc,
b National Engineering Laboratory for Modernistic Materials Surface Engineering Technology, Guangdong Institute of New Materials, Guangzhou, China,
Wenyou Ma
b National Engineering Laboratory for Modern Materials Surface Technology Technology, Guangdong Institute of New Materials, Guangzhou, Red china,
Bonnie Attard
c School of Metallurgy & Materials, University of Birmingham, Birmingham, UK,
Panpan Zhang
b National Engineering science Laboratory for Modern Materials Surface Engineering Technology, Guangdong Institute of New Materials, Guangzhou, Prc,
Tongchun Kuang
a School of Materials Science and Engineering, Southward China Academy of Technology, Guangzhou, Communist china,
Received 2017 Dec 4; Revised 2018 Mar 17; Accepted 2018 Mar 18.
Abstract
Selective laser melting (SLM) additive manufacturing of pure tungsten encounters about all intractable difficulties of SLM metals fields due to its intrinsic properties. The central factors, including pulverisation characteristics, layer thickness, and laser parameters of SLM high density tungsten are elucidated and discussed in item. The master parameters were designed from theoretical calculations prior to the SLM procedure and experimentally optimized. Pure tungsten products with a density of 19.01 thou/cm3 (98.fifty% theoretical density) were produced using SLM with the optimized processing parameters. A high density microstructure is formed without significant balling or macrocracks. The formation mechanisms for pores and the densification behaviors are systematically elucidated. Electron backscattered diffraction analysis confirms that the columnar grains stretch across several layers and parallel to the maximum temperature gradient, which tin can ensure practiced bonding betwixt the layers. The mechanical properties of the SLM-produced tungsten are comparable to that produced by the conventional fabrication methods, with hardness values exceeding 460 HV0.05 and an ultimate compressive strength of about ane GPa. This finding offers new potential applications of refractory metals in condiment manufacturing.
Keywords: Additive manufacturing, selective laser melting, tungsten, refractory metal, parameter design, densification, linear energy, light amplification by stimulated emission of radiation parameter, molten puddle, belongings
Nomenclature: 10 Engineering and Structural materials, 303 Mechanical / Physical processing, 210 Thermoelectronics / Thermal send / insulators, 106 Metal materials, 305 Plasma / Laser processing
Abstract
1. Introduction
Tungsten (Due west), as the highest melting point refractory metallic, has many unique physical and chemical properties, including high density, loftier thermal conductivity, high recrystallization temperature, low thermal expansion, and loftier force and hardness at room and elevated temperatures. Tungsten and its alloys have been applied in numerous fields, including as lighting engineering science, electronics, manufacturing, aerospace, military, medical field, and especially nuclear field [1–4]. For example, tungsten is of interest to the nuclear industry as a promising candidate for plasma-facing materials (PFM) in future nuclear fusion devices such as the International Thermonuclear Experimental Reactor (ITER) and high-functioning rocket nozzles, as intensive transient heat loads must exist withstood alongside the requirements of limited tritium retentiveness, and handling enormous particle flux of hydrogen, helium and neutrons [1,v,six].
However, due to its relatively high ductile-brittle transition temperature, tungsten products are conventionally manufactured through pulverization metallurgy (PM), spark plasma sintering (SPS), chemical vapor deposition (CVD), hot isostatic pressing (HIP), and thermoplastic routes [7,8]. Nevertheless, these methods have limitations in producing parts with circuitous internal and external structures. Therefore, a new approach of producing tungsten components is necessary.
Additive manufacturing using Selective laser melting (SLM) is capable of producing 3D parts in an incremental layer-by-layer fashion using the laser to cook, sinter, and bond powder particles together in a thin layer on a powder bed [9,10]. Due to many irreplaceable superiorities such as high resolution and dimensional tolerance, high fabric and resource efficiency, good part pattern and production flexibility, desirable mechanical performance of SLM technology [11–13], information technology has been gradually applied in customized medical and dental application fields, tooling inserts with conformal cooling channels and functional components with loftier geometrical complexity such equally porous and lattice constructs [14–16].
Materials well suited to SLM take practiced laser absorption and balanced properties of melting indicate, thermal conductivity, surface tension and viscosity [5]. SLM of pure tungsten encounters nearly all intractable difficulties of metal SLM manufacture. The high melting betoken (3695 Chiliad) of the element results in high cohesive free energy, while its high surface tension (2.361 N·m−i) promotes the germination of balling phenomenon and its high thermal electrical conductivity (173 W·1000−1 k−1) leads to rapid solidification and cooling. The high viscosity (about viii × 10−3 Pa·s), which derives from the loftier intrinsic cohesive energy, significantly reduces the flowability of the molten pool [five]. The oxidation sensitivity may lower wettability and lead to the formation of cracks when even the small amounts of oxygen are captivated past the molten pools.
Due to the aforementioned difficulties, quite a few works concerning condiment manufacturing tungsten were reported, and the high-density components have even so not withal been obtained [5,17]. Deprez et al. [17] produced a complex collimator from pure tungsten powder using SLM, achieving a density of 17.31 g/cmiii (89.92 of theoretical density). Zhou et al. [five] investigated the balling phenomena in selective light amplification by stimulated emission of radiation melted tungsten, obtaining a specimen density of 16 g/cm3 (82.nine% of theoretical density). Both of them were unable to produce loftier density tungsten parts and characterize the mechanical performance. In our work, high density pure tungsten parts were produced by SLM through optimization of the processing parameters. The densification, microstructure, and mechanical performances were characterized in detail. The results demonstrate that fifty-fifty the highest melting indicate metals can be produced via additive manufacturing, expanding the potential applications of this technique.
2. Experimental details
2.i. Materials and SLM process
The raw material used in this study was high purity, plasma spheroidized tungsten powder (purity 99.9%) supplied by Tekna Advanced Materials Inc. (Quebec, Canada). The size distribution of the powder was measured by a HORIBA Partica LA960 laser scattering particle size analyzer (Tokyo, Nihon), and the trace elements (carbon, oxygen, and nitrogen) were estimated by an Elementar Vario EL CUBE elemental analyzer (Langenselbold, Deutschland).
The experiments were carried out in an EOS M290 SLM organisation (EOS GmbH, Krailling, Germany). The laser parameters were orthogonally designed in which the laser power (P) and scanning speed (v) were set in the range of 200–370 W and 100–400 mm/southward, respectively. The linear free energy was divers as η =P/v (J/mm). The laser scanned in a zigzag pattern with 67° rotation between side by side layers to minimize the residual stress. When selecting the required layer thickness, a rest must be reached between achieving a fine resolution and allowing for good pulverisation flowability. This usually occurs in the range of 20–100 μm [18]. A layer thickness of 20 μm was selected every bit it allows for the complete melting of the powder and enables skilful bonding between the layers, since an increase in thickness would cut down the amount of free energy reaching the underlying layer, and reduce the thermal penetration depth for the re-melting of the underlying layers. Incomplete melting or poor bonding could result in regions of greater inhomogeneity with a higher prevalence of cracks and incomplete fusion in the material [nineteen]. A larger layer thickness could besides atomic number 82 to insufficient volumetric laser energy density, which promotes balling due to a lack of wetting of the molten pool with the previously deposited layer [20]. Tungsten powder has an extremely loftier melting point and is prone to balling during SLM, thus selecting the right layer thickness is crucial. Prior to SLM processing, the building platform was pre-heated to fifty °C (323 Thou). The oxygen content in the procedure chamber was kept beneath 0.1 vol. % through a continuous flow of argon, as balling tin can exist significantly reduced by improved oxygen control in the process sleeping room [18,xx].
2.2. Characterization
The roughness, S a , of the fresh surfaces for the as-fabricated specimens was evaluated by a BMT SMS Expert 3D model optical profilometer (Breitmeier Messtechnik GmbH, Ettlingen, Deutschland) with a measuring area of two mm × two mm, the average values and standard deviations were estimated from v measurements. The density of the as-made tungsten specimens was determined by both image analysis method using a Leica DMI5000M (Wetzlar, Germany) optical microscopy (OM) and the Archimedes method. The prototype analysis method evaluates the porosity within a specimen past computing the percentage area of porosity on the polished surfaces, while the Archimedes method calculates the specimen's density according to ASTM B962-08. The microstructural and surface morphology observations were performed using the OM and a Zeiss Merlin (Jena, Germany) field emission scanning electron microscope (Fe-SEM). Prior to microstructural observations specimens were etched in standard Murakami'due south solution (1:1:10 for KOH: 10003[Fe(CN)6]: H2O). X-ray diffraction (XRD) was conducted using a Bruker D8 Advance Diffractometer (Karlsruhe, Germany) with a Cu Kα radiations (wavelength, λ = 0.15418 nm), at xl kV and 40 mA in a 2θ range of thirty–90° using a pace size of 0.02°. Electron backscattered diffraction (EBSD) was carried out on a Hitachi Southward-3400N SEM system (Tokyo, Nippon) at 20 kV through an integrated Oxford/HKL EBSD detector, using a step size of 300 nm. Specimens for EBSD examination were mirror polished and then vibratory polished for 5 h in order to remove any scratches and plastic deformation from the surface layer. The EBSD data was analyzed using Channel 5 software (HKL Technology, Inc., Connecticut, CA, USA). Micro-hardness of the specimens was measured by a Shimadzu HMV-2T micro Vickers tester (Tokyo, Japan) with a load of 50 gf for 10 s. All hardness indentations were applied to the polished horizontal (parallel to the build plate) surfaces, and an average value and respective standard divergence were estimated from 10 measurements. The compressive tests for the equally-fabricated tungsten specimens were carried out on a SANS CMT 5105 universal material testing motorcar (MTS Systems China Co. Ltd., Shenzhen, China) with a strain rate range of 3.5 × 10−three/s–4.5 × 10−3/southward. The length to diameter (L/D) ratio of the specimens was 3 (every bit per ASTM E9-09), and all the specimens were simultaneously polished by a grinding car to ensure identical surface quality.
3. Results and discussion
3.1. Parameters pattern and analysis
iii.ane.1. Powders
In addition to light amplification by stimulated emission of radiation parameters, powder morphology and size distribution likewise have important effects on the SLM process. The morphology of the tungsten is of fundamental importance. Equally such a highly spherical pulverization was chosen, to improve the flowability, and promote the wetting between the melt pool and powder particles. Zhou et al. [5] produced pure tungsten by SLM using powder with an irregular morphology, which may have been one of the factors resulting in lower part density (16 g/cm3) obtained in their study. Wang et al. [21] demonstrated that spherical powder could increase light amplification by stimulated emission of radiation absorptivity and packing density in comparison to polyhedral powders. In addition to this, the size distribution of the powder also plays an important role during the SLM procedure. As depicted in Effigy i, the tungsten pulverisation was advisedly selected with a relatively small mean particle size of 17 ± 5 μm, and a narrow pulverization size distribution, with D50 and D90 values of xvi.24 and 23.67 μm, respectively. The levels of the main impurity elements O, C, N in the powders are measured to be 220, 580, and 470 ppm, respectively. The comparatively low oxygen content in the tungsten powder may also suppress the balling tendency during SLM.
SEM morphology (a) and size distribution (b) of the pure tungsten pulverization.
The temperature of the pulverisation bed is increased as it absorbs the light amplification by stimulated emission of radiation energy Eastward laser , equally expressed in the following rut balance equation [22]:
where r p , ρ , C p and ΔT are the particle radius, density, specific estrus and temperature increase in the powder particles, respectively. Obviously, the smaller particles are more easily heated due to the lower thermal capacitance. Even so, significantly undersized particles agglomerate more easily due to Van der Waals forces, reducing powder flowability and causing poor powder degradation [18].
3.ane.2. Laser parameters
Theoretically, an judge of the energy required to melt a volume of material Q p (J/mmiii), can be determined from the expression [23]:
where ρ is the density (kg m−three) of bulk material,C p is the specific oestrus (J·kg−1 M−1),T 1000 is the melting betoken (K),T 0 is the initial temperature (K), andL f is the latent oestrus of fusion (KJ·kg−1). The detailed values of these physical parameters are listed in Table 1, the free energy required to cook of pure tungsten powder preheated to 323 Grand is calculated to be about 8.595 J/mm3.
Table one.
Physical parameters used for theoretical calculation in laser parameters design.
| Physical parameter | Unit | Value |
|---|---|---|
| Density/ρ | kg·m−3 | nineteen.30 × 103 |
| Specific heat/C p | J (kg·Yard)-ane | 132 |
| Melting signal/T m | K | 3695 |
| Initial temperature/T o | Yard | 323 |
| Latent estrus of fusion/L f | J·kg−ane | 2.20 × xfive |
| Light amplification by stimulated emission of radiation absorptivity/α | – | 0.41 |
| Laser beam radius/r b | mm | 50 × 10−three |
| Laser melting radius/r | mm | 100 × 10−3 |
| Melting zone depth/H | mm | 30 × ten−3 |
Figure 2 illustrates the SLM process and the heat transfer in the molten pool. The intensity profile of the laser beam is assumed to accept a Gaussian distribution. The light amplification by stimulated emission of radiation energy flux can therefore be expressed equally [24]:
Schematic diagram of the SLM process and the heat transfer in molten pool.
whereP is the light amplification by stimulated emission of radiation power,r b is the light amplification by stimulated emission of radiation beam spot radius,r is the altitude from the axle middle, and α is absorptivity coefficient. The average laser energy flux E s (W/mm2) can be expressed as:
The free energy loss in the molten puddle tin exist divided into three parts, including convective estrus loss, radiation heat loss, and heat loss through evaporation [25]. Usually, the heat loss due to convection and radiation is mostly assumed to be about x% [24,26]. Previous calculations revealed that the latent heat of evaporation is relatively big, causing prominent energy loss, specially at high temperatures. The rut loss through evaporation was therefore estimated to be x% [25]. Therefore, the total rut loss during the SLM process was assumed to be twenty%. This interpretation of estrus loss is of a similar level with previous work of direct laser deposition H13 tool steel, in which about 75–85% of original power reached the molten pool surface [27]. Hence, the effective laser energy flux E in(West/mmtwo) used for melting the powders is:
Due east in =Due east s -E loss =Eastward southward (one - 0.2)
Therefore, the volumetric light amplification by stimulated emission of radiation energy per unit volume Q v (J/mmiii) of tungsten powders can be expressed as
where Δt is the laser exposure fourth dimension, 5 is the light amplification by stimulated emission of radiation browse speed, V m is the effective volume of molten pool. As shown in Figure two, we simplified the profile of the molten pool as a office of sphere, and the V m can exist estimated every bit:
where r is the effective radius of laser beam, H is the effective light amplification by stimulated emission of radiation melting depth and R is the radius of the sphere. Normally, the width of the laser melting runway on the powder is well-nigh twice every bit the radius of laser beam r b , then we suppose r = 2r b . Also, for the purpose of introducing laser re-melting of the nether layer, we designed the effective laser melting depth to be equally thick equally 2 layer thickness (i.due east. H = 40 μm). Since, it enables a reliable connexion between the layers and suppresses the balling phenomenon equally balling can occur when tungsten molten pools solidify too quickly to spread out completely. Furthermore, choosing a suitable laser re-melting depth would besides reduce the tendency of cracking, since re-melting can break up the oxide films and reduce the thermal stress [five,xviii,twenty,28]. Consequently, V m and Q v tin be rationally calculated. When Q v ≥Q p (Q p = 8.595 J/mm3), the light amplification by stimulated emission of radiation should theoretically completely melt the powders. To run into the requirement of Q v ≥Q p , the condition for linear energy η ≥ 0.42 J/mm needs to be satisfied, therefore, we designed the minimal η to be of 0.5 J/mm, in which the P, v and Q v are 200 W, 400 mm/southward and x.287 J/mmiii, respectively. A enough of light amplification by stimulated emission of radiation parameters were orthogonally designed with P, v and η setting in the range of 200–370 W, 100–400 mm/s, and 0.5–3.7, respectively.
3.2. Parameters optimization
Effigy three shows the optical images of SLM-produced pure tungsten parts. Figure 3(a) reveals the relationships between the sintering formability and the laser parameters, where the specimens produced with η i = 0.500, η ii = 0.625, η 3 = 0.667, η four = 0.750, η five = 0.833 and η half dozen = 1.000 J/mm (marked in yellow background color) nowadays meliorate sintering formability. Thus, the parameter window to obtain a good, structurally sound component ranges between η values of 0.5–ane.0 J/mm, with the corresponding P and v covering the range of 200–300 W and 200–400 mm/south, respectively. Farther increasing the linear energy leads to burning loss and surface corrugation, which destroyed the forming quality and hindered the powder recoating. Also, even the aforementioned linear energy, the different combination between P and five presented distinct forming qualities, for example, when the linear free energy was in the same value of η = ane.0 J/mm, the specimens prepared with p = 200 W and 5 = 200 mm/s show better sintering formability than these with p = 300 West and five = 300 mm/south. This is because the lower speed allowed more time for pools wetting and spreading, which compensated the intrinsic depression flowability of tungsten. Effigy 3(b) shows the optical images of an SLM-produced thin-wall (0.4 mm in thickness) tungsten part and some specimens with the η 3 linear energy. The shining and smooth surface without any macrocracks and balling certifies reasonable light amplification by stimulated emission of radiation parameters and high sintering quality.
Optical images of SLM made pure tungsten parts: (a) blocks fabricated with unlike linear energies; (b) 0.40 mm thin-wall tungsten role and specimens produced with η 3 = 0.667 J/mm.
3.3. Surface morphology characterization
Effigy four shows the untreated fresh surface morphology and roughness of SLM fabricated pure tungsten taken from horizontal cantankerous-sections. A relatively shine and dumbo surface with regular liquid fronts is obtained and exhibited in Figure iv(a). No balling or private unmelted particles tin can exist observed, showing that all the laser tracks have skilful metallurgical bonding with each other excepting some scattered black slags. Microcracks tin too be observed through the inset high-magnification SEM epitome in Figure 4(a). Normally, cracks in SLM processed parts are caused by thermal residual stresses, which ascend from two mechanisms, i.east. thermal gradient mechanism (TGM) and the cool-down phase of molten top layers [29]. Additionally, loftier brittleness and insufficient plasticity of pure tungsten also contribute to the formation of cracks. High-magnification SEM ascertainment on the liquid fronts is provided in Figure four(b), massive nanosized columnar crystals are perpendicular to the fronts moving directions, suggesting an extremely high cooling rate of molten pools. In fact, the ultra-high cooling rate of laser processing could reach the order of 1010 M/s [thirty], and the relative high thermal electrical conductivity (173 W·m−i One thousand−1) also contributes to the high cooling rate and grain refinement.
Surface morphology and roughness taken from horizontal surfaces of SLM made pure tungsten: (a) SEM image showing regular light amplification by stimulated emission of radiation tracks (η 3 = 0.667 J/mm) and microcracks (inset image); (b) corresponding high-magnification SEM image showing massive nanocrystals in the liquid fronts; (c) the respective 3D topography epitome of the untreated fresh surface and (d) relationships betwixt surface roughness and linear energy summarized from the 3D topography images.
Surface roughness is another of import gene reflecting the sintering formability. The representative (η three = 0.667 J/mm) 3D topography paradigm of surface is provided in Figure four(c), which as well reveals that a relativly smooth surface is obtained. Relationships between the boilerplate surface roughness and the linear free energy are depicted in Figure iv(d), which was summarized from a number of 3D topography analyses. The boilerplate Sa of the original surface of the specimens is in the range of nearly 7–ix μm, with the highest value of viii.97 μm in the η four specimen and the lowest value of 6.74 μm in the η three specimen. The relative smoothen surface implies the loftier germination qualities.
three.4. Densification behavior
Figure 5(a)–(f) show the representative polished OM images taken from the horizontal cross sections of specimens produced with η 1–η half-dozen. A plenty of pores tin can be observed in η i and η 6 specimens, while relativly fewer micropores presented in η 3 specimen. The human relationship betwixt η and density measured by both image analysis method and Archimedes method is shown in Figure 5(g). The relative densities measured by the image analysis method are about 98–99%, while the densities measured by Archimedes method are 18.86–19.01 g/cm3, which is 97.72–98.50% density of pure tungsten (xix.30 g/cm3). The pocket-size but systematic difference of density estimated from images and Archimedes methods might originate from open pores. Despite this pocket-size difference between these two methods, both results testify the SLM-produced specimens achieved a high density of more than 97.five% for all specimens. Both curves prove that an optimum density obtained at η 3. Moreover, in that location is a decline tendency of density when the η value decreases from η iii to η ane or increases from η 3 to η half-dozen, equally also revealed in Figure 5(a)–(f).
Optical micrographs showing the pores in the horizontal cross-sections of pure tungsten produced by SLM with different linear energies (J/mm): (a) η one = 0.500, (b) η ii = 0.625, (c) η iii = 0.667, (d) η 4 = 0.750, (due east) η 5 = 0.833, (f) η six = 1.000; (g) density graphs obtained from Archimedes and paradigm assay methods.
In that location are 2 main reasons accounting for pores and densification beliefs. First, turbulent flow of molten pools is incited by the high-energy density laser irradiation, and the pores are hands formed due to the protective gas rolled in the molten pools. On the one mitt, this could be explained by the dynamic viscosity (η v ) of the molten pool, which is defined as [31]:
whereone thousand, k, T, and σ are the atomic mass, Boltzmann constant, molten pool temperature and surface tension, respectively. The σ of tungsten is in negative linear relation with T, so increasing T tin lowerσ and atomic number 82 to a decrease inη five as a consequence. With the linear energy η increase from η 1 to η iii, the T of molten pools increases and the dynamic viscosity η v decreases, which increases the flowability of molten pool and improves the density in return. On the other hand, the molten pool is significantly influenced by the dynamics of the Marangoni effects. Marangoni flow increases with the laser free energy input, this increases the gas dragging probability toward the molten puddle which in plough leads to pores formation in the solidified molten pool [32]. Therefore, increasing η from η 3 to η 6 acquired the aggrandized porosity.
Second, the micropores may be ascribed to the balling phenomenon. During SLM process of pure tungsten, the droplet spread fourth dimension (86.3 μs) is virtually twice as its solidification time (46 μs) [5]. The solidification front end grows too fast, whereas the spreading time is express. And then the melt aerosol could be arrested and solidify as globular islands, which severely hinders the uniform subsequent pulverisation deposition. When the laser melts such an uneven powder layer, the movement of molten pool forepart undergoes a significant disturbance and even intermission. Consequently, it is hard to completely fill the inter ball pores on the surface of the previous layer, leading to the interlayer pores and a limited densification [33]. Insufficient light amplification by stimulated emission of radiation energy may cause the subtract in solidification fourth dimension and enhance balling; moreover, the formation of incomplete melted powders in the molten pool could decreases the flowability and wettability, which can also increase the porosity. Therefore, the porosity increases with η decreasing from η 3 to η one. Pores of SLM-produced pure tungsten are inevitable due to its intrinsic materials backdrop. It is worth mentioning that, the titanium and steel are like shooting fish in a barrel to exist produced by SLM, considering the solidification fourth dimension increases significantly with the increase of the melting temperature. The solidification fourth dimension tin can exceed the spreading time at the temperature of around 2100 K for both Ti and Fe [5], which ensures the complete spreading of the melt aerosol. And then an most fully dense part of Ti or Fe can be easily achieved, while pores of SLM-produced pure tungsten are inevitable in dissimilarity. In this work, the conscientious analysis and control of the laser parameters allowed achieving densities of up to 19.01 ± 0.02 g/cm3 (98.fifty ± 0.12% of theoretical density), which is much college than the previous reported results of 16 g/cm3 (82.nine%) [five], 17.31 one thousand/cm3 (89.92%) [17], and 18.53 grand/cm3 (96.01%) [21] for SLM-produced tungsten.
three.five. Microstructural observation and assay
Effigy six shows the microstructural morphologies and XRD patterns taken from horizontal cross-sections of SLM-produced pure tungsten. The dense microstructure along with a few micropores was observed by OM and provided in Figure vi(a) and (b). The spherical pores are commonly caused by gas entrapment in the molten pool, as discussed in Section 3.4. Grain boundaries could be conspicuously observed from the inset SEM morphology in Figure six(c), and massive nanosized spot-similar subgrains present within the grains in high-magnification SEM image. The nanosized subgrains would form in response to rapid solidification with extremely loftier cooling rates present in the laser procedure [34]. Likewise, as labeled in Effigy six(c), a few thermal cracks near or in the grain boundary may be caused by the thermal residue stress. Their formation was ascribed to the same tensile stress. In improver, the intrinsic backdrop of tungsten such as high brittleness, high oxidation sensitivity and low wettability also promote the formation of such cracks.
Microstructural observation and phase identification taken from horizontal cross-sections of SLM-produced pure tungsten: (a) depression-magnification and (b) high-magnification OM macrographs; (c) low-magnification (inset) and high-magnification SEM morphologies, and (d) XRD patterns of powder and SLM-produced specimens.
Figure 6(d) shows XRD patterns of raw tungsten pulverisation and SLM-produced tungsten parts. All the diffraction peaks characterize the standard body-centered-cubic (bcc) tungsten phase (ICDD #04-0806) in (1 1 0), (2 0 0), (ii 1 1), and (2 2 0) planes. Besides, the diffraction patterns of the SLM-produced specimens prove wider peaks in comparing to those of the precursor pulverization. According to the Debye-Scherrer formula, this top broadening is related to the modest size of crystalline domains formed during the rapid solidification after laser processing; it may also indicate residual stresses within the SLM-produced specimens acquired past lattice distortions [35,36].
Equally shown in Effigy 7, SEM morphology and EBSD maps recorded in the middle region along the building direction (Z direction) were further performed to report the microstructure and texture development of equally-fabricated tungsten. Cracks along the edifice direction can exist clearly observed in Figure 7(a). EBSD results reveal that the microstructure consists of columnar grains (Effigy 7(b)), which is slightly different from the strip microstructures in the horizontal cross-sections provided in Figure six(a) and (b). Every bit has been reported in many articles, columnar grains are formed in SLM due to the direction of oestrus transfer; they grow toward the molten pool along the Z-direction, since the growth velocity is much higher when the crystals growth management is aligned with the maximum temperature gradient [fifteen,37]. The columnar grain stretching across several layers is favorable for layer bonding. Because, the fracture and defects normally present among layers due to the layerwise edifice method, the stretched columnar grains could effectively reduce this fabrication orientation-based weakness. Additionally, microcracks are decumbent to form in the SLM process due to the intrinsic high brittleness of tungsten and the thermal gradients. When high power-density light amplification by stimulated emission of radiation irradiates on relative thin powder layer, the high thermal electrical conductivity leads to potent heat flux, which is parallel and negative to the building direction. This heat flux can cause re-melting of underlying layer (as designed in Section 3.one.2), which reduces temperature gradient at the layer interface. Therefore, the thermal stress and microcracks among layers decreased to a great extent. However, it was reported that columnar lamellar microstructures along Z management can provide long unrestricted sideslip paths for crack growth [38]. In this example, cleft can easily initiate and propagate in Z management, which was verified by the presence of cracks forth the building direction in Effigy 7(a).
SEM and EBSD analysis carried forth the building direction (Z direction): (a) SEM image showing a scissure along edifice management and (b) Inverse pole figure showing the microstructure and grain orientation map of SLM-produced tungsten.
three.6. Mechanical performance
Figure 8(a) illustrates the result of light amplification by stimulated emission of radiation linear energy on microhardness, which was measured from the cantankerous-sections of SLM-processed pure tungsten specimens. The microhardness of the specimens is about 445–467 HV0.05, and no statistically significant variation of hardness among the dissimilar linear energies processed specimens was observed. The hardness is maximal for the η 5 specimen, and exceeds 460 HV0.05 for both η iii and η v specimens. The SLM-processed tungsten specimens testify a superior hardness compared with conventional PM or SPS processed tungsten (typically 320–400 HV) [39]. The reasons are explained below: the residual stresses in SLM layerwise-build parts are quite large, merely they are not always disadvantageous. Because, on the premise of a sufficiently high densification without massive cracks or pores, a reasonable level of residual stress in SLM-processed parts may cause dislocation strengthening and hardness enhancement [7,33,twoscore].
(a) The effect of laser linear free energy on microhardness of SLM-processed pure tungsten and (b) compressive stress–strain curves of SLM fabricated pure tungsten with dissimilar light amplification by stimulated emission of radiation linear energies.
Figure eight(b) shows the compressive stress–strain curves of SLM-processed pure tungsten. The measured mechanical properties are summarized in Tabular array 1. The ultimate compressive strength (UCS) and compressive yield stress (CYS) of η three specimen are 1015 and 882 MPa, respectively, which reaches the highest value among all specimens. The η four specimen exhibits the lowest compressive strength with the corresponding UCS and CYS of 933 and 791 MPa, respectively. The compressive results are highly consequent with the density results (Figure five(g)). The mechanical backdrop of tungsten produced by other conventional fabrication techniques are also listed in Tabular array 2 for comparision. All quoted results were obtained in as-fabricated condition without subsequent treatment and under the same quasi-static compression examination. As revealed in Table 2, the compressive stength (particularly the η 3) is comparable for SLM and conventional fabrication methods, including CVD, HIP, PM, and SPS. Our results demonstrate that SLM is a new feasible methods for fabrication of pure tungsten parts with desirable functioning.
Tabular array two.
Comparing of the mechanical properties of pure tungsten fabricated by SLM and conventional processing methods.
| Specimens | CYS (MPa) | UCS (MPa) | Strain (%) | Density (g/cmthree) | South a (μm) | HV |
|---|---|---|---|---|---|---|
| SLM η 1 = 0.500 | 868 | 978 | 5.97 | 97.82% (18.88 ± 0.02) | seven.59 | 445 ± 39 |
| SLM η 2 = 0.625 | 864 | 984 | half-dozen.58 | 98.29% (18.97 ± 0.06) | eight.ten | 448 ± 25 |
| SLM η 3 = 0.667 | 882 | 1015 | 6.76 | 98.50% (19.01 ± 0.02) | half dozen.74 | 461 ± 18 |
| SLM η 4 = 0.750 | 791 | 933 | 8.65 | 97.98% (18.91 ± 0.05) | 8.97 | 452 ± 31 |
| SLM η v = 0.833 | 849 | 964 | 6.64 | 97.93% (18.xc ± 0.03) | seven.46 | 467 ± 29 |
| SLM η 6 = 1.000 | 860 | 962 | vi.36 | 97.72% (eighteen.86 ± 0.04) | 7.44 | 456 ± 41 |
| CVD [41,42] | – | 780–1480 | – | ≤99.79% | – | 419 (4.five GPa) |
| HIP [ii,43] | 1010 | 1180 | – | ≤98.00% | – | – |
| PM [44,45] | 900 | 1000–1200 | – | ≤98.twenty% | – | 344 |
| SPS [7,39,46] | 750 | 980 | – | ≤96.thirty% | – | 372 (4 GPa) |
iv. Conclusions
In summary, although, SLM manufacturing of pure tungsten encounters many intractable difficulties due to the cloth intrinsic properties, nosotros produced high-density thin-wall pure tungsten parts past SLM with optimized processing parameters. The main conclusions were summarized every bit follows:
-
(1)
Powders in highly spherical shape, sufficiently thin deposition layer thickness, and optimized linear free energy were vital controlling factors in SLM-produced tungsten. A reasonable friction match of these factors was in favor of a high-performance tungsten. Parameter calculation and design prior to experiments could give valuable guidance to the optimization of light amplification by stimulated emission of radiation parameters.
-
(2)
Different light amplification by stimulated emission of radiation linear energy (η) had significant influence on sintering formability, density, and mechanical properties of SLM specimens. In the η range of 0.5–3.7 J/mm, the η 3 = 0.667 J/mm presented better sintering formability and performance than the others. Specimens produced with η 3 reached the highest density of xix.01 ± 0.02 g/cm3 (98.50 ± 0.12% of theoretical density).
-
(3)
Dense microstructures without balling and with few microcracks were formed in SLM candy pure tungsten. The microcrack in SLM processed pure tungsten seemed inevitable, due to the intrinsic properties of fabric and the specific processing manner, particularly at grain boundary along the building direction. Only it seemed not seriously touch on the performance. The mechanical backdrop of SLM-produced tungsten were comparable to conventional fabrication methods, in which the highest hardness and the UCS reached more than 460 HV0.05 and 1 GPa, respectively.
The research outcomes suggest that careful tailoring of the laser processing parameters tin provide a valuable and cost effective way of optimizing the performance and density of the SLM-produced parts. This research gives us new insights into the awarding of refractory metals in additive manufacturing.
Funding
This piece of work was financially supported by the Natural Scientific discipline Foundation of Guangdong Province [grant number 2016A030312015]; Guangdong Scientific discipline and Technology Programme [grant numbers 2016B090916003; 2017B030314122; 2017A070702016]; Guangzhou Science and Technology Plan [grant numbers 201604016109; 201704030111]; Guangdong Academy of Sciences Projects [grant numbers 2016GDASPT-0206; 2017GDASCX-0202; 2017GDASCX-0111; 2018GDASCX-0402].
Disclosure statement
No potential disharmonize of involvement was reported by the authors.
References
[ane] Zhang X, Yan Q, Lang S, et al. . Grooming of pure tungsten via various rolling methods and their influence on macro-texture and mechanical backdrop. Mater Des. 2017;126:1–eleven. [Google Scholar]
[two] Huang L, Jiang L, Topping TD, et al. . In situ oxide dispersion strengthened tungsten alloys with high compressive strength and high strain-to-failure. Acta Mater. 2017;122:xix–31.x.1016/j.actamat.2016.09.034 [CrossRef] [Google Scholar]
[3] Koyanagi T, Kumar NAPK, Hwang T, et al. . Microstructural evolution of pure tungsten neutron irradiated with a mixed energy spectrum. J Nucl Mater. 2017;490:66–74.10.1016/j.jnucmat.2017.04.010 [CrossRef] [Google Scholar]
[iv] Valles 1000, Panizo-Laiz M, González C, et al. . Influence of grain boundaries on the radiations-induced defects and hydrogen in nanostructured and coarse-grained tungsten. Acta Mater. 2017;122:277–286.10.1016/j.actamat.2016.ten.007 [CrossRef] [Google Scholar]
[v] Zhou X, Liu X, Zhang D, et al. . Balling phenomena in selective light amplification by stimulated emission of radiation melted tungsten. J Mater Process Technol. 2015;222:33–42.ten.1016/j.jmatprotec.2015.02.032 [CrossRef] [Google Scholar]
[six] Parish CM, Wang K, Doerner RP, et al. . Grain orientations and grain boundaries in tungsten nonotendril fuzz grown under divertor-similar atmospheric condition. Scri Mater. 2017;127:132–135.10.1016/j.scriptamat.2016.09.018 [CrossRef] [Google Scholar]
[vii] Ma J, Zhang J, Liu W, et al. . Suppressing pore-boundary separation during spark plasma sintering of tungsten. J Nucl Mater. 2013;438:199–203.10.1016/j.jnucmat.2013.03.042 [CrossRef] [Google Scholar]
[eight] Erik Fifty, Schubert W-D. Tungsten: backdrop, chemistry, technology of the chemical element, alloys, and chemical compounds. Berlin: Springer Scientific discipline & Business organization Media; 2012. [Google Scholar]
[9] Olakanmi EO, Cochrane RF, Dalgarno KW. A review on selective laser sintering/melting (SLS/SLM) of aluminium alloy powders: Processing, microstructure, and backdrop. Prog Mater Sci. 2015;74:401–477.10.1016/j.pmatsci.2015.03.002 [CrossRef] [Google Scholar]
[10] Gu DD, Meiners W, Wissenbach G, et al. . Light amplification by stimulated emission of radiation additive manufacturing of metallic components: materials, processes and mechanisms. Int Mater Rev. 2013;57:133–164. [Google Scholar]
[11] Suryawanshi J, Prashanth KG, Scudino S, et al. . Simultaneous enhancements of forcefulness and toughness in an Al-12Si alloy synthesized using selective laser melting. Acta Mater. 2016;115:285–294.10.1016/j.actamat.2016.06.009 [CrossRef] [Google Scholar]
[12] Cunningham R, Nicolas A, Madsen J, et al. . Analyzing the effects of powder and post-processing on porosity and properties of electron beam melted Ti–6Al–4V. Mater Res Lett. 2017;v:516–525.10.1080/21663831.2017.1340911 [CrossRef] [Google Scholar]
[13] Challis VJ, Xu X, Zhang LC, et al. . Loftier specific strength and stiffness structures produced using selective light amplification by stimulated emission of radiation melting. Mater Des. 2014;63:783–788.10.1016/j.matdes.2014.05.064 [CrossRef] [Google Scholar]
[14] Amin Yavari South, van der Stok J, Chai YC, et al. . Bone regeneration functioning of surface-treated porous titanium. Biomaterials. 2014;35:6172–6181.10.1016/j.biomaterials.2014.04.054 [PubMed] [CrossRef] [Google Scholar]
[15] Tan C, Zhou Yard, Ma W, et al. . Microstructural evolution, nanoprecipitation behavior and mechanical properties of selective light amplification by stimulated emission of radiation melted high-performance form 300 maraging steel. Mater Des. 2017;134:23–34.10.1016/j.matdes.2017.08.026 [CrossRef] [Google Scholar]
[16] Leary Yard, Mazur Thousand, Elambasseril J, et al. . Selective laser melting (SLM) of AlSi12Mg lattice structures. Mater Des. 2016;98:344–357.ten.1016/j.matdes.2016.02.127 [CrossRef] [Google Scholar]
[17] Deprez K, Vandenberghe Due south, Van Audenhaege Yard, et al. . Rapid condiment manufacturing of MR uniform multipinhole collimators with selective light amplification by stimulated emission of radiation melting of tungsten powder. Med Phys. 2013;40:012501. [PubMed] [Google Scholar]
[18] Yap CY, Chua CK, Dong ZL, et al. . Review of selective laser melting: materials and applications. Appl Phys Rev. 2015;2:041101.x.1063/ane.4935926 [CrossRef] [Google Scholar]
[19] Spierings AB, Herres North, Levy G. Influence of the particle size distribution on surface quality and mechanical properties in AM steel parts. Rapid Prototyp J. 2011;17:195–202.10.1108/13552541111124770 [CrossRef] [Google Scholar]
[20] Li R, Liu J, Shi Y, et al. . Balling behavior of stainless steel and nickel powder during selective light amplification by stimulated emission of radiation melting process. Int J Adv Manuf Technol. 2012;59:1025–1035.ten.1007/s00170-011-3566-i [CrossRef] [Google Scholar]
[21] Wang D, Yu C, Zhou X, et al. . Dense pure tungsten made by selective laser melting. Appl Sci. 2017;7:430.10.3390/app7040430 [CrossRef] [Google Scholar]
[22] Qi H, Mazumder J, Ki H. Numerical simulation of heat transfer and fluid flow in coaxial laser cladding process for direct metallic deposition. J Appl Phys. 2006;100:11. [Google Scholar]
[23] Yadroitsev I, Krakhmalev P, Yadroitsava I, et al. . Energy input effect on morphology and microstructure of selective laser melting single rail from metallic powder. J Mater Process Technol. 2013;213:606–613.10.1016/j.jmatprotec.2012.11.014 [CrossRef] [Google Scholar]
[24] Thompson SM, Bian L, Shamsaei Due north, et al. . An overview of Straight Laser Degradation for additive manufacturing; Function I: send phenomena, modeling and diagnostics. Addit Manuf. 2015;8:36–62.10.1016/j.addma.2015.07.001 [CrossRef] [Google Scholar]
[25] Han L, Liou FW, Musti S. thermal behavior and geometry model of melt puddle in laser textile process. J Heat Trans. 2005;127:1005–1014.10.1115/1.2005275 [CrossRef] [Google Scholar]
[26] Wang L, Felicelli S. Assay of thermal phenomena in LENS™ degradation. Mater Sci Eng A. 2006;435–436:625–631.ten.1016/j.msea.2006.07.087 [CrossRef] [Google Scholar]
[27] He X, Mazumder J. Transport phenomena during directly metal deposition. J Appl Phys. 2007;101:053113.10.1063/1.2710780 [CrossRef] [Google Scholar]
[28] Das S. Physical aspects of process control in selective laser sintering of metals. Adv Eng Mater. 2003;v:701–711.10.1002/(ISSN)1527-2648 [CrossRef] [Google Scholar]
[29] Mercelis P, Kruth JP. Residual stresses in selective laser sintering and selective laser melting. Rapid Prototyp J. 2006;12:254–265.10.1108/13552540610707013 [CrossRef] [Google Scholar]
[xxx] Tan C, Zhu H, Kuang T, et al. . Light amplification by stimulated emission of radiation cladding Al-based baggy-nanocrystalline composite coatings on AZ80 magnesium alloy nether water cooling condition. J Alloys Compd. 2017;690:108–115.ten.1016/j.jallcom.2016.08.082 [CrossRef] [Google Scholar]
[31] Gu D, Shen Y. Effects of processing parameters on consolidation and microstructure of Due west–Cu components past DMLS. J Alloys Compd. 2009;473:107–115.10.1016/j.jallcom.2008.05.065 [CrossRef] [Google Scholar]
[32] Ng GKL, Jarfors AEW, Bi G, et al. . Porosity germination and gas bubble retention in light amplification by stimulated emission of radiation metal deposition. Appl Phys A. 2009;97:641.x.1007/s00339-009-5266-3 [CrossRef] [Google Scholar]
[33] Gu D, Hagedorn Y-C, Meiners W, et al. . Densification behavior, microstructure evolution, and wear functioning of selective laser melting processed commercially pure titanium. Acta Mater. 2012;60:3849–3860.10.1016/j.actamat.2012.04.006 [CrossRef] [Google Scholar]
[34] Zhang L-C, Attar H. Selective laser melting of titanium alloys and titanium matrix composites for biomedical applications. Rev Adv Eng Mater. 2016;18:463–475. [Google Scholar]
[35] Sun Z, Tan X, Tor SB, et al. . Selective laser melting of stainless steel 316L with low porosity and loftier build rates. Mater Des. 2016;104:197–204.ten.1016/j.matdes.2016.05.035 [CrossRef] [Google Scholar]
[36] Yihong K, Xipeng T, Shu Beng T, et al. . Fabrication and microstructural characterisation of additive manufactured Ti–6Al–4V parts by electron axle melting. Virtual Phys Prototyp. 2015;x:13–21. [Google Scholar]
[37] Vrancken B, Thijs L, Kruth JP, et al. . Microstructure and mechanical properties of a novel β titanium metal composite by selective laser melting. Acta Mater. 2014;68:150–158.x.1016/j.actamat.2014.01.018 [CrossRef] [Google Scholar]
[38] Davies P, Pederson R, Coleman M, et al. . The bureaucracy of microstructure parameters affecting the tensile ductility in centrifugally bandage and forged Ti-834 alloy during high temperature exposure in air. Acta Mater. 2016;117:51–67.x.1016/j.actamat.2016.07.015 [CrossRef] [Google Scholar]
[39] Liu G-Y, Ni Southward, Song M. Effect of indentation size and grain/sub-grain size on microhardness of loftier purity tungsten. T Nonferr Metal Soc. 2015;25:3240–3246.ten.1016/S1003-6326(15)63958-9 [CrossRef] [Google Scholar]
[xl] Gu D, Meiners West. Microstructure characteristics and germination mechanisms of in situ WC cemented carbide based hardmetals prepared by Selective Laser Melting. Mater Sci Eng A. 2010;527:7585–7592.ten.1016/j.msea.2010.08.075 [CrossRef] [Google Scholar]
[41] Yanwei L, Xiaodong Y, Chengwen T, et al. . Effect of the β phase on compressive mechanical property of CVD tungsten. Rare Metallic Mater Eng. 2011;twoscore:1138–1140.10.1016/S1875-5372(eleven)60045-3 [CrossRef] [Google Scholar]
[42] Huang H, Wu YQ, Wang SL, et al. . Mechanical backdrop of single crystal tungsten microwhiskers characterized by nanoindentation. Mater Sci Eng A. 2009;523:193–198.10.1016/j.msea.2009.05.060 [CrossRef] [Google Scholar]
[43] Zhou Q, Chen P. Fabrication and characterization of pure tungsten using the hot-shock consolidation. Int J Refract Metals Difficult Mater. 2014;42:215–220.ten.1016/j.ijrmhm.2013.09.008 [CrossRef] [Google Scholar]
[44] Lennon AM, Ramesh KT. The thermoviscoplastic response of polycrystalline tungsten in compression. Mater Sci Eng A. 2000;276:nine–21.10.1016/S0921-5093(99)00517-one [CrossRef] [Google Scholar]
[45] Wei Q, Jiao T, Ramesh KT, et al. . Mechanical beliefs and dynamic failure of loftier-forcefulness ultrafine grained tungsten under uniaxial compression. Acta Mater. 2006;54:77–87. [Google Scholar]
[46] Lee J-K, Kim S-Y, Ott RT, et al. . Result of reinforcement phase on the mechanical belongings of tungsten nanocomposite synthesized by spark plasma sintering. Int J Refract Metals Hard Mater. 2016;54:fourteen–18.10.1016/j.ijrmhm.2015.07.009 [CrossRef] [Google Scholar]
Articles from Scientific discipline and Applied science of Avant-garde Materials are provided here courtesy of National Institute for Materials Science and Taylor & Francis
Source: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5917440/
0 Response to "Review of Selective Laser Melting Materials and Applications"
Postar um comentário