Microstructural evolution and mechanical behavior of nickel aluminum bronze Cu-9Al-4Fe-4Ni-1Mn fabricated through wire-arc additive manufacturing

Additive Manufacturing 30 (2019) 100872

Contents lists available at ScienceDirect

Additive Manufacturing journal homepage: www.elsevier.com/locate/addma

Microstructural evolution and mechanical behavior of nickel aluminum bronze Cu-9Al-4Fe-4Ni-1Mn fabricated through wire-arc additive manufacturing

T

C. Dharmendraa, , A. Hadadzadeha,b, B.S. Amirkhizb,a, G.D. Janaki Ramc,a,1, M. Mohammadia ⁎

a

Marine Additive Manufacturing Centre of Excellence (MAMCE), University of New Brunswick, Fredericton, NB, E3B 5A1, Canada CanmetMATERIALS, Natural Resources Canada, 183 Longwood Road South, Hamilton, ON, L8P 0A5, Canada c Department of Metallurgical and Materials Engineering, Indian Institute of Technology Madras, Chennai, 600036, Tamil Nadu, India b

ARTICLE INFO

ABSTRACT

Keywords: Nickel Aluminum Bronze (NAB) Wire-arc additive manufacturing (WAAM) Electron microscopy κ-Phases Mechanical properties

As a step forward toward the development of the next generation of nickel aluminum bronze (NAB) components using wire-arc additive manufacturing (WAAM), square bars were printed in the vertical direction. The as-built microstructure was characterized using multi-scale electron microscopy techniques, where the differences in phase formation were compared to the reference cast-NAB based on the solidification characteristics. The as-cast microstructure typically consists of Cu-rich α-matrix, and four types of intermetallic particles referred to as κphases. In the WAAM-NAB, the formation of κI was suppressed due to high cooling rates. The microstructure was finer and the volume fraction of intermetallic particles was significantly lower than that of the cast-NAB. Based on energy dispersive spectroscopy (EDS) technique and diffraction pattern analysis using transmission electron microscopy (TEM), the phases formed in the interdendritic regions were identified as κII (globular Fe3Al) and κIII (lamellar NiAl), whereas numerous fine (5–10 nm) Fe-rich κIV particles were precipitated uniformly within the αmatrix. Electron backscatter diffraction analysis revealed weak texture on both parallel and perpendicular planes to the building direction with (100) poles rotated away from the build direction. The WAAM-NAB sample exhibited considerably higher yield strength (˜88 MPa) and elongation (˜10%) than the cast-NAB, but the gain in the ultimate tensile strength was marginal.

1. Introduction Nickel Aluminum Bronze (NAB) alloys are valued for their high strength combined with good ductility and toughness, high resistance to all forms of corrosion (especially, in marine environments), and excellent wear, cavitation, and galling resistance [1,2]. They display outstanding biofouling resistance, excellent cryogenic properties, and good damping capacity (nearly twice that of structural steels). Further, they are non-sparking [3], and have a lower density (due to high aluminum content) as compared to most copper alloys and are about 10% lighter than standard structural steels. NAB alloys have a long history of successful use in marine, architecture, and aerospace applications [4–6]. Within the family of NAB, the most commonly used is cast alloy Cu-9Al-4Ni-4Fe-1 Mn (C95800) that falls under the American Society for Testing and Materials standard (ASTM) B148 [7] and its wrought equivalent Cu-9Al-5Ni-4Fe alloy (C63200).

NABs form a sub-class of aluminum bronzes and in addition to aluminum as the primary alloying element contain significant amounts of nickel, iron, and manganese. In the binary copper-aluminum system, alloys containing up to 9% (wt.%) aluminum show a single-phase α microstructure under the equilibrium conditions. However, under normal (non-equilibrium) cooling conditions, aluminum content of above 8–8.5% results in a second phase known as β at high temperatures, which subsequently (below ˜ 565 °C) undergoes eutectoid decomposition to α + γ2 [8]. The formation of γ2 (Cu9Al4) is undesirable as it affects the ductility and toughness of the alloy, which is also prone to preferential corrosion [8]. The formation of undesirable γ2 can be suppressed by nickel and iron additions (˜ 5% each) that increase the solubility limit of aluminum up to 11%, and precipitation of more desirable κ-phases can be obtained in the microstructure [9]. Aluminum addition enhances the mechanical properties of NAB alloy via solid solution strengthening [10]. Nickel increases the yield strength,

Corresponding author. E-mail address: [email protected] (C. Dharmendra). 1 Dr. G.D. Janaki Ram was a Harrison McCain Visiting Professor at the University of New Brunswick in Fredericton, New Brunswick, Canada during the completion of this work. ⁎

https://doi.org/10.1016/j.addma.2019.100872 Received 7 March 2019; Received in revised form 10 September 2019; Accepted 11 September 2019 Available online 13 September 2019 2214-8604/ Crown Copyright © 2019 Published by Elsevier B.V. All rights reserved.

Additive Manufacturing 30 (2019) 100872

C. Dharmendra, et al.

improves corrosion resistance, and also acts as microstructure stabilizer [9]. The addition of iron causes grain refinement and also increases the yield strength [11]. Manganese addition improves the castability and approximately 1 wt.% manganese is usually added to NAB alloy to retard the β-phase transformation. This amount of manganese is considered as equivalent to 0.15% Al on the phase diagram [9]. Unlike many non-ferrous alloys, NABs can undergo a wide range of solid-state phase transformations during solidification resulting in the presence of numerous phases at room temperature. The absorption of aluminum from the matrix by κ-phases extends the apparent range of the α-field. As a result, under equilibrium conditions, no eutectoid formation occurs and β is not retained below 600 °C unless the aluminum content goes beyond 11%, compared to 9.5% in the Cu-Al binary system. The precipitation of κ-phases in the α-matrix increases the mechanical strength considerably without a significant reduction in ductility. This has been the most outstanding advantage of NABs over other aluminum bronzes. Various types of κ-phases were catalogued in the literature [6,12,13] based on their constitution and morphology, and are classified or designated as κI, κII, κIII, and κIV in the order of their precipitation during cooling from high temperatures. Apart from α and κ-phases, some amount of martensitic β (β′) may appear in the as-cast microstructures of Cu-9Al-4Fe-4Ni-1 Mn alloy, but it is uncommon in sand castings because of the relatively slow cooling rates. In general, casting defects such as segregation and porosity typically diminish physical properties, mechanical properties, and service performance. Due to long lead times and the complication in manufacturing large structures such as ship propellers, parts and components made of NAB are generally expensive. Hence, weld repair is usually preferred for extending the service life instead of replacing the parts. However, NAB alloys have some weldability issues such as porosity, hot cracking, and distortion [14] due to copper’s high thermal conductivity and coefficient of thermal expansion [15]. Friction stir processing (FSP) was used for cast-NAB as a weld repair technique [16], which enhanced the strength due to different strengthening mechanisms such as grain refinement, solid solutions, precipitation, and dislocation hardening. However, the formations of nano-level twins that are rich in martensite phases (β′) were observed in friction stir processed NAB alloys [16], which had a detrimental effect on the corrosion resistance. To improve the corrosion resistance, a post heat treatment is essential, which further increases the cost of production. In addition to heat treatment [17–19], some attempts were made to improve the mechanical properties and corrosion resistance of NAB alloys through changes in composition [20,21], plastic deformation [22–24], and surface treatment techniques such as friction stir processing [25], nickel ion implantation [26], thermal diffusion of Ni coating [27], laser surface melting [28], and laser surface alloying [29]. The limitation with surface modifying treatments is that they can only improve the corrosion resistance and mechanical properties of the surface layer instead of the bulk material. Heat treatment and plastic deformation techniques such as FSP have some limitations and are hard to use to process heavy castings such as ship propellers. Overall, NABs are metallurgically complex alloys in which small variations in the compositions can result in the formation of markedly different microstructures, which can, in turn, result in significant variations in seawater corrosion resistance [8]. In comparison with the conventional casting processes, superior mechanical properties can be achieved by additive manufacturing (AM) that results in certain microstructural benefits such as reducing segregation, uniform distribution of second phase particles, and finer grain sizes [30]. In addition, the defective and service-damaged components can be replaced by fabricating the required parts offshore or on a vessel instead of waiting for a long time to schedule and receive the new cast component. Among the various classes of AM, powder bed fusion techniques (Selective Laser Sintering, Direct Metal Laser Sintering, Selective Laser Melting, Electron Beam Melting) and powder metal deposition processes (Direct Metal Deposition, Lasform, Controlled Metal

Buildup, Laser Engineering Net Shaping) are the most suitable for the fabrication of small metallic parts [31]. In recent years, wire-arc additive manufacturing (WAAM) techniques are becoming attractive for the manufacturing sector due to their ability to produce large size components with full density, high deposition rate, high material utilization, and low equipment cost [32]. WAAM does not need a vacuum environment to operate as required in electron beam based processes [33] and the use of an electric arc offers a higher efficiency fusion source than laser-based methods [34]. WAAM process was employed as a fabrication process for various engineering materials such as Ti alloys [35–38], Al alloys [39], Ni alloys [40], and steels [41,42]. Wire-based deposition produces stronger adhesion of deposited layers as the wire melts fully, where geometry restrictions are quite minimal with WAAM. As this process uses a wire instead of an atomized powder like in other AM processes, WAAM can also be used for repair purposes where wires of matching composition can be used to deposit the material on damaged parts that have complicated or complex chemical compositions. In WAAM process, an electric arc is used as a heat source and employs either gas metal arc welding (GMAW) [43], gas tungsten arc welding (GTAW) [44] or plasma arc welding (PAW) [45] to melt the wire as the feedstock. GTAW and PAW processes are extremely sensitive to the arc length, and the wire is not fed coaxially in those two technologies which lead to variations during the process with a change in the welding direction [46]. This limits the applicability of those two processes as a rotary axis is required in the robotic welding systems to orient the wire-feed nozzle to match with the welding direction. Coaxial nature of GMAW process reduces machine complexity and provides high flexibility when combined with a positioning system such as a robot. To the author’s knowledge, at present, there has been no work reported on additive manufacturing of NAB alloys using other than the WAAM process. In an earlier work [47], WAAM process was used to deposit NAB on as-cast base metal in the form of straight thin walls of 10 mm width with typical length and height of 100 mm and 40 mm, respectively, manufactured under three different heat inputs (653, 874, and 1114 J/mm) for deposition. In another study [48], a bulk NAB alloy cubic sample (100 × 100 x 100 mm) was built up on the as-cast NAB alloy (substrate) by using the robotic WAAM system, and the influence of orientation on the mechanical properties was studied. As WAAM process is an inherently non-equilibrium thermal process, it is important to study its underlying physical metallurgical mechanisms to provide a guidance for the process optimization and improvement, and to adjust the process parameters according to target material characteristics. The subject of this paper is to employ lower heat input (170 J/mm) to deposit NAB alloy (Cu-9Al-4Fe-4Ni-1Mn) in form of square bars on 316 L stainless steel substrate using WAAM process, and to characterize the microstructure evolution and mechanical behavior. The objective of the present study is to identify the morphology and crystal structure of the phases, understand the microstructural development during deposition, and to correlate the microstructure with the mechanical properties. This study also intends to compare the microstructure evolution in NAB during WAAM with its typical cast microstructure as a reference. 2. Experimental work 2.1. Materials and methods A reference cast-NAB sample for microstructural characterization and mechanical testing was sectioned from a NAB salt-water impeller (a brackish water pump) that was 170 mm in diameter. The cast NAB specification corresponded to UNS C95800 and the nominal composition (in wt.%) of this material was: Cu: 79% minimum, Al: 8.5–9.5%, Ni: 4–5%, Fe: 3.5–4.5%, Mn: 0.8–1.5%. Berlin (Germany)-based GEFERTEC has invented a new-patented technology called 3D Metal Printing (3DMP) process [49] that uses a metal wire feedstock. Their new GTarc 60-5 WAAM equipment was 2

Additive Manufacturing 30 (2019) 100872

C. Dharmendra, et al.

standard metallography methods to a 1 μm finish, and then etched for 30 s using Klemm’s reagent. The microstructures of the samples were examined using Zeta-20 optical microscope and JEOL JSM6400 scanning electron microscope (SEM) fitted with an energy dispersive spectroscopy (EDS). Electron probe micro-analyzer (EPMA) characterization technique was used to identify the elemental distribution of the castNAB alloy. The individual elemental distribution maps were produced in an SEM (JEOL 733 Microprobe located at UNB’s Microscopy and Microanalysis facility) to identify the presence of elements in the matrix and in the second-phase (or intermetallic) particles. The WAAM-NAB samples were prepared by mechanical polishing to a thickness of 30 μm followed by further thinning via ion-beam milling using a Gatan Precision Ion Polishing System (Gatan model 695) operating at an accelerating voltage of 200 kV for transmission electron microscope (TEM) study. For characterization of phases present in the microstructure, FEI Tecnai Osiris TEM equipped with a 200 keV X-FEG gun was used. High angle annular dark field (HAADF) detectors were used in combination with EDS to generate elemental maps; where spatial resolutions in the order of 1 nm were obtained using a subnanometer electron probe. Selective area diffraction (SAD) and convergent beam electron diffraction (CBED) was used to obtain patterns from the matrix and intermetallic particles, respectively. Electron backscatter diffraction (EBSD) analysis was performed on WAAM-NAB samples along x–z and x–y planes (marked in Fig. 1). Field emission gun scanning electron microscope (FEG-SEM-FEI Nova NanoSEM-650) with an OIM 6.2 EBSD system (EDAX) was used to obtain the EBSD scan. The scans at low magnification (500×) were run over an area of 500 μm × 500 μm with a step size of 0.25 μm. At high magnification (2000×), the scan was performed on an area of 75 μm × 75 μm using a step size of 0.05 μm.

Table 1 WAAM process parameters for deposition of NAB. Wire feed rate

Travel speed

Voltage

Current

Heat input

6.5 m/min

480 mm/min

12.5 V

114 Amp

170 J/mm

used to fabricate all samples in this study. The machine uses gas-metal arc welding (GMAW) technology, in which a wire passes through the feedstock, is melted, and then deposited in successive layers. The GTarc 60-5 offers five motion axes and can produce metallic parts up to 0.8 m3 volume with 500 kg maximum mass. The machine is equipped with a SIEMENS control unit to correctly position the welding head enabling high levels of precision to handle CAD models. Nickel aluminum bronze filler wires corresponding to AWS A5.7 ERCuNiAl (nominal composition same as that of cast-NAB) were used in the WAAM process to produce the samples. The NAB was printed at wire feed rate of 6.5 m/min, travel speed of 480 mm/min, and voltage of 12.5 V at 114 A current. These deposition parameters are listed in Table 1 along with the heat input. Relatively low value of 170 J/mm was obtained with an assumption on process efficiency as 95%. If the efficiency drops, heat input on the samples decreases to much lower values. Fig. 1 shows the bulk NAB alloy square bars that were printed in vertical direction (160 mm height, two other sides are 25 mm) on stainless steel 316 L plate of 10 mm thick. In total, 60 layers were deposited making each layer thickness about 2.5 mm. 2.2. Microstructural characterization The samples of cast- and WAAM-NAB were cut from the cross-sectional plane and were mounted, mechanically ground, and polished by

2.3. Mechanical testing Mechanical properties were evaluated under quasi-static uniaxial tensile loading for both WAAM- and cast-NAB at room temperature.

Fig. 1. Photograph of nickel aluminum bronze square bars produced by WAAM. Inset shows the dimensions (not to scale; side: 25 mm and height: 160 mm) of the schematic bar. Blue rectangular box in the inset shows the location of the sample used for microstructural characterization. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 2. (a) Optical and (b) Secondary electron-SEM micrographs of cast-NAB sample. 3

Additive Manufacturing 30 (2019) 100872

C. Dharmendra, et al.

Sample dimensions for tensile testing corresponded to ASTM E-8 standard [50]. Cylindrical Samples for WAAM-NAB were machined such that the tensile axes were parallel to the print direction (z-axis in Fig. 1). A computer-controlled Instron model 1332 universal testing machine was used to conduct the uniaxial tensile tests. This machine was accompanied by a 25 mm Instron extensometer to accurately obtain displacement during the tensile test. Samples were pulled to failure at a constant cross head displacement rate (0.001 mm/min) and the data for engineering stress was simultaneously recorded throughout the test. Tests were carried out on multiple (3–5) samples to confirm repeatability of the tests, where typical tensile engineering stress-strain curve was plotted. Fracture surfaces were characterized using a JEOL 6400 SEM operating with a tungsten filament. In addition, optical

microstructural examination was performed on a plane normal to the tensile axis for further investigation of the fracture mechanism. 3. Results 3.1. Microstructures of cast-NAB Microstructural examination (Fig. 2) of the cast-NAB samples revealed a number of intermetallic particles dispersed in a matrix of relatively coarse α-dendrites. As can be seen in the SEM micrograph shown in Fig. 2(b), the intermetallic phases consisted of κI in rosette morphology, κII in the form of small roughly globular particles, κIII in lamellar morphology, and κIV in the form of numerous fine particles

Fig. 3. (a) SEM-Backscattered electron (BSE) image of cast-NAB, and corresponding EPMA elemental maps for (b) Cu, (c) Al, (d) Ni, and (e) Fe.

4

Additive Manufacturing 30 (2019) 100872

C. Dharmendra, et al.

precipitated homogeneously within the α-dendrites. It may be noted that, intermetallic phases κI to κIII are mainly present in the interdendritic regions. No evidence of retained β or acicular β′ martensite is observed in the as-cast samples. These microstructural features are very typical of a NAB alloy in the as-cast condition [6,13]. The precipitates κI to κIII primarily appear in the interdendritic regions due to the interdendritic segregation of alloying elements during solidification, while the κIV intermetallic homogeneously precipitates within the primary αdendrites. In the microstructural examination, the κI appeared as microsized rosette-shape particles as they form during solidification. On the other hand, the κII precipitates are seen as essentially finer, globular particles. The κIII precipitates exhibit a distinct lamellar morphology

because of their formation as a result of eutectoid decomposition of β phase to α + κIII. Finally, the κIV precipitates as numerous sub-micronsized globular particles homogeneously distributed within the α-dendrites. It may be noted that precipitates κI to κIV are essentially intermetallic phases. The results of EPMA elemental mapping studies on cast-NAB are shown in Fig. 3. In the as-cast sample, the dendrite cores can be seen to be relatively lean in Al, Ni, and Fe, whereas the interdendritic intermetallic phases were rich in Ni and/or Fe as well as Al. While κI, κII, κIV precipitates are rich in iron and aluminum, κIII precipitates are rich in nickel and aluminum. However, some amount of nickel is typically measured in κI, κII, κIV precipitates as is some iron in κIII precipitates. 3.2. Microstructures of WAAM-NAB A low magnification optical micrograph at a melt pool boundary of a WAAM-NAB sample is shown in Fig. 4(a). The bulk of the sample consisted of very fine α-dendrites with some dark-etched interdendritic regions, as can be seen in Fig. 4(b). Below each melt pool boundary, as marked in Fig. 4(a), a region of coarser microstructure in varied thicknesses up to 500 μm was observed. These coarser microstructural regions also consisted of bright-etched α-dendrites and dark-etched interdendritic regions, as can be seen clearly in Fig. 4(c). As in the ascast NAB samples, no β′ martensite was noticed in the WAAM samples. However, unlike in the as-cast NAB samples, optical microscopy did not reveal any intermetallic phases in the WAAM samples. Fig. 5 shows the SEM image of the WAAM-NAB at higher magnifications, which revealed some fine globular κII particles in the interdendritic regions. These fine κII particles precipitated in the dark etched regions seen in Figs. 4(b) and (c). Examination of the WAAM samples in TEM (Fig. 6) more clearly revealed the presence of various intermetallic phases in the interdendritic regions. In Fig. 6(a) and (b), α-dendrites and interdendritic regions can be clearly seen. The interdendritic regions displayed twotypes of intermetallic phases, κII in globular form and κIII in lamellar morphology in between fine α-lamellae (Fig. 6(b) and 6(c)). These intermetallic phases were found to be Fe3Al and NiAl, respectively, based on the diffraction pattern analysis as shown in Fig. 7. While some

Fig. 4. (a) Optical micrograph of WAAM-NAB sample in the bulk at a melt pool boundary (denoted by yellow dotted lines). The boxed regions marked (B) and (C) are shown at a higher magnification in (b) and (c), respectively. Note light and dark etched regions in (b) and (c). The primary α-dendrites are in light contrast. The dark etched regions are unresolved at optical level. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).

Fig. 5. SEM micrograph of WAAM-NAB sample at a region corresponding to the marked boxed region B in Fig. 4a. Note fine κII precipitates in the dark etched regions seen in Fig. 4(b) and (c).

5

Additive Manufacturing 30 (2019) 100872

C. Dharmendra, et al.

Fig. 6. STEM-HAADF images of WAAM-NAB: (a) and (b) show the general microstructure with interdendritic regions and primary α-dendrites, (b) and (c) show interdendritic α-lamellae with lamellar and globular intermetallic phases.

amount of Fe or precipitation of κII in NiAl precipitate is noticed (Fig. 7(b)), some Fe3Al (κII) particles (shown in Fig. 8) seem to be free from any NiAl precipitation. From Figs. 7(b) and 8, both NiAl and Fe3Al particles appear to exhibit semi-coherency with the matrix as indicated by misfit dislocations observed at their interfaces with the matrix. There is an Burger’s orientation relationship (lattice correspondence relation) between Fe3Al and surrounding NiAl, where [001]Fe3Al // [010]NiAl. Further, EDS elemental mapping (Fig. 9) studies also distinctly revealed the Fe-rich κII globular particles and Ni-rich κIII in lamellar morphology. The interdendritic regions were rich in Al, Ni, and Fe. The nucleation of κII occurs at relatively high temperatures in the β-phase with precipitating α-phase envelopes around it during cooling [6]. κII particles are in various ranges of sizes, mostly very fine to the size of

few nanometers (Figs. 6(b) and 7 (b)), and precipitated mainly in the regions of lamellar eutectoid mixture (mainly κIII). κII precipitates (Fe3Al, Iron aluminide) typically have DO3 crystal structure at low temperatures with a lattice parameter of 5.71 ± 0.06 Å [13]. κIII particles that are based on NiAl belong to B2 (ordered BCC) crystal structure with a lattice parameter of 2.88 ± 0.03 Å [13]. It must be noted from the EDS elemental map corresponding to Fe (Fig. 9(f)) that, the matrix contains a distribution of fine precipitates throughout the α-grains. To reveal them more clearly, TEM imaging (Fig. 10(a)) was done at a very high magnification of 630000x and corresponding EDS elemental maps for Al, Ni, and Fe are shown in Figs. 10(b)-(d), respectively. The κIV precipitates are in 5 to 10 nm in size, and believed to have a composition and crystal structure similar to

6

Additive Manufacturing 30 (2019) 100872

C. Dharmendra, et al.

Fig. 7. TEM-Bright field micrographs showing (a) Cu-matrix and corresponding SAD pattern in [001] zone axis of α with superlattice spots from the κIV precipitates, (b) the globular Fe3Al precipitates (κII) with corresponding CBED pattern taken at [111] zone axis, and (c) the lamellar NiAl (κIII) with corresponding CBED pattern taken at [1¯ 3¯ 1] zone axis.

κII based on Fe3Al [13]. Superlattice reflections from these precipitates can be noticed in diffraction pattern that corresponds to α-matrix in Fig. 7(a). Fig. 11 shows the EBSD microstructural and crystallographic characterization of WAAM-NAB along the build direction (z-axis). Fig. 11(a) shows the inverse pole figure (IPF-Z) map that reveals the grain structure (high and low angle grain boundaries), without any pronounced grain elongation along the build direction. This is in contrast to the characteristic of the directional growth along the build direction in AM processes [51]. The dotted lines in the IPF map show the orientation of grains that are elongated along the melt pool boundaries. Grains that

appear somewhat bigger in the map seem to be dominated by green color i.e. in < 101 > orientation. Rest of the grains are oriented either on < 001 > or < 111 > direction. Grain boundary misorientation map (Fig. 11(b)) reveals that high-angle grain boundaries are predominant. It can also be observed in Fig. 11(a) that the color associated with crystallographic orientations within individual grains is uniform. This may be indicating that low local stresses and low dislocation densities (less population of low-angle grain boundaries in Fig. 11(b)) are present in the WAAM-NAB. The average grain size is estimated to be 16.4 μm from the grain size distribution plot in Fig. 11(c). Fig. 11(d) shows the texture analysis of the EBSD data in the form of pole figures. It

7

Additive Manufacturing 30 (2019) 100872

C. Dharmendra, et al.

Fig. 8. TEM-Bright field micrograph showing Fe3Al precipitate (κII that is not associated with NiAl lamellae) by an arrow with corresponding CBED pattern taken at [001¯] zone axis.

demonstrates that, the WAAM-NAB had a weak to moderate texture denoted by the locations of the peaks in (001) tilted away from the build direction (y-axis in pole figures) and the four on-axis peaks in the (111) map. The maximum texture intensity is 1.85 times the random. The evolution of the texture may be related to the grain formation during repeated heating and cooling cycles, and strongly dependent on the heat input [52]. Fig. 12(a) shows the EBSD inverse pole figure (IPF) map of a WAAM-NAB sample on the section (x–y plane) perpendicular to the build direction (z-axis). The map shows only random orientations, similar to the texture of an as-cast material (mostly random texture). The dotted line indicates the melt pool track and somewhat larger dendrites are oriented towards < 101 > direction, similar to the EBSD scan on plane parallel to the build direction. The grain boundary misorientation map representing the grain-to-grain misorientation angle distributions in the α-phase is shown in Fig. 12(b). The distribution of misorientation angles shows a distinct random component and the population of the high-angle boundaries (> 15°) is the highest. From the grain size distribution plot (Fig. 12(c)), the average grain size measured was about 17.4 μm. Fig. 12(d) shows the pole figures measured on the x–y plane. Though the maximum texture intensity is the same to that on the x–z plane, changes in the rotations of (001) and (111) poles are distinctly visible. Changes in the direction of wire-feed deposition for two successive layers appear to modify the solidification conditions enough to cause low thermal gradients and to inhibit the development of a texture. The random texture is indicative of the solidification of the equiaxed grains that was promoted by heterogeneous nucleation in the presence of fine κIV particles in the matrix. To see the grain shapes more clearly, high magnification (2000x) EBSD images (IPF and grain size distribution) are presented in Figs. 12(e) and (f), where most of the αgrains are elliptical in shape. Networks of fine equiaxed grains (typically about 5 μm) are evolved in the interdendritic regions in random orientations. Overall, when compared to the as-cast NAB samples, the WAAM samples showed a distinctly finer microstructure without any κI intermetallic phases. The WAAM samples did contain κII, κIII, and κIV phases, but their amounts and sizes were significantly lower than in the cast condition.

are summarized in Table 2. As can be seen, the WAAM samples were found to exhibit considerably higher yield strength (YS) and percent elongation as compared to the as-cast samples. However, the increase in the ultimate tensile strength (UTS) was only marginal. Examination of the fracture surfaces revealed ductile fracture features in both cases, as can be seen in Fig. 14. In the as-cast samples, especially, some regions of flat fracture were seen occasionally (Fig. 14(b)). Examination of the microstructural features close to the fracture location of the WAAM samples (Fig. 15) revealed no preferential fracture along layer interfaces or in the areas (heat-affected zones (HAZ)) adjacent to the meltpool boundaries. 4. Discussion 4.1. Solidification characteristics Solidification in cast Cu-9Al-4Fe-4Ni-1 Mn alloy occurs by the formation of α-dendrites, where it may also involve a peritectic reaction L + α →β [53]. Some pre-primary Fe-Al intermetallic particles can also form directly from the melt, especially under relatively slow cooling conditions, and are known to aid in the nucleation of α-crystals [54]. During solidification, the alloying elements show a strong tendency to segregate into the liquid and the last-to-solidify liquid develops considerably higher concentration of solute elements. As a result, some rosette-like κI particles generally form in the interdendritic regions during cooling to about 1040 °C. It may be noted that, the severity of segregation of the alloying elements depends on the cooling rate and faster cooling helps to minimize the segregation if it is beyond the critical rate of 101 K/s [55]. After solidification, the alloy successively passes through α + β, α + β+κ, and α + κ regions on the cooling path to the room temperature. The β phase essentially forms in the interdendritic regions. Some precipitation of relatively coarser κII globular particles forms at high temperatures as the alloy cools through the α + β+κ region. Here again, the amount of precipitation in the form of κII depends on the cooling rate, but typically occurs at about 930 °C. Subsequently, as the alloy cools into the α + κ region, a eutectoid reaction β → α + κ takes place (at about 800 °C) [8]. Unless the cooling rate is very high, the β phase is entirely consumed in the eutectoid reaction. The eutectoid mixture exhibits a characteristic lamellar structure and consists of a nickel-rich κIII phase in between α-lamellae. The scale of the eutectoid mixture is also dependent on the cooling rate. Finally, as the alloy cools further, precipitation of very fine κIV particles disperse homogeneously in the primary α-dendrites. It may be noted

3.3. Mechanical properties Fig. 13 shows the typical tensile stress-strain curves of the cast- and WAAM-NAB samples. The results of tensile testing on multiple samples

8

Additive Manufacturing 30 (2019) 100872

C. Dharmendra, et al.

Fig. 9. (a) STEM-HAADF images of WAAM-NAB in an interdendritic region, (b) composite map of the main elements, (c)-(f) show the corresponding EDS elemental maps for Cu, Al, Ni, and Fe, respectively. The lamellar precipitates are NiAl (κIII) and the fine globular precipitates are Fe3Al (κII) that are formed in the core of the NiAl (κIII) precipitates. Note the precipitation of Fe-rich κIV in the matrix in (f).

that, the eutectoid reaction can get fully suppressed if the cooling rates are very high. In such a case, the β phase undergoes a martensitic transformation (diffusion-less) leading to some acicular β′ martensite in the room-temperature microstructure. If the diffusion-less transformation occurred, all phases (κII and κIV) consist of the same chemical

composition [13] though they differ in morphology. However, under equilibrium or slow cooling conditions, both κII and κIV particles correspond to the same phase, albeit with slightly different composition, which may be attributed to different temperatures for precipitation during cooling. 9

Additive Manufacturing 30 (2019) 100872

C. Dharmendra, et al.

Fig. 10. (a) STEM-HAADF image of WAAM-NAB in α-matrix region taken with the beam parallel to [001]; (b)-(d) show corresponding EDS elemental maps for Al, Ni, and Fe, respectively. Fe-precipitates (κIV) are in size of 5–10 nm. Note that thickness of TEM thin specimen is ˜100 nm.

solute atoms becomes very difficult not only in solid state, but also in liquid phase.

As compared to conventional casting, the cooling rates during WAAM are significantly higher (around 102 °C/s) [30]. This has a strong bearing on the microstructural development in alloy Cu-9Al-4Fe4Ni-1 Mn. The high cooling rates in WAAM results in a significantly finer solidification structure as compared to conventional casting. WAAM process is identical to that of multi-pass welding as arc (i.e. heat source) locally melts the NAB as it passes, and the deposited layer solidifies before the upper portion of it partially re-melts again during the next scan for successful bonding between the layers. The lower portion below the bottom of the melt pool drives the solidification process and thus, the microstructural evolution. In general, under slow cooling rates, solute atoms have enough time to diffuse, where micro-segregation does not occur. At relatively high cooling rates (non-equilibrium conditions), diffusion of solute atoms into solid get suppressed, which leads to interdendritic micro-segregation. However, if the cooling rate is further increased to critical cooling rates such as in WAAM, the velocity of the growing dendrite increases and approaches its atomic diffusive speed [56]. In such situation of solute trapping, diffusion of

4.2. κ-Precipitates In WAAM process, the high cooling rates in newly deposited layer minimize the thermal gradient with lower portion of the layer beneath it. This can reduce the segregation of alloying elements during solidification and suppress the formation of rosette-like κI particles in WAAM samples. Even if some precipitation occurs in the upper portion of the layer, those precipitates get dissolved, when that portion reheated to solidus temperature again during subsequent deposition of the next layer. It may be important to note that, specifying a single cooling rate value for the whole AM deposition process is likely inappropriate as it is a strong function of temperature and the location at which it is measured as each layer experiences multiple thermal cycles and temperature peaks. In AM, typically, two different temperature ranges are important to consider. Firstly, cooling rate from the liquidus to solidus

10

Additive Manufacturing 30 (2019) 100872

C. Dharmendra, et al.

Fig. 11. EBSD results from a WAAM-NAB sample at 500x on plane (x–z) parallel to the build direction (z-axis, which is vertical in the images): (a) inverse pole figure (IPF) map, (b) grain boundary misorinetation map, (c) grain size distribution plot, and (d) pole figures. Dotted lines in (a) indicate melt pool boundaries. The build direction is marked as BD in pole figures.

temperature range, which determines the features of solidification microstructures such as cells and dendrites [57], and another cooling rate is during the solid-state phase transformations. Perhaps, the most wellknown range for NABs is the 850 to 500 °C, where the microstructure is significantly affected by the cooling rate [8,13,58]. While the κI particles contribute very little to the strength, they have a marked detrimental effect on the ductility and toughness of the alloy [8]. Hence, the absence of coarse κI particles in WAAM samples is a significant advantage. The TEM diffraction data shown in Figs. 7 and 8 indicate that there are two structurally distinct κ-phases in the interdendritic region; one (κII) based on Fe3Al and the other (κIII) based on NiAl. This agrees with the findings of Weill-Couly et al. [59] that κII particles are Fe-rich, while the lamellar eutectoid decomposition product κIII is Ni-rich. There is no comprehensive information in the literature on the phase relationships

in the Cu-Al-Ni-Fe quaternary system (i.e. ternary phase diagrams at moderate or high cooling rates) to comment on the potential composition range of the κII and κIII intermetallic phases. The size of κII precipitates (in few nanometers) in WAAM-NAB is significantly lower than that of the cast-NAB (5–10 μm). In cast-NAB, it was reported that both κII (Fe3Al) and κIII (NiAl) particles exhibit Kurdjumov-Sachs (K–S) orientation relationships with the α-matrix (Cu) [13], which means that the close-packed planes in the two structures (precipitate and matrix) are parallel or nearly-parallel. The subsequent evolution of texture during solidification depend on this crystallographic orientation relationships. Fine Fe-rich κIV particles uniformly precipitated throughout the αgrains. However, their size is very small to the level of 5–10 nm. Their shape is neither acicular nor globular, and appears likely to be rhombohedron from Fig. 10(d). Short available time for diffusion under fast

11

Additive Manufacturing 30 (2019) 100872

C. Dharmendra, et al.

Fig. 12. EBSD results from WAAM-NAB on x–y plane (perpendicular to build direction i.e. z-axis): (a) inverse pole figure (IPF) map at 500x, (b) grain boundary misorinetation map, (c) grain size distribution plot, (d) pole figures, and (e) IPF map at higher magnification (2000x) and corresponding (f) grain boundary misorientation map. Dotted line in (a) represents for scan track. The build direction is marked as BD in pole figures.

12

Additive Manufacturing 30 (2019) 100872

C. Dharmendra, et al.

decomposition of the β phase into α + κIII phases takes place. Because the reheat thermal cycle involves a slower cooling rate as compared to the cooling rate in a newly deposited weld bead, the κII precipitates as well as the eutectoid mixture in the HAZ may appear slightly coarser. It may be noted that reheating to temperatures below the β solvus, but well into the α + β+κ phase field, can also result in noticeable coarsening of the microstructure. In this case, the κII precipitates do not undergo any dissolution (some coarsening can be expected instead), where the reverse eutectoid reaction (α + κIII → β) takes place. During cooling, the β phase once again undergoes transformation to a coarser α + κIII lamellar mixture. In any case, the microstructural changes occurring in the HAZ of alloy Cu-9Al-4Fe-4Ni-1 Mn are not seriously detrimental as the HAZ bands are discontinuous. 4.3. Microstructure evolution Low heat input employed in the WAAM process might have increased the cooling rate, prevented the significant grain growth, and yielded to fine grains as evidenced in the EBSD maps (Figs. 11 and 12). Although the scale (fine/coarse) of microstructure is noticeably different (Fig. 4), the phase constitution (i.e. presence of κII and κIII) in the HAZ and in the unaffected weld metal regions is essentially the same. From the EBSD IPF map in Fig. 11(a), it may be noted that the region between the melt pool boundaries exhibit slightly finer grains. The presence of fine grains in alternating bands is observed more clearly in the optical microscopy, which is presented in Fig. 4. Texture formation in WAAM-NAB showed that the texture was almost random. This may be attributed to the nucleation, which lead to the formation of randomly oriented nuclei of the solid phase. The growth of some of the nuclei, however, takes place by selecting a specific orientation, here probably < 001 > , thereby produced somewhat sharp texture in the (100) pole figure (Fig. 11(d)). It is noted that, < 001 > orientation, parallel to the build direction, is the easiest growth direction for the columnar grains in most FCC metals and alloys [61]. {001} is a less close-packed plane, which can provide more comfortable locations for the moving atoms in the liquid phase to adhere. Even in such scenario, texture on (001) for WAAM-NAB is not very significant. Development of solidification texture might be controlled due to the low heat input employed and the corresponding low thermal gradients. EBSD images at high magnification (Fig. 12(e) and (f)) clearly revealed the grain structure evolution of WAAM-NAB. The grain size is finer than the castNAB (shown in Fig. 2) and small equiaxed grains nucleated on the grain boundaries of primary-α, which are interdendritic regions. Nucleation of these fine grains is attributed to the heterogeneous nucleation from the globular κII particles and also to the higher cooling rates where shorter growth times are available for the growth of the nucleated grains.

Fig. 13. Typical tensile stress-strain plots of cast and WAAM samples.

Table 2 Tensile test results of NAB for cast and WAAM samples. Material

Yield Strength (MPa)

Ultimate Tensile Strength (MPa)

% Elongation

Cast-NAB WAAM-NAB

300 ± 15 388 ± 12

670 ± 8 685 ± 10

16 ± 2 26 ± 1

cooling rates of WAAM process probably restricted the solid-state transformation of κIV. These κIV particles contribute to strength significantly. It may be possible to induce satisfactory increase in their volume fraction and size in WAAM samples by conducting an annealing heat treatment in the temperature range of 500 to 600 °C. Further work is needed in this regard. It is obvious as to why the cooling rates in WAAM are high enough to suppress the formation of κI precipitates, but not so in the case of κII, κIII, and κIV precipitates. It is known that, the rate at which the weld metal cools gradually drop off with temperature [60]. Therefore, any precipitation events in both high and low temperature regimes are more likely to get suppressed during weld metal cooling. While high cooling rates are responsible for suppression of precipitation reactions at high temperatures (κI formation in the present case), the cooling rates are not typically high enough and the diffusivities are not typically low enough to suppress any precipitation reactions in intermediate temperature ranges. For this reason, the formation of κII, κIII, and κIV precipitates was not suppressed in the WAAM-NAB and solid-state phase transformations did occur. During WAAM, each layer is subjected to significant reheating when the next layer is being deposited. The peak temperatures experienced in the reheated region decrease as a function of distance away from the fusion boundary. The region that (re)heated above the phase transformation temperature undergoes a noticeable microstructure change, which is seen as the heat-affected zone (HAZ) underneath each weld bead (Fig. 4(a)). Normally, the HAZ of each weld bead appears thicker at the bottom of the weld bead as the temperatures would be the highest at the center of the weld pool. In alloy Cu-9Al-4Fe-4Ni-1 Mn, the HAZ can be expected to experience peak temperatures above the β solvus temperature (˜830 °C) [8]. Consequently, any κII, κIII, and κIV precipitates in the reheated region get partially or completely dissolved during the heating part of the reheat thermal cycle. Further, some coarsening of α and β regions can take place (Fig. 4(c)). During cooling, the κII precipitates initially form in the β regions and then eutectoid

4.4. Mechanical properties As noted in Section 3.3, the WAAM samples in as-fabricated condition exhibited considerably higher yield strength as compared to the as-cast samples. The increase in yield strength in WAAM samples can be attributed to the finer and homogeneous solidification structure (grain boundary strengthening), presence of fine κIV precipitates, and the absence of κI that is detrimental for mechanical properties. It may be noted that the strength levels displayed by the WAAM samples are particularly impressive given that the samples did not contain significant volume fraction of κIV strengthening precipitates in as-fabricated condition. Finer microstructural features, such as finer cell or dendrite spacing as observed from the EBSD images, enhanced yield strength of the WAAM-NAB significantly. In the perspective of NAB component engineering design, yield strength is more relevant than UTS [7,8]. The WAAM samples may prove to be even more superior to

13

Additive Manufacturing 30 (2019) 100872

C. Dharmendra, et al.

Fig. 14. SEM fractographs of (a and b): Cast, and (c and d): WAAM samples. Encircled region in (b) shows flat fracture features due to the presence of coarse intermetallic phase.

conventional castings after a simple annealing treatment in the temperature range of 500 to 600 °C [8], which would result in homogeneous κIV precipitation in the α-matrix. The WAAM samples that are deposited under low heat input showed significantly finer and lower volume of intermetallic second-phases, and the absence of κI as compared to the cast samples. As a result, the WAAM samples exhibited significantly higher tensile ductility than the as-cast samples. Fractographic examination as well as microstructural examination close to the location of fracture in WAAM samples did not show any cracking features such as micro-void coalescence or secondary cracks. The gain in the tensile ductility was very striking in spite of the fact that the WAAM samples were tested in the z-direction (layer interfaces oriented perpendicular to the loading direction), which is the orientation in which additive manufactured samples typically show their lowest tensile ductility [30,47,62]. The strength and ductility levels of as-built WAAM-NAB may be even higher than the cast alloy if it is deposited in the longitudinal or transverse directions. The current study shows that the nickel-aluminum bronze alloy Cu9Al-4Fe-4Ni-1 Mn can be readily used for part fabrication in WAAM

process without any serious concerns about weld metal solidification cracking or HAZ liquation cracking. The parts may be considered for use in as-fabricated condition as the tensile properties of WAAM samples easily meet the minimum properties specified in ASTM B148 for alloy C95800 in as-sand-cast condition [7]. If necessary, a stress-relieving treatment in the temperature range of 300–350 °C may be carried out for improving stress-corrosion cracking resistance. As mentioned earlier, an annealing treatment in the temperature range of 500–600 °C may be considered for increasing the strength of WAAM parts in alloy Cu-9Al-4Fe-4Ni-1 Mn. For best corrosion resistance in seawater, an annealing treatment at 675 °C for a minimum of 6 h is generally recommended for alloy C95800 sand castings. The precipitates including globular or rosette shaped κII and lamellar κIII are hard phases and less ductile. The improved corrosion resistance after annealing treatment is known to be due to globularization of the lamellar κIII and other intermetallic phases [63]. In future, it is appropriate to consider this heat treatment as well for WAAM parts in alloy Cu-9Al-4Fe-4Ni-1 Mn.

14

Additive Manufacturing 30 (2019) 100872

C. Dharmendra, et al.

WAAM tensile samples are tested along the normal direction (build direction). The ductile fracture mode was determined to be the main failure mode of both NAB alloys. 7 Heat treatments in the range of 500–600 °C may result in increasing the volume fraction of fine-κIV precipitates, which may increase the strength levels of WAAM-NAB. 8 Overall, the prospects of WAAM to produce NAB seem very bright and it can significantly widen the scope and applicability of additive manufacturing to produce NAB parts and components, mainly for the marine industry. Further work towards the effect of various heat treatments on microstructure, mechanical properties, and corrosion resistance would be highly beneficial. Declaration of Competing Interest The authors declare that there is no conflict of interest. Acknowledgements The Authors would like to thank Natural Sciences and Engineering Research Council of Canada (NSERC) grant number RGPIN-201604221, New Brunswick Innovation Foundation (NBIF) grant number RIF 2018-005, Atlantic Canada Opportunities Agency (ACOA)- Atlantic Innovation Fund (AIF) project number 210414, Mitacs Accelerate Program grant number IT10669 for providing sufficient funding to execute this work. The authors would like to acknowledge Alexander Riemann of Gefertec GmbH for fabricating the nickel aluminum bronze via wire-arc additive manufacturing technique, Brian Guidry and Vince Boardman at UNB-Mechanical Engineering for tensile samples preparation and mechanical testing, Dr. Douglas Hall and Steven Cogswell at UNB’s Microscopy and Microanalysis Facility for SEM and EPMA, Dr. Mark Kozdras at CanmetMATERIALS for facilitating the research, and Pei Liu and Catherine Bibby for TEM sample preparations and Dr. Jian Li for assisting with EBSD analysis.

Fig. 15. (a) Low and (b) high magnification optical micrographs at the location of fracture in a WAAM tensile specimen. No preferential cracking along layer interface or HAZ bands.

5. Conclusions The material chosen for this study was microstructurally complex. The intent of this work was to explore the feasibility of adopting wirearc additive manufacturing (WAAM) to produce Nickel Aluminum Bronze (NAB) alloy bars, and to study the morphology, crystallography, and the distribution of the phases present using optical, scanning electron microscopy, and transmission electron microscopy techniques. The microstructural features and mechanical properties of WAAM-NAB are compared with the cast-NAB. The findings of this study are as follows:

References [1] J.R.C. Strang, Cast valve materials for seawater service nickel-aluminum bronze and its rivals, Proceedings of the Valve World Conference (2006). [2] C. Powell, H. Stillman, Corrosion Behavior of Copper Alloys Used in Marine Aquaculture, International Copper Association, Ltd., Madison, New York, 2009. [3] E.A. Culpan, A.G. Foley, The detection of selective phase corrosion in cast nickel aluminum bronze by acoustic emission techniques, J. Mater. Sci. 17 (1982) 953–964. [4] R.J.C. Dawson, Aluminium Bronze, Eng. Mater. Des. 22 (12) (1978) 25–28. [5] V.A. Callcut, Aluminium bronze for industrial use, Met. Mater 5 (3) (1989) 128–132. [6] E.A. Culpan, G. Rose, Microstructural characterization of cast nickel aluminium bronze, J. Mater. Sci. 13 (1978) 1647–1657. [7] ASTM B148-18, Standard Specification for Aluminum-Bronze Sand Castings, ASTM International, West Conshohocken, PA, 2018, https://doi.org/10.1520/B0148-18 www.astm.org. [8] I. Richardson, Guide to Nickel Aluminum Bronze for Engineers, Copper Development Association Publication, Helmel Hempstead, 2016. [9] M. Cook, W.P. Fentiman, E. Davis, Observations on the structure and properties of wrought copper-aluminium-nickel-iron alloys, J. Inst. Metals 80 (1952) 419–429. [10] J.A. Wharton, R.C. Barik, G. Kear, R.J.K. Wood, K.R. Stokes, F.C. Walsh, The corrosion of nickel-aluminum bronze in sea water, Corros. Sci. 47 (2005) 3336–3367. [11] H.T. Michels, R.M. Kain, Effect of Composition and Microstructure on the Seawater Corrosion Resistance of Nickel-aluminum Bronze, Paper No. 03262, NACE Corrosion/2003, NACE International, Houston, Texas, 2003. [12] D.M. Lloyd, G.W. Lorimer, N. Ridley, Characterization of phases in a nickel-aluminium bronze, Met. Technol. 7 (1980) 114–119. [13] F. Hasan, A. Jahanafrooz, G.W. Lorimer, N. Ridley, The morphology, crystallography, and chemistry of phases in as-cast nickel-aluminum bronze, Metall. Trans. 13A (1982) 1337–1345. [14] J. Cheanvarin, The Case Study of Repairing the Propeller Blade of H.T.M.S. Srichang, Royal Thai Dockyard, Bangkok, 1994. [15] Welding of Aluminum Bronzes, Copper Development Association, Publication No. 85, 1988. [16] S.K. Menon, F.A. Pierce, B.P. Rosemark, K. Oh-Ishi, S. Swaminathan, T.R. McNelley, Strengthening mechanisms in NiAl bronze: hot deformation by rolling and friction stir processing, Metall. Mater. Trans. A 43 (2012) 3687–3702. [17] J. Anantapong, V. Uthaisangsuk, S. Suranuntchai, A. Manonukul, Effect of hot

1 WAAM process is suitable for additive manufacturing of NAB alloy with relatively low heat input (170 J/mm). In the current study, nearly full dense square bars (two) of 25 mm side and 160 mm height were successfully deposited in multiple layers with excellent layer bonding and no defects at the layer interfaces. 2 The process of WAAM significantly altered the microstructure. While the as-cast microstructure consists of copper-rich α-phase and all four types of κ-phases, κI phase is not precipitated in WAAM-NAB due to higher cooling rates. 3 In WAAM-NAB, κII and κIII phases are precipitated in the interdendritic regions whereas κIV is uniformly nucleated in the α-matrix. κII particles are globular and Fe3Al based; κIII is an eutectoid mixture in lamellar structure and NiAl based. 4 No retained β is observed in the WAAM-NAB. As the rate of cooling in weld metal drops off with decrease in temperature, eutectoid reaction in the intermediate temperature range is not suppressed, and the β-phase is entirely consumed and transformed to κIII. 5 WAAM-NAB exhibited fine equiaxed grain structure and relatively weak texture. The (001) poles, which is typically an easy growth direction for FCC structure tilted away from the build direction. There is no significant difference in the texture evolution between x–z plane and x–y plane. 6 Reheating associated with multi-layer deposition cause no detrimental microstructural changes in WAAM-NAB. It exhibits relatively finer microstructure, and show higher yield strength and percent elongation. The gain in room temperature tensile properties which may attribute to the absence of κI is quite impressive even when 15

Additive Manufacturing 30 (2019) 100872

C. Dharmendra, et al.

[18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30]

[31] [32] [33] [34] [35]

[36]

[37] [38]

working on microstructure evolution of as-cast nickel aluminum bronze alloy, Mater. Des. 60 (2014) 233–243. Z. Wu, Y.F. Cheng, L. Liu, W.J. Lv, W.B. Hu, Effect of heat treatment on microstructure evolution and erosion-corrosion behavior of a nickel-aluminum bronze alloy in chloride solution, Corros. Sci. 98 (2015) 260–270. R.P. Chen, Z.Q. Liang, W.W. Zhang, D.T. Zhang, Z.Q. Luo, Y.Y. Li, Effect of heat treatment on microstructure and properties of hot-extruded nickel-aluminum bronze, Trans. Nonferrous Met. Soc. China 17 (2007) 1254–1258. M. Wysiecki, L. Labuc, Optimization of chemical composition of five-component aluminium bronze for the production of ship’s propellers, Int. J. Mater. Prod. Technol. 5 (1990) 339–348. J. Labanowski, T. Olkowski, Effect of chemical composition and microstructure on mechanical properties of BA1055 Bronze Sand castings, Adv. Mater. Sci. 9 (2009) 23–29. D.T. McDonald, C.J. Barr, K. Xia, Effect of equal channel angular pressing on lamellar microstructures in nickel aluminum bronze, Metall. Mater. Trans. A 44 (2013) 5556–5566. C.J. Barr, D.T. McDonald, K. Xia, Significantly enhanced tensile strength and ductility in nickel aluminum bronze by equal channel angular pressing and subsequent heat treatment, J. Mater. Sci. 48 (2013) 4749–4757. Y.T. Lv, L.Q. Wang, Y.F. Han, X.Y. Xu, W.J. Lu, Investigation of microstructure and mechanical properties of hot worked NiAl bronze alloy with different deformation degree, Mater. Sci. Eng. A 643 (2015) 17–24. D.R. Ni, B.L. Xiao, Z.Y. Ma, Y.X. Qiao, Y.G. Zheng, Corrosion properties of frictionstir-processed cast NiAl bronze, Corros. Sci. 52 (2010) 1610–1617. Z.B. Qin, Z. Wu, X.S. Zen, Q. Luo, L. Liu, W.J. Lu, W.B. Hu, Improving corrosion resistance of a nickel-aluminum bronze alloy via nickel ion implantation, Corrosion 72 (2016) 1269–1280. Q. Luo, Z. Wu, Z.B. Qin, L. Liu, W.B. Hu, Surface modification of nickel-aluminum bronze alloy with gradient Ni-Cu solid solution coating via thermal diffusion, Surf. Coat. Technol. 309 (2017) 106–113. C.H. Tang, F.T. Cheng, H.C. Man, Effect of laser surface melting on the corrosion and cavitation erosion behaviors of a manganese-nickel-aluminum bronze, Mater. Sci. Eng. A 373 (2004) 195–203. C.H. Tang, F.T. Cheng, H.C. Man, Laser surface alloying of a marine propeller bronze using aluminum powder: part I: microstructural analysis and cavitation erosion study, Surf. Coat. Technol. 200 (2006) 2602–2609. T. Debroy, H.L. Wei, J.S. Zuback, T. Mukherjee, J.W. Elmer, J.O. Milewski, A.M. Beese, A. Wilson-Heid, A. De, W. Zhang, Additive manufacturing of metallic components – process, structure and properties, Prog. Mater. Sci. 92 (2018) 112–224. D. Herzog, V. Seyda, E. Wycisk, C. Emmelmann, Additive manufacturing of metals, Acta Mater. 117 (2016) 371–392. W.E. Frazier, Metal additive manufacturing: a review, J. Mater. Eng. Perform. 23 (2014) 1917–1928. J.O. Milewski, ‘Lasers, Electron Beams, Plasma Arcs’, Additive Manufacturing of Metals, Springer International Publishing, 2017, pp. 85–97, https://doi.org/10. 1007/978-3-319-58205-4 ISBN:978-3-319-58205-4. M.A. Jackson, A. Van Asten, J.D. Morrow, S. Min, F.E. Pfefferkorn, A comparison of energy consumption in wire-based and powder-based additive-subtractive manufacturing, Procedia. Manuf. 5 (2016) 989–1005. B. Baufeld, E. Brandl, O. Van der Biest, Wire based additive layer manufacturing: comparison of microstructure and mechanical properties of Ti-6Al-4V components fabricated by laser-beam deposition and shaped metal deposition, J. Mater. Process. Technol. 211 (2011) 1146–1158. B.A. Szost, S. Terzi, F. Martina, D. Boisselier, A. Prytuliak, T. Pirling, A comparative study of additive manufacturing techniques: residual stress and micro-structural analysis of CLAD and WAAM printed Ti-6Al-4V components, Mater. Des. 89 (2016) 559–567. E. Brandl, A. Schoberth, C. Leyens, Morphology, microstructure, and hardness of titanium (Ti-6Al-4V) blocks deposited by wire-feed additive layer manufacturing (ALM), Mater. Sci. Eng. A 532 (2012) 295–307. F. Wang, S. Williams, P. Colegrove, A.A. Antonysany, Microstructure and

[39] [40] [41] [42] [43] [44]

[45] [46] [47] [48]

[49] [50] [51] [52]

[53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63]

16

mechanical properties of wire and arc additive manufactured Ti-6Al-4V, Metall. Mater. Trans. A 44 (2013) 968–977. H. Geng, J. Li, J. Xiong, X. Lin, F. Zhang, Optimization of wire feed for GTAW based additive manufacturing, J. Mater. Process. Technol. 243 (2017) 40–47. F. Xu, Y. Lv, Y. Liu, F. Shu, P. He, B. Xu, Microstructural evolution and mechanical properties of Inconel 625 alloy during pulsed plasma arc deposition process, J. Mater. Sci. Technol. 29 (2013) 480–488. C. Shen, Z. Pan, Y. Ma, D. Cuiuri, H. Li, Fabrication of iron-rich Fe-Al intermetallics using the wire-arc additive manufacturing process, Addit. Manuf. 7 (2015) 20–26. T. Abe, H. Sasahara, Dissimilar metal deposition with a stainless steel and nickelbased alloy using wire and arc-based additive manufacturing, Precis. Eng. 45 (2016) 387–395. D. Ding, Z. Pan, D. Cuirui, H. Li, A practical path planning methodology for additive manufacturing of thin-walled structures, Robot. Com-Int. Manuf. 34 (2015) 8–19. Y. Ma, D. Cuirui, C. Shen, H. Li, Z. Pan, The effect of location on the microstructure and mechanical properties on in-situ alloying and additive layer manufacturing of titanium aluminides using gas tungsten arc welding, Mater. Sci. Eng. A 631 (2015) 230–240. F. Martina, J. Mehnen, S.W. Williams, P.A. Colegrove, F. Wang, Investigation of the benefits of plasma deposition for the additive layer manufacture of Ti-6Al-4V, J. Mater. Process. Technol. 212 (2012) 1377–1386. N. Rodriguez, L. Vázquez, I. Huarte, E. Arruti, I. Tabernero, P. Alvarez, Wire and arc additive manufacturing: a comparison between CMT and TopTIG processes applied to stainless steel, Weld World 62 (2018) 1083–1096. D. Ding, Z. Pan, S.V. Duin, H. Li, C. Shen, Fabricating superior NiAl Bronze components through wire arc additive manufacturing, Materials 9 (2016) 652–663. C. Shen, Z. Pan, D. Ding, L. Yuan, N. Nie, Y. Wang, D. Luo, D. Cuiuri, S.V. Duin, H. Li, The influence of post-production heat treatment on the multi-directional properties of nickel-aluminum bronze alloy fabricated using wire-arc additive manufacturing process, Addit. Manuf 23 (2018) 411–421. http://www.gefertec.de/en/3dmp-process. E8/E8M-15a Standard Test Methods for Tension Testing of Metallic Materials, ASTM, 2015. L.L. Parimi, G. Ravi, D. Clark, M.M. Attallah, Microstructural and texture development in direct laser fabricated in 718, Mater. Char. 89 (2014) 102–111. L.E. Murr, E. Martinez, K.N. Amato, S.M. Gaytan, J. Hernandez, D.A. Ramirez, P.W. Shindo, F. Medina, R.B. Wicker, Fabrication of metal and alloy components by additive manufacturing: examples of 3D materials science, J. Mater. Res. Technol 1 (1) (2012) 42–54. P.J. Macken, A.A. Smith, The Aluminium Bronzes, Copper Development Association Publication No. 31, 1966. W.L.J. Crofts, D.W. Townsend, A.P. Bates, Soundness and reproducibility of properties of sand-cast complex aluminium bronzes, J. Inst. Brit. Foundrymen 57 (1964) 89. G. Krauss, Solidification, segregation, and banding in carbon and alloy steels, Metall. Mater. Trans. B 34 (6) (2003) 781–792. D.M. Herlach, B. Feuerbacher, Non-equilibrium solidification of undercooled metallic parts, Adv. Space Res. 11 (7) (1991) 255–262. H. Yin, S.D. Felicelli, Dendrite growth simulation during solidification in the LENS process, Acta Mater. 58 (4) (2010) 1455–1465. A. Jahanafrooz, F. Hasan, G.W. Lorimer, N. Ridley, Microstructural development in complex nickel-aluminum bronzes, Metall. Trans. A 14A (1983) 1951–1956. P. Weill-Couly, D. Arnaud, Influence De La composition et de la structure des cuproaluminiums sur leur comportment en service, Fonderie 28 (1973) 123–135. S. Kou, Welding Metallurgy, 2nd edition, (2002) ISBN: 978-0-471-43491-7. S. Mishra, C. Därmann, Role and control of texture in deep-drawing steels, Int. Met. Rev. 27 (1982) 307–320. H. Asgari, C. Baxter, K. Hosseinkhani, M. Mohammadi, On microstructure and mechanical properties of additively manufactured AlSi10Mg_200C using recycled powder, Mater. Sci. Eng. A 707 (2017) 148–158. E.A. Culpan, G. Rose, Corrosion behavior of cast nickel aluminum bronze in sea water, Br. Corros. J. 14 (3) (1979) 160–166.