Abstract
Shaped metal deposition is a relatively new additive layered manufacturing method. It is a novel technique to build net-shaped or near-net-shaped metal components in a layer-by-layer manner via applying metal wire and selection of a heat source such as laser beam, electron beam, or electric arc. It is a manufacturing method used for production of complex featured and large-scaled parts, especially in aerospace and metal-die industries. This method can lower the cost of fabricated parts by reducing further machining and finishing processes and shortening lead time. This article presents a comprehensive literature review on shaped metal deposition, and it mainly aims to highlight some of the areas which were reported by the researchers in this field to give an extensive overview of shaped metal deposition processes, classification of its methods, and their applications. The presented literature review covers extensive details on microstructure, mechanical properties, and residual stresses induced in the metallic parts produced by various shaped metal deposition techniques as well as fabrication of dual-material parts. Additionally, grain refinement of the deposition morphologies using various techniques, especially the arc pulsation process, was mentioned. This study demonstrates that shaped metal deposition method using wire can be considered as a distinctive low-cost method for fabricating large-scaled components due to high deposition rates, high efficiencies, and dense part production capabilities. However, the accuracy and surface finish are less compared to laser and electron beam melting methods.
Introduction
Additive manufacturing (AM) is an innovative net-shaped or near-net-shaped manufacturing technology used for producing final solid objects by depositing successive layers of material in powder or wire form via melting them using a focused heat source directed from an electron beam, laser beam, or plasma or electric arc.1,2 AM is a novel manufacturing technology, which was developed in 1987 by three-dimensional (3D) systems company in the United States using stereolithography (SLA) technology. 3 In 1990, this technology emerged in Europe and now it is pioneered in the same continent. 3 Several new AM technologies were established and more new commercial systems introduced to the market after 1991. 4 AM technology basically offers significant reduction in lead time, cost, and waste material in the form of machining chips and less toxic waste from cutting fluid.1,3 AM can help to significantly reduce high buy-to-fly (BTF) ratios of cast, forged, and machined parts in the aerospace industry. BTF ratio reflects material efficiency of manufactured components, and it is often referred by the aerospace community to establish an amount of material needed to purchase rather than manufacture final ‘flying’ part. BTF ratio is the weight ratio between the raw material used for a component and the weight of the component itself.1,5 Traditional machining methods typically produce components with BTF ratios of 5:1, but sometimes greater than 20:1 due to large amount of waste and difficult-to-recycle material. 6 Figure 1 shows an example of comparison of BTF ratio between traditional and AM technologies. The finished part has a weight of 250 lb and to machine from a forged billet would require a starting billet weight of 5000 lb which gives a BTF ratio of over 20. This method is extremely wasteful in terms of material. In contrast, using some processes in AM technology, the substrate has a weight of 200 lb as a rolled sheet and 75 lb of wire which is added with 25 lb machined away to get the final profile. This gives BTF ratio near to 2 and saving materials of 4825 lb.
Comparison of traditional machining process versus AM process in terms of BTU.
AM processes can be grouped into three categories according to feedstock material types: (1) powder-bed processes, (2) powder-fed processes, and (3) wire-fed processes. 7 Table 1 shows the metal AM classification based on heat sources and material additive methods, and it also shows a comparison of the characteristics of these processes.1,8–16 But the most suitable classification of AM techniques can be done according to the type of the heat sources and additive materials. In that sense, Figure 2 depicts the classification of various AM processes with the combination of different heat sources with different material feeding methods. Some of these techniques are not popular in the industries especially the techniques used ionized gas as a heat source to melt the added material (powder or wire) to create the desired part using ion fusion forming (IFF) process. 17 Also the techniques used the explosive energy as a heat source during adding the metal layers in order to fabricate a part using sheet laminating additive manufacturing (SLAM) method. 13
Classification and comparison of additive manufacturing processes.
DMLS: direct metal laser sintering; SLS: selective laser sintering; SLM: selective laser melting; LENS: laser-engineered net shaping; LAM: laser additive manufacturing; DLD: direct laser deposition; LMDS: laser metal deposition shaping; LMD: laser metal deposition; SMD-DMD: shaped metal deposition–direct metal deposition; EBSM: electron beam selective melting; EBM: electron beam melting; EBAM: electron beam additive manufacturing; SMD-EBF3: shaped metal deposition–electron beam free form fabrication; SMD-WAAM: shaped metal deposition–wire and arc additive manufacturing; UC: ultrasonic consolidation; SLAM: sheet lamination additive manufacturing; ++: excellent; +: good; 0: neutral; −: negative.
Classification of metallic AM processes.
More than 20 different AM techniques have been developed until now, and they emerged to the industry to produce prototypes and net-shaped products. The most attractive feature of AM is thus flexibilities in designing of products. 18 AM gives an ability to the designers to adopt more nature-based inspiration to their design since AM gives ease of manufacture capability. 18 In metal industry, AM processes can easily be used to fabricate large parts with relatively high complexities. In fact, manufacturing of some complex shapes and geometric features is extremely difficult using traditional methods (milling, turning, forging, etc.).
AM makes it possible to produce parts or batch of parts with high complexities without tooling. Therefore, AM techniques have been preferred in many areas. Figure 3 highlights the current AM applications in different fields.
AM applications in different industrial areas.
The main objective of this article is to make a comprehensive survey on one of the AM methods, namely, shaped metal deposition (SMD). This work offers a survey of various investigations of the SMD process overall: its potential advantages/disadvantages, control structure used SMD techniques, critical review on the microstructural analysis and effects on mechanical properties, and effects of thermal history on induced residual stresses in SMD parts. In addition, challenges in SMD method such as multi-material use and grain refinement are also discussed. In this survey, a general classification of SMD methods is made on the basis of the used control system, heat source, and material supply techniques for the first time.
SMD process description, challenges, and applications
SMD process is performed by adding molten wire of metallic material layer by layer in order to obtain a designed component shape.19,20 SMD technique helps to effectively minimize the amount of material required for the final shape of the part as aforementioned in Figure 1.
SMD can be considered as a net-shaped or near-net-shaped part production technique compared to the modern additive layer manufacturing techniques. This process can be used especially for rapid manufacturing of metal components where no additional tooling is available. SMD method could be most beneficial for the components produced from expensive alloys such as Ti-6Al-4V, Inconel 718, and TiAl.21–24 SMD technology produces fully dense parts and promises great advantages of using any material which can be welded, particularly difficult-to-machine and formed alloys. However, it should be clear that the microstructure of the SMD components differs from those prepared with conventional techniques since SMD process repeatedly subjects high temperatures and high cooling rates to the deposited components and this may lead to anisotropic phenomenon.25–27 SMD components may have some superior properties compared to the casting components with the same design. In fact, castings have a weight penalty due to their less consistent mechanical properties while SMD components have internal potentials to outperform polycrystalline castings of the same chemistry due to the more consistent solidification conditions made possible by the deposition process as compared to the solidification conditions of the casting processes, resulting in the ultimate strength and strain values of SMD components slightly higher than that of the as-cast material.28,29
Wire and arc additive manufacturing (WAAM) represents a turning point in the metal AM technologies since it combines the electric arc heating sources with a metallic wire feeding system to create 3D metallic parts by depositing beads of weld metal in a layer-by-layer manner. Mechanical properties of a component produced via an SMD process (WAAM or direct metal deposition (DMD)) vary greatly depending on the deposition parameters, such as arc plasma energy, speed, and wire feeding. Generally, mechanical properties of the deposited parts are competitive to cast and even wrought material properties and they may possess some desirable properties appropriate for aerospace industries. 30
In general, SMD group of AM processes offer high flexibilities and cost-effective alternatives to conventional manufacturing methods for both large-production runs and one-off prototype creations.31,32 It is expected that the SMD process has a major impact on design changes and production of large dense parts, and it can also be considered as a repair method to be used for high-performance alloy components such as metal dies across a wide market base. 33 In addition, this process promises to produce large-scale and high-quality parts with very high deposition rates.34,35 SMD techniques have been presented to the aerospace manufacturing industry as a unique low-cost solution for large structural component manufacture due to their high deposition rate and efficiency. These techniques can significantly improve product development time, capital investment, and BTF ratios. 36 The main advantages and limitations of SMD methods are summarized as follows:
Deposition process using WAAM techniques can be done using chamber or outside chamber and there is no need for vacuum environment such as in electron beam melting (EBM) process;5,37
High deposition rates, low-cost parts, and less time of production due to less tooling time;19,35
It is able to manufacture large structural components, especially for aerospace and defence industry components;24,36
It has been gaining considerable interest in the recent years due to its high deposition rate and higher efficiency;36,41
It always needs further machining processes to get a better surface finish; 4
It is limited to mass production; 42
The WAAM techniques cannot be used to produce small components and very intricate part geometries. 1
Classification of SMD methods
AM processes are generally categorized based on the type of heat sources and the material supply. A typical SMD system integrates four basic units: (1) deposition path motion unit, (2) melting source unit, (3) material supply unit, and (4) control system and software unit which enables depositing materials according to the defined path. Figure 4 shows the main units in the SMD system and also shows a new classification for the SMD methods according to the heat source used for melting the metal wire which is separately fed into the melting pool in either hot or cold state. Such systems are controlled via robots, computer numerically controlled (CNC) machines, manipulator arms, and gantry type routers. Currently, two typical heat sources have been used in SMD systems, that is, electric arc and power beam obtained by laser and electron beam. These heat sources are used for melting the supplied material in the form of wire. Although powder material supply type of AM techniques is popular, wire-based SMD processes are gaining more favour because of their higher deposition rates and higher efficiencies compared to others. 36 In the following sections, the most used SMD techniques are briefly introduced: (1) SMD using an electron beam heat source (electron beam free form fabrication (EBF3)), (2) laser beam heat source (DMD-laser wire feed), and (3) SMD using a WAAM.
Shaped metal deposition units.
SMD using electron beam heat source (EBF3)
EBF3 process was first introduced by NASA Langley research centre.43,44 This process uses electron beam as a melting source and a wire feedstock which is fed into the melt pool in vacuum environment. It uses engineering alloys such as Ti-6Al-4V and Al-2219 in standard wire form. Figure 5(a) shows the shapes of the components produced at NASA Langley.1,43 The EBF3 system consists of electron-beam gun, wire feeder, and positioning system enclosed in and vacuum chamber as shown in Figure 5(b). This process uses a wire-feed system to deposit good-quality parts that are better than cast and similar to wrought materials.43,44,46 The work 47 introduces an acoustic emission (AE) technique as a non-destructive testing method that can be used as an online monitoring tool for finding place and characterization of different types of active defects during any process especially for electron beam welding (EBW) defects during EBW of titanium alloys to get the information about nature and location of the defect formed. Taminger and Halfey 45 used EBF3 for feasible solutions to the problems on deposition rate, process efficiency, and material compatibility for insertion into the production environment.
SMD using a laser beam heat source (DMD–laser wire feed)
Lasers can be categorized according to different criteria such as the pumping method or operation mode, but the most common classification is based on the type of active medium used to activate the lasing action. So the lasers are classified as gas, solid state, semiconductor or diode, liquid, fibre, and free-electron X-ray. 48
Gas laser provides various practical advantages such as gases are relatively inexpensive and there is no practical limit in the volume of the gas that is required and there is no risk of damaging the gain medium due to thermally induced deformation. 49 Some of the well-known gas lasers include helium–neon (λ = 632.8 nm), argon (λ = 334–514 nm), CO2 (λ = 10.6 µm), or ArF (λ = 191 nm). In solid-state lasers, the active medium consists of a nonconductive solid of crystal or glass composition doped in a small percentage with ions. The main solid-state lasers include Nd:YAG (λ = 1.06 µm), Yb:YAG (λ = 1.03–1.05 µm), and Er:YAG (λ = 2.94 µm). 49 The semiconductor lasers are widely known as diode lasers. Diode lasers are widely used in material processing due to the relative ease of laser combination for delivery of suitable high-power output and also their compact size and crucially their high-energy efficiency. 48 But among all the different types of laser devices, there are four that are commonly used for manufacturing applications (welding, cutting, drilling, surface modification, and metal depositing): 50 CO2, diode laser, Nd:YAG laser, and fibre laser. The term ‘fibre laser’ usually refers to those lasers in which the medium is a diode optical fibre. It may be emitted either continuous or pulsed. Fibre lasers have important advantages such as 51 high efficiency, compact size, ease of operation, and good beam quality.
Laser metal deposition (LMD) can be defined as a process in which a metal is overlaid on another using a laser beam. A high-power laser beam is focused over the surface of a substrate, where the intense energy is absorbed and a melt pool is generated. An external material, commonly in the form of powder or wire, is then fed to the melt pool, where it melts and adds to the volume of the pool. 52 Four different methods can be used for material delivery in LMD: 52 (1) pre-placed powder (powder-bed), (2) paste (powder-binder mixture paste feed), (3) blown powder (injection powder), and (4) wire feeding method.
DMD technique works by introducing wire feedstock into the melt pool zone, which results from the heat generated by the focused laser beam.1,30Figure 6 shows a schematic drawing of the process, top and side view images of the real process, and some of the parts fabricated using this technique. Some of the processes such as selective laser beam melting (SLBM or SLM), selective laser sintering (SLS), laser direct forming (LDF) and direct laser deposition (DLD) 53 , which use the laser beam as a heat source and powders as filler materials, face major practical difficulties during deposition parts:1,14,46,54–56
Lower capture efficiency of the metal powders;
Some of the gases (e.g. O2, H2, N2, and CO) and other impurities may contaminate the metal powder feedstock. This leads to poor quality of the deposited parts (poor overhung surfaces);
Existence of porosities and lack of fusion problems;
Low laser energy efficiency, since less than 10% of the plug input energy is converted to beam energy;
Lower deposition rates and build volumes of limited size.
In the last decade, considerable amount of attempts have been conducted using laser-wire feedstock combination in order to overcome the mentioned drawbacks. 34 Advantages of this process are higher deposition rates in depositing large components, more stable melt pool results with better control of the weld pool shape, no porosity problems, and metal wires are always cheaper than metal powders.34,41,55 Laser additive manufacturing (LAM) technology is mainly preferred in the manufacturing of complex geometry parts where the productivity and the amount of the parts are in concern. 57
An experimental study of laser and wire feedstock system through the studying of hardness and dimensions of single beads of Ti-6Al-4V filler wire was conducted by Brandl et al. 58 It was shown that the mechanical properties of SMD components are greatly dependent on the microstructure, which is in turn greatly dependent on the thermal history and certain process parameters. Syed and Li 59 presented a study that compared the wire and powder feed system indirect laser deposition process. The authors showed that the wire-based laser process is gaining more acceptances due to higher efficiency and higher deposition rates compared to the powder feed system. Brandl et al. 30 used two different SMD methods: DMD using Nd:YAG laser beam+wire feedstock system and the WAAM using 3D tungsten inert gas (TIG) welding+wire feedstock system. They aimed to increase the knowledge regarding material properties in one system Ti-6Al-4V alloy as a deposition material with respect to aerospace material specifications. Mechanical properties, such as static tension and high cycle fatigue, with heat treatment were investigated. The results showed that the specimens fabricated by the SMD techniques could achieve static tensile properties similar to the material specifications AMS 4928 of wrought material and/or ASTM F1108 of cast material. On the other hand, Gong et al. 60 investigated the effect of existing defects in fatigue performance of Ti-6Al-4V samples in an as-built surface finish condition. They also showed that the defects play a serious role in reducing fatigue properties and fatigue life of specimens.
More recently, some of the studies focused on investigation of microstructure and mechanical properties of single-wall multi-layer Ti-6Al-4V structures deposited with DMD techniques. The microstructure examinations denoted that long columnar grains were found growing parallel to the build direction across the multiple layers. The solid areas can be hardened by dislocation or boundary hardening.58,61 Flexible techniques were used, which allow cladding consumables deposited in either wire or powder form fed into the molten pool using DMD techniques. These techniques can be used on new or worn components and are typically used to rebuild worn or damaged surfaces and to hard-face wear, corrosion, oxidation susceptible materials, or depositing super alloys on turbine blades and valve seats.62,63
In the study by Strano et al., 64 the authors submitted an experimental work consisting of developing a model for part surface quality prediction and model for simultaneous energy consumption prediction. Pareto set was used to represent the best compromises between the surface roughness and energy saving objectives. In the study by Bakhadyrov et al., 65 a machine vision setup for surface voids detection was shown. It can be used for surface quality inspection of parts built with layered manufacturing (LM) techniques. The work by Jin et al. 66 used an adaptive process to deal with different intricate product models by introducing non-uniform rational basis spline (NURBS) curves to represent the contours of the sliced layers instead of using standard transform language (STL) files, which are inherently inaccurate in geometrical representation and need much more storage space to represent a highly complex model. The authors studied a series of optimization algorithms and strategies to optimize the tool-path generation. A number of case studies have been tested, for example, an ear model and a bone model. They showed that using this optimal strategy, the average reduction in the building time is 16 s for a single layer in the ear model and 46 s for a single layer in the bone model.
A temperature measurement system was developed by Hua et al. 67 since the molten pool temperature is an important controlling parameter during a laser rapid forming process which immediately governs metallurgical quality and forming accuracy of the part. The thermal monitoring techniques were described and the influences of the processing parameters on the thermal behaviour of the molten pool were also investigated.67,68 Almir 34 used the laser deposition process using wire as an additive material. In the work the first step was to develop a basic understanding of the process characteristics of the used system, regarding deposition stability and formation of defects. Control strategies were developed such that they enable automatic control of the deposition process with process stability and good geometrical fit of the production parts.
SMD using WAAM
WAAM technique can be sub-grouped into three major processes based on the type of electric arc torch used to melt the wire: (1) 3D TIG welding (TIG or gas tungsten arc welding (GTAW)+wire feedstock) as shown in Figure 7, (2) 3D metal inert gas (MIG) welding (MIG or gas metal arc welding (GMAW)+wire feedstock), and (3) 3D plasma welding (plasma+wire feedstock). The roots of wire-added AM processes went back to 1920s patented by Baker, 69 which stated a new method for producing 3D metallic parts in successive layers using manual arc welding. In the work by Shockey, 70 a WAAM technique was investigated in facing and cladding applications for improving the lifetime of components subjected to abrasive wear. White 71 used wire+arc welding method to repair worn surfaces of large metal pressure rollers. The development of shape welding (SW) technology was performed later in many experimental works. In the beginning of the 1980s, some of shape welded products (rotors and cylindrical courser) were conducted. 72 In the study by Schmidt et al., 73 the authors used SW process to build large reactor coolant components for up to 10.5 m in length and 5.8 m in diameter which showed the capabilities and successful applications of WAAM process.
Deposition process using 3D TIG method.
At the beginning of the1990s, WAAM technology gained a significant importance in the Welding Engineering Research Centre of Cranfield University. The works by Ribeiro and colleagues74,75 introduced a new fabrication approach by developing a manufacturing system for Rolls-Royce plc. Their system was capable of creating solid metal parts directly from the computer-aided design (CAD) models.
Spencer et al. 76 studied the surface roughness, microstructural effects, and bead geometries for samples produced by the MIG-SMD system and the effect of heat built-up during deposition. SMD technique was developed by Rolls-Royce plc, but it was not widely adopted for commercial production. In more recent years, SMD process was further developed by investigating the additive SMD process and it was practically developed in a European project entitled as Rapid Production of Large Aerospace Components (RAPOLAC). 19 RAPOLAC was initiated in 2005 by Advanced Manufacturing Research Centre (AMRC) at The University of Sheffield in collaboration with seven partners from across Europe. Main aims of the project were to develop the SMD process based on cold wire-fed GTAW, build large parts with a variety of aerospace materials, material properties control, and process parameters optimization. 19 Further developments that were made within RAPOLAC project have been recently presented by Escobar-Palafox et al. 77 They supposed that deposition condition is attained faster with closed-loop control than in manually deposited parts, and the surface conditions of the produced parts can also be considerably improved in quality. Martina et al. 78 investigated the feasibility of using plasma wire deposition (PWD) in AM technology. They used 1.2-mm Ti-6Al-4V material to deposit and measure walls of 140-mm long.
Control of SMD process
This section presents the literature concerning development of the closed loop, sensing, real time, and other control systems. These systems are used for controlling the deposition process and improving the deposited part quality. It was found that the surface quality of the deposited parts can be improved using temperature control of the part at the starting point of each pass. 19 Further investigations were done in the Robotics Institute and the Engineering Design Research Centre of Carnegie Mellon University. Robotic deposition systems were used in several thermal deposition processes in order to directly fabricate prototype metal shapes. 39 In RAPOLAC, the fully automatic welding system was improved using a closed-loop welding controller, which uses image processing and a video feedback control. The working variables have been identified in four fundamental key parameters such as welding current, wire feed speed (WFS), travel speed (TS), and the thickness of each layer (the step height (SH) between two subsequent layers). 22 For the frame of the RAPOLAC European project, further studies were investigated for the control of SMD systems. The proposed system aimed to show the software interface of an experimental system adopted to investigate control strategies for rapid manufacturing of aerospace components.19,20,35 Wang and Kovacevic 79 offered a WAAM system based on a 3D TIG technique for producing shaped parts made of aluminium alloys as Al-5356 and Al-4043. In their study, they used online monitoring and control systems for controlling process parameters in order to achieve good-quality parts. Zhang et al. 38 used STL files for slicing and planning RP system using the 3D MIG technique. A CAD surface or a solid model in the standard of initial graphics exchange specification (IGES) format was used to facilitate the controlling on the part to be deposited. It was shown that the control of the deposition parameters helps to control ignition, minimize ignition times, and crater filling control. Figure 8 shows the effects of application of control system in start and end portions. 80
A welding deposition system (MIG-welding SMD) was established with a new interface between the modelling data and the deposition process. 81 The interface is based on the IGES format, which is compatible with the most solid modelling environment. The system was computer controlled to provide the welding torch with the desired path and angle. 81 It was shown that the control of the size, flux, and thermal states of the droplet effect on improving the quality of the deposition. Besides an adaptive robotic control (TIG welding), SMD process was developed by Ding et al. 82 in order to enhance the weld quality with full penetration of Al depositions.
Another WAAM technology, which was provided by ready-to-use additive manufacturing (RUAM) project, opened new routes to manufacture large parts. This work includes a robotic welding SMD system. The researchers improved the process using modern technologies such as cold metal transfer (CMT) process and plasma welding.34,35 The study by Xiong et al. 83 applied a neural network and second regression analysis for predicting bead geometry of depositing parts, which was manufactured using a robotic SMD process representing 3D TIG welding technique. The main findings of their study were that employing this analysis model can effectively be used to predict the desired bead geometry with high precision. Yilmaz et al. 84 developed a new cold feed wire TIG system that is used for mid- and small-size metal components. The main findings of their work are that they designed, constructed, and controlled a new SMD machine used for deposition, with different features of metal parts with working area of 400 mm3. They used an online programming method to control the deposition process using an open-source system, which gives the user high flexibility to tune the process parameters during the deposition process.
Analysis of microstructural and mechanical properties of SMD parts
The following literature mainly focuses on the works concerning microstructure and mechanical properties of samples produced by SMD processes represented by 3D welding techniques for different materials. It is obvious that microstructural properties both affect the further mechanical properties and quality of the parts. As previously discussed, that SMD techniques use wire fed into a local melt pool by arc beam, electron beam, or laser beam. After that the solidification will occur and that is producing microstructures which are significantly affected by thermal cycles.33,58,82–85 The deposited material is thermally affected by the deposition of subsequent layers, resulting in complex thermal histories, which lead to problems of residual stresses and distortions. Furthermore, SMD process parameters can also influence the microstructure and its specifications.41,82 Many studies investigated the microstructure of deposited materials and their mechanical properties of different materials. A large body of literature focused on Ti-6Al-4V which is one of the most important Ti-alloys. 86 In the works by Baufeld and colleagues,25,26,33 an SMD technique using a robotic TIG welding, which are enclosed in an airtight chamber filled with pure argon gas, was presented. Argon is controlled to 99.999% purity and moisture is also controlled at minimum amount. Deposition material was Ti-6Al-4V wire with 1.2 mm in diameter. The microstructure examinations showed that the layered structures are reflecting SMD layers and large elongated grains which are shown in Figure 9. These grains grew epitaxial, inclined in a direction to the layers following the temperature gradient resulting from the movement of the component relative to welding head. These grains are also visible on the etched cross-sections of the components as shown in Figure 10.
Microstructural heterogeneity or banding in the titanium AM is a result of the repeated thermal cycles experienced by the solidified material, with a reducing peak temperature, as the heated source is passed across the surface each time a new layer is consolidated as shown in Figure 10.
The formation of banding in wire plus laser AM process for Ti-6AL-4V alloy was discussed by Kelly and Kampe. 87 Banding is caused by cyclic thermal effects and inhomogeneous distribution of microstructure with change in the local variation in hardness (see Figure 11). Generally, fluctuation of hardness in single beads results from differences in thermal history and chemical composition. 58 SMD components exhibit two different appearing regions: a bottom region, which is a coarse Widmanstätten structure with circular lamellae, and a top region, with a fine, needle-like structure exists in cross-section, as shown in Figure 10. Height of the top region is not very sensitive as the height of the whole component. The microstructure of the both regions consists of α phase laths in a β matrix, but their morphology is different in both regions as indicated in Figure 12.
Cyclic changes in hardness values at a regular distance along the deposition direction for: (a) Ti-6Al-4V 86 and (b) SS308LSi.
Thus, it has been reported that the SMD components show periodic bulges that reflect the separate layers of each deposit and the large thermally grooved columnar grain. 27 The fine Widmanstätten structure at the top and a coarse Widmanstätten structure in the bottom region results from the repeated heat treatment by SMD.
Mechanical tests (tensile test and micro-hardness test) showed that the ultimate tensile strength (σu) of the specimens in the horizontal orientations is slightly higher than the specimens with vertical orientation. Ultimate tensile strength (σu) and ductility depend only slightly on the orientation, location, and also heat treatment. Micro-hardness tests showed that there is no obvious dependence on the wall height. The heat treatment with the β-phase increases the σu and decreases the ductility. The mechanical properties vary to some extent with the variation of welding parameters.25–27 The work 24 showed that anisotropic mechanical properties are expected due to anisotropic microstructure. The results presented by Baufeld et al.27,33 claimed that the microstructure exhibits large, columnar prior-β grains with a gradient in the individual α-lath thickness between the deposited layers, except for in the last three layers of Ti-6Al-4V alloy. The cooling rate in a newly deposited layer decreases as the number of additional layers increases. 87 The thermal cycle is responsible for the formation of the characteristic layer in a layer n due to the deposition of the next layer; n + 1 and n + 3. They showed that upon deposition of layer n + 3, only a narrow region near the top of layer n will see an excursion into β-phase field as shown in Figure 13.1,87 Skiba et al. 24 used SMD technique by TIG welding to produce components of 308 stainless steels to gain steel components with high corrosion resistance and better mechanical properties. They showed that the microstructure of the components consists mainly of austenite and ferrite with values comparable to that of the material prepared by conventional techniques. There is no anisotropy of properties with orientation. They also showed that the components do not suffer from formation of a M23C6 type of carbide.
Zhao et al. 88 used SMD process represented by 3D MIG welding for producing parts of H08Mn2Si steel wire. In their study, researchers combined experiments and numerical model of 3D transient heat transfer. An interaction exists between the pre-heating of fore layer and post-heating of rear layer. It has been an inter-layer metallurgy bonding requirement satisfied by re-melting process between every two adjacent layers. It was also shown that the direction of deposition affects the temperature gradient state, where the same deposition direction has larger temperature gradient than the reverse depositing direction. Also the heat diffusion condition could be improved by optimizing the deposition direction, in case of keeping other parameters constant. Clark et al. 28 used SMD process represented by 3D MIG welding to produce developmental ring rolled combustion casing incorporating an internal circumferential SMD flange, which can be used for aero engine made of Inconel 718 alloy. They showed that the microstructural conditions of the features are highly dependent on welding parameters and practices.
CMT and cold wire feed in SMD techniques
A new innovative technique was first developed in the end of 1990s, 89 namely, CMT. It is a GMAW-based solution, which is based on dip transfer mode mechanism. CMT process was investigated by some of the researches33,36,82,89 and they proved that CMT gives excellent quality, lower thermal heat input, and nearly spatter free weld. Furthermore, CMT overcomes common difficulties encountered during conventional short circuiting GMAW such as unstable process behaviour and severe spatter formation. It is suitable for high deposition rate SMD of Ti-6Al-4V for large scales as represented in RUAM concept parts which were also produced using CMT technology. 35 Hoye et al. 32 introduced a new method in SMD via 3D TIG technique with independent cold wire feed (CWF) unit in the same sense.
Residual stresses deformation in metal AM
Residual stresses and distortion are the two major drawbacks which are induced by the thermal cycles during the welding-based AM processes. They lead to non-uniform plastic deformation due to non-uniform thermal cycles, which may cause critical mechanical effects of the welded parts and produce errors in the accuracy of geometries. So that, great efforts were spent on developing predictive models of thermo-mechanical performance of welding process. Guyot 90 introduced empirical method about welding distributions through systematically performed experiments, whereas Rosenthal 91 generated the first analytical solution to compute the heat transfer of welding process. Makino et al. 92 used a holographic-hole drilling technique to determine the residual stress state using the laser-engineered net shaping (LENS) AM process. Nowadays, continuous improvement in finite element analysis (FEA) allows researchers to study the welding process on a detailed level. Finite element (FE) technique is the most important tool used for solving the effects of thermo-mechanical phenomena. WAAM process suffers from similar problems of residual stresses and distortions. During the deposition process in the area where the heat source locates is heated up sharply and fused locally, two types of residual stresses are generated. 93 First, the compressive stress is formed in front of the fusion zone due to thermal expansion of the heated material which is being constrained by the surrounding cold material. Second, the tensile stresses are generated in the region behind the heat source due to the contraction of cooling material, and this material contraction is restrained by the surrounding cold material which consequently results in tensile stresses. In particular, tensile residual stresses can reduce the mechanical performance and the life span of the produced components since they reduce effective fatigue life and tensile strength of the structure. 94 Some of the works focused on numerical thermo-mechanical modelling of laser used for cladding, 95 hardening, 96 and laser-based solid free form fabrication (SFF). 97 Labudovic et al. 98 developed a 3D model using both analytical and numerical methods for analysing heat flow and residual stresses in laser-based SFF.
Figure 14 shows temperature readings for 20 deposition layers from a representative thermocouple inserted during fabrication H13 tool steel using LENS technique. Each peak represents the thermocouple response as the heat source passes over or near the thermocouple from the initial insertion to subsequent layer depositions. It was obvious that after the initial peak in thermocouple which is approximately 1500 °C, heat is quickly conducted away at about 15 s to a nominal value of 150 °C for the first layer. Each subsequent pass reheats the previous layers. Initial layer still receives the thermal excursion less as the number of the layers increases until the thermocouple nominally reads 500 °C.
Temperature traces at one position as 20 layers are deposited on top of thermocouple inserted into H13 shell build. 9
Figure 15 shows the effects of thermal history on residual stress during part deposition at various powers and travelling speeds. It is obvious that the residual stress is reasonably higher near to the substrate, but this value decreases after 1 in of deposition. This means that the substrate efficiently dissipates the heat and results in a large thermal mismatch and higher residual stress state. In the first layer, the temperature gradient is apparently larger than the other layers due to the nonexistence of pre-heating effects on this layer. 76 Due to deposit of successive layers, the substrate and the deposited component are heated, so the thermal mismatch is reduced, and the residual stress values diminish. 68
Effect of volumetric exposure (J/cm3) on residual stress distribution along Z direction. 68
Volumetric exposure (ratio of power to travelling speed) strongly affects the resulted residual stress values, as can be seen in Figure 15. At higher volumetric exposure, the values for residual stress is higher but at higher power and when the thermal behaviour is no longer influenced by the substrate, the values for residual stress dramatically decrease. The residual stress values mainly depend on large thermal mismatch at the interface. Therefore, the thermal mismatch at the interface must be reduced or eliminated, for example, by pre-heating the substrate before and during the deposition process. The WAAM process provides high deposition rates, so that high amount of heat input occurs, which leads to generate big residual stresses and distortions. For that reason, it is very necessary to study and analyse the thermo-mechanical behaviour of the WAAM processes. Mehnen et al. 35 investigated the possibility of building optimal patterns for building parts with minimum residual stress and distortion through studying on the thermo-mechanical performance of WAAM process in RUAM project. They employed a commercial finite element method (FEM) software in this task. Ding et al. 82 employed thermo-elastic-plastic FEM to simulate and analyse thermo-mechanical performance of the large-scale wall shaped by WAAM process. Residual stresses and unwanted distortions may lead to dimensional inaccuracies during finishing processes by machining since these stresses are relieved during the process of removal of substrate plate. 99
Dual material use in SMD process
There has been a new trend in the material structures where dissimilar types of materials may be required at certain locations in the same structure, enabling the structure to exhibit different functional material properties. So that this type of structure may be produced from a combination of materials with different properties such as ductile-hard material and wear corrosion–resistant material. As a result, elemental constituents and alloys are blended to create a new type of material. In addition, a major challenge of difficult-to-join material systems was solved with successful fabrication of some multi-material systems.5,14,100 Some of the AM processes are suitable for manufacturing multi-material parts with new alloy-based parts, better surface properties, novel mechanical properties, and new alloy design.5,43,101–103 Several AM technologies such as 3D printing, SLA, SLS, ultrasonic consolidation (UC), LENS, LMD, EBM, and some others have been proven to possess multi-material capabilities of dissimilar material manufacturing, but all these technologies use powders as a deposition material.5,100–102,104,105 Nowadays, attempts for fabricating multi-materials structures using SMD technologies are very limited since these require a lot of technical skills and skilfulness of researchers. However, NASA has patented new technology, namely, EBF3, which can be used to produce dual-material products. 43
Grain refinement attempts in SMD processes
In the aforementioned microstructural and mechanical properties section, the parts manufactured by SMD techniques tend to be with large and elongated grains as shown in Figures 9 and 10. So the mechanical properties vary with direction of test since it may result in anisotropic phenomenon. Hence, there is a need to study and control the grain sizes during SMD methods by the common methods of grain refining such as inoculation method through adding of nucleating agent or inoculants to the liquid metal to get very fine equiaxed grains.106–108 Other researchers used external excitation techniques for the purpose of grain refinement, such as weld pool stirring, arc oscillation, ultrasonic vibration, and arc pulsation.46,106,109,110Figure 16(a) shows the effect of the arc oscillation by vibrating the weld torch. 104 Figure 16(b) shows the effect of the arc oscillation on the grain refinement. In the pulsed arc technique, the low-current cycle or background current is sometimes less than 50% of the peak current, so that a sudden reduction in heat input occurs which leads to refinement of the grains as shown in Figure 17.112–116 Wang et al. 31 investigated the effects of direct current pulse GTAW parameters on the morphology of Ti-6Al-4V. They concluded that there is no grain refinement in the WAAM parts of Ti-6Al-4V alloy deposited at both low- and high-frequency pulsed current due to the manner in which the titanium solidifies by planar and cellular rather than dendritic growth.
(a) Effect of arc vibration amplitude on grain size in Al-2.5 Mg and (b) effect of oscillation process in refining the dendrites in the fusion zone. 106
While the work 117 stated that AC was much more effective than DC in refining the fusion zone (FZ) of Ti-6Al-4V welds, the pulsed AC causes a greater agitation in the weld pool. Authors of the works35,36 employed inter-pulse GTAW system as a heat source for the fabrication of large structural Ti-6Al-4V components.
Conclusion and further studies
SMD is a relatively new technology among AM methods where SMD method enables to create net-shaped or near-net-shaped components by depositing a filler material (mostly metal wires) layer by layer. Melting heat sources are currently electric arc, laser beam, and electron beam. This study intended to provide an overview of SMD processes, the main types of SMD techniques, its process principles, applications, analyses of microstructures and mechanical properties of the deposited parts, control methods of this technology, feasibility of using multi-materials, and possibility of grain refining by SMD process. The outcomes of the collected and analysed literature works can be summarized as follows.
Conclusion
WAAM techniques have been presented to the aerospace manufacturing industry as a unique low-cost solution for large-scale structural component manufacture due to their high deposition rate, high efficiency, and high density especially. This technology is enclosed on the materials which are weldable such as super alloys, for example, Inconel 718 and Ti-6Al-4V. It is important to know that deposition parameters have a vital role in this technology because of their effects on microstructure of producing components, thereby affecting mechanical properties of the final parts.
The innovative GMAW process with CMT is suitable for depositing a large-scale manufacture based on its excellent quality, lower thermal heat input, and minimum spatter during deposition.
The works in the literature show that the mechanical properties (ultimate tensile strength and ductility) of 3D welding SMD components depend on the orientation, location, and heat treatment. Also, depending upon the microstructure test results, it is obvious that the surface of the components exhibits a layered structure from the welding and also shows large, elongated grains inclined in a direction to the moving welding torch.
Many of the researchers investigated dual-material structures using some of AM processes, since this trend will enable the structures to exhibit different functional properties, in required locations, combine metal with different properties, and to get a new alloy through blending different elements or/and constituents to create new materials with special properties. There is still a clear lack of literature about using dual materials by SMD techniques.
SMD of aerospace alloys such as Ti-6Al-4Vcomponents usually tends to form two distinct regions, namely, top and bottom regions. The top region consists of Widmanstätten microstructure of fine α-lamellae and the bottom region consists of coarse lamellae. The anisotropy phenomenon appears obviously during SMD of Ti-6Al-4V alloy and it disappears when depositing 308 stainless steel. Anisotropic mechanical properties are expected due to anisotropic microstructure.
SMD process represented by 3D TIG welding technique produces fully dense parts, but usually the accuracy and surface quality are not as good as for other SMD techniques such as DMD and EBF3. Pulsed current TIG process does not affect the morphology of Ti-6Al-4V parts, where there is no evidence of grain refinement through using high- and/or low-frequency pulsed current. The reason may belong to the fact that the grain refinement occurrence in titanium alloys is difficult due to relatively narrow region of constitutional super-cooled at the S/L interface.
In the WAAM techniques, the residual stresses induced in the deposited parts may lead to adverse action during sequential finishing processes and removing of substrate plate.
Further studies
In order to increase the grain refinement at the solidified parts produced in SMD processes, the possibility of increasing cold-wire-feed rate must be further investigated with controlling other parameters with and without using pulsed currents for various materials. As well as further investigation focuses on the effects of using pulsed current in the welding morphology and grain size of different materials used in SMD techniques.
Dual-material deposition would be tried via dual wire feedstock using WAAM techniques.
In order to increase the understanding of the SMD process, further studies must be carried out on the serious and common weld defects, especially the carbide sensitisation (M23C6), which may appear during deposition of austenitic stainless steels and nickel-based alloys.
Further works are required in order to investigate the possibility of producing high-volume and large-size metallic components using SMD techniques and comparing the used SMD technique with the traditional methods as concerning the total costs, rate of productivity, and the quality of the fabricated component(s).
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.