Effect of process parameters on mechanical properties of 5554 aluminum alloy fabricated by wire arc additive manufacturing

1 Introduction

The manufacturing landscape has evolved significantly in the past decade, with traditional methods categorized as machining and non-machining now further delineated into three distinct categories: machining, non-machining, and additive manufacturing, reflecting the rapid technological advancements. Among these methods, additive manufacturing stands out, offering numerous advantages including direct translation of design to part, design flexibility, waste reduction, accelerated product development, and the ability to produce lightweight components with intricate structures [1]. Although additive manufacturing may entail higher costs and more production time compared to traditional methods, it is preferred for products requiring high added value. Particularly in sectors such as aerospace and space exploration, where lightweight and low-volume production are paramount, additive manufacturing proves indispensable. Moreover, its applications extend to diverse sectors including medicine, dentistry, automotive, construction, machinery, sculpture, and jewelry making. This innovative technology continues to reshape industries, promising to play a significant role in the future [2], [3].

Wire and arc additive manufacturing (WAAM) technology represents a novel production method that combines traditional welding techniques with additive manufacturing technology, experiencing rapid development in recent years. This innovative approach offers several distinct advantages, including enhanced automation capabilities, accelerated production speeds compared to other metal additive manufacturing methods, improved raw material utilization efficiency, direct production without the need for additional processing, and the ability to work with challenging-to-process metals [4]. Compared to powder-based metal additive manufacturing methods, the use of wire raw material in the WAAM method offers distinct advantages [5], [6]. The lower cost of wire raw material and its ability to provide higher material deposition rates contribute to the rapid advancement of this technology. Unlike powder-based methods, where part sizes are limited by the size of the powder bed, WAAM leverages robot kinematics, theoretically enabling the production of parts without constraints imposed by equipment dimensions. Furthermore, the WAAM method promotes environmental sustainability by achieving up to 100 % raw material efficiency and eliminating the formation of a dusty production environment that may pose risks to human operators. These inherent characteristics position the WAAM production method as a promising solution for various manufacturing applications [7].

During WAAM, part fabrication occurs through the successive deposition of molten metal wire, layer by layer. Several studies in the literature have investigated the production parameters for aluminum parts using the WAAM method. Wang et al. [8] conducted an experimental study utilizing aluminum magnesium alloy wire, producing a structure resembling a ladder and obtaining samples from inner layers and interlayers. Through tests conducted on these samples, they examined defect development and microstructure. Their findings indicated that as the deposition rate increases, the average grain size also increases, porosity gradually rises, and hardness distributions remain relatively consistent; however, hardness decreases due to heat accumulation-induced grain size increase, subsequently leading to a decrease in tensile strength. Chen et al. [9] fabricated a wall-shaped sample piece using Al–Si alloy wire (4047) via the WAAM method. Their objective was to investigate the effects of co-heat input on Al–Si alloy material production using WAAM. Through analysis of internal structure and mechanical properties, they concluded that simultaneous increases in wire feed and feed rate impact effective wall thickness and layer length, with both parameters contributing to increased wall thickness. Additionally, they found that wire feed speed predominantly influences shape formation, while grain size increases with escalating wire feed and feed rate.

Wang et al. [10] conducted an examination of the microstructure and mechanical properties of wall-shaped parts produced using two distinct Al–Cu alloy wires, leading to several notable conclusions. Their findings revealed that samples fabricated with wire containing higher copper content exhibited smaller grain sizes compared to those made with wire containing lower copper content. This disparity can be attributed to the presence of more grain-reducing elements in the wire with lower copper content. While increasing the copper content of the wire enhances its mechanical properties at room temperature, these values tend to decrease at elevated temperatures. Zhong et al. [11] employed the WAAM method to fabricate wall-shaped samples using AA2050 Cu-containing aluminum wire. Upon completion of production, the microstructure and mechanical properties of the resulting wall made of 2050 Al–Cu alloy were examined. Analysis revealed differing microstructures across layers, with initial layers exhibiting a vertical line-type structure and subsequent layers characterized by particles of uniform size, attributed to temperature gradients and cooling rates during welding. Additionally, second-phase particles observed in the study were primarily composed of Cu along grain boundaries. Following heat treatment applied to the samples post-production, hardness values were found to decrease while yield stress values increased.

Köhler et al. [12] investigated the production process utilizing the WAAM method with Al-4047 and Al-5356 aluminum alloy wires, conducting a comprehensive analysis of properties and micro-durability for the smooth walls generated through this technique. Their study revealed that the use of Al-5356 resulted in narrower and taller structures with less deviation in deposition direction. Although tensile strengths were slightly lower in the vertical direction, vertical elongation was found to be 40 % lower than horizontal elongation. Kazmi et al. [13] explored single bead deposition using a robotics Gas Metal Arc Welding (GMAW) process based on RSM (Response Surface Methodology), aiming to analyze the impact of input variables (current, tool speed) on output responses (bead width, height, depth of penetration, and surface roughness). Defects-free deposition of ER-4043 aluminum alloy was achieved, with an evaluation of process parameter effects on morphology, microstructure, microhardness, and wear behavior of the deposited bead. Their findings indicated that current and tool speed significantly influenced bead quality, shape, and size, with current affecting all output responses and tool speed notably affecting surface roughness and bead height. Tawfik et al. [14] aimed to minimize defects and enhance mechanical properties of Al–Mg aluminum alloy through their study on the effects of WAAM process travel speed (TS) on microstructure characteristics, microhardness, and tensile properties of deposited components. Their conclusions highlighted minimal influence of TS on microhardness, with inner-layer regions exhibiting coarse grains and interlayer regions featuring refined equiaxed grains. Grain growth trend was notably slowed with increasing TS, and deposited Al–Mg alloy demonstrated ultimate tensile strength values ranging from 212.5 to 245.5 MPa with TS varying from 200 to 500 mm min⁻¹. Wei et al. [15] conducted an investigation into the impact of arc oscillation patterns on the formation, distribution of pores, and mechanical properties of WAAM 2319 aluminum alloy through experimental analysis. Their findings revealed that arc oscillation enhances the vortex of liquid metal in the horizontal direction, thereby improving the spread of liquid metal and resulting in changes to the shape and size of the molten pool. Additionally, arc oscillation influences the thermal distribution of the molten pool, thereby affecting the solidification process and the formation of pores. Notably, arc oscillation effectively enhances the tensile strength and plasticity of Cold Metal Transfer (CMT) WAAM 2319 aluminum alloy in the vertical direction. Furthermore, the tensile properties in the horizontal direction surpass those in the vertical direction, with concurrent improvements in plasticity.

Derekar et al. [16] conducted a study on the effects of different deposition conditions – specifically, heat input, interlayer temperature, and interlayer dwell time – on porosity formation and distribution. They compared results from pulsed-MIG (Metal Inert Gas) and CMT processed samples regarding hydrogen dissolution and metal deposition techniques. Their investigation revealed that pulsed-MIG consistently exhibited higher pore content than CMT. Pulsed-MIG displayed a higher total pore volume fraction under conditions of low heat input, low interlayer temperature, and longer dwell time control methods compared to high heat input, high interlayer temperature, and shorter dwell time conditions. The opposite trend was observed for CMT. Zhou et al. [17] explored the influence of electric arc travel speed on macro-morphology, microstructure, and mechanical properties. Their findings demonstrated that increasing electric arc TS led to a decrease in the size and volume fraction of equiaxed grain. Additionally, micro-hardness and tensile strength analysis revealed that samples fabricated at a TS of 350 mm min⁻¹ exhibited finer equiaxed grain and higher ultimate tensile strength (273.5 MPa) and yield strength (182.9 MPa) compared to those fabricated at 250 mm min⁻¹. Hauser et al. [18] investigated pore behavior in WAAM of AW4043/AlSi5 (wt.%) and developed a post-process monitoring approach. They observed that increasing shielding gas flow rate led to an increase in porosity in aluminum parts due to rapid solidification of the melt pool through forced convection. Higher convection rates appeared to limit the escape of gas inclusions, with gas inclusions escaping from the melt pool leaving cavities on the surface of each deposited layer.

In the WAAM method, achieving high-quality and efficient production hinges upon accurately setting process parameters. Consequently, the success of the WAAM process is contingent upon identifying these parameters correctly. Moreover, process parameters may vary depending on the materials and geometries involved, necessitating their selection to align with the properties of the produced part and the material utilized. Experimental studies have elucidated the effects of various production parameters in the WAAM method. The current study breaks new ground as it investigates an unprecedented example in the literature: the use of 5554 aluminum wire, deemed suitable for the WAAM method, and draws inferences from analyzing samples produced with this wire under different process parameters. Notably, this study stands out by examining the combined effects of process parameters. While previous studies have typically focused on individual parameters, research encompassing the effects of multiple parameters remains scarce. Therefore, this study represents a comprehensive exploration in this regard.

2 Experimental approach

As part of the study, samples were manufactured using the CMT welding method, known for its minimal risk of thermal distortion. For the creation of straight wall samples as planned, the preferred wire was a 1.2 mm diameter Mg-dominated aluminum wire with the serial number 5554. The chemical composition of this wire type is detailed in Table 1. Furthermore, Table 2 provides the technical specifications of this wire.

The MIG/MAG (Metal Inert Gas/Metal Active Gas) welding method stands as one of the most widely employed welding techniques today, boasting notably high welding speeds. However, for this study, CMT method was chosen due to its characteristic low heat input, supported by the welding machine. As depicted in Figure 1, a gas welding machine (Fronius TPS 500i, located in Bursa, Turkey) integrated with the torch at the end of a robotic arm mechanism (Kuka Robotics KR Iontec 20 R2100, also situated in Bursa, Turkey) was utilized during sample production. This welding machine facilitates MIG/MAG gas welding processes, accommodating currents ranging from 3 to 500 A and voltage values up to 39.0 V.

Figure 1:

Robot arm and welding machine equipment.

To prepare the produced samples for mechanical testing, they were initially separated from the bottom plate. Subsequently, they were transformed into thin sheet metal using a milling and grinding machine. Following these procedures, a template (as illustrated in Figure 2) was employed to cut the samples from the sheet metal using wire erosion for tensile testing, microstructure examination, and hardness testing.

Figure 2:

Test specimen dimensions’ template for straight wall sample.

In accordance with the specified wire and process parameters, the production of six samples was successfully completed, as depicted in Figure 3. Using this template, the samples will be detached from the plate, allowing for direct utilization in the tensile test. Additionally, samples earmarked for hardness and microstructure analysis will be transferred to bakelite, where requisite procedures will ensue. Studies in the literature have consistently shown that wire feeding speed, robot arm speed, and interlayer waiting time significantly influence the mechanical properties and internal structure of products manufactured via the WAAM method [19]. Consequently, following preliminary evaluations conducted with the technical personnel responsible for these three parameters, six different parameter sets were established. To facilitate a comparative assessment of these parameters, Table 3 was formulated and employed as the process parameters in production.

Figure 3:

Post-production views of straight wall samples with WAAM.

The process of preparing the sample for hardness, microstructure, and macrostructure tests typically begins by placing the sample in the bakelite mounting device. Samples placed in bakelite powder within the chamber of the bakelite mounting device (Metkon/Ecopress 52, located in Bursa, Turkey) undergo processing under high pressure at 180 °C for approximately 12 min. Following this procedure, sanding and polishing processes are employed to enhance the surface quality of the sample obtained. Initially, the samples are sanded using 240, 320, 500, 800, 1,200, 2,000, and 2,500 grit water sandpaper, after which the polishing process commences with the addition of 6-, 3-, and 1-μm solution. This process continues until the surface of the sample achieves a shiny appearance. Subsequently, the surfaces of the samples are cleaned by washing with alcohol and rendered shiny for etching. The etching process stands as one of the most crucial steps for micro and macrostructure measurement and necessitates the preparation of the correct recipe. In this process, Keller liquid, commonly utilized in aluminum etching, is applied to the samples for 30 s. Once the etching process concludes, the samples are primed for micro and macrostructure measurement. Following etching, samples prepared for microstructure imaging were analyzed using an industrial microscope (Nikon Eclipse LV150NL, located in Bursa, Turkey). In this laboratory device, samples are positioned on the microscope, and images are projected onto the computer monitor utilizing 200 µm and 500 µm lenses.

Vickers microhardness measurements were conducted using a Metkon DUROLINE-M microhardness tester, located in Bursa, Turkey. A 100 g-force load and 10-s dwell time were utilized due to the thin wall thickness of the samples. To assess local variations within the samples, the average of results obtained from three different locations on each sample was calculated, and an average value was employed for comparisons.

The purpose of the tensile test is to determine the elastic and plastic deformations resulting from the static load applied to the materials. The tensile tester comprises two jaws capable of moving up and down relative to each other, to which the sample piece is attached, along with units capable of providing movement or force to these jaws and measuring these quantities. One of the jaws holding the test piece is moved at a constant speed, applying varying tensile force to the test piece, while the amount of extension resulting from this force is recorded using the extensometer camera.

Tensile samples were machined perpendicular to the deposition direction in accordance with ASTM E8/8M standards. Tensile tests were conducted using a computerized tensile testing machine (Shimadzu Autograph AG-X Plus, located in Bursa, Turkey) with a constant crosshead speed of 5 mm min⁻¹.

3 Results and discussion

Hardness tests were conducted on each test sample piece embedded in bakelite. Table 4 presents the values obtained by averaging three different hardness test results performed on each sample, utilizing HV 100/10 values, to observe local differences.

The following graphs (see Figure 4) were utilized to interpret the changes in hardness depending on the parameters. These graphs illustrate the variations in hardness values based on wire feeding speed, robot speed, and dwell time between layers.

Figure 4:

Vickers hardness (HV) depending on, a) wire feed speed, b) robot speed and c) dwell time.

Upon comparing the parameters, it was observed that increasing the wire feeding speed (resulting in increased heat input) led to higher hardness values (Figure 4a). Conversely, increasing the robot speed resulted in decreased hardness values (Figure 4b). Additionally, a slight decrease in hardness values was observed with an increase in dwell time (Figure 4c).

In general, the effects of three different parameters were examined, revealing that high heat input during the formation of aluminum alloys improved the hardness values, while low heat input had a negative effect on hardness. However, contradicting results have been reported in the literature. For instance, Su et al. analyzed the effect of heat input (changing WFS/TS) on the microhardness of WAAM-CMT Al–Mg alloy samples. They concluded that microhardness values varied with changes in WFS (from 7 m min⁻¹ to 8 m min⁻¹) but showed no significant change with different TS along the mid-height direction of samples and three typical regions along the width direction of samples. Lower WFS and higher TS were conducive to weld pool cooling, resulting in the formation of small grains, which benefited the improvement of microhardness [20]. It is predicted that the reason for this contradiction may stem from the chemical composition of the wire used, or that the microhardness values of the process parameters increase up to a certain point and then decrease.

In the tensile test experiments, a total of four tests were conducted for each wall, comprising two horizontal and two vertical samples. This approach allowed for the observation of local differences with various test samples taken from the wall, and the average of these values was considered for the most accurate result. To facilitate comparisons between walls, maximum stress, maximum strain, and energy absorption values were analyzed and visualized as column charts. Upon examination of the maximum stress values in Figure 5, it is observed that among the horizontal samples, the highest stress value Is Sample 4, while the lowest is Sample 2. Conversely, when comparing vertical samples, the highest value is in Sample 2, while Sample 5 exhibited the lowest tensile behavior.

Figure 5:

Maximum stress (MPa) depending on WAAM method parameters.

The occurrence of anisotropy poses a significant challenge for production in the additive manufacturing method. Anisotropy is evident when examining the results of the tensile test, one of the distinguishing test methods. The disparity in results between vertical and horizontal samples indicates this phenomenon. Upon reviewing Figure 5, it is apparent that the most significant differences in results between vertical and horizontal samples are observed in Samples 2 and 4, suggesting the presence of anisotropy in these samples.

Considering Figure 6 prepared for the strain values, it can be observed that the highest strain values in horizontal samples were approximately 25 % in Samples 3 and 6, while the lowest strain values were observed at 17 % in Samples 2 and 5. As for the vertical samples, the highest strain is approximately 28 % in Sample 1, while the lowest strain is 17 % in Sample 4. Upon comparing the strain values vertically and horizontally, the most notable difference is seen in Sample 1, suggesting that it exhibits anisotropic characteristics.

Figure 6:

Maximum strain (%) depending on WAAM method parameters.

During the tensile test, a material curve is plotted for each sample. By calculating the area under this curve through integration, the energy absorption, which serves as a measure of the sample’s toughness, can be obtained. A column chart illustrating the energy absorption values calculated in both horizontal and vertical directions for each sample was generated, as depicted in Figure 7. From the graph, it is evident that Sample 3 exhibits the highest energy absorption behavior among horizontal samples (47,595 kJ m⁻³), whereas Sample 2 demonstrates the lowest energy absorption rate (30,340 kJ m⁻³). Conversely, in vertical samples, Sample 2 displays the highest energy absorption behavior (44,139 kJ m⁻³), while Sample 4 exhibits the lowest energy absorption (32,108 kJ m⁻³). Upon examination of the three charts, it is evident that Sample 3 exhibits the most stable and highest properties in the tensile test. This inference was made with attention to the differences in values between vertical and horizontal samples.

Figure 7:

Energy absorption depending on WAAM method parameters.

Prior to etching, porosity formation was assessed by scanning the surfaces of the samples using a microscope. Through the examinations (see Figure 8), dense and large porosities were detected in Sample 5, with a measured porosity density per unit area of 4.1 %. Conversely, Sample 3 exhibited the least amount of porosity formation, with a porosity density per unit area of 0.13 %.

Figure 8:

Porosity examination of samples.

Upon examining the tensile test behavior of Sample 5, which exhibited high porosity according to the porosity results, it was evident that this adversely impacted its mechanical properties due to the presence of pores. Conversely, upon scrutinizing the tensile test behavior of Sample 3 from the same perspective, it was observed that its mechanical properties were enhanced owing to the lower pore content, yielding comparable results in both the vertical and horizontal directions.

4 Conclusions

The results of the effects of WAAM method parameters on mechanical performance and microstructure of Mg-dominated 5554 series aluminum can be summarized as follows.

1. The mechanical performance and microstructural can be successfully optimized with WAAM method parameters. The highest and most accurate post process properties was at the wire feed speed 7.5 m min⁻¹, robot speed 0.5 m min⁻¹ and dwell time between layers 30 s.

2. Increasing the input energy density (high WFS/low TS) improved the Vickers microhardness values. This is due to optimum deposition-cooling loop and better grain size formation.

3. To check the anisotropic status from the results of the tensile test, the values in the vertical and horizontal samples were compared with each other. It was observed that the important thing was not high values but equal results in all directions. In this respect, it was determined that the best parameter set was in Sample 3.

4. Information from the porosity results also confirmed other mechanical tests. It was observed that the samples with high porosity were damaged at lower stress, ones with least porosity illustrate more accurate consequences.

5. After all this experimental work, it was deduced that heat treatments must be carried out in order to improve mechanical properties (hardness and toughness) in additive manufacturing with the WAAM method.