1. Introduction
Additive manufacturing (AM) technology is one of the technologies of the modern manufacturing industry. Its technical demand has been increasing as it can address problems that occur from the manufacturing of products with complex geometry, customized production, and traditional manufacturing methods (e.g., safety accidents and high manufacturing costs) in various industrial sectors, including aerospace, automotive, shipbuilding, and marine industries1-6). The wire arc additive manufacturing (WAAM) process is one of the AM technologies used to manufacture, clad, and repair metal parts with complex three-dimensional (3D) geometry based on the gas metal arc welding (GMAW) technology. This AM technology utilizes metal wire as feedstock and manufactures products by directly depositing the metal material onto the substrate using arc heat as the thermal source. During the deposition process, the high-temperature arc heat stably generated from the shielding gas atmosphere provides thermal energy that forms a molten pool on the substrate or previously deposited metal material, and the metal wire is continuously supplied to the molten pool in a stable manner7-10). This process is repeated to produce a metal part with the desired 3D structure11,12). In addition, among various GMAW technologies, cold metal transfer (CMT) welding based on the short-circuit transfer method can reduce heat input by 30 to 40% compared to metal inert gas (MIG) welding, and it can also reduce spatters13). The WAAM process has six degrees of freedom in metal deposition due to the integration of the wire feeder, robot, and arc generation system. It has a material recycling rate of over 95% and a production speed of 2 to 10 kg/h compared to conventional manufacturing methods, making it the most suitable process for various industries that require productivity and economic efficiency14-17). Laser-powder bed fusion (L- PBF) and laser-direct energy deposition (L-PED) processes, which are representative deposition processes, have benefits (e.g., low heat input, low thermal distortion, and dimensional precision) compared to the WAAM process. In industries that require production efficiency, economic feasibility, and an increase in product size, however, the market related to the WAAM process is expected to grow continuously due to its benefits compared to the PBF and L-DED processes, including 10 to 20 times higher deposition speed, low equipment cost at a 1/5 level, and low production cost18-21).
Among the ER70S-6 solid wires, the KC-28 solid wire is commercially produced by KISWEL for welding high-strength steel components for automobiles. It has been evaluated as a product with high reliability and quality by maintaining excellent mechanical properties in various fields, including butt welding or fillet welding of ship and marine structures as well as welding of bridges and steel structures. When the KC-28 solid wire used for welding is applied to the WAAM process, however, some mechanical properties are degraded by unique thermal cycle accumulation, interlayer defect formation, and residual stress accumulation compared to conventional welding methods. A. Aprilia et al.22) analyzed the decarburization phenomenon of the ER70S-6 steel produced through the WAAM process and identified changes in decarburization depth depending on heat treatment conditions. M. Abbaszadeh et al.23) analyzed the tensile properties and microhardness of ER70S-6 produced through the WAAM process. They linked this to microstructural analysis and reported that the selected deposition parameters induce isotropic mechanical behavior. Y. Kabaldin et al.24) produced heterogeneous metal composites that used ER70S-6 and ER309LSI through the WAAM process, and analyzed the effects of process parameters and post-heat treatment on structural and mechanical properties. They reported the possibility of securing both sloping structures and excellent mechanical properties. There are various previous studies related to the ER70S-6 solid wire, but they reported only the effects of post-heat treatment or a single deposition parameter. In the WAAM process, however, various deposition parameters (e.g., the arc type, shield gas composition, and cooling conditions) are closely connected to each other, and a failure to comprehensively consider these factors may have a serious impact on the mechanical properties of the product, defect generation, and microstructures. In addition, there are few studies related to cooling to control the defects caused by the high heat generated during the WAAM process. Since the application of the results of laboratory-scale research to industries causes various problems, their applicability to industries must also be considered.
Therefore, in this study, optimal deposition conditions were derived and1 by comparing and analyzing various deposition parameters to enhance the applicability of the WAAM process to various industries.
2. Experimental Method
In this study, the WAAM process was performed using KISWEL’s KC-28 solid wire. The commercial wire with a diameter of 1.2Ø was used in the deposition process. Table 1 shows the chemical compositions of the KC-28 solid wire and substrate. The table also includes the mechanical properties of the KC-28 solid wire provided by KISWEL.
Table 1
Chemical compositions and mechanical properties of solid wire and substrate used for experiments (wt%)
Fig. 1 shows the actual image and schematic related to the experimental setup used in the WAAM process. The WAAM process was performed by constructing an integrated system with the CMT welding equipment (TPS 4000-CMT, Fronius) and robot system (KR 20 R1810-2, KUKA) shown in Fig. 1(a). For the CMT torch, the arc transfer mode must be selected according to the material type, wire diameter and type, and the shielding gas ratio. In addition, welding conditions must be selected according to the welding wire feed rate. In the case of the WAAM process based on the GMAW welding mechanism, the GMAW welding process and input parameters are the same, but additional settings for the interlayer distance between beads, inter-pass distance, and deposition path are required to implement the product geometry25). As can be seen from the schematic of Fig. 1(b), deposition was performed using the zigzag method that applies opposite start and end positions by layer to prevent the deflection of the specimen.
The WAAM process parameters set in this study are shown in Table 2. As can be seen from the table, the wire feed rate was set to 7.7, 9.0, and 9.9 m/min. For CMT welding, the current and voltage are determined by the filler metal wire feed rate. Therefore, the welding currents were 200, 220, and 252 A while the welding voltage was set to 16.9, 18.2, and 18.5 V. The torch angle and travel angle were fixed at 90° in the deposition experiment. The contact tip to work distance (CTWD) was set to 15 mm and the shielding gas rate to 20 L/min (80% Ar + 20% CO2) in the experiment.
Table 2
Parameters used in the WAAM process
Fig. 2 shows the schematic of the inter-pass distance, bead height, and depositing method between layers. The inter-pass distance is the gap between the center of a single bead formed through the WAAM process and the center of the next bead. The inter-pass distance was moved by 68 to 76% based on the results of measuring the bead width through the bead on plate (BOP) experiment under each condition to measure the bead height deviation. In the case of interlayer deposition, interlayer cooling was performed by setting the dwell time between depositions to zero, two, four, and six minutes. For AM technologies, the cooling temperature and time significantly affect the quality of the product. If the interlayer temperature is not properly maintained due to insufficient cooling time, the product may flow down or its quality is degraded by the unstable arc caused by the thermal deformation of the substrate.
Fig. 3 shows the deposition path to produce the specimens of the product and the schematic of sampling to analyze the mechanical properties of the product. SCM440 in a size of 150 mm (D) × 200 mm (H) × 25 mm (T) was used for the substrate. To conduct the tensile test and observe microstructures using the solid wire, deposition was performed in the form of a 2D-wall in a size of 130 mm (D) × 110 mm (H). The tensile test was conducted in two directions (horizontal and vertical), and specimens for the test were prepared in accordance with the ASTM E8 sub size. Ten specimens were prepared in each direction to analyze mechanical properties, and the tensile test was conducted at a cross-head speed of 3 mm/min at room temperature.
The specimens for microstructure analysis were cut into a size of 15 mm (D) × 20 mm (H) × 10 mm (T), and changes in microstructure were compared and analyzed according to the height of the product. A mixed solution of ethanol 95 ml + hydrochloric acid 5 ml + picric acid 3 g was used for microstructure analysis, and microstructures were observed using an optical microscope.
Specimens of 5 (pass) × 1 (layer), 3 (pass) × 3 (layer), and 5 (pass) × 5 (layer) were additionally prepared to compare and analyze the effective rectangle areas of products according to the inter-pass distance and interlayer distance between beads. In addition, the dwell time was set to zero, two, four, and six minutes in the experiment to compare changes in effective rectangle area according to the dwell time.
3. Results & Discussion
3.1 Comparison by deposition condition
Fig. 4 shows the results of comparing bead geometry according to the wire feed rate. Overall, no defect was observed inside the beads. As the wire feed rate increased, both the bead width and bead height showed a tendency to increase. The bead width increased as the wire feed rate increased because the molten pool was expanded by the increase in welding current and heat input26). As can be seen from the graph, however, while the bead width increased proportionally to the increase in wire feed rate, the bead height increased by approximately 0.14 mm as the wire feed rate increased from 7.7 to 9.0 m/min and only by approximately 0.03 mm as it increased from 9.0 to 9.9 m/min.
Fig. 5 shows the results of comparing the height deviation of the product, i.e., surface roughness, according to the inter-pass distance. For five deposited beads, the highest and lowest parts were measured to compare and analyze the height deviation. At a wire feed rate of 9.0 m/min, the results were 0.4 mm for 68%, 0.82 mm for 72%, and 1.02 mm for 76%, showing that the deviation tended to increase as the inter-pass distance increased. At a wire feed rate of 9.9 m/min, the results were 0.75 mm for 68%, 0.69 mm for 72%, and 0.98 mm for 76%, showing that the deviation was lowest at an inter-pass distance of 72%. Therefore, when deposition is performed under the aforementioned process conditions to secure stable beads, the conditions of 9.0 m/min at 68% and 9.9 m/min at 72%, which exhibited the lowest height deviation, are judged to be ideal conditions.
Fig. 6 shows the results of comparing the effective rectangle areas of products according to the inter-pass distance. Since deposition on the substrate shows somewhat different tendencies in bead size and geometry compared to continuous deposition on the product, different inter-pass distances were applied to the first and second layers, and the effective rectangle area was compared and analyzed as shown below. A wire feed rate of 9.9 m/min and inter-pass distances of 68 and 72% were applied to the first layer while a wire feed rate of 9.0 m/min and inter-pass distances of 68 and 72% were applied to the second layer in the experiment. These conditions were selected based on the surface roughness analysis results under the heat input conditions for the first and second layers mentioned above. The effective rectangle area was found to be 119.74 mm2 when the inter-pass distance of the first and second layers was 68% and 122.63 mm2 when it was 72%. The largest effective rectangle area of 123.90 mm2, however, was observed when different inter-pass distances were applied to the first and second layers (72% for the first layer and 68% for the second layer). During the deposition process, deposition on the substrate is judged to affect the product considering the alloy composition of the base metal and the size of the substrate. This means that bead formation is affected by the material in direct contact. In relation to this, Kim et al.27) reported that the average amount of deposition per layer decreased by 12% in the iron-nickel mixture section compared to IN625 single deposition.
Fig. 7 shows the deposited area/effective rectangle area ratio according to the dwell time in the WAAM process and the surface temperature according to the number of layers. For the WAAM process, such defects as flowing and sagging of the product and pores frequently occur due to the high heat generated during the deposition process. The accumulation of excessive heat forms an overheated molten pool, thereby causing pores and interlayer non-fusion. High heat accumulation during Ti alloy deposition causes brittle cracks by forming α’-martensite in large quantities. If the cooling rate is high, the molten pool rapidly solidifies and fine equiaxed grains are formed, thereby improving mechanical strength and increasing the risk of pore formation due to hydrogen entrapment. If the cooling rate is low, the formation of large dendrites is induced, leading to hot tear and microsegregation17,28,29). Repeated heating cycles cause thermal gradients due to the heating of the top of the product and the cooling of the bottom, resulting in structural distortion and delamination due to tensile stress concentration and unbalanced contraction30). In addition, an increase in dwell time is expected to improve mechanical properties, reduce residual stress, and enhance microstructure uniformity, but it decreases productivity in the process31). Therefore, heat management of the product is a very important factor, which should be considered in setting various deposition conditions. The dwell time was set to zero, two, four, and six minutes in the experiment. The deposited area/effective rectangle area ratio was found to be 77.87% for a dwell time of zero minute, 78.46% for two minutes, 80.72% for four minutes, and 83.34% for six minutes, showing that the ratio tended to increase as the dwell time increased. In relation to the temperature change according to the dwell time, the temperature was measured at corresponding positions after the arc was completely extinguished. The median values that correspond to the purple box were measured at the positions immediately before the arc was activated. As can be seen from the graph, the product surface temperature of the first layer was similar (approximately 450°C). As the number of layers increased, however, the surface temperature of the product increased. It was found that the temperature of the product decreased more significantly as the dwell time increased. Sufficient dwell time can reduce defects that may occur during the deposition process, but it has a negative impact on rapid production, which is the benefit of the WAAM process. From an industrial perspective, product temperature management is a very important factor for a high production speed and a low defect rate. Considering variables, such as the size of the substrate used and the deposition speed, the optimal dwell time was determined to be four minutes for the deposition of specimens in the form of a 2D-wall in the experiment of this study.
3.2 Microstructure analysis under the optimal deposition conditions
Fig. 8 shows a specimen deposited in the form of a 2D-wall by applying the solid wire to the WAAM process and the image of X-ray analysis results. A wire feed rate of 9.9 m/min and an inter-pass distance of 72% were applied to the first layer while a wire feed rate of 9.0 m/min and an inter-pass distance of 68% were selected from the second layer. The waiting time between beads was set at 30 seconds and the dwell time at four minutes. As can be seen from Fig. 8, deflection was observed at both ends of the product, but deposition was completed with no significant defect. The X-ray image also shows that there were few internal defects.
Fig. 9 shows the microstructure of the heat affected zone (HAZ) at the top and bottom of a 2D-wall specimen. The area of HAZ was wider and relatively coarse columnar grains were observed at the bottom of the product compared to the top. This appears to be because the bottom of the product was repeatedly exposed to the high heat generated from the WAAM process. In the case of the top of the product, fine equiaxed grains were observed in narrow HAZ due to the relatively low heating cycle compared to the bottom.
Fig. 10 shows the microstructures of the fusion zone (FZ) and HAZ according to the deposition height of the 2D-wall specimen. No significant change in grain size was observed from the FZs of the top, middle, and bottom of the product, and fine grains were densely distributed. In the case of HAZ, however, the size of grains decreased and columnar grains changed into equiaxed grains as the deposition height increased32). Columnar grains were formed at the bottom because the high heat generated by the arc was rapidly transferred downward and the temperature gradient at the solid-liquid interface was high. At the top, however, fine equiaxed grains were observed due to the multinucleation caused by significant supercooling as heat was slowly transferred in various directions.
3.3 Comparison of mechanical properties under the optimal deposition conditions
Fig. 11 shows the tensile test results in the two directions (horizontal and vertical) of the ER-70S-6 wire deposited by the WAAM process. In the tensile test results, the yield strengths in the two directions were 325.63 and 336.34 MPa, respectively, and similar tensile strength results (518.38 and 522.22 MPa) were observed33-35). The elongation in the horizontal direction (35.96%), however, was approximately 8.58% higher compared to the vertical direction (27.38%). This results from the anisotropy of the microstructure, which is mostly observed from the deposition process. Owing to relatively high heat transfer in the z-axis direction, columnar grains are formed in the heat flow direction36). In addition, relatively uniform microstructures are observed in the x- and y-axes due to the equiaxed grains, but stress concentration easily occurs in the z-axis during deformation due to the columnar structure. This appears to have caused the elongation of the z-axis to be lower than that of the x- or y-axis37). Repeated heating cycles induce directional microstructures by changing the thermal gradient and cooling rate during the deposition process by layer. This forms equiaxed grains in the horizontal direction when the thermal gradient is low and the cooling rate is high, but columnar grains grow dominantly in upper layers in the vertical direction with high thermal gradients and low cooling rates38,39). For the equiaxed structure, stress is distributed due to the uniform grain distribution. It is favorable for plastic deformation due to the delay in crack propagation and a large number of grain boundaries40,41). In the case of the columnar structure, however, the long direction of the structure coincides with the deposition direction. Thus, a tensile load causes stress concentration at the grain boundary and cracks propagate early along the grain boundary, thereby reducing the elongation42).
Fig. 11
Graph of tensile test results for horizental and vertical directions deposited with solid wire
The mechanical properties of the specimens prepared using the WAAM process were relatively lower than those provided by KISWEL, which manufactured the solid wire. This appears to be because the deposition process formed non-uniform and coarse microstructures under the influence of repeated heating cycles unlike the welding process, and the cooling rate relatively lower compared to the welding process under the influence of the accumulation of the high heat generated from previous layers affected mechanical properties43).
4. Conclusions
In this study, the wire feed rate, inter-pass distance, and dwell time were compared and analyzed as parameters to derive the optimal deposition conditions of the ER70S-6 solid wire in the wire arc additive manufacturing (WAAM) process. Research was conducted on microstructure and mechanical properties for various parameters, and the following conclusions were drawn.
As the wire feed rate increased, the bead width and bead height showed a tendency to increase. A wire feed rate of 9.0 to 9.9 m/min was selected as the optimal deposition condition. The condition of 9.0 m/min exhibited the lowest deviation of 0.4 mm when the inter-pass distance was 68% while the condition of 9.9 m/min showed the lowest deviation of 0.69 mm when it was 72%. In addition, the largest effective rectangle area of 123.90 mm2 was observed when an inter-pass distance of 72% was selected for the first layer and 68% from the second layer. In the case of the dwell time, the surface temperature of the first layer showed similar results regardless of the dwell time. As the height of the product increased, however, the change in dwell time affected the surface temperature of the product. As the dwell time increased, the deposited area/effective rectangle area ratio also showed a tendency to increase.
The solid wire applied to the WAAM process rapidly solidified to form microstructures with few defects, such as cracks and pores. In addition, there was no crack between the product and SCM440, which was used as substrate. For the fusion zone (FZ), fine grains were observed regardless of the deposition height. The heat affected zone (HAZ), however, was wider and relatively coarse grains were observed at the bottom of the product compared to the top. This appears to be because the repeated exposure of the bottom of the product to the high heat generated during the WAAM process affected the size and area of grains.
The yield strengths of the product in the horizontal and vertical directions were 325.63 and 336.34 MPa and the tensile strengths were 518.38 and 522.22 MPa, respectively, indicating no significant difference in mechanical properties. The elongation in the horizontal direction (35.96%), however, was approximately 8.58% higher than that in the vertical direction (27.38%). This results from the anisotropy of the microstructure, and it is judged that the vertical direction in which the grain boundary was parallel to the stress direction exhibited a relatively low elongation compared to the horizontal direction. In addition, the mechanical properties of the product were measured to be relatively lower than those provided by the manufacturer of the solid wire used in this study. This appears to be because the WAAM process formed non-uniform and coarse microstructures under the influence of repeated heating cycles, unlike the welding process, and the relatively lower cooling rate-due to the accumulation of high heat generated from previous layers-affected the mechanical properties.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
