Abstract
Aluminum alloy 5083 has low density, mechanical properties, high corrosion resistance, and high welding capability. Due to high thermal conductivity, specific heat, and latent heat, as well as the relatively high coefficient of thermal expansion of aluminum, achieving a connection with good mechanical properties is always very important. This study aimed to investigate nanoparticles’ surprising and unexpected effects in forming various microstructures, which directly improve the mechanical properties of the welding process as a new and surprising idea. The effects of siliconized Zn-graphene oxide complex nanoparticles on weld geometry, mechanical properties, and microstructure of AA5083 were investigated. The gas metal arc welding process welded the samples, and different amounts of nanoparticles were investigated. The results reveal that utilizing nanoparticles can be affected by the weld geometry and the properties of weld metal. 0.25 g of nanoparticles had low face reinforcement, and the bead width and penetration depth dramatically increased. The presence of nanoparticles inside the molten zone and its effect on grain size improved mechanical properties, according to the Hall–Petch relationship. The ultimate tensile strength and yield strength of the S0.25 sample increased by 58.24 and 28.28%, respectively, compared to the welded sample without the presence of nanoparticles. Also, the ductility of this sample during the failure test showed that elongation increased by 36.75%.
1. Introduction
Aluminum alloys are used in many industrial fields because of their low density, high thermal and electrical conductivity, good corrosion resistance, and good workability [1–8]. AA5083, one of the 5xxx series aluminum alloys, is used in shipbuilding due to its mechanical properties and high corrosion resistance. One of the essential features of aluminum 5083 is its high weldability so that it can be welded with most welding processes, such as electric arc [9, 10]. Although different methods are used in welding aluminum alloys, metal inert gas and tungsten inert gas are generally used to join them [10]. Gas metal arc welding (GMAW) is a fusion welding process that uses a shielding gas to protect the weld pool [11]. GMAW has a very high efficiency compared to shielded metal arc welding, and it offers many advantages, for instance, continuous electrode, high deposition rate, and high speed, as well as without slag. Furthermore, GMAW is one of the most practical processes for welding different types of metals and alloys [12–15], especially for AA5083 [16–19]. Moreover, utilizing nanoparticles is an effective and awe-inspiring option to achieve higher mechanical properties, increased corrosion resistance, more proper microstructures, and more ductility in welding joints. Several studies have been conducted to investigate the mechanical properties and microstructures of the weld metal; for instance, Maurya et al. studied the effect of different types of carbonaceous particle reinforcement on the mechanical properties of Al6061. They also stated that samples with carbonaceous reinforcement had higher mechanical properties, and graphene had the best performance [20]. In another study, Fattahi et al. used graphene/aluminum composite nanoparticles to investigate aluminum joints’ mechanical and microstructure properties and reported increased tensile strength and weld microstructure [21]. Also, in similar studies, the effect of nanoparticles has been investigated [22–25]. Although the previous research conducted in the field of welding with the presence of nanoparticles generally proves the usefulness of this method to achieve a better welding joint, it should be noted that the investigation of the use of different nanoparticles was proposed as a way to obtain a weld metal with desirable properties due to their inherent properties and compositions and the effects they can have on microstructure and mechanical properties. In this study, as a new and challenging idea, the nanoparticles of the siliconized Zn-graphene oxide complex (functionalized graphene oxide nanoparticles) were used to connect AA5083 plates with the MIG process. Also, in addition to investigating the effects of these nanoparticles on microstructure and mechanical properties, the impact of oxygen in the nanoparticles on the penetration depth of the weld metal was investigated.
2. Siliconized Zn-Graphene Oxide Complex Nanoparticles
Carbon nanostructures are uniquely positioned in nanoscience and technology due to their high elastic modulus and unique electrical, chemical, and mechanical properties. They are used in various fields, such as sensors, energy storage, and composites [26, 27]. Graphene oxide is a single-layer structure of graphite oxide with oxygenated functional groups such as carbonyl, hydroxyl, and epoxide. It also has a unique structure, high flexibility, and excellent physical and chemical stability [28–30]. The properties and chemical composition of graphene oxide nanopowder are shown in Table 1. Reactivity and interactions between nanoparticles and surroundings affect their application significantly. One of the unavoidable issues associated with nanoparticles is their inherent instability over long periods, so coating nanoparticles using organic and mineral molecules and modifying surface nanoparticles can increase the potential of using them in various fields and applications. Usually, electrostatic chemical adsorption (ligand addition) and covalent bonding (ligand substitution) are processes used to modify the surface of nanoparticles [31, 32]. Also, silanization is performed to achieve a more stable complex during the synthesis of nanoparticles (the cover containing hydroxyl groups is coated with silane molecules). Silanization creates a more stable complex by forming a stronger bond on the surface [33, 34]. In this study, the surface of nanoparticles was functionalized by Zn+2 cations. Also, due to the presence of oxygenated functional groups in graphene oxide, the APTES silane molecule was used for silanization (Figures 1 and 2).


3. Experimental
In the present study, the plates of AA5083 were welded by using the KARA TCK 600P MIG machine (Figure 3) to 20 × 10 × 6 mm thick. SFA/AWS A5.10: ER5183 as a welding wire with a diameter of 1.2 mm was used and developed to provide the highest strengths possible in the as-welded condition of alloy 5083 and other similar high magnesium alloys. The chemical compositions and mechanical properties of the welding wire and AA5083 plate are reported in Table 2. Pure argon was used as a shielding gas, and the flow rate was kept to 20 liters per minute. Direct current with the electrode positive (DCEP) was used for electrical cleaning, higher deposition rate, and breaking of oxide layers [13]. Several experiments were performed to achieve optimal values of the parameters (current was selected at 261.5 amps, arc voltage was set at 24.2 volts, travel speed was determined at 36.2 cm/min, and the deposition rate was established at 73.4 g/min). Nanoparticles scattered when the shielding gas was used, creating a longitudinal groove on the connecting edges to overcome this challenge (Figure 4, the distance from the top edges was 1 mm), and different amounts of nanoparticles (0.25, 0.50, and 0.75 g) were placed inside the groove.


4. Results and Discussion
4.1. Welding Geometry
The samples were welded, and Figure 5 shows the surface of welded, the weld cross-sectional area, and the microstructures of the samples. As the results show, in sample S0.75, welded with 0.75 g of nanoparticles, the geometry of the weld containing penetration depth, bead width, and weld height was 7.4 mm, 3 mm, and 9 mm, respectively. The weld cross-sectional area had inclusion in addition to a lack of fusion in the weld joint. Many spatters accompanied the arc, and a few nanoparticles were agglomerated in the boiling pool, which was observed as cavities in the cross section of the weld metal. Also, there were hot cracks in the cross section of the weld. The S0.50 sample that contained 0.5 g of nanoparticles had less spatter, and the surface of the weld was much better than that of S0.75. The weld cross-sectional area in S0.50 was similar to that in S0.75, but its inclusions were smaller. The geometry of the weld contains penetration depth, bead width, and weld height which were 7.8 mm, 1.9 mm, and 11 mm, respectively. In the S0.25 sample containing 0.25 g of nanoparticles, the weld had a smooth surface without cutoff and inclusion, and the weld’s height was lower than in the previous samples. The weld pool in this sample was more comprehensive, but the penetration depth was much higher. The weld’s penetration depth, bead width, and height were 8.9 mm, 12 mm, and 1.5 mm, respectively. While in the sample without nanoparticles, the value of these parameters was 4.5 mm, 10 mm, and 3 mm, respectively. S0.25 was something more than the other samples, and the thing that makes this sample unique is its penetration depth; another thing was to be the dramatically smooth surface of the weld. Also, the VT visual inspection and NDT inspections confirmed that this sample was much better. Therefore, S0.25 was considered a basis for comparison in subsequent experiments. Figure 5 shows the surface of welded, the weld cross-sectional area, and the microstructures of the samples. Spatter is a factor that can cause cavities, insufficient penetration, and disruption in welding cycles. The spattering of large particles usually occurs due to the low intensity of the current compared to the diameter of the welding wire, or the length of the arc is very high (very high voltage), which causes droplets not to be transferred in a straight axis. This type of spraying is prevalent in welding and causes turbulence in the shielding gas flow. But because welding parameters with the same values were used to apply the same conditions during the welding of samples with different amounts of nanoparticles, it can be concluded that the cause of less sputtering and also the smooth welding surface in S0.25 with 0.25 g of nanoparticles are related to the absorption of incoming heat by nanoparticles and its transfer to the bottom area of the weld pool. As a result, the temperature of the weld pool is reduced, and spraying is also reduced. This heat transfer process has not been completed entirely in samples with 0.50 g and 0.75 g of nanoparticles.

4.2. Microstructural Characterization
The microstructure of specimens is shown in Figure 6. A microstructural analysis of AA5083 shows that the matrix of this alloy is a fine-grained structure consisting of an alpha-solid solution phase, and particles of solid magnesium solution are distributed in the aluminum matrix. The aluminum matrix includes three types of sediments: silicon-magnesium with the composition of Mg2Si, magnesium-rich sediments with Mg2Al3 composition, and iron-rich deposits with Al6(Fe-Mn) composition; also, in addition to Fe, Mn, and Al, some Cr can be possibly to found in it [35]. To etch the samples, HBF4 solution was used [36]. After etching, light microscope images were taken. The matrix contains two Al-Al and Al-Mg interfaces, where the Al-Mg interfaces are observed as holes (Figure 6).

The FESEM images, elemental analysis, and atomic analysis of specimens are shown in Figure 7. Magnesium tends to develop on grain boundaries; it accumulates in the prone places around the cores of nanoparticles. The tiny dimensions of nanoparticles in large numbers have played the role of nucleation, and it caused to spread magnesium throughout the weld metal by heterogeneous nucleation.

Figure 8 shows the FESEM images, elemental analysis, and atomic analysis for three points of the base metal (as reference). It contained Mg as the main element, Fe, Mn, and Ti. Also, Figure 8 shows the FESEM images and elemental and atomic analyses of the weld metal in the S0.00 sample. The S0.00 sample contained different amounts of Al, Mg, Mn, Fe, and Ti (compounds of the base metal).

Figure 9 shows the elemental and atomic analysis of the points determined in the S0.25 sample in two areas of the weld metal. The results showed that Al, Mg, Mn, and Fe related to the base metal contained different amounts of C, O, and Zn. Graphene is a two-dimensional single/multilayer of carbon atoms with a compact hexagonal, carbon-filled surface structure [37]. So, the nanoparticles in this study can be a source of C, O, and Zn. Also, these elements inside the weld metal confirmed the success of adding nanoparticles to the weld metal.

All samples were welded under the same condition. Sample S0.00 did not report any weight percentage of oxygen. Therefore, the amount of oxygen shown in the EDS analysis of sample S0.25 is related to the nanoparticle (graphene oxide) that decomposed and released due to high heat. Oxygen in S0.25 can cause a concentrated arc and reverse Marangoni flow inside the weld pool [38, 39]. Full-bodied turns and the change of reverse Marangoni flow lead to more heat, which can significantly affect the microstructure of molten and heat-affected areas. As microscopic images of different regions of welded samples were shown (Figures 10 and 11), grains in sample S0.00 were elongated in the heat-affected zone. In contrast, in sample S0.25, grains were deformed almost spherical and axially. It can be considered that the reversal of the Marangoni flow and the presence of more heat inside the weld pool have caused a more significant thermal effect in the region’s heat-affected zone and spheroidization of the grains. Furthermore, the penetration depth in S0.25 is greater than S0.00, confirming the presence of the active oxygen element and the change of the Marangoni flow into the weld pool (Figures 10 and 11). The presence of nanoparticles in the molten zone and the mechanism of heterogeneous nucleation and grain growth have caused sample S0.25 to have more refined grains than sample S0.00. Also, due to the high cooling rate and heterogeneous heat of the GMAW process, grains can change from columnar to elongated, and of course, 0.15% of Ti in electrode compounds can affect grain refinement [40, 41].


4.3. Mechanical Properties and Fractography Analysis
In this study, the tensile test was used to investigate the mechanical properties of samples. Also, its results were used to determine the behavior of the material, such as the range of elastic and plastic, elongation, ultimate tensile strength (UTS), and yield strength (YS) in samples [42]. According to the ASTM-E8-subsize standard, samples were tested by using a SANTAM STM-600 traction machine under the ISO/IEC170258M standard at 22°C and 27% humidity (capacity was 10 tons, and speed was 1 mm/min). The results showed that the S0.25 sample broke under 8.68 KN of force in the weld zone (the weld cross-sectional area was 34.85 mm2), and the S0.00 sample, which did not use nanoparticles, broke under a force of 3.79 KN. Figure 12 shows the stress-strain diagram, and Figure 13 shows the values of UTS and YS in MPa. The UTS and YS in the S0.25 sample were 58.84% and 28.24%, respectively, which were higher than those of the S0.00 sample. Since the tensile and yield strength represent the material’s strength, the weld strength in S0.25 has increased compared to the weld sample without nanoparticles. The UTS and YS in the base metal (no thermal effects and no sediments at the grain boundaries) were 295.61 MPa and 218.94 MPa, respectively. Defects such as porosity in the weld metal can affect the final strength of the samples. Also, local changes in temperature cause changes in the rate of dissolution and redeposition (especially β-shaped needle deposits, which are one of the most influential factors in strengthening aluminum alloys). Generally, four mechanisms of strain hardening, solute hardening, precipitation hardening, and grain size hardening are used to strengthen aluminum alloys. Therefore, the presence of the Mg element with the solute-hardening mechanism can cause a high strength of AA5083 [8, 43, 44]. The results of metallography and SEM images of the cross section of the sample welded with 0.25 g of nanoparticles (S0.25) and comparing it with the sample welded without nanoparticles (S0.00) showed that the presence of nanoparticles in the molten zone and the mechanism of heterogeneous nucleation and grain growth as well as creating suitable places for magnesium deposition have caused magnesium to be distributed on the entire surface of the weld metal. So the presence of nanoparticles inside the molten zone and its effect on grain size improved mechanical properties, according to the Hall–Petch relationship [45]. The elongation parameters and cross-sectional area show the ductility of materials, so the elongation rate was measured to compare the ductility of samples during the failure test. The relative elongation rate of S0.25 (with a thickness and width of 6.01 × 5.80 mm2) was 16%, and also, the relative elongation rate of S0.00 (with a thickness and width of 6.03 × 5.57 mm2) was 6.5%. A comparison of the results showed that ductility in S0.25 is higher, and the elongation rate in this sample was 36.75% and higher than in S0.00. Generally, the decrease in the percentage of elongation in S0.00 can occur due to a wide variety of reasons, such as porosity and hot cracks. Also, the reduced load-bearing surface will result in premature failure. In welding of AA5083 by the GMAW process, due to the soft nature of the aluminum alloy, the fracture surface in the weld metal has dimples with smooth oval edges, and the size and depth of the holes created in the fracture section indicate the degree of fracture softness [16]. Figure 14 shows the scanning electron microscopy images and the weight percentage of the elements present on the broken surface of the S0.25 and S0.00 samples. Examination and comparison of the amplitude on the fracture surface showed that in S0.25, dimples indicate the ductile fracture mechanism. At the same time, S0.00 had more profound dimple growth, and the failure surface mainly consisted of interconnected cavities.



5. Conclusions
In this study, sheets of AA5083 with 6 mm thickness were welded by gas metal arc welding under the protection of argon gas (MIG process). The effect of functionalized graphene oxide nanoparticles was investigated on the weld geometry, microstructure, and mechanical properties of AA5083. The visible results were as follows:(i)Utilizing nanoparticles significantly affected the weld geometry, especially the penetration depth. Face reinforcement was low in the S0.25 sample containing 0.25 g of nanoparticles, and the weld pool in this sample was vaster. The penetration depth was much higher than in the previous samples.(ii)The oxygen in nanoparticles inverted the direction of Marangoni flow in the welding pool. In addition, the presence of nanoparticles has caused the arc’s focus, so these mechanisms led to an increased penetration depth in the S0.25 sample.(iii)Nanoparticles as centers of growth, as well as creating suitable places for magnesium deposition, caused magnesium to spread over the entire surface of the weld metal. So, according to the Hall–Petch relationship, grain refinement improved mechanical properties. The tensile test showed that the ultimate tensile strength and yield strength increased by 28.24% and 58.24%, respectively, compared to the welded sample without nanoparticles.(iv)A comparison of the ductility of samples during the failure test showed that the rate of elongation in the S0.25 sample was 36.75% and higher than in the S0.00 sample.
Data Availability
The data used to support the findings of this study are included within the article.
Conflicts of Interest
The authors declare that they have no conflicts of interest.
Authors’ Contributions
Farhad Rahmati performed experiments, prepared the manuscript, and analyzed data; Masood Aghakhani designed and supervised the whole project and revised and analyzed data; and Farhad Kolahan analyzed and revised the manuscript. All the authors read and approved this paper.
Acknowledgments
The authors would like to thank Dr. Eshagh Karimi, Dr. Shahab Zangeneh, Mr. Farzad Pahnaneh, and the Razi University of Kermanshah for their assistance throughout the research.