Experimental and Numerical Investigation of the Thermal and Force Regulation Mechanism of Bypass Coupling Double-Sided Arc Welding Based on 6061 Aluminum Alloy

Abstract

In this study, 6061 aluminum alloy was proposed for welding using bypass coupling double-sided arc welding (BCO-DASW) to further improve its welding quality and efficiency. To gain insight into the thermal and force regulation mechanism of the BCO-DASW of 6061 aluminum alloy, the dynamic effects of the high-temperature plasma inside the arc with various parameters were fully compared and investigated through the combined method of the physical experiment and the numerical simulation. The thermal flow field of the hybrid arc was analyzed numerically. Furthermore, its working adaptability and mechanical behaviors were studied experimentally. The results show that a single penetration of the 6 mm sheets can be achieved without visible defects when the center offset of the arcs is within 3 mm on both sides of the base metal during BCO-DASW. Through the thermal analysis, it was found that, compared with the MIG process, the introduction of the bypass arc lead to a temperature decrease at the bottom of the hybrid arc due to energy redistribution. Furthermore, through the kinetic analysis, it was found that not only could the level of arc pressure be reduced, but also the action range of the arc pressure could be regulated up to 4.6 mm. The thermal force regulation mechanism worked together to enhance the stability of the molten pool and achieves good joint strength during the BCO-DASW of 6061 aluminum alloy. This research not only has great significance in further improving the welding quality and efficiency of aluminum alloy, but also deeper understanding of the energy regulation mechanism during aluminum alloy welding.

1. Introduction

Because of its high specific strength, strong machinability, excellent corrosion resistance, and other performance advantages [1], 6061 aluminum alloy is widely used in modern essential equipment manufacturing sections, including aerospace [2,3], shipbuilding [4,5], transportation [6,7], etc. MIG (metal inert gas arc welding) technology is often applied in recent manufacturing methods of aluminum alloy complex structures of medium plate due to its high quality, high efficiency, and high flexibility [8]. However, molten pool depression and the thermal crack are prone to occur due to the difficulty of regulating the thermal force effect of the conventional process during the 6061 aluminum alloy welding [9]. Based on this problem, it is hard to thoroughly melt through medium-thickness sheets of aluminum alloys. However, practical working situations often require medium and thick sheets of 6061 aluminum alloys to be penetrated to meet the service strength requirements. MIG-TIG (tungsten inert gas arc welding) hybrid welding is an efficient process that can redistribute arc energy, and thus higher welding quality and efficiency can be achieved compared to the single-arc MIG. Moreover, double-sided arc welding (DSAW) has obvious advantages for achieving thorough penetration in medium-thickness sheets and for further increasing the welding quality and efficiency compared to one-sided welding. To further optimize these processes for different working requirements, a lot of research work has been carried out in the field of science and engineering.

Kanemaru et al. [10] investigated the interaction mechanism between arcs during MIG-TIG hybrid welding. They found that the TIG arc enhanced the welding process stability and controlled the heat input. Cui et al. [11] also studied the effects of the TIG arc on the hybrid process. They thought that with the use of the TIG arc, the degree of metal vapor decreased, and the hybrid arc temperature increased compared to the single MIG process. The MIG-TIG hybrid process with a bypass-current was first proposed by Zhang et al. [12]. Due to its special technical characteristics such as low heat input, it has been successfully applied to metal droplet control [13,14], dissimilar metals joining [15,16], and metal additive manufacturing [17,18]. Furthermore, its process characteristics have attracted wide attention in the field of research and engineering in recent years [19,20,21]. On the other hand, Sridhar et al. [22] explored the influence of process parameters on the steel joint by double-sided continuous welding. They reported that the increase in the welding current is the fundamental factor leading to a reduction in the tensile strength of the weld. Xiong et al. [23] analyzed the performance of the high-strength steel joint with double-sided TIG cross-welding. They demonstrated that the welded structure had good tensile strength and hardness due to good preheating and cooling mechanisms. Zheng et al. [24] obtained a good joint of aluminum/steel dissimilar metals using MIG-TIG double-sided welding. They found that the strength of the aluminum/steel joint decreased with increasing heat input. Wei et al. [25] achieved a penetration of 8 mm sheets of 5083 aluminum alloy using double-sided TIG cross-welding. They found that double-sided welding can achieve higher welding quality than single-sided welding.

In this work, bypass coupling double-sided arc welding (BCO-DASW) based on the 6061 aluminum alloy was proposed and studied to further enhance welding quality and efficiency. Compared with the conventional process, the thermal and force at the front side of the sheets can efficiently be regulated by the altered behavior of charged particles in the MIG-TIG hybrid arc. Furthermore, the back arc can regulate the direction of the welding current flowing into the base metal, making it more efficient in serving the base metal melting. Two-sided arcs work in synergy to achieve higher quality, higher efficiency welding of medium and thick sheets of 6061 aluminum alloy. To deeply understand its energy regulation mechanism, the effects of different parameters on the welding stability are tested and compared to uncover the rules of the BCO-DASW for the welding of 6061 aluminum alloy. Furthermore, to gain insight into the fundamental contribution of the enhancement mechanism of welding reliability, the thermal fluid behaviors of the BCO-DASW were investigated with the help of the numerical analysis and the current density and the arc pressure were studied via a physical experiment. The research not only contributes to improving the welding quality and efficiency of aluminum alloys in medium-thickness sheets but also provides significant benefits in enriching the energy regulation theory of metal manufacturing.

2. Experimental and Calculated Details

2.1. Experiment Set-Up

Figure 1 shows the experiment set-up of the BCO-DASW of 6061 aluminum alloy in this work. The working principle of this process is that, on the front side of the sheets, the MIG-TIG hybrid process is used to achieve thermal force regulation and improve the stability of the welding process, and on the back side of the sheet, the TIG arc is coaxially combined with the MIG-TIG hybrid arc to achieve higher quality and higher efficiency welding, as shown in Figure 1a. I_w is the main current flowing through the welding wire, I_b is the bypass current flowing through the tungsten, and I_m is the working current flowing through the base metal. The value of I_w is set by the main power source and the value of I_b can be adjusted in real time using the auxiliary power supply to meet the work requirements. The parameters of these variables are listed in

Table 1. Process parameters.

Case	Iw (A)	Ib (A)	MIG-Gas (L min−1)	TIG-Gas (L min−1)
1	130	-	15	-
2	110	20	15	5
3	90	40	15	5

. The parameters are set based on the principle of constant total heat input. Figure 2 shows the experiment content under different process parameters. The experiment first explores the case without the back arc to obtain the optimal parameters, as shown in Figure 2a. Then, the experiment using the BCO-DASW hybrid arc with various offsets along and perpendicular to the welding direction was conducted separately to explore the process adaptability, as shown in Figure 2b,c. In addition, the measurement experiment of current density and arc pressure are shown in Figure 1b,c. The working principle is to prefabricate an orifice on the water-cooled copper block. One end of the orifice is applied to conduct the hybrid arc, while the other end of the orifice is equipped with a pressure sensor or current probe. When the hybrid arc is stably burned, the analog signal of the arc pressure or current density is received by the pressure sensor or current probe and converted into a numerical signal through a digital/analog conversion card to be transmitted to the central control and processing system. It is worth noting that the fast measurement method was applied because the metal droplets would cover the orifice. It was conducted by first activating the bypass arc switch and subsequently activating the main arc switch. Therefore, the hybrid arc could be immediately energized and stabilized to obtain valid data before the metal droplets fill the orifice. The chemical composition of the base material used for the experiment is shown in

Table 2. Chemical composition of base material.

Base	Fe	Cu	Mn	Ti	Si	Cr	Mg	Zn	Al
6061	≤0.7	0.15~0.4	0.15	0.15	0.4~0.8	0.04~0.35	0.8~1.2	0.25	Balanced

. Other process parameters applied in the experiment such as the main arc voltage is set at 22 V and welding speed is 60 cm/min. The thickness of the substrate used in this work is 6 mm, and the sheets did not need to be beveled. The joint with a width of 10 mm was stretched through a universal testing machine (Instron 5967) at a rate of 2 mm/min. Finally, the actual image of the arc during the 6061 aluminum alloy welding is extracted by a high-speed camera system with a frame rate of 4000 fps.

Figure 3a shows the 3D computational domain of the hybrid arc behaviors established in this work. The whole domain is a cylinder with a base radius of 10 mm and a height of 12 mm. Based on the setup of the physical experiment, the diameter of the welding wire is 1.2 mm and the diameter of the tungsten electrode is 2.4 mm. The distance between the end of the tungsten electrode and the upper surface of the sheet is 5 mm. The welding wire is perpendicular to the upper surface of the sheet. The distance between the welding wire and the tungsten electrode is 3 mm and the angle between them is 45°, as shown in Figure 3b. The computational domain adopts a tetrahedral mesh with a total of 549,921 discretized cells.

2.2. Simulation Procedure

2.2.1. Basic Assumptions and Control Equations

To rationally analyze and calculate the physical behaviors and the thermal fluid force mechanism of the hybrid arc during 6061 aluminum alloy welding, the following basic assumptions need to be stated in advance: (1) the flow of the high-temperature plasma is laminar; (2) the arc plasma is optically thin and satisfies the local thermodynamic equilibrium (LTE); (3) the gas phase in the computational domain is pure argon gas (99.99%), whose thermo-physical properties, such as the density, the specific heat capacity, the thermal conductivity, and the viscosity, etc., are taken from the reference [26]; (4) the interaction in the anode zone at the welding wire end and the cathode zone at the tungsten end is not considered; and (5) energy and mass losses due to metal droplets and vapor are not considered [27]. Accordingly, the control equations can be given:

Mass continuum equation [28]:

$$\nabla \cdot (\rho \overrightarrow{v})=0$$

(1)

where ρ is density; $\overrightarrow{v}$ is velocity vector.

Momentum conservation equation [28]:

$$\frac{\partial (\rho \overrightarrow{v})}{\partial t}+\nabla \cdot (\rho \overrightarrow{v}\overrightarrow{v})=-\nabla P-\nabla \cdot \tau +\overrightarrow{J}\cdot \overrightarrow{B}+\rho g$$

(2)

where t is time; P is pressure; $\overrightarrow{B}$ is magnetic field strength; $\overrightarrow{J}$ is current density; and τ is viscous shear. g is the gravity acceleration.

Energy conservation equation [28]:

$$\frac{\partial (\rho H)}{\partial t}+\nabla \cdot (\rho \overrightarrow{v}H)=\frac{{(\overrightarrow{J})}^2}{{\sigma }_e}+\nabla \cdot (\frac{k}{Cp}\nabla H)+\frac{5k_b}{2e}\overrightarrow{J}\cdot \nabla H-S_R$$

(3)

where H is enthalpy; Cp is the Specific heat; k is thermal conductivity; σ_e is conductivity; k_b is Stefan–Boltzmann constant; e is electron charge; and S_R is thermal radiation energy loss.

Current conservation equation [29]:

$$\nabla \cdot \overrightarrow{J}=-\nabla \cdot ({\sigma }_e\nabla \phi )=0$$

(4)

where φ is the potential.

Magnetic field intensity [29]:

$$\overrightarrow{B}=\nabla \times A$$

(5)

$${\nabla }^{2}A=-{\mu }_0\overrightarrow{J}$$

(6)

where A is the magnetic vector and µ₀ is the vacuum permeability.

2.2.2. Boundary Conditions

The boundary surfaces of the numerical model in this work are given in detail in Figure 2a. The velocity, temperature, and electric and magnetic potential vectors on these boundary surfaces are detailed in

Table 3. Boundary conditions.

Boundary Surface	Temperature (K)	Velocity (m/s)	Electric Potential (V)	Magnetic Potential (Wb m−1)
Outlet	400	$\partial \overrightarrow{v}/\partial n=0$	$\partial \phi /\partial n=0$	0
Surface_Bottom	1000	-	-	0
MIG_Inlet	400	VMIG_shielding	$\partial \phi /\partial n=0$	$\partial A/\partial n=0$
TIG_Inlet	400	VTIG_shielding	$\partial \phi /\partial n=0$	$\partial A/\partial n=0$
Surface_MIG_top	2000	-	${\sigma }_e\frac{\partial \phi }{\partial n}=\frac{I_w}{\pi R_w^2}$ [32]	$\partial A/\partial n=0$
Surface_TIG_top	3000	-	${\sigma }_e\frac{\partial \phi }{\partial n}=-\frac{I_b}{\pi R_b^2}$ [32]	$\partial A/\partial n=0$
Surface_MIG_side	1000	-	$\partial \phi /\partial n=0$	0
Surface_TIG_nozzle	1000	-	$\partial \phi /\partial n=0$	0
Surface_TIG_side	2000	-	$\partial \phi /\partial n=0$	0

R_w and R_b are the conductive working radius of the tungsten and the welding wire end, respectively.

. It should be noted that the boundary surface temperatures were set, tested, and adjusted with reference to the literature [30,31].

2.2.3. Numerical Calculation Method

The computational procedure for the thermal fluid force effects of the hybrid arc physical behaviors during 6061 aluminum alloy welding is based on the commercial CFD software FLUENT (version 2019R1). The UDF (User-Defined Functions) for momentum (electromagnetic forces) and energy (joule heat and thermal radiation) source terms and the UDS (User-Defined Scalars) for the electric potential and magnetic potential were designed and compiled using the C++ procedure. The control equations are solved by the SIMPLEC algorithm, and the energy and momentum equations as well as the UDS are discretized in the second-order windward format.

3. Results and Discussion

3.1. BCO-DASW Mechanism Analysis

Figure 4 compares the experiment results of the BCO-DASW without the back arc. It can be found that when the main current (I_w) was 130 A (case 1 in Table 1), the molten pool had an obvious depression phenomenon, which led to the face side of the weld collapsing and the root side of the weld being poorly formed. Furthermore, keeping the total current unchanged, when the bypass current (I_b) was set at 20 A (case 2 in Table 1), it can be found that the depression phenomenon of the molten pool still existed. When the bypass current was 40 A (case 3 in Table 1), the molten pool depression phenomenon was eliminated, and the weld surface was bright and evenly. This indicates that the MIG-TIG hybrid process can redistribute the heat input to improve the stability of the weld pool compared to the conventional arc.

Figure 5 compares the experimental results of the BCO-DASW with different offsets perpendicular to the welding direction. When the offset is 0 mm, it was found that the weld is uniformly formed and the sheets are well penetrated. When the offset was 1 mm, the centerlines of the front and back of the weld are slightly shifted, but the penetration of the sheets was achieved. When the offset was 2 mm, the centerline of the front and back of the weld shifted and the sheets were still penetrated. However, when the offset was 3 mm, the centerlines of the front and back of the weld significantly shifted and no penetration of the sheets was achieved. Figure 6 compares the experimental results of the BCO-DASW with different offsets along the welding direction. It was also found that, when the offset dimension exceeded 3 mm, the BCO-DASW method was unable to achieve sheet penetration. This indicates that BCO-DASW can achieve penetrating 6 mm 6061 aluminum alloy sheets in a single pass without beveling. Furthermore, no visible defects were found in the weld. Furthermore, the BCO-DASW method had a reasonable offset of 3 mm both along and perpendicular to the welding direction.

Analysis of the BCO-DASW process mechanism for 6061 aluminum alloy is displayed in Figure 7. In the initial stage of welding, both ends of the substrate started to melt under the action of the BCO-DASW hybrid arc. The heat from the arcs at both sides of the substrate gradually spread to the center of the base metal, as shown in Figure 7a. With the thermal accumulation, the volume of the molten pools at both sides of the base material gradually increased, while the heat-affected zone produced a superimposed action effect, as shown in Figure 7b. With the further accumulation of heat, the molten pools at the front and back of the sheet started to contact each other and melted through the base metal, as shown in Figure 7c,d. Combined with the above experiment, the center line of the back arc and the top arc did not deviate too much. The excessive distance made the heat-affected zones of the base material deviation, and thus not enough heat accumulates in the middle of the sheet, and the base material did not melted through the substrate. Furthermore, the main current is the key to the formation of the base material penetration. However, the aluminum alloy molten pool collapsed due to the excessive current. The introduction and regulation of the bypass arc ensured the penetration of the base metal and eliminated the depression of the molten pool. Accordingly, it was necessary to carry out an in-depth study on the thermal force regulation mechanism of the bypass arc for the DASW hybrid process.

3.2. Thermal Regulation Mechanism of Hybrid Arc

Figure 8 compares the calculated results of the arc morphology with different parameters from Table 1. The computational results shown in the blue dashed box used the parameter set of case 1, the computational results shown in the green dashed box used the parameter set of case 2, and the computational results shown in the green dashed box used the parameter set of case 3. Moreover, this color correspondence strategy was also applied to the following discussion and analysis. In addition, the variable λ, defined in this work as the regulation coefficient is introduced here, which is equal to the ratio of the bypass current to the main current. By comparing the X-Z section of the calculation results, it can be found that the MIG arc tended to be pulled by the bypass arc. Furthermore, this tendency increased significantly as the λ increased from 0 to 0.44. In addition, the temperature of the bypass arc also increased significantly with the increase in λ, especially in the center of the bypass arc. On the other hand, from the Y-Z cross-section of the calculated results, the hybrid arc showed a symmetrical distribution whether the bypass current was loaded or not. By further comparing the arc morphology of the Y-Z section, it was be found that the hybrid arc tended to shrink more compared to the MIG arc with the increase in the λ, as shown in Figure 9. This effect ultimately lead to a decrease in the width of the metal molten pool, which is consistent with reports in reference [33].

Table 4. Arc profile comparison between experimental and calculated results.

λ	Experimental Results	Binarization	Calculated Results
0
0.18
0.44

shows the actual images of the arc with the process parameters in Table 1, extracted by the high-speed camera system during the 6061 aluminum alloy welding experiment. It was found that the volume of the high-temperature plasma tended to decrease significantly with the increase in λ, and the numerical calculation showed the same conclusion. The arc images obtained from the physical experiment were binarized and compared with the calculated results under the corresponding parameters, and these results are also shown in Table 4. It should be pointed out that it is difficult to obtain the true value of the arc profile obtained by binarization. Therefore, a qualitative analysis of the arc behavior trend was carried out here. Isotherms used to extract the arc profile in the calculation results is 8500 K. It can be found that, under the same process parameters, the trends of the arc behavior obtained from the calculation results matched the experiment results, respectively. This indicates that the numerical model developed in this work is reliable. However, there are still errors between the experimental and calculated results. This is because the effect of metal droplets is not considered here for the moment. In addition, there are inevitable errors in the data processing.

Figure 10 and Figure 11 compare the temperature field distribution at the bottom of the arc with the λ increase. It can be found that, when λ is equal to 0, the thermal field of the hybrid arc bottom has the highest peak temperature and the widest range of action comparing λ = 0.18 and λ = 0.44. From the physical experiment, the molten pool depression of 6061 aluminum alloy occurred. In contrast, the thermal field at the bottom of the composite arc had the lowest peak temperature and narrowest range of action. At this point, the results of the physical experiment demonstrated that the molten pool depression phenomenon was eliminated. Specifically, when λ = 0, the maximum temperature of the hybrid arc was 6877 K, and the action range of the high-temperature plasma perpendicular to the welding direction was 15.2 mm; when λ = 0.18, the maximum temperature of the hybrid arc was 6200 K, and the action range of the high-temperature plasma perpendicular to the welding direction was 11 mm; and when λ = 0.44, the maximum temperature of the hybrid arc was 5433 K, and the action range of the high-temperature plasma perpendicular to the welding direction was 10.6 mm. It can be found that when the distribution coefficient is increased by 0.44, the maximum temperature of the hybrid arc decreases by 1444 K, and the action range decreases by 4.6 mm. This indicates that the BCO-DASW has a highly efficient energy regulation mechanism for the welding process of medium-thick plate 6061 aluminum alloy. Moreover, combined with the above experimental results, there is no obvious defect in the weld channel under the BCO-DASW process, which indicates that the redistribution of heat inhibits the tendency of thermal cracking of 6061 aluminum alloy, and thus improves the weld quality.

3.3. Force Regulation Mechanism of Hybrid Arc

Figure 12 compares the flow behaviors of charged particles inside the hybrid with different λ. In general, the gas phase particles entered the arc zone and then were activated into high-temperature plasma gas clusters under the excitation of the electromagnetic field. It was found that, when λ was equal to 0, the maximum velocity of the charged particles was 205.6 m/s; when λ was equal to 0.18, the maximum velocity of the charged particles was 150.6 m/s; and when λ was equal to 0.44, the maximum velocity of the charged particles was 107.4 m/s, as shown in the orange dashed box. Furthermore, the high-velocity charged particles were mainly concentrated in the middle of the arc. When the bypass arc was introduced, the conductive area of the hybrid arc increased along the welding direction, increasing the speed of some charged particles entering the arc, as shown in the black dashed box. Specifically, when λ was equal to 0, its average speed was about 21 m/s; when λ was equal to 0.18, its average speed was about 41 m/s; and when λ was equal to 0.44, its average speed was about 59 m/s.

Figure 13 and Figure 14 compare the current density and arc pressure measurement results at the bottom of the hybrid arc with different λ. It was found that the current density at the bottom of the arc tends to decrease gradually as the λ increases no matter along or perpendicular to the welding direction. When λ was equal to 0, the peak arc bottom pressure was 650 Pa; when λ was equal to 0.18, the peak arc bottom pressure was 512 Pa; and when λ was equal to 0.44, the peak arc bottom pressure was 249 Pa. In addition, along the welding direction, the action range of the peak arc bottom pressure was about 4 mm; perpendicular to the welding direction, the arc bottom pressure decreased with the increase in distance from the center. Overall, the arc pressure and the current density had the same tendency along the welding direction and perpendicular to the welding direction, respectively. When λ was elevated by 0.44, the maximum decrease in arc pressure was about 401 Pa, which was located in the center region of the arc. This was because when λ is 0, the particle flow velocity at the center of the arc was about 1.9 times higher than that when λ was equal to 0.44. In addition, at the front of the arc, on the other hand, the arc pressure and current density were slightly higher when λ was equal to 0.44 compared to the other two cases. This is due to the introduction of the bypass arc leading to an increase in the conductive area of the arc zone, and thus the velocity of the particles moving under the tungsten pole is enhanced by a factor of 2.8. However, due to the overall lower velocity of the particles at the front of the arc, the pressure level is also overall lower. Combined with the above analysis, the BCO-DASW hybrid arc has a good thermal fluid force regulation mechanism, and thus has good adaptability to the welding process for medium-thickness plates of 6061 aluminum alloy.

Finally, Figure 15 shows the fracture morphology of the tensile specimen under the BCO-DASW process when λ is equal to 0.44. The specimen fractured at a distance of 6 mm from the weld at the base metal, away from the heat-affected zone, and the port showed a typical ductile fracture. From the stress–strain curve, it was found the first half of the loading is a linear transformation. When it reached 160 MPa, the curve started to flatten which meant that the specimen yielded at this time and continued loading until 210 MPa, the joint fracture point. The tensile experiment results show that the strength of the joint welded by the BCO-DASW is higher than the base material, which indicates that the thermal and force regulation mechanisms have favorable strength-enhancing effects.

4. Conclusions

In this work, a research approach combining a physical experiment and a numerical analysis was adopted for thoroughly exploring the thermal force regulation mechanism of bypass coupling double-sided arc welding based on 6061 aluminum alloy. The following conclusions were obtained:

(1)

Compared to the conventional single arc, the temperature at the bottom of the hybrid arc decreases by up to 1444 K due to the energy redistribution by the hybrid arc.

(2)

Compared to the single MIG arc, the center pressure of the hybrid arc bottom is reduced by a maximum of 401 Pa due to a 1.9 times decrease in charged particle flow velocity in the arc region. Thermal and force regulation mechanisms work synergistically to eliminate the collapse of the molten pool and inhibit the tendency of the thermal cracking of 6061 aluminum alloy to occur.

(3)

Failure of the specimen in the tensile test occurs in the base metal, which indicates that the joint strength of the BCO-DASW method is high enough under the action of thermal and force regulation mechanism.