Controlling Welding Residual Stress and Distortion of High-Strength Aluminum Alloy Thin Plates by a Trailing Hybrid High-Speed Gas Fluid Field

This paper presents an investigation of the welding residual stress and distortion of LY12 high-strength aluminum alloy (6061) by improving the local welding thermal and mechanical fields. A trailing hybrid high-speed gas fluid method was proposed and applied to decrease the welding residual stress and distortion of 6061 aluminum alloy efficiently. Firstly, the temperature and stress fields were calculated using the finite element simulation method, considering a trailing hybrid high-speed gas fluid field. The distance between the aerodynamic load and the heat source action was a key factor determined by the simulation method. In addition, the reasonable effective range of gas pressure was obtained. Subsequently, welding and distortion tests were conducted on the self-developed device under conventional welding and high-speed gas fluid field conditions. The results showed that an aerodynamic load under 30 MPa of gas pressure was available near the area at a distance of 20–28 mm from the heat source for thin plate welding distortion. The peak longitudinal residual tensile stresses in the weld’s mid-length section decreased by 77.73%, the peak residual compressive stresses decreased by 69.23% compared with conventional welding, and the deflection distortion disappeared almost entirely. The maximum deflection of the distortion was only 1.79 mm, which was 83.76% lower than the 11.02 mm of the conventional welding distortion. This validates that the method can simultaneously and greatly eliminate the welding residual stress and distortion.


Introduction
In the context of lightweight materials becoming increasingly mainstream, sheet metal structures are widely used in ships, machinery, chemicals, aerospace, and other fields. However, in general, warpage instability distortion readily occurs in thin plate welding, showing wave-shaped characteristics [1,2]. Welding distortion will seriously affect the accuracy of the dimensions and flatness of the sheet metal structure, making it a huge potential risk in the service process in areas with high safety requirements. Therefore, most sheet metal structures need to be straightened due to the welding distortion, which will increase the manufacturing cycle and production costs, especially for LY12 high-strength aluminum alloy (6061), which has a significant coefficient of linear expansion, leading to a larger distortion after welding. Thus, it is very critical to control its welding distortion.
Traditionally, controlling welding distortion is referred to as post-welding straightening, which requires a great amount of energy; therefore, it is limited in its application in industrial production [3][4][5][6][7][8]. Distortion control during the welding process can improve production efficiency; however, the current mechanism for controlling the weld distortion of outflow pipe

The Principle of Controlling the Distortion of Thin Plate Welding with High-Speed Welding
The welding process is a local transient inhomogeneous heating process [18][19][20]. When the weld is heated by a heat source, the metal in the area where the heat source acts is instantaneously heated to the melting point and forms a molten pool. However, the other solid areas do not reach the melting point, and the area adjacent to the molten pool expands by heat [21][22][23][24][25][26]. However, the metal in its vicinity is still in a colder state; thus, the molten pool expands due to the heat, and the metal is compressed and deformed by the action of the metal in its vicinity. When the heat source leaves the region and the molten pool begins to solidify, the metal in the region should contract, but it is restrained in its contraction process by the action of its nearby metal. At this time, the metal in this area is deformed in a tensile plastic manner, and after complete cooling, residual stresses are generated inside the plate. Therefore, the compressive strain generated during heating of the heat source is reduced, or the tensile strain generated during the cooling and solidification phase should increase [27][28][29][30][31]. In addition, when the residual compressive stress inside the plate is less than the critical instability stress of the plate, the purpose of eliminating the instability distortion can be achieved.
Based on this principle, Japanese scholars proposed the theory of intrinsic strain [32], which is expressed as: where [ε*] is the intrinsic strain, [εY] is the compressive strain, and [εL] is the tensile strain. That is, the intrinsic strain is composed of the compressive strain generated during the heating of the heat source and the tensile strain generated during the cooling and solidification process. During heating of the heat source, [εL] is much smaller than [εY]; thus, [ε*] is always negative. As long as [ε*] is not 0, residual stresses are generated. According to inherent strain theory, the purpose of controlling the welding stress and distortion is to decrease the inherent strain. The tensile strain in the weld area under a high-speed gas fluid loading consists of two components. On the one hand, there is the tensile strain due to the aerodynamic loading. On the other hand, there is the tensile strain influenced by shrinkage during the cooling of the weld. Both effects counteract the compressive strain generated by the thermal expansion of the weld metal. The ultimate goal is to reduce the inherent strain of the thin plate structure and control the welding distortion of the thin Based on this principle, Japanese scholars proposed the theory of intrinsic strain [32], which is expressed as: where [ε*] is the intrinsic strain, [ε Y ] is the compressive strain, and [ε L ] is the tensile strain. That is, the intrinsic strain is composed of the compressive strain generated during the heating of the heat source and the tensile strain generated during the cooling and solidification process. During heating of the heat source, [ε L ] is much smaller than [ε Y ]; thus, [ε*] is always negative. As long as [ε*] is not 0, residual stresses are generated. According to inherent strain theory, the purpose of controlling the welding stress and distortion is to decrease the inherent strain. The tensile strain in the weld area under a high-speed gas fluid loading consists of two components. On the one hand, there is the tensile strain due to the aerodynamic loading. On the other hand, there is the tensile strain influenced by shrinkage during the cooling of the weld. Both effects counteract the compressive strain generated by the thermal expansion of the weld metal. The ultimate goal is to reduce the inherent strain of the thin plate structure and control the welding distortion of the thin plate.

Parametric Characterization of the Aerodynamic Load
According to aerodynamics, high-speed gas fluid can produce mechanical effects [33]. When the distance between the high-speed gas fluid and the workpiece is small, the highspeed gas fluid will produce a stable laminar flow, which can be regarded as an ideal fluid, and it will produce a concentrated force on the surface of the workpiece. According to the theoretical basis of aeroelasticity [34,35], the relationship between the size of the aerodynamic load, F, and other parameters can be deduced, and its expression is as follows: where P 1 and P 2 are the gas pressures at the inlet and outlet of the outflow pipe, respectively; T 1 and T 2 are the gas temperatures at the inlet and outlet of the outflow pipe, respectively; Materials 2022, 15, 6451 4 of 17 V 1 is the gas flow rate at the inlet of the outflow pipe; ρ is the gas density; r is the radius of the outlet of the outflow pipe.
There is a range of loading distances behind the heat source, and when the loading distance exceeds this range of action, the mechanical effect of the aerodynamic load will not be able to regulate the residual stresses generated by the weld. As shown in Figure 2, the force applied to the weld area by the high-velocity gas behind the heat source can be characterized by the gas pressure: P1, P2, P3, . . . , Pn; the loading distance: d1, d2, d3, . . . , dn; the radius of the area of action: S1, S2, S3, . . . , Sn. The appropriate aerodynamic load can be matched by finetuning the values of each parameter. aerodynamic load, F, and other parameters can be deduced, and its expression is as follows: where P1 and P2 are the gas pressures at the inlet and outlet of the outflow pipe, respectively; T1 and T2 are the gas temperatures at the inlet and outlet of the outflow pipe, respectively; V1 is the gas flow rate at the inlet of the outflow pipe; ρ is the gas density; r is the radius of the outlet of the outflow pipe.
There is a range of loading distances behind the heat source, and when the loading distance exceeds this range of action, the mechanical effect of the aerodynamic load will not be able to regulate the residual stresses generated by the weld. As shown in Figure 2, the force applied to the weld area by the high-velocity gas behind the heat source can be characterized by the gas pressure: P1, P2, P3, …, Pn; the loading distance: d1, d2, d3, …, dn; the radius of the area of action: S1, S2, S3, …, Sn. The appropriate aerodynamic load can be matched by finetuning the values of each parameter.

Finite Element Modeling
Numerical finite element calculation of TIG welding test was carried out using an LY12 aluminum alloy sheet, as shown in Figure 3a. In order to accurately measure the deflection, grids 10 × 10 mm were carved on the specimen. The model size was 320 × 200 × 1 mm, as shown in Figure 3b, and the element type was an eight-node hexahedral element with 48,477 nodes and 32,000 individual elements. In order to ensure the accuracy of the calculation and improve the efficiency of the calculation, the overall number of computational meshes was reduced in the form of sparse transition, and only the critical areas of the weld and aerodynamic loading locations were meshed refined, as shown in Figure  3c.

Finite Element Modeling
Numerical finite element calculation of TIG welding test was carried out using an LY12 aluminum alloy sheet, as shown in Figure 3a. In order to accurately measure the deflection, grids 10 × 10 mm were carved on the specimen. The model size was 320 × 200 × 1 mm, as shown in Figure 3b, and the element type was an eight-node hexahedral element with 48,477 nodes and 32,000 individual elements. In order to ensure the accuracy of the calculation and improve the efficiency of the calculation, the overall number of computational meshes was reduced in the form of sparse transition, and only the critical areas of the weld and aerodynamic loading locations were meshed refined, as shown in Figure 3c. As shown in Table 1, the material parameters (modulus of elasticity (E), linear expansion parameter (α), yield limit (σs), specific heat (c), thermal conductivity (K), etc.) of the LY12 aluminum alloy sheet were dependent on the temperature, while the material's Poisson ratio (μ) and mass density (ρ) were constant values.  As shown in Table 1, the material parameters (modulus of elasticity (E), linear expansion parameter (α), yield limit (σ s ), specific heat (c), thermal conductivity (K), etc.) of the LY12 aluminum alloy sheet were dependent on the temperature, while the material's Poisson ratio (µ) and mass density (ρ) were constant values. Table 1. Material performance parameters [36]. The welding heat source in this study was modeled using a double-elliptical heat source [37], as shown in Figure 4. The temperature field of the second half of the arc was affected due to the accelerated cooling effect of the high-speed gas fluid behind the arc during the welding process. However, the effect on the temperature field of the front half of the arc was very small; therefore, a modified double-ellipsoidal heat source model was chosen, with the specific expressions as shown in Equations (3)-(5).  As shown in Table 1, the material parameters (modulus of elasticity (E), linear expansion parameter (α), yield limit (σs), specific heat (c), thermal conductivity (K), etc.) of the LY12 aluminum alloy sheet were dependent on the temperature, while the material's Poisson ratio (μ) and mass density (ρ) were constant values. The welding heat source in this study was modeled using a double-elliptical heat source [37], as shown in Figure 4. The temperature field of the second half of the arc was affected due to the accelerated cooling effect of the high-speed gas fluid behind the arc during the welding process. However, the effect on the temperature field of the front half of the arc was very small; therefore, a modified double-ellipsoidal heat source model was chosen, with the specific expressions as shown in Equations (3)-(5). The first half of the ellipsoidal heat source distribution is expressed as: The first half of the ellipsoidal heat source distribution is expressed as: The latter half of the ellipsoidal heat source distribution is expressed as: The total heat flow density is: where Q is the total power of the welding mobile heat source; f 1 and f 2 represent the ellipsoidal heat distribution functions of the front and rear parts, and f 1 + f 2 = 2; a, b, and c are the ellipsoidal shape parameters; λ is the heat source model correction factor, which is related to the flow rate of the high-speed gas and heat dissipation coefficient.

Development and Compilation of the Aerodynamic Load Subroutines
Since the aerodynamic load's impact was not an inherent boundary condition of the Marc software, a secondary development of the software was performed in order to enable the application of this boundary condition. The compilation of the moving load subroutine was completed with Fortran77 as the compilation language, and the impact effect of the aerodynamic load on the LY12 aluminum alloy sheet was realized.

Temperature Field with the Action of Welding with the Trailing of the Hybrid High-Speed Gas Fluid Field
Under the same welding specification parameters, the welding temperature field was simulated and calculated for both conventional welding and welding with a trailing high-speed gas flow. For the latter, the distance of the center of action of the aerodynamic load from the heat source was the key parameter, which was initially set to 12 mm. The calculated contour distribution of the welding temperature field at different moments under both conditions are shown in Figure 5.
( ) 1 2 x, , = + q y z q q (5) where Q is the total power of the welding mobile heat source; f1 and f2 represent the ellipsoidal heat distribution functions of the front and rear parts, and f1 + f2 = 2; a, b, and c are the ellipsoidal shape parameters; λ is the heat source model correction factor, which is related to the flow rate of the high-speed gas and heat dissipation coefficient.

Development and Compilation of the Aerodynamic Load Subroutines
Since the aerodynamic load's impact was not an inherent boundary condition of the Marc software, a secondary development of the software was performed in order to enable the application of this boundary condition. The compilation of the moving load subroutine was completed with Fortran77 as the compilation language, and the impact effect of the aerodynamic load on the LY12 aluminum alloy sheet was realized.

Temperature Field with the Action of Welding with the Trailing of the Hybrid High-Speed Gas Fluid Field
Under the same welding specification parameters, the welding temperature field was simulated and calculated for both conventional welding and welding with a trailing highspeed gas flow. For the latter, the distance of the center of action of the aerodynamic load from the heat source was the key parameter, which was initially set to 12 mm. The calculated contour distribution of the welding temperature field at different moments under both conditions are shown in Figure 5.  As can be seen from the figure, when the welding was at t = 20 s and t = 60 s, the temperature field of the welding under trailing hybrid high-speed gas fluid conditions was slightly different from conventional welding conditions, and the temperature field error was less than 2% and not significant. The forced convection effect generated by the high-speed gas fluid field on the temperature field of the high-strength aluminum alloy sheet was small; thus, the slight change in the temperature field with the trailing hybrid high-speed gas fluid almost did not affect the welding stress field. As can be seen from the figure, when the welding was at t = 20 s and t = 60 s, the temperature field of the welding under trailing hybrid high-speed gas fluid conditions was slightly different from conventional welding conditions, and the temperature field error was less than 2% and not significant. The forced convection effect generated by the high-speed gas fluid field on the temperature field of the high-strength aluminum alloy sheet was small; thus, the slight change in the temperature field with the trailing hybrid high-speed gas fluid almost did not affect the welding stress field.

Determination of the Aerodynamic Load Action Distance Range
The distance of the aerodynamic load from the heat source should be based on the yield strength of the base or weld metal at different locations from the heat source at the weld centerline, and the yield strength of the material is related to the temperature. The nodal temperatures of the weld metal at different distances from the heat source were collected in the path followed by the direction shown in Figure 6. As can be seen from the figure, when the welding was at t = 20 s and t = temperature field of the welding under trailing hybrid high-speed gas fluid co was slightly different from conventional welding conditions, and the temperat error was less than 2% and not significant. The forced convection effect generate high-speed gas fluid field on the temperature field of the high-strength alumin sheet was small; thus, the slight change in the temperature field with the trailin high-speed gas fluid almost did not affect the welding stress field.

Determination of the Aerodynamic Load Action Distance Range
The distance of the aerodynamic load from the heat source should be base yield strength of the base or weld metal at different locations from the heat sour weld centerline, and the yield strength of the material is related to the tempera nodal temperatures of the weld metal at different distances from the heat sou collected in the path followed by the direction shown in Figure 6. weld center path In order to analyze the relationship between the temperature and yield str different positions, the two curves were combined for comparison. The nodal tures of the weld metal at different distances from the heat source are shown in (curve 1), and the yield strength of the material corresponding to the node tempe different distances from the heat source was obtained according to the thermo In order to analyze the relationship between the temperature and yield strength at different positions, the two curves were combined for comparison. The nodal temperatures of the weld metal at different distances from the heat source are shown in Figure 7 (curve 1), and the yield strength of the material corresponding to the node temperature at different distances from the heat source was obtained according to the thermophysical property parameters of the material, as shown in Figure 7 (curve 2). As can be seen from Figure 7, the farther the distance from the heat source, the lower the node temperature. Within 12 mm from the heat source, the node temperature decreased faster, and when the distance from the heat source was greater than 12 mm, the node temperature decreased more slowly. The farther the distance from the heat source, the lower the temperature of the node, thus the higher the yield strength of the material. Within 12 mm from the center of the heat source, the yield strength of the material was only within 1 MPa. Ranging from 12 to 24 mm, the yield strength value of the material increased sharply. When the distance from the heat source was greater than 24 mm, the yield strength of the material steadily increased, and by 32 mm, the yield strength reached over 100 MPa.
According to the relationship between the distance from the heat source and the yield strength, this experiment could initially determine the range of the aerodynamic load action distance. When the material yield strength fluctuated within only 1 MPa, the material in this region was close to the "mechanical melting point", and unrecoverable plastic distortion easily took place. However, when the yield strength of the material reached 100 MPa or more, an aerodynamic load device was required to provide a high level of energy, which is difficult to perform. Therefore, the initial determination of the aerodynamic load action distance was 12-32 mm, as shown in the rectangular shadow area in Figure 7. In order to explore the action law of the aerodynamic load and reduce the number of simulation calculations, distances from the heat source of 12, 16, 20, 24, 28, and 32 mm were selected as the positions of action of the aerodynamic load. Table 2 lists the calculated range of the aerodynamic load meter at different loading distances.
property parameters of the material, as shown in Figure 7 (curve 2). As can be seen Figure 7, the farther the distance from the heat source, the lower the node temper Within 12 mm from the heat source, the node temperature decreased faster, and whe distance from the heat source was greater than 12 mm, the node temperature decr more slowly. The farther the distance from the heat source, the lower the temperatu the node, thus the higher the yield strength of the material. Within 12 mm from the c of the heat source, the yield strength of the material was only within 1 MPa. Ranging 12 to 24 mm, the yield strength value of the material increased sharply. When the dis from the heat source was greater than 24 mm, the yield strength of the material ste increased, and by 32 mm, the yield strength reached over 100 MPa. According to the relationship between the distance from the heat source and the strength, this experiment could initially determine the range of the aerodynamic loa tion distance. When the material yield strength fluctuated within only 1 MPa, the ma in this region was close to the "mechanical melting point", and unrecoverable plasti tortion easily took place. However, when the yield strength of the material reache MPa or more, an aerodynamic load device was required to provide a high level of en which is difficult to perform. Therefore, the initial determination of the aerodynamic action distance was 12-32 mm, as shown in the rectangular shadow area in Figure  order to explore the action law of the aerodynamic load and reduce the number of lation calculations, distances from the heat source of 12, 16, 20, 24, 28, and 32 mm selected as the positions of action of the aerodynamic load. Table 2 lists the calcu range of the aerodynamic load meter at different loading distances.

Variations in the Welding Residual Stress under an Aerodynamic Load
In order to analyze the effect of the different loading distances on the longitudinal residual stress, the longitudinal residual stress values in the mid-length section at different loading distances were extracted from the finite element simulation results for comparison with conventional welding conditions. For an aerodynamic load of 30 MPa and loading distances (d) of 20, 24, 28, and 32 mm, the longitudinal residual stress results are shown in Figure 8.
As can be seen from the figure, welding with high-speed gas fluid field conditions for the control of residual stresses in the weld was much more effective. In particular, in the weld zone, which is the area of direct action of the aerodynamic load, the residual stress decreased much more significantly. When the loading distances (d) were 20, 24, 28, and 32 mm, the stresses in the center of the weld were −70.34, −69.87, 11.47, and 146.66 MPa, respectively. The peak tensile stress decreased by 74.90%, 73.76%, 58.93%, and 15.01%, respectively, compared with conventional welding. Under the loading distances (d) of 20 and 24 mm, the residual stress at the center of the weld converted from tensile stress to compressive stress. residual stress, the longitudinal residual stress values in the mid-length section at d ent loading distances were extracted from the finite element simulation results for c parison with conventional welding conditions. For an aerodynamic load of 30 MPa loading distances (d) of 20, 24, 28, and 32 mm, the longitudinal residual stress result shown in Figure 8. As can be seen from the figure, welding with high-speed gas fluid field condit for the control of residual stresses in the weld was much more effective. In particula the weld zone, which is the area of direct action of the aerodynamic load, the resi stress decreased much more significantly. When the loading distances (d) were 20, 24 and 32 mm, the stresses in the center of the weld were −70.34, −69.87, 11.47, and 14 MPa, respectively. The peak tensile stress decreased by 74.90%, 73.76%, 58.93%, 15.01%, respectively, compared with conventional welding. Under the loading dista (d) of 20 and 24 mm, the residual stress at the center of the weld converted from te stress to compressive stress.
As the loading distance increased, the longitudinal residual stress had a similar tribution to that under conventional welding conditions. The peak tensile stress and compressive stress also increased with the increases in the loading distance, and the of tensile stress on the specimen gradually decreased. This was due to the fact that w the loading distance increases, the temperature of the metal in the area of action decre and the yield strength increases. However, the aerodynamic load was kept constant MPa, and the plastic ductile effect on the metal in the area of action was reduced. Th fore, the controlling effect of the longitudinal residual stress was reduced. When d and 24 mm, the aerodynamic load of 30 MPa was greater than the yield strength o material in the area of action. At this time, plastic distortion of the metal in area of ac occurred, and the elastic distortion recovery of the surrounding metal was limited. result, the stress inside the plate could no longer be further changed; thus, the differe in the longitudinal residual stress under d = 20 mm and d = 24 mm were smaller.

Analysis of the Welding Deflection Distortion under an Aerodynamic Load
The welding distortion of thin plates at different loading distances (d) is show Figure 9. As can be seen from the plot, when the loading distance (d) was less than 20 the required aerodynamic load was small enough to control the distortion of the L As the loading distance increased, the longitudinal residual stress had a similar distribution to that under conventional welding conditions. The peak tensile stress and peak compressive stress also increased with the increases in the loading distance, and the area of tensile stress on the specimen gradually decreased. This was due to the fact that when the loading distance increases, the temperature of the metal in the area of action decreases and the yield strength increases. However, the aerodynamic load was kept constant at 30 MPa, and the plastic ductile effect on the metal in the area of action was reduced. Therefore, the controlling effect of the longitudinal residual stress was reduced. When d = 20 and 24 mm, the aerodynamic load of 30 MPa was greater than the yield strength of the material in the area of action. At this time, plastic distortion of the metal in area of action occurred, and the elastic distortion recovery of the surrounding metal was limited. As a result, the stress inside the plate could no longer be further changed; thus, the differences in the longitudinal residual stress under d = 20 mm and d = 24 mm were smaller.

Analysis of the Welding Deflection Distortion under an Aerodynamic Load
The welding distortion of thin plates at different loading distances (d) is shown in Figure 9. As can be seen from the plot, when the loading distance (d) was less than 20 mm, the required aerodynamic load was small enough to control the distortion of the LY12 aluminum alloy sheet, but the controlling effect was not more evident. When the loading distance was greater than 32 mm, the aerodynamic load required to control the distortion of the LY12 aluminum alloy sheet was too large. Thus, when the loading distance was between 24 and 32 mm, the applied aerodynamic load could effectively control the welding distortion of the LY12 aluminum alloy thin plates. In addition, when the loading distance was 28 mm and the aerodynamic load was 30 MPa, the controlling effect was much better, and the weld displacement decreased by 94.20% compared with conventional welding.
of the LY12 aluminum alloy sheet was too large. Thus, when the loading distance was between 24 and 32 mm, the applied aerodynamic load could effectively control the welding distortion of the LY12 aluminum alloy thin plates. In addition, when the loading distance was 28 mm and the aerodynamic load was 30 MPa, the controlling effect was much better, and the weld displacement decreased by 94.20% compared with conventional welding. (e) (f)

Welding Test
In order to verify the validity of the simulation results, the test materials and dimensions of the specimen were the same as those of the finite element calculation model. Moreover, Table 3 lists the chemical composition of the materials, and the test apparatus consisted of a welding torch and a compressed gas pipe, as shown in Figure 10. AC TIG welding was selected for the test, and the welding was conducted in the form of surface deposition on the thin plates. Full penetration was achieved due to the thin dimensions of the thickness. The welding machine model was a Panasonic YC-300WX4, and the welding parameters are listed in Table 4. In order to verify the validity of the simulation results, the test materials and dimensions of the specimen were the same as those of the finite element calculation model. Moreover, Table 3 lists the chemical composition of the materials, and the test apparatus consisted of a welding torch and a compressed gas pipe, as shown in Figure 10. AC TIG welding was selected for the test, and the welding was conducted in the form of surface deposition on the thin plates. Full penetration was achieved due to the thin dimensions of the thickness. The welding machine model was a Panasonic YC-300WX4, and the welding parameters are listed in Table 4.   The aerodynamic load generator model was a PGH30-0.1XT with an adjustable gas pressure range from 0.1 to 60 MPa. Nitrogen of 99.99% purity was used as the test compressed gas, and argon of 99.99% purity was applied as the welding shielding gas.
The initial loading distance was 12 mm, and the loading distance was gradually increased in 1 mm increments until it reached 32 mm. Figure 11 shows the relationship between the specimen's size and the range of the loading distance (d). The welding was performed by controlling the pressure output of the oil-lubricated compressor to obtain different gas pressures. The test results showed that the specimen was well formed, and the distortion was well straightened when the loading distance (d) was 20 mm and the gas pressure was 30 MPa.  The aerodynamic load generator model was a PGH30-0.1XT with an adjustable gas pressure range from 0.1 to 60 MPa. Nitrogen of 99.99% purity was used as the test compressed gas, and argon of 99.99% purity was applied as the welding shielding gas.
The initial loading distance was 12 mm, and the loading distance was gradually increased in 1 mm increments until it reached 32 mm. Figure 11 shows the relationship between the specimen's size and the range of the loading distance (d). The welding was performed by controlling the pressure output of the oil-lubricated compressor to obtain different gas pressures. The test results showed that the specimen was well formed, and the distortion was well straightened when the loading distance (d) was 20 mm and the gas pressure was 30 MPa.

Longitudinal Residual Stress Test under a High Speed Gas Fluid Field during Welding
The residual stress test was conducted using the strip-cutting method, as shown in Figure 12. It was carried using the static strain gauge model CM-1A-20 to measure the stress of the welded specimen's mid-length section under conditions of conventional Figure 11. Adjustment range of the loading distance (d).

Longitudinal Residual Stress Test under a High Speed Gas Fluid Field during Welding
The residual stress test was conducted using the strip-cutting method, as shown in Figure 12. It was carried using the static strain gauge model CM-1A-20 to measure the stress of the welded specimen's mid-length section under conditions of conventional welding and a high-speed gas flow field (d = 20 mm, P = 30 MPa). Since the residual stresses on both sides of the weld were symmetrically distributed along the weld's centerline, it was sufficient to mount strain gauges on only one side of the welded specimen. Gauges were spaced close together near the weld where the residual stresses varied drastically, while the stress in the farther area varied slowly. The strain gauges were arranged schematically, as shown in Figure 13. Strips approximately correspondingly wide were cut from the specimen. 320 Figure 11. Adjustment range of the loading distance (d).

Longitudinal Residual Stress Test under a High Speed Gas Fluid Field during Welding
The residual stress test was conducted using the strip-cutting method, as shown in Figure 12. It was carried using the static strain gauge model CM-1A-20 to measure the stress of the welded specimen's mid-length section under conditions of conventional welding and a high-speed gas flow field (d = 20 mm, P = 30 MPa). Since the residual stresses on both sides of the weld were symmetrically distributed along the weld's centerline, it was sufficient to mount strain gauges on only one side of the welded specimen. Gauges were spaced close together near the weld where the residual stresses varied drastically, while the stress in the farther area varied slowly. The strain gauges were arranged schematically, as shown in Figure 13. Strips approximately correspondingly wide were cut from the specimen.  The residual stresses presented in the welded specimen were released by cutting the welded specimen along the mid-length section and then cutting both sides of the strain gauge along its long edge direction. The change in the elastic strain caused by the release of stress was indicated by a static strain device, and then the longitudinal residual stress was calculated using Hooke's law. The longitudinal residual distribution of the welded specimens under different conditions can be seen in Figure 14. Note that with the application of a trailing aerodynamic load during welding, the peak tensile stress had a larger reduction. The overall residual stress obviously decreased compared to that of conventional welding. Among them, the peak tensile residual stress location of the trailing aerodynamic load was different from that of conventional welding, slightly far from the center of the weld. Moreover, the longitudinal residual stress at the center of the weld and in the vicinity of the weld area significantly decreased. All the test results were in excellent agreement with the results of the FEM numerical simulation, which illustrates the accuracy of the above numerical simulation results. This was due to the fact that sufficient stretching was caused to the metal in the area of action by the application of the aerody- The residual stresses presented in the welded specimen were released by cutting the welded specimen along the mid-length section and then cutting both sides of the strain gauge along its long edge direction. The change in the elastic strain caused by the release of stress was indicated by a static strain device, and then the longitudinal residual stress was calculated using Hooke's law. The longitudinal residual distribution of the welded specimens under different conditions can be seen in Figure 14. Note that with the application of a trailing aerodynamic load during welding, the peak tensile stress had a larger reduction. The overall residual stress obviously decreased compared to that of conventional welding. Among them, the peak tensile residual stress location of the trailing aerodynamic load was different from that of conventional welding, slightly far from the center of the weld. Moreover, the longitudinal residual stress at the center of the weld and in the vicinity of the weld area significantly decreased. All the test results were in excellent agreement with the results of the FEM numerical simulation, which illustrates the accuracy of the above numerical simulation results. This was due to the fact that sufficient stretching was caused to the metal in the area of action by the application of the aerodynamic load, effectively counterbalancing the longitudinal compressive plastic distortion produced during the welding process. Furthermore, the inherent strain of the metal in the welded joint area was reduced, resulting in lower residual stresses in the sample. According to the results of the comparison, the maximum tensile residual stress under a high-speed gas flow field condition was 61.02 MPa, which decreased by 77.73% over that of conventional welding, and the peak residual compressive stress was −24.87 MPa, which decreased by 69.23%. specimens under different conditions can be seen in Figure 14. Note that with the application of a trailing aerodynamic load during welding, the peak tensile stress had a larger reduction. The overall residual stress obviously decreased compared to that of conventional welding. Among them, the peak tensile residual stress location of the trailing aerodynamic load was different from that of conventional welding, slightly far from the center of the weld. Moreover, the longitudinal residual stress at the center of the weld and in the vicinity of the weld area significantly decreased. All the test results were in excellent agreement with the results of the FEM numerical simulation, which illustrates the accuracy of the above numerical simulation results. This was due to the fact that sufficient stretching was caused to the metal in the area of action by the application of the aerodynamic load, effectively counterbalancing the longitudinal compressive plastic distortion produced during the welding process. Furthermore, the inherent strain of the metal in the welded joint area was reduced, resulting in lower residual stresses in the sample. According to the results of the comparison, the maximum tensile residual stress under a highspeed gas flow field condition was 61.02 MPa, which decreased by 77.73% over that of conventional welding, and the peak residual compressive stress was −24.87 MPa, which decreased by 69.23%.

Flexural Distortion Test under a High-Speed Gas Fluid Field during Welding
Due to the large coefficient of linear expansion and small modulus of elasticity for LY12 aluminum alloy, the residual stress took place in the weldment heated unevenly by the heating source, which caused it to easily deform in flexure after welding. In order to verify the ability to control the welding distortion of an LY12 aluminum alloy sheet under high-speed gas fluid field conditions, a comparison was made to analyze the distortion of the LY12 aluminum-alloy-welded specimens under conventional welding and high-speed gas fluid field conditions. As shown in Figure 15, a handheld laser 3D scanner was used to scan the out-of-plate distortion of the welded specimens of the LY12 aluminum alloy under two conditions. The model was an HSCAN771, with an accuracy of 0.01 mm, the test temperature was 20 • C, and the 3D point cloud data regarding the displacement of the specimen were obtained after scanning, as shown in Figure 16a,c.

Flexural Distortion Test under a High-Speed Gas Fluid Field during Welding
Due to the large coefficient of linear expansion and small modulus of elasticity for LY12 aluminum alloy, the residual stress took place in the weldment heated unevenly by the heating source, which caused it to easily deform in flexure after welding. In order to verify the ability to control the welding distortion of an LY12 aluminum alloy sheet under high-speed gas fluid field conditions, a comparison was made to analyze the distortion of the LY12 aluminum-alloy-welded specimens under conventional welding and high-speed gas fluid field conditions. As shown in Figure 15, a handheld laser 3D scanner was used to scan the out-of-plate distortion of the welded specimens of the LY12 aluminum alloy under two conditions. The model was an HSCAN771, with an accuracy of 0.01 mm, the test temperature was 20 °C, and the 3D point cloud data regarding the displacement of the specimen were obtained after scanning, as shown in Figure 16a,c.    (c) (d)  The laser scanning results under different conditions were compared with the finite element simulation results, as shown in Figure 16. The figure shows that the test results were in good agreement with the finite element simulation results, and the LY12 aluminum alloy conventionally welded specimens deformed in deflection, while almost no distortion occurred during the welding under high-speed gas field conditions. It was found that the maximum distortion occurred along the transverse edge of the aluminum alloy rectangular plate. With a ruler, experiments were conducted to measure the deflection distortion value of the plate's edge regarding the width of the welded thin plate specimen under different conditions.
The results showed that the plate edge deflection (δ) of the welded specimen was 11.02 mm under that of the conventional welding conditions, while the deflection was only 1.79 mm under conditions of a trailing high-speed gas fluid field. Compared to conventional welding, the deflection decreased by 83.76%. As shown in Figure 17, the results were consistent with the laser 3D scanning results, further proving that under trailing high-speed gas fluid field conditions, welding distortion can be effectively controlled. The laser scanning results under different conditions were compared with the finite element simulation results, as shown in Figure 16. The figure shows that the test results were in good agreement with the finite element simulation results, and the LY12 aluminum alloy conventionally welded specimens deformed in deflection, while almost no distortion occurred during the welding under high-speed gas field conditions. It was found that the maximum distortion occurred along the transverse edge of the aluminum alloy rectangular plate. With a ruler, experiments were conducted to measure the deflection distortion value of the plate's edge regarding the width of the welded thin plate specimen under different conditions.
The results showed that the plate edge deflection (δ) of the welded specimen was 11.02 mm under that of the conventional welding conditions, while the deflection was only 1.79 mm under conditions of a trailing high-speed gas fluid field. Compared to conventional welding, the deflection decreased by 83.76%. As shown in Figure 17, the results were consistent with the laser 3D scanning results, further proving that under trailing high-speed gas fluid field conditions, welding distortion can be effectively controlled.

Conclusions
The influence of welding with a trailing hybrid high-speed gas fluid field on the welding residual stress and distortion of LY12 aluminum alloy was studied. The following conclusions were drawn:

Conclusions
The influence of welding with a trailing hybrid high-speed gas fluid field on the welding residual stress and distortion of LY12 aluminum alloy was studied. The following conclusions were drawn: (1) A new method was proposed to control the welding distortion of high-strength aluminum alloy sheet by "welding with a trailing hybrid high-speed gas fluid field". It was based on the original perspectives of mechanical mechanisms and the concept of "flexible" control. Otherwise, a welding distortion control model was established with a trailing hybrid high-speed gas fluid field, which was also parametrically characterized; (2) Through the numerical simulations of the temperature field of the welding process under conditions of welding with a trailing hybrid high-speed gas fluid field, the effective action distance (d) between the aerodynamic load and the heat source was determined to range from 12 to 32 mm. Various aerodynamic loads were set according to the corresponding loading distance (d). The longitudinal residual stresses in the mid-length section were measured at an aerodynamic load loading distance of 20 mm and a gas pressure of 30 MPa; (3) The aerodynamic load applied to the weld could effectively reduce the residual stress of the weld. The control effect was good when the loading distance was 20-28 mm and the aerodynamic load was 30 MPa. The maximum tensile residual stress under a high-speed gas flow field condition was 61.02 MPa, which decreased by 77.73% over that of conventional welding, and the peak residual compressive stress was −24.87 MPa, which decreased by 69.23%. (4) The maximum deflection of LY12 aluminum-alloy-welded parts under conditions of welding with a trailing hybrid high-speed gas fluid field was 1.79 mm, which was 83.76% lower than the 11.02 mm of conventional welding.