Effect of Precise TIG Welding Pool Temperature Control on Microstructure and Mechanical Properties of 7072 Aluminum Alloy Joints

Abstract

This study investigates the effect of TIG weld pool temperature on the microstructure and mechanical properties of crack-repaired joints in 7072 aluminum alloy. To address poor temperature stability and slow response in indirect temperature control during TIG welding by adjusting the parameters, a new closed-loop molten pool temperature control method is proposed. Experimental comparisons were conducted to evaluate its effectiveness. The results show that using real-time molten pool temperature as direct feedback allows the welding current to be adjusted dynamically. This approach enables precise control of heat input. It also achieves real-time tracking and stable regulation of the molten pool temperature during welding. Temperature-controlled welding yields a more uniform joint microstructure with reduced porosity. Notably, the joint exhibits optimal comprehensive mechanical properties at 1825 °C, with a tensile strength of 328 MPa and an elongation of 9.9%. Overall, precise control of the TIG weld pool temperature effectively improves the quality and performance uniformity of aluminum alloy crack repairs.

1. Introduction

Aluminum alloys not only have the advantages of corrosion resistance, high electrical conductivity, and light weight [1] but also possess unique properties such as fatigue resistance and high reflectivity [2], making them widely used in aerospace, electronics, automotive, optical equipment, and other fields [3]. During long-term service of Al alloys as engineering components under the combined action of cyclic loads and environmental corrosion, they are prone to microcrack initiation, which gradually evolves into macrocracks and eventually induces structural fracture failure [4]. There are many techniques capable of repairing surface cracks on aluminum alloys, including TIG welding, laser forming repair, cold spray, friction stir welding, etc. Thanks to the application of these technologies in aluminum alloy crack repair, the service reliability and safety of engineering components have been greatly improved. Compared with 5xxx (Al–Mg) and 6xxx (Al–Mg–Si) alloys, 7xxx series alloys exhibit higher specific strength. However, their application is limited by weldability issues. Problems such as hot cracking, porosity, and softening in the fusion zone are often observed. Among them, 7072 aluminum alloy shows moderate strength. It is widely used in welded joints of armored vehicles. It is also applied in low-temperature pressure vessels. Components for liquid propulsion engines in aerospace systems are manufactured from this alloy. In addition, it is used in bridge girder structures for highway and railway systems [5].

Tungsten Inert Gas (TIG) welding employs a tungsten electrode to generate an arc as the heat source [6]. Under the protection of inert gases such as argon, the base metal and filler metal are melted to achieve metallurgical bonding. This process is widely applied in fields including construction, automotive, shipbuilding, railway, and aerospace [7]. Compared with cold spray and laser forming repair, TIG welding crack repair has the advantages of simple operation, high-quality welds, flexibility, and suitability for repairing cracks in complex shapes [8]. Special attention must be paid to the energy input during the TIG welding process, as the heat-affected zone (HAZ) can experience grain coarsening and precipitation phase dissolution at high temperatures, forming a softened zone (with hardness reduced by 30–50%) [9]. More importantly, improper temperature control can cause excessive temperature gradients in the molten pool, affecting the uniformity of the heat-affected zone microstructure and, consequently, the performance of the joints. For example, imprecise control of the molten pool temperature often leads to undercutting, porosity, cracks, and spatter during the TIG welding process [10].

At present, many studies on TIG welding of aluminum alloys focus on optimizing welding process parameters (welding current/voltage, tungsten diameter, gas flow rate and wire feed speed), which influence heat input and weld pool temperature to some extent. Faraji et al. [11] found that the welding current slightly increases the temperature, while the welding speed decreases it and widens its distribution, as determined through numerical simulation, experiments, and statistical analysis. Yu et al. [12] proposed a weld penetration state recognition method based on infrared thermal imaging and an artificial neural network (ANN). This method involves adjusting welding current parameters while synchronously monitoring the welding temperature field using an infrared thermal camera. To minimize interference from the dynamic behavior of the molten pool, the solidified weld region behind the molten pool was selected as the region of interest (ROI). Temperature field distribution features were extracted from the ROI to construct an ANN classification model. Experimental results demonstrated that the model achieved a recognition accuracy of over 96% for three weld penetration states: incomplete penetration, full penetration, and excessive penetration. Yu et al. [13] developed a low-cost infrared online monitoring system applicable to the welding process. By adjusting the welding speed and capturing real-time temperature distribution data perpendicular to the welding direction, the system identifies anomaly features in the temperature field that correlate with welding defects. The results demonstrate that the method effectively detects three typical types of welding defects—undercut, humping bead, and lack of fusion—thereby confirming a consistent correspondence between specific defect types and the characteristic signatures of infrared thermal signals.

However, this indirect temperature control mode has two inherent limitations. First, the mapping between welding parameters and the actual molten pool temperature is nonlinear. It is easily affected by factors such as ambient temperature and workpiece size. As a result, temperature stability becomes poor. Second, the response of temperature regulation is slow. It is difficult to track the dynamic variation in molten pool temperature in real time during welding. In existing studies, direct temperature measurement methods are mainly used for temperature monitoring and process diagnosis. In this work, a closed-loop temperature control method for TIG welding is proposed. The real-time molten pool temperature is used as direct feedback. The welding current is actively adjusted by the control system. As a result, the molten pool temperature is maintained near a preset target value. Precise control of the molten pool heat input is therefore achieved.

Through experiments, this study verifies the impact of direct control of the molten pool temperature on welding quality. Controlled and uncontrolled weld pool temperature experiments were designed for comparison. In the uncontrolled tests, welding was performed at fixed currents of 170 A, 175 A, and 180 A. For the controlled tests, a closed-loop temperature regulation system was used to maintain the weld pool at three predetermined average temperatures (1750 °C, 1825 °C, and 1900 °C), which corresponded to the mean pool temperatures measured during the uncontrolled experiments at each current level. The microstructure and mechanical properties of welded joints obtained under temperature-controlled and non-temperature-controlled conditions were systematically studied through metallographic analysis, SEM and EDS characterization, tensile tests, and hardness measurements. It is found that precise control of the molten pool temperature can reduce hot cracks and porosity, enhance the strength and toughness of welds, optimize microstructure, and reduce thermal deformation, significantly improving welding quality. This strategy provides a unique process approach for high-precision welding of Al alloy components.

2. Materials and Methods

2.1. Experimental Material and Setup

As-received 7072 aluminum alloy is used as the welding plate and ER5356 as the welding wire in this study. The plate was 200 mm × 100 mm × 5 mm, and the diameter of the welding wire was 1.2 mm. The weldability of 7xxx high-strength aluminum alloys is mainly limited by their high susceptibility to hot cracking and heat-affected zone softening [14]. Using 5xxx-series welding wire can improve the quality of welded joints [15]. The chemical compositions of ER5356 welding wire and 7072 aluminum alloy plates are shown in

Table 1. 7072 Aluminum Alloy and ER5356 Welding Wire Chemical Composition (wt.%).

Materials	Si	Fe	Cu	Mn	Mg	Zn	Ti	Cr	Al
7072	0.1	0.33	1.5	0.03	2.68	5.8	0.03	0.23	Bal.
ER5356	≤0.25	≤0.1	≤0.1	0.05~0.2	4.5~5.5	≤0.1	—	0.05~0.2	Bal.

.

The welding device used in the study is a Panasonic YC-350WX5HGE IGBT inverter-controlled AC/DC argon arc welding machine (Tangshan, China). It has a constant-current power source with a drooping characteristic, and the current remains relatively stable as the arc length changes. By conducting closed-loop temperature control experiments and performing comparative analysis of data and results with open-loop experiments using fixed welding currents, this study investigates the enhancing effect of molten pool temperature control on the forming quality of TIG welding. The uncontrolled weld-pool-temperature experiments were designed with welding current as the variable. A DIKAI OPTOELECTRONICS IT-8H3 infrared thermometer(Wuhan Dikai Optoelectronics Technologies Co., Ltd., Wuhan, China) was used to measure the real-time welding temperature, with a range of 800~2600 °C. The experimental process parameters are shown in

Table 2. Parameters of welding experiments in this study.

Parameters	Non-Temperature-Controlled	Temperature-Controlled
Welding current/A	170, 175, 180	/
Welding temperature/°C	/	1750, 1825, 1900
Wire feeding speed/(cm·min−1)	260
Welding speed/(mm·s−1)	3
Argon flow rate/(L·min−1)	15
Tungsten electrode height/mm	3

.

The molten pool temperature closed-loop control system consists of four core threads: the temperature acquisition thread, the control algorithm thread, the welding machine control thread, and the data visualization thread, as shown in Figure 1. During the TIG welding process, an infrared pyrometer is employed to measure the molten pool temperature in real time. The system periodically sends temperature query commands to the sensor, which returns temperature readings at a fixed sampling rate. Once the temperature data are successfully acquired, the temperature signal is transmitted to the control algorithm thread. In the control algorithm thread, the real-time measured temperature is compared with the preset target temperature. The temperature deviation is calculated based on the difference between the measured value and the target value, and the welding current is adjusted through the control algorithm to compensate for this deviation. The calculated current command is then packaged and transmitted to the welding machine control thread. In the welding machine control thread, the updated current parameters are temporarily stored and verified. After receiving confirmation from the welding machine that the current setting has been successfully applied, the new current command is transmitted to the welding machine via RS485 communication. The welding current is continuously updated, and the corresponding temperature response is fed back to the system, thereby forming a complete closed-loop control process.

2.2. Experimental Process

The experimental parameters were determined based on systematic pre-experimental studies. Initially, single-factor experiments on welding current revealed that when the current fell below 170 A, insufficient heat input led to inconsistent weld penetration and incomplete fusion defects. Conversely, when the current exceeded 180 A, excessive heat input resulted in a significant expansion of the heat-affected zone and workpiece deformation. Therefore, current levels of 170 A, 175 A, and 180 A were selected as the gradients for the non-temperature-controlled experiments to cover the complete process window from critical fusion to overheating-induced deformation. Subsequently, real-time monitoring of the weld pool temperature under these current parameters showed a stable distribution within the range of 1700 °C to 2000 °C.

Single-factor experiments indicated that when the welding current was below 170 A, insufficient heat input resulted in inadequate melting of the base metal; the deposited molten filler metal formed a small wetting angle, exhibited poor fluidity, and showed excessive filler metal overflow. When the current exceeded 180 A, excessive heat input led to a significant expansion of the heat-affected zone and induced workpiece deformation. Therefore, welding currents of 170 A, 175 A, and 180 A were selected as the current gradients for the non-temperature-controlled experiments. Real-time monitoring results further showed that under these current conditions, the weld pool temperature was stably distributed within the range of 1700–2000 °C. To investigate the effect of precise weld pool temperature control under comparable heat input conditions on weld microstructural uniformity and mechanical properties, target temperatures of 1750 °C, 1825 °C, and 1900 °C were selected for the temperature-controlled experiments, enabling a systematic comparison between temperature-controlled and non-temperature-controlled processes.

This approach allows for a systematic comparison of the effects between temperature-controlled and non-temperature-controlled processes under similar heat input conditions. The temperature profiles are shown in Figure 2. As can be seen from Figure 2, under non-temperature-controlled conditions, variations in welding current cause the temperature to continuously increase over the 60 s welding process, leading to differences in microstructure and mechanical properties across different regions of the weld. In contrast, under temperature-controlled conditions, the temperature remains stable, expecting a homogeneous microstructure and superior weld performance. Due to the unstable heat input during the arc ignition and extinction stages of welding, which are prone to defects such as lack of fusion, these stages were excluded from the data analysis. Only the welding segments in which the arc and molten pool reached a stable state were selected for investigation.

Before welding, prefabricated cracks (1 mm wide and 2 mm deep) were cut into the aluminum alloy plate using wire cutting. The welding equipment and the welding process are shown in Figure 3a. After 60 s of welding to repair the cracks, cross-sectional samples of the weld joints were cut by wire cutting, then ground, polished, and etched. The morphology and microstructure of the samples were observed using a metallographic microscope and a ZEISS Sigma 500 scanning electron microscope (SEM) (Carl Zeiss AG, Jena, Germany) equipped with an energy dispersive-spectrometer (EDS). Tensile tests were carried out according to the international standard ISO 6892-1:2019 [16], “Metallic materials-Tensile testing-Part 1: Method of test at room temperature”, and the results were averaged from at least three tests (Figure 3b). After the tensile tests, the samples were cut and polished (Figure 3c), and the fracture morphology after stretching was observed using SEM. The specific technical approach is described as follows. A non-contact infrared thermometer (DIKAI OPTOELECTRONICS IT-8H3, Wuhan Dikai Optoelectronics Technologies Co., Ltd., Wuhan, China) is used to continuously measure the temperature at the center of the welding molten pool in real time. The sampling frequency is 200 Hz, and the measurement accuracy is ±5 °C. The measured temperature data are transmitted to the main control unit and compared in real time with a preset welding temperature threshold, such as 1750 °C. If the measured temperature deviates from the target value, the main control unit immediately sends an adjustment command to the welding power supply. The welding current is actively regulated with a predefined step size of 2 A. For example, the current is fine-tuned from 150 A to 152 A. In this way, closed-loop and precise control of the molten pool temperature is achieved. This control strategy ensures stable heat input. It also accurately extends the molten pool lifetime. As a result, welding formation quality is effectively improved.

3. Results

3.1. Macro-Morphology of TIG Welded Joints

Figure 4 shows the surface and cross-sectional Optical Microscopy images of the joints under different welding parameters. For non-temperature-controlled joints, as the welding current increases, the arc penetrates deeper into the prefabricated cracks and fuses with the base material. As the arc penetration increases, the fusion depth also increases until it fully penetrates the base material. At the same time, the ends of the joints collapse along the welding direction, affecting weld quality. Among the three welding current parameters, the residual height on the top side is greater than that on the bottom side under 170 A and 175 A welding currents. Under 180 A, the residual height on the top of the weld is nearly flat, while residual height appears on the bottom. The top region, directly exposed to the arc heat source, allows the molten pool to fully develop and achieve a smooth surface. The bottom is limited by the base material thickness, resulting in insufficient heat accumulation and preventing the liquid metal from fully spreading. Low current leads to insufficient heat input, causing overflow of deposited metal and resulting in increased weld reinforcement and uneven morphology; high current enhances heat input, allowing molten pool driving forces to dominate the forming process, leading to weld surfaces that become flat or even concave [17].

For comparison, temperature control eliminates the increase in molten pool temperature during continuous welding by reducing heat dissipation and heat accumulation, improving the uniformity of different parts of the weld seam and the surface smoothness of a single-layer molten pool, resulting in an enhancement of the dimensional accuracy of the weld seam.

Figure 5 shows the OM morphology at high magnification under temperature-controlled conditions, where obvious pores are present in the weld zone (WZ). The evolution of microstructure can be observed from Figure 5, which shows that the weld zone mainly exhibits an equiaxed grain structure, with relatively fine grains and a uniform distribution. From the weld zone to the heat-affected zone, the grain morphology in the fusion zone (FZ) transitions from planar crystals to cellular crystals and then to columnar dendrites. A small number of columnar grains are present near the heat-affected zone (HAZ). The images of the weld cross-sections at the same magnification were processed and measured using Image-Pro Plus 6.0 software, and the ratio of the pore area to the total image area was calculated to obtain the weld porosity (

Table 3. Porosity of the weld joints at different positions.

Experiments	Parameters	Welding Position
		Front	Central	Rear
Non-temperature-controlled	170 A	1.24%	0.35%	0.56%
	175 A	1.12%	0.32%	0.45%
	180 A	0.96%	0.31%	0.20%
Temperature-controlled	1750 °C	0.27%	0.11%	0.12%
	1825 °C	0.21%	0.05%	0.04%
	1900 °C	0.09%	0.05%	0.14%

). It can be seen that there is a significant difference in porosity across the different weld regions (front, central, and rear). First, regardless of temperature control, the front section has the most pores, and the central section has the fewest. Second, the temperature-controlled welds have fewer pores across all three sections than the non-temperature-controlled welds, especially in the front section, where porosity is reduced by 6–10 times. This is due to the consistency of heat input under temperature-controlled conditions, resulting in minimal weld defects. Finally, the optimal parameter is a temperature-controlled condition of 1825 °C, at which the porosity is lowest.

In the TIG welding process, both process parameters and local heat transfer conditions collectively determine the thermal history of the material, which directly influences the evolution of its microstructure. Key solidification parameters include the temperature gradient (G) and the solidification rate (R) at the solidification front [18]. Figure 5a presents a solidification map correlating G and R, where combinations of G × R (controlling structure size) and G/R (determining morphology) dictate whether cellular, planar, equiaxed dendritic, or columnar dendritic microstructures form. Higher cooling rates (G × R) yield finer features, while lower rates result in coarser structures [19]. During solidification, downward heat flow occurs at the boundaries of the molten pool. As a result, the area near the fusion line, adjacent to the base material, exhibits a significant temperature gradient due to higher thermal diffusion efficiency, causing the molten pool metal to solidify preferentially and form a columnar crystal structure. In contrast, the central region of the weld, due to continuous heat input, experiences a reduced temperature gradient and slower solidification rate, ultimately forming equiaxed or coarse columnar crystal structures (Figure 6b).

It is worth noting that precise temperature control directly affects the amount of heat input, increasing the molten pool’s lifetime. Heat dissipation slows, thereby regulating the cooling rate of the weld, resulting in a smaller crystallization temperature gradient, which is conducive to refining coarse columnar grains [20]. The small size of fine equiaxed grains can hinder the movement of dislocations and grain boundary sliding, thereby increasing the dislocation density of the material. This, in turn, enhances the material’s deformability, reduces the number of grain boundaries, and makes them more stable. Therefore, the refinement of equiaxed grains achieved through temperature-controlled welding contributes to the improvement of strength and ductility throughout the weld.

Figure 6. Distribution of welded crystalline structures: (a) effect of temperature gradient G and growth rate R on the morphology and size of solidification microstructure; (b) schematic diagram of weld grain growth; (c) various symbols in the schematic diagram. The schematic diagrams are adapted and summarized from the literature [15,21].

Figure 6. Distribution of welded crystalline structures: (a) effect of temperature gradient G and growth rate R on the morphology and size of solidification microstructure; (b) schematic diagram of weld grain growth; (c) various symbols in the schematic diagram. The schematic diagrams are adapted and summarized from the literature [15,21].

An increase in the solidification rate accelerates the cooling of the weld pool, making it difficult for gases dissolved in the weld pool (such as hydrogen) to escape in time, thereby increasing the likelihood of pore formation. At the same time, a higher solidification rate causes the surface layer of the weld pool to solidify rapidly, hindering the diffusion of gas to the outside, leading to gas bubbles being trapped within the weld pool, where they aggregate and grow, eventually forming larger pores [22]. In addition, rapid solidification may cause local overcooling of the weld, reducing the wettability and fusion capability of the weld metal with the base metal, thus inducing lack-of-fusion defects. These defects can result in uneven microstructural distribution in the joint, creating distinct stress concentration areas that are more prone to becoming crack initiation sites under external loads [23].

Grain size measurements of the fusion zone were carried out on optical micrographs using Nano Measurer 2.0 software, as shown in Figure 7a. Under the uncontrolled-temperature welding condition (175 A), the fusion zone was dominated by columnar grains, with an average grain size of approximately 84.1 μm. In contrast, under the temperature-controlled welding condition (1825 °C), the average grain size in the columnar-grain-dominated region of the fusion zone was significantly reduced to about 28.9 μm, indicating pronounced grain refinement. Quantitative analysis demonstrates that temperature-controlled welding effectively regulates the thermal gradient (G) and solidification rate (R), thereby suppressing the preferential growth of columnar grains near the fusion line and promoting the nucleation and growth of equiaxed grains, ultimately resulting in significant grain refinement.

3.2. Micro-Morphology of TIG Welded Joints

Figure 8 shows SEM images under six welding parameters. The fusion line is represented by a black solid line; the area between the two lines is the partially melted zone (PMZ). Above the fusion line is the fusion zone, and below it is the heat-affected zone. As temperature increases, the PMZ width widens, and the second phase is distributed along grain boundaries, forming a white, mesh-like structure. Near the PMZ, the second phase (SP) coarsens and aggregates near the heat-affected zone, forming a continuous network structure (Figure 8b) in the joints at 175 A. In contrast, at 1825 °C, the distribution of the second phase near the PMZ is more dispersed (Figure 8e), indicating that temperature control effectively suppresses grain boundary segregation.

EDS elemental analysis was performed near the fusion zone of the welded joints obtained under the condition of 1825 °C, as shown in Figure 9. For the received joints, the fusion zone contains elements of Al, Mg, Si, Fe, Mn, as well as a small amount of Cu and Zn elements. For Mg and Zn elements, they are significantly reduced due to high-temperature evaporation. For the joints obtained at 1825 °C, it can be observed that, compared with the HAZ, the weld zone contains more grain boundaries and more secondary phases, and these grain boundaries are not continuous. This is mainly attributed to the rapid non-equilibrium solidification and the combined action of multiple factors, in which the epitaxial competitive growth of adjacent grains creates additional grain boundaries. Meanwhile, the transient nature of the welding thermal cycle promotes the discontinuous precipitation of secondary phases. Elemental analysis shows that, as indicated by point 1 in Figure 9b, the weld zone mainly consists of an aluminum matrix with a small amount of dissolved magnesium. Points 2 and 3 mainly contain elements such as Al, Fe, and Si. It can be inferred that the secondary phase formed is the Al-Fe-Si phase, based on the atomic ratios of these elements. The Al-Fe-Si phase usually has a complex crystal structure with strong atomic bonding, making it difficult for dislocations to move within it, resulting in high hardness but poor plasticity, thereby exhibiting brittleness. Point 4 in Figure 9b shows that the secondary phase exists in granular or short rod shapes and is distributed along grain boundaries. Based on the atomic ratios of elements, it can be inferred to be the β-AlFeSi phase. Based on elemental-atomic ratio analysis, the secondary phase at point 3 can be approximated as the α-AlFeSi phase (Al₈Fe₂Si), which exhibits a fine-grained, skeletal, or Chinese-script morphology [24]. Compared to the β-AlFeSi phase, the α-AlFeSi phase exhibits more favorable morphological and microstructural characteristics, primarily due to its more rounded grain morphology and more uniform distribution. These attributes effectively mitigate stress concentration, thereby reducing the tendency for cracking during aluminum alloy processing [25]. Consequently, the α-AlFeSi phase has a lesser detrimental effect on the alloy’s overall ductility and toughness, thereby enhancing its comprehensive mechanical properties [26]. Differences in structure and bonding between the α phase and β phase. Under non-equilibrium solidification conditions during welding, or in the presence of impurity element enrichment, two main intermetallic compounds are formed: the α-AlFeSi phase and the β-AlFeSi phase. The α-AlFeSi phase has a body-centered cubic structure with a relatively uniform atomic arrangement, where Fe–Si covalent bonds form the framework and Al–Fe/Si bonds serve as connecting links. Its bonding is weakly anisotropic, and the phase appears as rounded or skeletal morphologies. The β-AlFeSi phase has an orthorhombic layered structure, with Fe–Si covalent bond chains aligned along specific crystallographic directions, resulting in significant structural anisotropy. This phase tends to grow in needle-like or plate-like morphologies and is inclined to distribute continuously along grain boundaries, making it prone to stress concentration under external loads [27]. By promoting the formation of the beneficial α phase while inhibiting the formation of the harmful β phase, effective dispersion strengthening can be achieved, significantly improving the joint’s strength and hardness and greatly reducing stress concentration and crack sensitivity [28]. The formation of these finely distributed strengthening phases improves the mechanical properties of welded joints.

3.3. Mechanical Properties of Welded Joints

Tensile tests were conducted on six welded joints, and the tensile strength and elongation are shown in Figure 10. The tensile strength and elongation of the original 7072 base material are 545 MPa and 12%, respectively. After welding, the weld tensile strength is less than 60% of that of the base material, and the fracture occurs in the HAZ, far from the weld. The joints produced by temperature-controlled welding show overall improvements in tensile strength and elongation compared to those produced by non-temperature-controlled welding.

Under non-temperature-controlled conditions, as shown in Figure 10a, the tensile strength and elongation of the joints showed continuous improvement with increasing welding current from 170 A to 175 A, reaching 295 MPa and 9.8% at 175 A. Then, the tensile strength and elongation began to decline after 175 A. Figure 10b demonstrates that under controlled-temperature conditions, the welds achieved higher tensile strength and elongation than under uncontrolled welding, with increases of 12.7%, 11.2%, and 14.2%, respectively. At 1825 °C, the maximum tensile strength reached 328 MPa. When the temperature was above 1825 °C, the weld tensile strength and elongation reduced. This phenomenon can be attributed to the formation and growth of coarse equiaxed dendrites in the weld as thermal input increased. These large grains significantly reduce the effective bonding area between grains, thereby weakening intergranular bonding and degrading the overall mechanical properties of the welds [29].

Notably, the tensile strength and elongation of welded joints vary significantly in different areas under the same current conditions, as shown in Figure 11, mainly due to uneven heat input distribution in the molten pool. Figure 11a,c,e, respectively, demonstrate the evolution of weld tensile properties at different locations under identical welding current conditions. The entire welding process can be divided into three stages: In the initial stage, the low initial molten-pool temperature results in high cooling rates and steep temperature gradients. The rapid solidification leads to defects such as porosity and incomplete fusion, severely compromising joint integrity. During the intermediate stage, thermal input and dissipation reach equilibrium, resulting in a uniform temperature distribution and moderate cooling rates. This optimal cooling rate facilitates the formation of fine equiaxed crystals, achieving the best uniformity in weld microstructure and residual stress levels, thereby maximizing joint tensile strength and ductility. In the final stage, thermal accumulation leads to localized overheating, significantly reducing cooling rates and prolonging the high-temperature dwell time.

By comparison, the tensile properties of joints under the controlled-temperature conditions are approximated at the weld’s front, central, and rear sections at different temperatures in Figure 11b,d,f. The weld achieves a peak tensile strength of 328 MPa at 1825 °C. This demonstrates that the system maintains a uniform molten pool temperature and constant heat input, resulting in a more homogeneous temperature field distribution in the melt pool. Along the welding direction, the temperature gradient remains consistent, ensuring that the material undergoes identical thermal cycling processes. The uniform temperature field also leads to synchronous thermal expansion and contraction across different material regions, resulting in comparable thermal stress levels at various locations. Under identical heat input, phase transformation and grain growth proceed simultaneously, effectively suppressing defects such as porosity accumulation caused by localized overheating or cooling. Both processes are thermally activated and occur within overlapping temperature ranges during the welding thermal cycle. The material undergoes continuous non-isothermal heating and cooling. Once the critical transformation temperature is reached, phase transformation is initiated. At the same time, the increase in temperature enhances atomic diffusion and grain boundary mobility, thereby promoting grain growth synchronously [30].

3.4. Tensile Fracture Analysis

After fracture, all specimens exhibited a double-cup shape with pronounced necking. Taking the specimen at 180 A as an example, as shown in Figure 12a, the fracture occurred in the HAZ, which was far away from the weld zone. The fracture displayed a uniform ductile fracture morphology at the weld zone, featuring small-scale, deep dimples characteristic of ductile fracture. This indicates that the weld zone absorbed substantial strain before fracture, demonstrating strong resistance to sudden rupture. Figure 10b revealed large, shallow equiaxed and flat-shaped pits in the HAZ, with numerous small pits surrounding the larger ones [31], suggesting limited energy-absorption capacity and restricted plastic rheological behavior during fracture. Studies have reported that HAZ is most sensitive to current, attributed to grain coarsening under high temperatures, which reduces dislocation hindrance at grain boundaries and decreases the material’s resistance to plastic deformation [32].

For welding joints without temperature control, fracture morphology is significantly affected by welding currents. As shown in Figure 13a–c, the fracture surfaces exhibit unevenly sized dimples with numerous discontinuous tear ridges and notches. The bottom of these dimples contains fragmented reinforced-phase particles due to stress concentration, which serve as nucleation centers for dimples. Figure 13a demonstrates that the fracture of the joint at 170 A appears relatively flat, revealing a structure composed of tear ridges and fine dimples, indicating a mixed ductile-brittle fracture mechanism [33]. With the welding current increased to 175 A, the number of dimples begins to increase, forming extensive tear ridges and ultimately leading to a ductile fracture. When the current is further increased to 180 A, welding thermal cycling causes HAZ grain coarsening, with needle-like β″ phases transforming into rod-like β′ phases or lamellar β phases, accompanied by reduced dislocation density and an uneven distribution. During tensile testing, stress concentration in the HAZ softening zone accelerates hole nucleation and growth, forming large dimples [34]. This is due to unstable thermal input, which can cause local overheating, severe grain coarsening, and brittle phase aggregation [35].

Figure 14 shows the energy dispersive X-ray spectroscopy (EDS) analysis of the fracture surface of the 1825 °C tensile specimen. It confirmed that the fracture occurred in the HAZ adjacent to the base material. At the grain boundaries within the heat-affected zone, impurity elements Fe and Si, as well as the primary alloying element Mg, exhibit a pronounced tendency to segregate. This localized elemental segregation alters the compositional distribution in the grain boundary region, directly weakening grain boundary cohesion and potentially further promoting the precipitation of coarse brittle phases. On one hand, the coarse needle-like or blocky brittle second phases formed from Fe, Si, and Al are the β-AlFeSi phase; on the other hand, under the non-equilibrium thermal cycles induced by welding, the rapid enrichment of Mg and Si at grain boundaries facilitates the formation of coarse equilibrium Mg₂Si phases. These coarse phases not only lose their strengthening effect but also embrittle the grain boundaries. The distribution of such brittle phases along grain boundaries severely disrupts the continuity of the matrix, acting as stress concentration sites and potential crack initiation points, thereby degrading the overall mechanical properties of the material. Figure 15a revealed residual fragmented second-phase particles at the bottom of the dimples. The EDS result in Figure 15b confirmed that these particles are the Mg₂Si phase. The formation of these dimples can be attributed to the fact that under tensile stress, dispersed Mg₂Si particles detached from the matrix, subsequently serving as nucleation centers for micro-pores and leading to their expansion.

3.5. Microhardness

Figure 16 shows the Vickers microhardness profile across the cross-section of the welded sample, revealing a characteristic “W”-shaped distribution transverse to the welding direction [36]. This phenomenon is consistent with reports from TIG welding studies on aluminum alloys [37]. The average microhardness of the base metal was 63 HV_0.2–66 HV_0.2, while the Heat-Affected Zone (HAZ) measured 57 HV_0.2–65 HV_0.2 and the weld zone 57 HV_0.2–60 HV_0.2. Notably, the heat-affected zone exhibited significant changes in microhardness, likely due to the dissolution and growth of precipitates, resulting in lower hardness values than those of the base metal [38]. The lowest hardness value in the welded joint is in the fusion zone, where a few porosity defects are present (Figure 16b). This may be one of the reasons for the decrease in hardness. Comparative analysis shows that the microhardness fluctuations in the temperature-controlled weld zone are significantly smaller than those in the non-temperature-controlled weld, and the distribution is also more uniform.

Figure 17 shows a significant difference in microhardness across different positions in the non-temperature-controlled welds. This is attributed to the continuous heat input, leading to a more intense temperature gradient. Such unstable heat input promotes an increase in defects in the rear weld region, further exacerbating the dispersion of hardness values.

As shown in Figure 17, the closed-loop feedback control system for molten pool temperature maintains stability during welding while reducing differences in cooling rates across the weld, resulting in a more uniform grain size and, consequently, a more uniform hardness distribution. Both the temperature-controlled and non-temperature-controlled joints exhibited significant softening across the entire weld zone and heat-affected zone, with hardness values generally lower than those of the base material. The joint’s hardness reaches its minimum within the heat-affected zone. For the heat-affected zone (HAZ), the Hall–Petch equation (Equation (1)) indicates a negative correlation between hardness and the square root of the grain diameter [39].

$${\mathsf{\sigma }}_{\mathrm{y}}={\mathsf{\sigma }}_0+{\displaystyle \frac{\mathrm{k}}{\sqrt{\mathrm{d}}}}$$

(1)

where σ_y denotes the yield strength, σ₀ represents the frictional stress of the single-crystal material, k is the Hall-Petch constant, and d is the average grain diameter. Both σ₀ and k are grain size-independent constants. As grain size increases, hardness tends to decrease, suggesting that the larger grain size in the weld zone may contribute to reduced hardness [40]. Overall, the welded joint obtained at 1825 °C exhibits superior microhardness, reaching 60 HV_0.2, indicating optimal mechanical properties.

4. Conclusions

The main goal of this study is to evaluate how closed-loop molten pool temperature control affects TIG welding. We focus on welding stability and joint performance. The influence of temperature feedback control on these properties is systematically investigated:

• Under controlled-temperature welding conditions, the closed-loop feedback control system for molten pool temperature effectively stabilized heat input, significantly improving the joint’s overall performance. The optimal comprehensive mechanical properties were achieved at 1825 °C. Under such conditions, reduced porosity, enhanced fracture morphology, and refined and homogenized microstructure were achieved. These structural optimizations lead to comprehensive improvements in mechanical properties: the weld zone exhibits higher average hardness with significantly reduced hardness variation across different positions; the tensile strength and elongation are also enhanced, demonstrating consistent mechanical properties along the welding direction.
• By contrast, non-temperature-controlled welding cannot regulate heat input effectively. Prolonged heat accumulation pushes the temperature in the middle and rear of the joint to an excessively high level. It triggers defects like grain coarsening and increased porosity. The macroscopic morphology and microstructural uniformity of the joint along the welding direction are both degraded by these issues. This instability shows that stable welding quality cannot be guaranteed only by finding the optimal current settings. It further highlights the necessity of molten pool temperature control.
• From an engineering perspective, the molten-pool-temperature-based closed-loop TIG welding strategy helps stabilize the heat input during welding. It effectively reduces microstructural non-uniformity, porosity, and property dispersion caused by local overheating. As a result, the consistency and reliability of welded joints are improved. This makes the method suitable for industrial manufacturing and component repair applications. It should be noted that, in practical industrial environments, the performance of this strategy may still be affected by environmental disturbances, increased workpiece size, and system response characteristics. Therefore, the control parameters need to be further optimized and calibrated for specific working conditions. Overall, the proposed temperature control strategy provides a feasible engineering approach for achieving high-quality and stable welding.