Effects of Tool Rotational Speed on the Microstructure and Properties of Friction Stir Welded AZ61 Magnesium Alloy Joints

Abstract

Magnesium alloys, characterized by high specific strength and low density, have high potential for applications in transportation and aerospace. Nevertheless, ensuring the reliable joining of thin-walled components remains a major technical challenge. This study examines how rotational speed affects the microstructure and mechanical properties of friction stir welded AZ61 magnesium alloy hollow profiles (3 mm thick), with particular focus on the underlying mechanisms. The results show that higher rotational speed during friction stir welding promotes dynamic recrystallization and weakens the basal texture. It also affects microstructural homogeneity, where an optimal rotational speed produces a relatively uniform hybrid microstructure consisting of refined recrystallized and un-recrystallized regions. This balance enhances both texture strengthening and microstructural optimization. The weld joint fabricated at a rotational speed of 1500 rpm showed the best overall mechanical properties, with ultimate tensile strength, yield strength, and elongation reaching peak values of 286.7 MPa, 154.7 MPa, and 9.7%, respectively. At this speed, the average grain size in the weld nugget zone was 4.92 μm, and the volume fraction of second-phase particles was 0.67%. This study establishes a critical process foundation for the reliable joining of thin-walled magnesium alloy structures. The optimized parameters serve as valuable guidelines for engineering applications in lightweight transportation equipment and aerospace manufacturing.

1. Introduction

Magnesium and its alloys, known as the lightest structural metals, show great potential for broad applications in the automotive, rail, marine, and aerospace industries [1,2]. As a primary metal joining method, welding is essential for broadening the industrial use of magnesium alloys. Magnesium alloys are mainly joined by fusion welding and solid-state welding techniques. However, fusion welding often causes defects such as hot cracking, porosity, vaporization of alloying elements, and grain coarsening, which severely degrade the mechanical integrity of welded joints [3,4]. Friction Stir Welding (FSW) is an advanced solid-state joining technique that employs frictional heat and mechanical deformation to form high-integrity welds below the base material’s melting point. Unlike conventional fusion welding, FSW operates below the melting temperature, thereby avoiding typical fusion-related defects such as second-phase formation, porosity, brittle microstructures, and solidification cracks. This solid-state process thus produces joints with superior metallurgical quality and mechanical properties [5,6,7,8].

Because welded structures are required to endure long-term service under harsh conditions, including dynamic fatigue loads and corrosive environments, joint reliability has emerged as a primary research focus. For example, in rail transit equipment, studies have demonstrated that optimizing the FSW process can yield aluminum alloy joints with significantly improved fatigue life [9,10]. At the same time, enhancing corrosion resistance through microstructural control has also become a major research focus [11,12]. Taken together, these studies suggest that aligning process optimization with specific industrial service requirements is a critical step toward the widespread and mature application of FSW technology.

Optimizing process parameters in friction stir welding (FSW) of magnesium alloys remains a key research focus, as these parameters directly determine the joint’s microstructure and mechanical properties. The mechanical and microstructural characteristics of FSW joints are mainly controlled by three critical parameters: rotational speed, welding speed, and axial force [13,14]. Sachin [15] et al. systematically studied the effects of rotational speed and welding speed on the microstructural evolution and mechanical properties of 6 mm thick AZ61A magnesium alloy FSW butt joints. The experimental results show that the highest joint strength is obtained at a rotational speed of 2000 rpm and a welding speed of 20 mm/min, achieving a welding efficiency of 89% compared to the base material. Mohamed [16] et al. systematically investigated the effects of key FSW process parameters—including rotational speed, welding speed, axial force, and tool pin penetration depth—on weld quality using an orthogonal experimental design. The experimental results indicated that rotational speed is the most critical parameter influencing weld quality in the FSW process. Dr. S. Ugender [17] systematically studied the effects of key FSW process parameters on the mechanical properties and microstructural evolution of AZ31 magnesium alloy joints. The experimental results showed that rotational speed strongly affects grain refinement in the FSW material, with higher speeds leading to more significant grain size reduction. Hou Jing [18] et al. systematically studied the effects of high-speed FSW parameters on the microstructure and mechanical properties of AZ31B magnesium alloy joints. The experimental results show that the high-speed FSW process can produce defect-free joints, and that a combination of high rotational speed and low welding speed significantly improves the tensile properties of the welds.

Currently, few studies have reported on FSW of hollow thin-walled magnesium alloy profiles with plate thicknesses of 3 mm or less for rail transit applications. These thin-walled hollow structures are ideal for lightweight equipment design; however, their unique geometric features present new challenges for Friction Stir Welding (FSW): the relatively thin walls (e.g., ≤3 mm) result in low stiffness and thermal mass, rendering the process highly sensitive to heat input. Excessive heat input may cause overheating, distortion, or even burn-through, whereas insufficient heat input can result in poor material plastic flow and welding defects [13]. Current studies indicate that tool rotational speed strongly affects the performance of FSW joints, playing a particularly important role in both weld formation quality and mechanical properties in thin-plate applications. Therefore, this study performs FSW experiments on 3 mm thick walled plates of AZ61 magnesium alloy hollow thin-walled profiles to investigate the effects and mechanisms of rotational speed on the microstructure and mechanical properties of the welded joints, aiming to provide reference data and theoretical guidance for FSW of thin-walled components in applications such as rail transit, automotive, and aerospace industries.

2. Experimental Materials and Methods

The hollow thin-walled extruded magnesium alloy profiles used in this study have a rectangular cross-section of 110 mm × 80 mm and a thickness of 3 mm, as illustrated in Figure 1. The extrusion was performed using a 1800-ton horizontal extrusion press at Shanxi Yinguang Huasheng Magnesium Industry Co., Ltd. (Yuncheng, China). The extrusion temperature was set to 400 °C, with an extrusion ratio of 23.1. The 110 mm × 110 mm × 3 mm plates used for welding were laser-cut from profiles, and only those with high flatness that allowed for tight abutment were selected. The plates were subsequently ground with sandpaper to produce a smooth surface and remove the oxide layer. In addition, 5 mm of material was milled from the faying edges before welding. The chemical composition and mechanical properties of the plates are presented in

Table 1. Chemical Composition and Mechanical Properties of the Base Material.

BM	Chemical Composition/wt.%							UTS	YS	EL
	Al	Zn	Mn	Fe	Si	Cu	Ni	/MPa	/MPa	/%
AZ61	6.45	0.58	0.23	0.0023	0.012	0.0019	0.00030	296	169	18

.

Friction stir welding experiments were performed on a Liutai LT-XJ16-1510-05 gantry-style FSW machine (Guangdong Liutai Co., Ltd., Guangzhou, China), as illustrated in Figure 2 and Figure 3. The welding tool had a probe length of 2.7 mm and a shoulder diameter of 10 mm. The FSW process parameters are summarized in

Table 2. FSW Process Parameters of AZ61 Magnesium Alloy.

Rotational Speed/rpm	Welding Speed/(mm/min)	Plunge Depth/mm	Tool Tilt Angle/°
1000–1900	300	0.15	2.5

. The processing parameters were defined as follows: the welding speed (300 mm/min), plunge depth (0.15 mm), and tool tilt angle (2.5°) were kept constant, whereas the tool rotational speed varied between 1000 and 1900 rpm. The prepared plates were positioned on the welding machine’s loading platform in a butt-joint configuration and securely clamped using a dedicated FSW fixture. Once the plates were fixed, welding was carried out. After welding, the metallographic and mechanical test samples were sectioned using wire cutting. The metallographic samples were then ground and polished sequentially with 200–2000# water sandpaper, followed by etching in a bitter acid solution (4.2 g bitter acid + 10 mL acetic acid + 70 mL anhydrous ethanol + 10 mL deionized water) for 30 s. The microstructure was examined, and the quantity and distribution of the second phase were analyzed using scanning electron microscopy (JEOL (Tokyo, Japan)). The samples were polished with a Gatan 697 argon ion polishing system (Gatan, Pleasanton, CA, USA), and the grain size and orientation of the welded joints were characterized using electron backscatter diffraction (Thermo Scientific, America). The tensile properties were measured using a TSE105D microcomputer-controlled electronic universal testing machine (New Sansi (Shanghai) Enterprise Development Co., Ltd. (Shanghai, China)) at a crosshead speed of 2 mm/min. Three specimens were prepared for each type of welded joint, and the average value was calculated. The dimensions of the tensile specimens (in mm) are presented in Figure 4.

3. Results

3.1. Mechanical Properties

The mechanical properties of tensile specimens from AZ61 thin-plate welded joints fabricated at different rotational speeds are presented in Figure 5. The tensile strength of the base material (BM) is 296.0 MPa. For FSW joints fabricated at rotational speeds ranging from 1000 to 1900 rpm, the tensile strength varies between 242.6 and 286.7 MPa, reaching a maximum of 286.7 MPa at 1500 rpm, corresponding to a welding efficiency of 96.9%. The elongation of the base material is 18%. For FSW joints fabricated at rotational speeds ranging from 1000 to 1900 rpm, the elongation varies from 5.9% to 9.7%, reaching a maximum of 9.7% at 1500 rpm, which corresponds to 53.9% of the base material’s elongation. The yield strength of the base material is 160.0 MPa. For FSW joints fabricated at rotational speeds ranging from 1000 to 1900 rpm, the yield strength varies between 130 MPa and 154.7 MPa, remaining above 147.5 MPa at rotational speeds from 1300 to 1500 rpm. The maximum yield strength of 154.7 MPa is achieved at 1500 rpm, corresponding to 91.5% of the base material’s yield strength.

3.2. Microstructural Characterization of Welded Joints

To investigate the mechanism by which rotational speed affects the mechanical properties of welded joints, joints fabricated at rotational speeds of 1100 rpm, 1500 rpm, and 1900 rpm were selected for microstructural analysis. The metallographic structures of welded joints at different rotational speeds are presented in Figure 6. At a low rotational speed of 1100 rpm (Figure 6a), the welded joint displayed a relatively small plasticized zone. The base of the ‘basin-shaped’ structure was narrow, and a distinct, abrupt transition appeared at the boundary between the thermomechanically affected zone (TMAZ) on the advancing side and the nugget zone (NZ). At a rotational speed of 1500 rpm (Figure 6b), the increased heat input caused the plasticized zone to expand and the base width of the ‘basin-shaped’ structure to increase noticeably. In addition, the microstructural transition between the nugget zone (NZ) and the thermomechanically affected zone (TMAZ) became more gradual, leading to enhanced microstructural homogeneity. At a rotational speed of 1900 rpm (Figure 6c), excessive heat input caused further expansion of the plasticized zone, resulting in the maximum base width of the ‘basin-shaped’ structure. However, microstructural homogeneity may have deteriorated due to overheating. Overall, at all rotational speeds, the weld seams exhibit a distinct “basin-shaped” morphology. As the rotational speed increases, the bottom size of this “basin-shaped” structure also increases. This is attributed to higher heat generation at elevated rotational speeds, which enlarges the plasticized zone of the welded joint. The low-density microstructure on both sides of the FSW joint is asymmetrical. The boundary between the advancing side of the weld seam (AS-TMAZ) and the weld nugget (NZ) is more distinct than that between the retreating side (RS-TMAZ) and the weld nugget. At low rotational speeds, the microstructure on the advancing side of the NZ and TMAZ exhibits an abrupt transition, whereas the retreating side shows a more gradual transition between the NZ and TMAZ [19,20,21].

Figure 7 and

Table 3. Average Size and Volume Fraction of Second-Phase Particles in the Weld Nugget.

Second Phase in the Weld Nugget Zone		Rotational Speed/rpm
	B M	1100	1500	1900
Average Size/μm	1.33	2.32	2.14	2.29
Volume Fraction/%	1.10	1.26	0.67	0.60

show SEM micrographs of the nugget zone and the average sizes of secondary phases in the base metal and welded joints fabricated at different rotational speeds. The base metal (AZ61) had an average secondary phase size of 1.33 μm, with the phases mainly located along grain boundaries. Compared with the base metal, the nugget zones of all welded joints exhibited larger secondary phases, with average sizes of 2.32 μm, 2.14 μm, and 2.29 μm at rotational speeds of 1100 rpm, 1500 rpm, and 1900 rpm, respectively. The increase in secondary phase size is primarily attributed to thermal effects during the FSW process. While the mechanical stirring of the tool can fragment the original coarse secondary phases, the elevated temperatures in the nugget zone promote a dominant thermodynamic process. At the lower rotational speed (Figure 7a), the heat input was relatively low but sufficient to induce coarsening of the secondary phases. Meanwhile, the mechanical stirring was relatively weak, resulting in limited fragmentation efficiency. As a result, the average secondary phase size increased to 2.32 μm. At a rotational speed of 1500 rpm (Figure 7b), both stirring intensity and heat input were increased, achieving a favorable balance. This balance facilitated fragmentation of coarse secondary phases in the nugget zone, refining them and enhancing distribution homogeneity. The intense material flow further improved uniformity, resulting in a slightly smaller average secondary phase size of 2.14 μm compared with the lower-speed condition. At the highest rotational speed (Figure 7c), the intensified stirring and excessive heat input led to significantly elevated temperatures in the nugget zone. This, in turn, caused coarsening of the secondary phases, likely due to the coalescence and growth of smaller particles [22,23,24,25]. Different rotational speeds also affect the volume fraction of second-phase particles in the weld nugget, as presented in Table 3. The volume fraction of second-phase particles in the base material is 1.10%. After friction stir welding, the volume fraction is 1.26% at a rotational speed of 1100 rpm. When the rotational speed increases to 1500 rpm and 1900 rpm, the volume fraction decreases to 0.67% and 0.60%, respectively.

The IPF (Inverse Pole Figure) maps of the AZ61 base material and the weld nugget of FSW joints fabricated at different rotational speeds are presented in Figure 8. Figure 8a presents the microstructure of the base material, while Figure 8b–d show the microstructure of the weld nugget in joints fabricated at rotational speeds of 1100 rpm, 1500 rpm, and 1900 rpm, respectively. The corresponding average grain sizes are 12.86 μm, 7.12 μm, 4.92 μm, and 6.57 μm. As shown in Figure 9, the average grain size in the weld nugget is smaller than that of the base material. At a relatively low rotational speed (1100 rpm), both the mechanical stirring and heat input generated by the tool are weak. Although the original grains are fragmented, the low temperature limits dynamic recrystallization, resulting in insufficient grain refinement. As the rotational speed increases to 1500 rpm, both mechanical stirring and heat input are enhanced. The elevated temperature in the weld nugget promotes dynamic recrystallization, leading to significantly finer grains. At a relatively high rotational speed (1900 rpm), although mechanical stirring is intensified and heat input is further increased, the excessively high temperature promotes growth of recrystallized grains. Furthermore, due to insufficient refinement and dispersion of the second-phase particles at this temperature, their pinning effect on grain growth is weak, leading to a larger average grain size. At a rotational speed of 1500 rpm, the temperature in the weld nugget is moderate, which favors the formation of finely dispersed second-phase particles and effectively inhibits the growth of recrystallized grains. The temperature is sufficient to promote dynamic recrystallization without causing coarse grains, resulting in a fine and uniform microstructure.

To evaluate the degree of recrystallization in the base material and weld nugget at different rotational speeds, the proportion of high-angle grain boundaries (HAGBs > 15°) in each sample was statistically analyzed, as shown in Figure 10. The base material exhibits the highest proportion of HAGBs. After friction stir welding, the HAGB proportion in the weld nugget is lower than that of the base material. At rotational speeds of 1100 rpm, 1500 rpm, and 1900 rpm, the HAGB proportions are 78.6%, 88.4%, and 89.6%, respectively. The intense stirring action of the tool causes the original grains to elongate and fragment, while dislocations accumulate due to plastic deformation, promoting the continuous formation of low-angle grain boundaries (2° < LAGBs ≤ 15°) within the grains. As the rotational speed increases, the higher heat input facilitates the gradual transformation of low-angle grain boundaries (LAGBs) into high-angle grain boundaries (HAGBs) via boundary migration at elevated temperatures [23]. Therefore, the proportion of HAGBs serves as an indicator of the degree of dynamic recrystallization, and a sufficiently high recrystallization level is achieved when the rotational speed exceeds 1500 rpm.

To further quantify the degree of dynamic recrystallization and the morphological characteristics of the recrystallized grains, the grain orientation spread (GOS) of the microstructure in the weld nugget was measured, as shown in Figure 11. The GOS value represents the crystallographic orientation spread within a grain and effectively reflects its stored deformation energy. A lower GOS value corresponds to lower internal deformation energy and indicates a higher degree of recrystallization. In Figure 11, the GOS values increase from blue to orange. Regions with lower GOS values appear blue, while higher values shift toward orange. A larger blue area indicates a higher degree of recrystallization in that region [24,25]. As illustrated in Figure 11 and summarized in

Table 4. Recrystallization Volume Fractions at Different Rotational Speeds.

	BM	1100 rpm	1500 rpm	1900 rpm
Recrystallization Volume Fraction (%)	80.12	49.05	74.88	79.64

, the base material exhibits the highest degree of dynamic recrystallization, resulting from the high-temperature, large-deformation conditions during hot extrusion. After friction stir welding, the degree of dynamic recrystallization in the weld nugget is lower than that of the base material. At a low rotational speed of 1100 rpm, the recrystallized volume fraction is only 49.05%, whereas it increases to 74.88% and 79.64% at rotational speeds of 1500 rpm and 1900 rpm, respectively. Higher rotational speeds increase welding heat input and strain rate. The intensified stirring effect induces greater cumulative strain in the matrix and more frequent grain fragmentation–recrystallization cycles, thereby promoting nucleation of dynamic recrystallization and enhancing its overall degree. Among all samples, the mixture of fully recrystallized and un-recrystallized structures was most uniform at 1500 rpm, with smaller individual grain sizes. This uniform and refined microstructure facilitates grain-interface coordination during plastic deformation and enhances the mechanical properties of the alloy.

The {0001} basal plane pole figures and the corresponding maximum pole densities for the base metal and the nugget zone of the welded joint are shown in Figure 12 and

Table 5. Maximum Pole Densities at Different Rotational Speeds.

	BM	1100 rpm	1500 rpm	1900 rpm
Maximum Pole Density	13.42	30.58	25.87	23.17

, respectively. The texture of the base material is relatively weak and dispersed, with a maximum pole density of only 13.42. After friction stir welding, the basal plane texture in the weld nugget becomes more concentrated, and its intensity increases significantly. At rotational speeds of 1100 rpm, 1500 rpm, and 1900 rpm, the maximum pole densities are 30.58, 25.87, and 23.17, respectively. As the rotational speed increases, the basal plane texture strength in the weld nugget shows a slight weakening trend. Higher rotational speeds increase welding heat input and intensify the stirring effect, which enhances shear deformation and deformation temperature in the plastic zone. This promotes dynamic recrystallization and generates numerous newly formed grains with random orientations, thereby reducing the basal plane texture intensity [26].

4. Discussion

This study systematically investigates the effect of rotational speed on the microstructure and properties of friction stir welded joints in AZ61 magnesium alloy thin-walled hollow profiles. The discussion that follows is organized into three main aspects: microstructural evolution, macroscopic mechanical properties, and a comparative analysis with other alloy systems.

4.1. Effect of Rotational Speed on the Microstructure and Mechanical Properties of FSW Joints

Rotational speed controlled microstructural evolution in the nugget zone by regulating thermomechanical coupling. EBSD analysis (Figure 8, Figure 9, Figure 10 and Figure 11) showed a non-monotonic trend in grain size, with the finest grains (4.92 μm) observed at 1500 rpm. At this rotational speed, the moderate heat input was sufficient to induce substantial dynamic recrystallization (recrystallization fraction of 74.88%) without significant grain coarsening. Simultaneously, this speed achieved an optimal balance between the fragmentation and coarsening/dissolution of secondary phases, promoting improved plasticity. Moreover, extensive dynamic recrystallization (DRX) produced a higher number of grains with random orientations, resulting in a weakened basal texture. The combination of grain refinement, texture weakening, and favorable secondary phase distribution collectively formed the microstructural basis for the superior properties of the joint fabricated at 1500 rpm. Mechanical property measurements (Figure 5) show that the ultimate tensile strength, yield strength, and elongation of the joints all peaked at 1500 rpm. This behavior is mainly attributed to effective control of weld macrostructure and internal defects. At the lower rotational speed of 1100 rpm, both heat input and mechanical stirring were insufficient. This likely caused incomplete material plastic flow and may have induced bonding defects, such as tunnel defects, leading to the observed reductions in both strength and ductility [13]. At a rotational speed of 1500 rpm, heat input and stirring action achieved an optimal balance, enhancing material flow and enabling the formation of a high-quality weld that was dense and defect-free. This observation aligns with the trend reported by Sachin et al. [15] for the AZ61A alloy, although their optimal rotational speed (2000 rpm) was higher due to a thicker sheet (6 mm versus 3 mm in this study). This suggests that a lower level of thermomechanical input is sufficient to achieve optimal properties in thin-sheet welding. At an excessively high rotational speed (1900 rpm), the resulting high heat input caused abnormally elevated temperatures in the weld zone. While this facilitated complete dynamic recrystallization, it also induced grain coarsening (Figure 9) and the coalescence and growth of secondary phase particles (Figure 7, Table 3). Consequently, contributions from grain boundary and dispersion strengthening were reduced, leading to degraded mechanical properties. This observation is consistent with Mohamed et al. [16], who concluded from orthogonal experiments that “rotational speed is a key parameter” and that it further clarifies the underlying micro-mechanisms.

4.2. Comparison of Mechanical and Microstructural Properties of AZ61 FSW Joints Under Different Processing Parameters

Compared to the published literature, the optimized parameters identified in this study (1500 rpm, 300 mm/min) highlight the strong influence of component geometry and alloy composition. For example, in contrast to the 1120 rpm used by Ugender et al. [17] for 5 mm thick AZ31, or the “high-rotational-speed” approach employed by Hou Jing [18] for AZ31B, the medium rotational speed found optimal for thin-walled AZ61 profiles in this work indicates a distinctly different processing window. This difference arises from the greater sensitivity to heat input of thin-walled structures and the distinct behavior of the larger volume fraction of β-Mg₁₇Al₁₂ phases in AZ61 alloy during the thermal cycle. Therefore, for thin-walled AZ61 components, precise control of heat input is more crucial than employing extreme processing parameters. The optimized rotational speed of 1500 rpm identified in this study offers a direct reference for developing FSW procedures for such components. Future studies could investigate the fatigue performance of joints under simulated rail transit vibratory loading [9] and their corrosion resistance in specific environments [11] using this parameter set, thereby enabling a comprehensive evaluation of service reliability.

5. Conclusions

(1)

This study systematically investigates the influence of rotational speed on the microstructure and mechanical properties of friction stir welded joints in 3 mm thick AZ61 magnesium alloy thin-walled hollow profiles. The results demonstrate that the geometric characteristics of the thin-walled hollow profiles substantially constrain the optimal processing window, rendering it distinctly different from that of conventional solid thick plates. Within the investigated parameter range, 1500 rpm was identified as the optimum rotational speed, at which the heat input and mechanical stirring action reached an optimal balance, resulting in microstructural refinement and performance enhancement.

(2)

Based on the current findings, future work should expand the processing window and explore optimization strategies—such as variable-speed welding or specialized tool design—to address the challenges arising from the high thermal sensitivity of thin-walled structures. Furthermore, assessing the long-term service performance of the joints under fatigue, dynamic loading, and corrosive conditions, along with extending the process to other magnesium alloy systems, represents crucial directions for advancing the engineering application of this technology.