Introduction

Impulse friction stir welding (IFSW) has been developed and patented in 2001 [1]. Like in the traditional friction stir welding (FSW), IFSW joins materials in a solid condition, which preserves all the advantages of FSW in IFSW joints [2]. These processes are commonly applied in the production of light structures from aluminum alloys used in various fields of engineering [3, 4].

A principal feature distinguishing IFSW from FSW is the fact that the axial force acting on the material during the impulse process is not constant. As a result, the friction force, the torque, the contact conditions on the tool/billet interface, and the thermal power vary during the welding process. The main advantage of IFSW is elevation of the mechanical properties of the welded joint above those of the matrix metal. An additional pulse imposed during the welding raises the level of plastic strain in the material and thus promotes refinement of the structure of the weld. This affects positively the mechanical properties of the welded joint, especially the fatigue resistance [1]. It has been shown in [5] that the endurance of welded joints of aluminum-magnesium alloy 6082-T6 after IFSW is 40% higher than after the traditional FSW.

It is known that the oxide layer formed under FSW on the interface of two parts of a welded joint fractures during the tool welding and is transferred to the region of the weld [6,7,8,9,10,11,12]. The oxide blisters retained in the welded joints consist of mall-size particles of Al2O3 (< 100 nm) forming dense accumulations negatively affecting the strength and fatigue properties of the welded joint. It has been shown in [9] that the fatigue resistance of welded joints of aluminum alloys 5058 and 2024 containing oxide inclusions is 35 and 55% lower, respectively, than that of the welded joints bearing no oxide inclusions (after 2 × 106 loading cycles). By the data of [7], the distribution of the oxide layer affects considerably the ultimate strength and the elongation of aluminum alloys.

Precipitation hardening of alloy 6082 is a result of precipitation of intermetallic phases from the solid solution. The process involves precipitation and growth of an equilibrium Mg2Si β-phase, its intermediate phases and clusters in the aluminum matrix. By the data of [13, 14], alloy 6082 contains intermetallic inclusions of different compositions and sizes, i.e., coarse (1 – 10 μm) and very fine (0.1 – 1 μm) particles, in addition to the equilibrium β -Mg2Si phase. Addition of copper into an Al – Mg – Si alloy promotes formation of a Q-phase, which affects positively the kinetics of artificial aging [15]. A study of these phases in the structure of a friction stir welded joint has shown the presence of four regions with specific distributions of the particles of the hardening phase (the size, the density, the morphology). The differences are a result of the action of different maximum temperatures in different regions of the weld in the welding process [16]. It has been shown in [17, 18] that increase in the power of the heat introduced during FSW into Al – Mg – Si alloys causes growth of a stable Q-phase and formation of dispersed Al – Mg – Si intermetallics in the zone of dynamic recrystallization and of a metastable four-component Q′-phase in the thermomechanically affected zone, which worsens the strength properties.

The aim of the present work was to determine the effect of the force and frequency of a friction stir welding pulse on the microstructure (the grain size, the distribution of the oxide layer and of the particles of the hardening phase) and mechanical properties (the hardness, the strength, the ductility) of a welded joint of an aluminum alloy of the Al – Mg – Si system.

Methods of Study

We applied impulse and traditional friction stir welding to sheets with a size of 125 × 400 mm and a thickness 2 mm from aluminum alloy 6082 in state T6. The alloy had the following chemical composition in wt.%: 0.89 Si, 0.40 Fe, 0.31 Cu, 0.40 Mn, 1.18 Mg, the remainder Al. The mechanical properties of the base metal were as follows: σr = 323 MPa; σ0.2 = 257 MPa; δ = 6.3%.

The structure of the base metal in the state as delivered contains grains extended in the rolling direction (about 38 mm in the longitudinal direction and about 20 μm in the transverse direction) (Fig. 1).

Fig. 1.
figure 1

Structure of the base metal (alloy 6082-T6).

Full size image

Friction stir welding was conducted in a triaxial miller equipped with a special tool and a hydraulic system providing rotation and motion of the tool with simultaneous feeding of pulses; the cylindrical tip had a diameter of 4 mm and the shoulder had a diameter of 12.5 mm. In the process of FSW the tool was immersed into the material to a depth of 1.8 mm. The plates were pressed by massive grips to the steel substrate at a minimum distance between them. The FSW parameters were as follows: welding speed 400 mm/min, speed of rotation of the tool 710 rpm, angle of slope of the axis of the tool 2°. The parameters of the IFSW process are given in Table 1.

Table 1. Parameters ofWelding Processes, Characteristics of Hardening Phase in DRZ and Mechanical Properties ofWelded Joints of Alloy 6082-T6
Full size table

We performed a metallographic analysis for determining the regions (zones) of the welded joints and studying the microstructure. The samples for the light microscopy were prepared by color etching using a reagent containing 1 g NaOH, 4 g KMnO4, and 100 ml H2O. The microstructure of the base metal and of the zones of the weld was studied using a Leica DM 2500 M light microscope at a magnification of up to × 500. The grain size was determined automatically with the help of the QX-Grain software in the zones of dynamic recrystallization, thermomechanical effect and heat effect, and in the base metal. The microstructure was also studied with the help of a MIRA II XMH Tescan scanning electron microscope at the following parameters: voltage 15 kV, accelerating voltage 124.4 eV, angle of deflection 40.6°. The microscopic x-ray spectrum analysis of inclusions of the hardening phase was conducted using an INCA Oxford spectrometer (50 mm2 ) attached to the microscope.

The Vickers hardness was measured over the line passing at a distance of 1.1 mm from the lower surface of the weld at a load of 0.9807 N applied for 15 sec. The transverse specimens with rectangular section were tested for tensile strength using a Walter + bai AG universal machine at a loading speed of 1.2 mm/min until breakage of the specimen. Two spring clips were mounted on the specimen so that the distance between them was 50 mm.

Results

Figure 2 presents the surfaces of the welded joints formed by FSW and IFSW under the shoulders of the tool. It can be seen that feeding of a pulse during the FSW changes the appearance of the weld. When the pulse intensity is increased, the sizes of the flakes grow, while increase in the frequency of the pulse increases their number. Increase in the pulse intensity to 10 kN produces a sheet-thinning defect as a result of knocking-off of the material in the welding process. We detected no other defects on the upper and lower surfaces of the welded joint.

Fig. 2.
figure 2

Surfaces of welded joints from alloy 6082-T6 after FSW (specimen A ) and after IFSW (specimens B, C and D ).

Full size image

Analyzing the structure of a cross section of the weld formed by IFSW (Fig. 3) we detected the same zones (regions) as in the joint formed by FSW. The zone of dynamic recrystallization (DRZ) was located in the center of the cross section and had a fine-grained structure. Under the action of severe plastic deformation and high temperature the thermomechanically affected zone (TMAZ) acquired a strained structure. The heat-affected zone (HAZ) of the structure was coarse-grained, because the material in this zone had experienced a purely thermal action. Note that in contrast to the FSW, the DRZ and the TMAZ became narrower and changed under the IFSW.

Fig. 3.
figure 3

Macrostructure of a welded joint of alloy 6082-T6 in cross section after IFSW: IS and RS) invasion and retreat sides; HAZ and TMAZ) heat-affected and thermomechanically affected zones, respectively; DRZ) dynamic recrystallization zone

Full size image

The study of the macro- and microstructures in cross sections of the welded joints formed by FSW and IFSW has shown that the oxides getting into the joint under impulse welding are more refined and more uniformly distributed in the structure (Fig. 4). After the FSW the oxide layer remains in the DRZ in the form of a wavy band, whereas after the IFSW it is distributed in the form of fine oxide particles. Increase in the pulse intensity under the IFSW intensifies this effect due to higher straining of the metal, which should affect positively the mechanical properties of the weld [16].

Fig. 4.
figure 4

Macrostructure of the weld in cross section of IFSW joint (a, c, e) and microstructure of the singled out regions (b, d, f ): a, b ) specimen A; c, d ) specimen B; e, f ) specimen C.

Full size image

The results of the study of the microstructure in the DRZ reflect formation of fine equiaxed grains of matrix α-solid solution after the IFSW (Fig. 5), like under the FSW [18,19,20,21]. Figure 6 presents the variation of the grain size in the DRZ, TMAZ and HAZ of welded joints depending on the parameters of the used IFSW. At a low pulse intensity, the grain size ranges within 2.5 – 3.0 μm. At a constant pulse frequency, the grain size decreases with growth in the pulse intensity to 2.5 kN; at a pulse intensity of 10 kN it increases to about 3.7 μm. The average grain size in the TMAZ and HAZ differs inconsiderably both for the FSW and for the IFSW. Note that after the IFSW at a low pulse intensity (1 – 2.5 kN), the grain size in the DRZ is smaller than after the FSW.

Fig. 5.
figure 5

Structure of a welded joint of alloy 6082-T6 in the region of dynamic recrystallization after FSW (specimen A ) and IFSW (specimens B, C, D ).

Full size image
Fig. 6.
figure 6

Grain size of a welded joint of alloy 6082-T6 after IFSW at different pulse intensities (given at the columns in kN): DRZ) dynamic recrystallization zone; TMAZ) thermomechanically affected zone; HAZ) heat-affected zone.

Full size image

Figure 7 presents the distribution of hardening phases in the structure of DRZ after FSW and IFSW. It can be seen that these phases have the form of light coarse inclusions of an irregular extended shape with a size of 1 – 5 μm. The x-ray spectrum microanalysis has shown that the hardening particles in the DRZ are unlike inclusions with different concentrations of Si, Mn, Mg, Fe and Cu. The chemical compositions and the sizes and shapes of the inclusions allow us to speak of the presence of intermetallic particles of AlFe(Mn)Si, a β-phase (Mg2Si) and/or a Q-phase (of the Al – Mg – Si – Cu system) in the DRZ. The parameters of the particles of the hardening phases in the DRZ of the FSW and IFSW joints are presented in Table 1. It follows from the Table that the volume fraction and the average size of the particles of hardening phases in the DRZ of the joints formed by IFSW with a low pulse (below 1 kN) are lower than after the FSW. Increase in the pulse intensity of the IFSW to 10 kN causes considerable growth in the volume fraction of coarse particles and growth of stable phases and intermetallics of the Al – (Fe(Mn) – Si system in the structure of the alloy.

Fig. 7.
figure 7

Structure of the dynamic recrystallization zone in a welded joint of alloy 6082-T6 (electron microscope): a) specimen A, FSW; b ) specimen B, IFSW.

Full size image

The results of the measurement of microhardness of welded joints are presented in Fig. 8. It has been shown in [14, 22] that heat-hardened aluminum alloys exhibit W-shape curves of distribution of microhardness over cross sections of welds. A profile of such a shape contains a plateau in the center of the weld, the width of which commonly corresponds to the width of the shoulders. With distance from the center of the weld the hardness decreases to a minimum value in the HAZ and then grows to the hardness of the base metal. In our work we obtained similar curves of hardness distribution in the welds formed by FSW and IFSW with low pulse intensity (1 kN). The hardness of the base metal is about 120 HV0.2 and decreases from the edge to the center of the weld (to about 80 HV0.1 ). The lowest hardness value has been detected in the HAZ at a distance of 6 – 7 mm from the center of the weld. The results show that the pulse parameters in the IFSW affect considerably the hardness in the DRZ. The highest hardness has been obtained in the welded joints formed at a low pulse intensity (up to 1 kN) and amounts to 97 – 100 HV0.1 (79% of the hardness of the base metal). Under the IFSW with a higher pulse intensity the hardness in the DRZ is 36% lower than that of the base metal.

Fig. 8.
figure 8

Distribution of microhardness HV0.1 in welded joints of alloy 6082-T6 after FSW and IFSW (h is the distance from the center of the weld): 1 ) 0 kN; 2 ) 1.0 kN, 8.9 Hz; 3 ) 2.5 kN, 8.9 Hz; 4 ) 10.0 kN, 5.9 Hz.

Full size image

The results of the tensile tests of FSW and IFSW joints are presented in the Table. Comparative analysis shows considerable increase in the ultimate strength and elongation and some decrease in the yield strength of the welded joints formed by the IFSW as compared to the same characteristics after the FSW. However, further increase in the pulse intensity is accompanied by decrease in the strength and ductility parameters of the joints, which may be caused by a sheetthinning defect. In the joints welded at 1 kN and 8.9 Hz the place of failure is located in the HAZ on the side of invasion, whereas the other specimens fail over the joining line, i.e., in the zone of dynamic recrystallization (Fig. 9).

Fig. 9.
figure 9

Appearance of specimens A – D after tensile tests until failure of welded joints of alloy 6082-T6 produced by FSW and IFSW.

Full size image

Conclusions

  1. 1.

    The structure of the welded joints of sheets of aluminum alloy 6082-T6 (of the Al – Mg – Si system) produced by impulse friction stir welding contains three regions typical for the traditional friction stir welding. However, application of a pulse reduces the width of the zones of dynamic recrystallization and thermomechanical effect in the welded joint.

  2. 2.

    The oxide layer is stirred more efficiently in the joints formed by IFSW than in the case of FSW.

  3. 3.

    Application of a pulse promotes refinement of the structure of the welded joint, especially in the zone of dynamic recrystallization, due to the action of additional plastic strain during the friction stir welding.

  4. 4.

    The parameters of the IFSW process affect substantially the kind of the structure of the welded joint of the aluminum alloy including the sizes and the distribution of the particles of the hardening phases.