Introduction

Friction stir welding (FSW) is a solid-state joining process and is considered to be the most significant progress in metal joining in the past few decades due to its energy efficiency, environmental friendliness, and versatility (Ref 1). FSW is of significant importance especially for joining of dissimilar metals with wide difference in thermal and mechanical properties. However, joining of dissimilar metals involves severe flaws like the formation of brittle intermetallic compounds (Ref 2,3,4). Joining of dissimilar metals is more challenging than that of similar metals due to variation in properties, but is often required to meet the growing industrial needs (Ref 5). FSW came in the limelight for joining of low and high melting point metals as it helps to solve the complex functional problems in joining of dissimilar materials. Aluminum has good heat transfer property, good formability and low weight that are vital for aerospace industries, automobile parts, and some naval components. Stainless steel possesses high strength, high corrosion resistance, and high toughness. Stainless steel is a significant structural material extensively used in automotive and aerospace industry. FSW of stainless steel and aluminum is attractive, because such joints provide the design engineers an opportunity to explore their applications in the field of cryogenics where cryogenic liquid is stored in Al chambers and transferred through SS pipeline (Ref 6,7,8). Nuclear industry is also looking forward to SS-Al assembly to fabricate neutron sensitive ion chambers (Ref 9). Combinations of aluminum and stainless steel are potential candidates for developing industrial products especially in automobile industry due to their high strength-to-weight ratio leading to higher fuel efficiency (Ref 10,11,12). FSW of Al 6061-T6 and SS (STS304) at tool rotational speeds of 300-600 rpm and travel speed of 48 mm/min were reported (Ref 13). Ghosh et al. (Ref 3) used varying tool rotational speeds from 560 to 900 rpm for joining of 6061 Al and 304 stainless steel. Chen (Ref 14) used traverse speeds of 54, 72, and 90 mm/min in joining of aluminum to steel and the minimum traverse speed gave the best impact values and acceptable values of tensile strength. Tool rotational speed and tool feed rate are significant parameters in making a sound joint. Habibnia et al. (Ref 2) employed variations in offset as 0, 0.8, and 1.5 mm for joining of 304 SS and 5050Al. The offset was provided with an aim to reduce overheating and pin erosion. They reported that at zero offset cracks were formed, at 0.8 mm offset tunnel defects were observed and for 1.5 mm tool offset, rotational movement of tool abraded the edge of steel sheet that effected a uniform distribution of steel particles in the stir zone due to small size of the steel particles and consequently lesser defects resulted in the stir zone (Ref 2). The possible intermetallic phases formed at the interface were reported to be FeAl, Fe2Al5, FeAl2, FeAl3, and Fe4Al13 (Ref 15,16,17). However, brittle Al-rich intermetallic compounds were formed at the interface of aluminum and steel joints due to intermixing of materials during welding and it was reported that the thicker intermetallic layers lead to failure of the joint under stressed conditions (Ref 18). Ghosh et al. (Ref 9) have performed the joining of CP Al-304SS at 1000 rpm for traverse speed of 50 mm/min. However, report on the effect of tool rotational speed on joining of dissimilar material CP Al and 304SS is scanty in the literature. In the present study, FSW of commercial pure aluminum and 304SS with tool rotational speeds from 200 to 1000 rpm in steps of 200 rpm at 60 mm/min traverse speed was carried out. In this study, microstructure of different zones, formation of intermetallic phases, microhardness, tensile strength, and fracture morphologies of the FSW joints has been investigated.

Experimental Procedure

Commercially pure aluminum (CP Al) and 304 stainless steel (304SS) sheets having dimensions of 160 mm length × 60 mm width × 3 mm thickness were used in this investigation. The chemical composition of CP Al is Al-0.02Fe-0.2Si-0.003Mn-0.002S-0.002P-0.005N (wt.%) and of 304SS is Fe-18.2Cr-8.4Ni-1.34Mn-0.04C-0.4Si (wt.%). The sheets were cleaned with ethanol and then mounted on the machine to make a butt joint. CP Al was placed on the retreating side and 304SS on advancing side. The welding direction was parallel to the rolling direction of the sheets. A tungsten carbide tool with a shoulder diameter of 25 mm, conical probe of 6 mm diameter, and pin length of 2.7 mm was used for the welding process. The welding tool was tilted 2° angle forward from the vertical and shifted toward the Al side with an offset of ~ 1.25 mm as shown in Fig. 1. The two major reasons for providing the offset were to avoid tool breakage while plunging at the interface due to higher strength of stainless steel side with respect to aluminum side and also to avoid superheating of aluminum due to the difference in the material flow behavior of the two materials. FSW was performed at tool rotational speeds of 200-1000 rpm in steps of 200 rpm, with a travel speed of 60 mm/min. Both base metals and welded samples were prepared by conventional metallographic techniques. CP Al side was first etched using Keller’s reagent, and then 304SS side was etched using aqua regia. Microstructures of the welded samples were examined with an optical microscope. The polished interface of the welded joint was examined using a scanning electron microscope SEM (JEOL JSM-5510) in backscattered mode. Energy-dispersive spectroscopy (NORAN System Six) was used to determine the chemical composition of the interface. X-ray diffraction (Philips PW 1830) technique was used to conform the phases predicted. To evaluate the mechanical properties, tensile and micro-microhardness tests were carried out. Tensile test of the welded joints was carried out in an Instron 4204 machine at a constant crosshead speed of 1.67 × 10−3 mm/s. The tensile samples were prepared according to the ASTM E8M-11 standard. Microhardness was measured along the transverse section of the welded specimen using a diamond micro-indenter with a load of 100 gf and dwell time of 20 s. Indentations were made at mid-section of the welded joints. The fracture surfaces of the welded samples were examined by SEM (Hitachi S 3400) in secondary electron mode.

Fig. 1
figure 1

Schematic illustration of tool offset position for FSW

Full size image

Results and Discussion

Microstructural Characterization

The surfaces of the friction stir welded samples processed at varying rotational speeds are shown in Fig. 2. The welded surface at lower rotational speed experienced less amount of heat generation during stirring action (Fig. 2a and b). However, at higher rotational speeds flashes were formed due to greater heat generation as shown in Fig. 2(c), (d), and (e).

Fig. 2
figure 2

Surface view of the welded joints at varying tool rotational speeds (a) 200 rpm, (b) 400 rpm, (c) 600 rpm, (d) 800 rpm and (e) 1000 rpm

Full size image

The optical photomicrographs of friction stir welded joints are shown in Fig. 3. Three zones are formed during welding: heat affected zone (HAZ), thermo-mechanically affected zone (TMAZ), and stir zone (SZ). The microstructure of HAZ region at the stainless steel side does not show much difference as compared to base material. The grains are elongated in the TMAZ due to deformation. At the stir zone, the grains are finer due to the stirring action of the pin. At the retreating side, more reaction takes place due to shifting of pin toward the CP Al side. So, more plastic deformation occurs on the CP Al side. Stainless steel particles of different size were distributed over the CP Al side due to stirring action of the pin. These particles are irregular in shape and unevenly distributed. Cracks are formed on the steel particles as a result of high deformation and mixing of materials (Ref 6). The grains formed in the stir zone of aluminum side are fine, equiaxial, and of recrystallized structure. At higher rotational speeds, the grains are coarser due to severe plastic deformation followed by recrystallization on the aluminum side (Ref 19, 20). The average grain sizes of the different welded zones for varying tool rotational speeds are shown in Fig. 4. Hassan et al. (Ref 21) reported that the grain structure is very finer at lower heat inputs, whereas at higher heat inputs the grain structure is coarser due to grain growth. Hence, the grain size at lower rotational speeds is finer as compared to that of higher tool rotational speeds.

Fig. 3
figure 3

Optical photomicrograph of varying tool rotational speeds at (a) HAZ of SS side at 400 rpm, (b) TMAZ of aluminum side at 400 rpm, (c) TMAZ of SS side at 600 rpm, (d) SZ of aluminum side at 800 rpm, (e) SZ of SS side at 600 rpm and (f) interface region at 1000 rpm

Full size image
Fig. 4
figure 4

The average grain size of different welding zones for varying rotational speed at 60 mm/min traverse speed

Full size image

The SEM images of the friction stir welded joints are shown in Fig. 5. Intermetallic compounds are formed at the stir zone due to the thermal and mechanical action that took place in the material during stirring of the pin. The images were taken at the interfacial region of the welded sample. Figure 5(a) shows intermetallic layer thickness (marked by arrow) at lower rotational speed which is much less compared to that at higher rotational speed (Fig. 5f).This demonstrates that the intermetallic layer thickness increases with increase in rotational speed (Ref 22). Figure 5(b) exhibits an insignificant amount of aluminum particles at 304SS side besides the interface, but steel particles were observed at the CP Al side (Fig. 5c).This is due to the fact that 304SS was placed at the advancing side and CP Al at the retreating side, the tool rubbed the surface of the 304SS and pulled the steel particles over to the CP Al side. The other reason could be the 1.25 mm offset of the tool toward the CP Al side; hence, there was no significant structural change in the stainless steel side. In Fig. 5(d) and (e), steel particles are observed in broken condition that can be attributed to rotation of the pin and the plastically deformed aluminum trying to fill the gaps.

Fig. 5
figure 5

SEM–BSE images of varying tool rotational speeds (a) interface region of welded joint at 200 rpm, (b) interface region of welded joint at 400 rpm, (c) SS particle on the aluminum side at 400 rpm, (d) interface region of welded joint at 600 rpm, (e) interface region of welded joint at 800 rpm and (f) reaction region at 1000 rpm

Full size image

The peak temperature at the interface was described by Arbegast et al. (Ref 23) and expressed by the equation

$$ \frac{T}{{T_{\text{m}} }} = K\left[ {\frac{{\omega^{2} }}{{V \cdot 10^{4} }}} \right]^{\alpha } $$

where T is the temperature during welding, Tm is the melting point of the alloy, V is the traversing speed of the tool (mm/s), α is a constant and considered as 0.04, and K is a constant considered as 0.65 (Ref 3, 23). The temperatures at the interface for varying tool rotational speeds are given in Table 1. The temperatures at the interface indicate the prospect of formation of intermetallic phases like FeAl3, Fe2Al5, FeAl2, and Fe4Al13 that is vindicated from Al-Fe phase diagram (Ref 24).

Table 1 Temperature at the interface during friction stir welding
Full size table

The quantitative analysis of friction stir welded samples is shown in Fig. 6. Figure 6(a) shows that the steel particles were distributed in the aluminum side and the particles comprised of Al (71-78 wt.%), Fe (18-21 wt.%) (at point 1), Al (~ 97 wt.%) Fe (~ 3 wt.%) (at point 2), Al (~ 16 wt.%), Fe (~ 72 wt.%), Cr (~ 8 wt.%), Ni (~ 4 wt.%) (at point 3) and Al (~ 42 wt.%), Fe (~ 51 wt.%), Cr (~ 5 wt.%), Ni (~ 2 wt.%) (at point 4), so the phases observed are FeAl3, Al, FeAl, and FeAl, respectively. Figure 6(b) shows that the cracks were formed within the steel particles and intrusion of aluminum particle had taken place. The composition of the particles is Al (~ 19 wt.%), Fe (~ 58 wt.%), Cr (~ 7 wt.%), Ni (~ 4 wt.%) (at point 1), Al (~ 93 wt.%), Fe (~ 7 wt.%) (at point 2), Al (~ 98 wt.%), Fe (~ 2 wt.%) (at point 3), Al (~ 38 wt.%), Fe (~ 53 wt.%), Cr (~ 6 wt.%), Ni (~ 3 wt.%) (at point 4), Al (~ 12 wt.%), Fe (~ 78 wt.%), Cr (~ 6 wt.%), Ni (~ 4 wt.%) (at point 5), and Al (~ 13 wt.%), Fe (~ 65 wt.%), Cr (~ 8 wt.%) and Ni (~ 4 wt.%) (at point 6), so the formation of FeAl3, Al, Al, FeAl, Fe, and FeAl3 phases, respectively, are confirmed (Ref 9).The compositions of the intermetallic phases were determined from Al-Fe, Al-Cr binary phase diagrams (Ref 24).The phases have major quantity of Al with considerable quantity of Fe and very less of Cr and Ni which indicates insignificant influence of chromium and nickel in the formation of phases. The line profile of the interface region is shown in Fig. 6(c) and summarized in Fig. 7. On 304SS side, the material flow was within the interface region and the interface took a non-linear shape. The formation of non-linear shape, according to Lee et al. (Ref 25), is due to stirring of the steel side by the outer edge of the pin. Figure 7 shows an increase in aluminum content (in counts) at the cracks of the SS side due to the extrusion of the pin and the plastically deformed aluminum trying to fill the gaps. The intermetallic phase formed at the interfacial zone (vide Fig. 6d) consists of Al (~ 66 wt.%), Fe (~ 24 wt.%), Ni (~ 2.4 wt.%), and Cr (~ 3.5 wt.%). This composition indicates the formation of FeAl3 intermetallic phase, which is Al-rich phase (Ref 22, 26). Yazdipour et al. (Ref 27) reported the formation of FeAl3 intermetallic compound at the interface while joining aluminum 5083-H321 and 316L.

Fig. 6
figure 6

Energy-dispersive spectroscopy images (a) SS particle on aluminum side at 400 rpm, (b) interface region at 600 rpm, (c) line scan of interface region at 800 rpm, and (d) interface region at 1000 rpm

Full size image
Fig. 7
figure 7

SEM–EDS line scan showing composition gradients across the interface; taken along the line at 800 rpm

Full size image

Figure 8 shows x-ray diffraction patterns of the friction stir welded joints at varying rotational speeds. The phases like FeAl2, Fe4Al13, and Fe2Al5 were observed in the welded joints. While FeAl3 intermetallic phase was observed in SEM micrographs, Fe4Al13 and Fe2Al5 intermetallics were not observed presumably due to low volume fraction. Sahim (Ref 28) reported the formation of FeAl3 phase in FSW of aluminum and AISI 304. Ghosh et al. (Ref 3) have reported that the formation of phases: FeAl2, Fe2Al5, and Fe3Al between Al 6061 and SS304 in friction stir welded joints. However, Elrefaey et al. (Ref 16) reported the formation of intermetallic phases like Fe4Al13 and Fe2Al5 in friction stirred lap joint of aluminum and steel. The structure and morphology of intermetallic phases were investigated by Bozzi et al. (Ref 29) for joining of Al and 6016/IF-steel, and the precipitates were confirmed to be either FeAl2 (rhombohedric phase), Fe2Al5 (orthorhombic phase) or FeAl3 (monoclinic phase). Agudo et al. (Ref 30) observed that most of the FeAl3 precipitates contained a large amount of microtwins arising probably due to high stress generated by the volume expansion associated with the formation of the FeAl3 phase.

Fig. 8
figure 8

XRD pattern of welded sample for varying tool rotational speeds at tool traverse speed of 60 mm/min

Full size image

Mechanical Properties

Tensile tests were performed at room temperature for the base metals and the joint samples. The results are shown in Fig. 9. Maximum tensile strength of ~ 119.7 MPa was obtained at 400 rpm tool rotational speed, which was about 90% of the CP Al. Ghosh et al. (Ref 9) have reported an ultimate tensile strength (UTS) value of 82% of the parent Al (UTS-74 MPa) for CP Al-304SS FSW joint. Yilmaz et al. (Ref 31) have obtained ~ 88 MPa bond strength in friction welding of CP aluminum to 304SS. Yazdipour et al. (Ref 27) have reported tensile strength of ~ 75% of the base material while joining of aluminum 5083-H321 and 316L. However, with increase in the rotational speed, the tensile strength decreased gradually and the minimum tensile strength of ~ 78 MPa was achieved at 1000 rpm tool rotational speed.

Fig. 9
figure 9

Tensile properties of friction stir welded joints and base material at room temperature for varying tool rotational speeds

Full size image

Figure 10 shows the summary of tensile strength results with varying tool rotational speeds from 200 to 1000 rpm. The maximum tensile strength was achieved at a tool rotational speed of 400 rpm. At higher tool rotational speeds, more deformation and high mixing of materials occur that leads to the formation of brittle intermetallic compounds. To eliminate excess melting phenomena in the weld nugget and in order to get good tensile strength, the tool rotational speed should be low. The intermetallic layer thickness for tool rotational speed of 1000 rpm was ~ 4.80 µm, whereas ~ 0.7 µm for 400 rpm. At 200 rpm, intermetallic layer was not observed due to low heat input and mixing of materials. Yilmaz et al. (Ref 31) reported that the tensile stress is highly influenced by the intermetallic layer thickness at the reaction zone of the welded joint and increase in layer thickness leads to a decrease in tensile strength.

Fig. 10
figure 10

Tensile strength and percentage elongation at varying tool rotational speeds

Full size image

The fracture surface of friction stir welded joints of CP Al side and 304SS side are shown in Fig. 11 and 12, respectively. The fractograph shows dimples at the center of the joint which indicates ductile fracture. The flat flakes observed at the bottom of the joint indicate brittle fracture.

Fig. 11
figure 11

Fractured surface of tensile sample on aluminum side

Full size image
Fig. 12
figure 12

Fractured surface of tensile sample on SS side

Full size image

Microhardness was measured across the cross section at the center of the friction stir welded joints, and successive indentations were taken at an interval of 10 µm. The results are shown in Fig. 13. The hardness profiles were carried out at different zones, and interaction between successive indentations was minimized to the greatest possible extent. The microhardness variation of different welded joints by varying rotational speeds is shown in Fig. 14. The hardness reaches the peak at the interface of the welded sample. Mahto et al. (Ref 32) have reported of the maximum hardness value at the interface zone, indicating more brittle nature of the interface due to the stirring action of the pin leading to excessive plastic deformation and recrystallization. The grain size and brittle intermetallics influenced hardness values. The hardness decreased slightly in TMAZ of CP Al at the retreating side, which indicated second phase particle dissolution and coarsening caused by thermo-mechanical effect due to dynamic recovery and recrystallization (Ref 1). The decrease in hardness at the HAZ is due to the dissolution of the precipitates during welding. The increase in the hardness of the SS (near the interface) is directly related to high deformation. The maximum hardness value of the welded joint was observed for the joints processed at 1000 rpm.

Fig. 13
figure 13

Schematic sketch of the arrangement of microhardness indentation across the interfaces of friction stir welded sample

Full size image
Fig. 14
figure 14

Hardness profile of varying tool rotational speeds (a) 200 rpm, (b) 400 rpm, (c) 600 rpm, (d) 800 rpm and (e) 1000 rpm

Full size image

Conclusions

  1. 1.

    The grain structure of CP Al side underwent severe plastic deformation accompanied by dynamic recrystallization. The grain structural change in stainless steel side was minimal due to the offset of tool toward the CP Al side. Fine particles of stainless steel were distributed over the CP Al side.

  2. 2.

    The intermetallic layer thickness increased with increase in rotational speed. The intermetallic phases like FeAl2, Al13Fe4 and Al5Fe2 were detected in x-ray diffraction patterns. The intermetallic compound formed at the weld nugget was FeAl3.

  3. 3.

    The tensile strength of the welded joints showed the maximum value of about ~ 90% of CP Al at 400 rpm tool rotational speed and the tensile strength decreased gradually with the increase in tool rotational speed due to an increase in the volume fraction of brittle intermetallics. However, at a tool rotational speed of 200 rpm, heat generation and mixing of material were not enough to make a good joint.

  4. 4.

    Hardness value reached a maximum at the stir zone due to the formation of brittle intermetallics. With an increase in rotational speed, hardness value of the stir zone increased due to an increase in temperature promoting the formation of more intermetallics.