1 Introduction

As the lightest structural materials, magnesium and its alloys have excellent specific strength, good castability, hot formability, excellent machinability, good electromagnetic interference shielding, and recyclability [1,2,3]. Mg alloys are emerging as important engineering materials, especially in aerospace, automobile, electronic communications, and military fields. Because of the poor corrosion and wear resistance, the surface properties of Mg alloy are the key factors affecting its application [4, 5], and they are gradually receiving attention. Previous studies indicate that microstructures, especially the precipitations and grain size, have an impact on the corrosion and wear resistance of Mg alloys [4, 6,7,8]. Therefore, it is meaningful to understand the relationships between the microstructure evolution and surface properties of Mg alloys, so as to improve the corrosion and wear resistance and expand their application.

Friction stir processing (FSP) is based on the principle of friction stir welding (FSW), which was invented at The Welding Institute (TWI) in 1991 [9]. Instead of joining, FSP is a promising method of microstructure modification. By severe plastic deformation, FSP results in grain refinement and microstructural homogenization as well as the improvement in properties [10,11,12,13,14]. In addition, as reported in previous studies [15,16,17], multipass FSP causes the materials to be repeatedly subjected to the thermal–mechanical effects and leads to significant changes in microstructure and properties for FSP Mg plates. Many studies found that the subsequent passes could further refine the grains in the stir zone, and an effective improvement in both strength and ductility can be achieved in multipass FSP of Mg–Al–Zn alloys [15, 16, 18,19,20]. Lee et al. [21] suggested that multipass FSP can improve microstructural uniformity by accumulating a higher degree of strain to recrystallize the initial grains fully. Most previous studies have focused on the changes in microstructure and mechanical properties of Mg alloys during multipass FSP. Few studies have reported about the effect of multipass FSP on the surface corrosion and wear resistance of Mg alloys.

Mg–Zn–Zr alloy is a typical age-hardening alloy and has the characteristics of high strength and heat treatability, and it has wide potential application in aerospace industries [22]. The surface properties have an impact on its application in aerospace [23]. Recently, it was found that, because of the significant grain refinement, as well as the break-up and redistribution of precipitates in the FSW process, the surface corrosion and wear resistance of ZK60 alloy can be enhanced by FSW [23]. However, the effect of subsequent passes on the surface corrosion and wear resistance of Mg–Zn–Zr alloy during multipass FSP has not yet been reported. Therefore, in the current work, microstructure evolutions on the top surface of multipass FSP ZK60 plates were studied. The effects of multipass FSP on the surface corrosion and wear resistance of ZK60 plates were evaluated. The present work is helpful for further understanding the surface properties of Mg–Zn–Zr alloy, as well as choosing the appropriate technologies to improve the surface properties of ZK60 alloy.

2 Experimental Procedures

2.1 Microstructure Observation

The initial material was cast ZK60 Mg alloy (Mg-5.2 wt% Zn-0.5 wt% Zr) with an aged state of 448 K for 10 h. The thickness of the ZK60 plates was 6 mm. After being polished by abrasive paper and cleaned with acetone, several ZK60 plates were subjected to FSP. A cylindrical thread probe with a length of 4 mm and a diameter of 5 mm was used. The diameter of the tool shoulder was 20 mm. To avoid the effects of processing parameters on the microstructure evolution and surface properties of ZK60 alloy, multipass FSP was performed with the same rotation speed and traveling speed of 1000 rpm and 90 mm/min, respectively. The pin travel direction was the same with respect to the previous pass with a 100% overlap. FSP experiments using from one pass (1-PASS) up to three passes (3-PASS) were tried. Thus, four types of specimen were used in the current study: base metal (BM) specimen, 1-PASS specimen, 2-PASS specimen, and 3-PASS specimen.

Optical microscopy (OM) and scanning electron microscopy (SEM) were used to examine the microstructure and precipitates on the top surface of the multipass FSP ZK60 alloy. Before the examination, metallographic specimens were ground and polished by standard methods. Then, polished samples were etched with an acetic–picric solution (2 ml of distilled water, 2 ml of glacial acetic acid, 14 ml of ethanol, and 0.84 g of picric acid) to examine the microstructure.

2.2 Corrosion Tests

Potentiodynamic polarization and electrochemical impedance spectroscopy tests were carried out in a corrosion cell containing 450 ml of 3.5-wt% NaCl solution at room temperature by using a corrosion measurement system. Prior to the experiment, the samples were polished using 3000 grit emery paper to achieve a mirror finish. A three-electrode system was implemented: a saturated calomel electrode (SCE) and a platinum plate were used as the reference electrode, together with an auxiliary electrode. Potentiodynamic polarization tests were carried out at a scan speed of 0.5 mV/s and commencing from − 1.8 to − 0.9 V. All potentials in the study are quoted with respect to the SCE. The electrochemical impedance spectroscopy was measured under the frequency range from 10 mHz to 100 kHz. The voltage amplitude was ± 5 mV rms. Each test was repeated at least three times to obtain reliable results.

Corrosion resistance specimens were extracted from four types of specimen with dimensions of 10 × 10 × 6 mm. Nonworking surfaces were impregnated with epoxy resin. The immersion surface in the specimens was ground with 3000# grit SiC paper and then cleaned with alcohol and dried by flowing air. Immersion experiments were carried out in 3.5-wt% NaCl solution at room temperature. The emitted hydrogen volume was detected by using a burette during the immersion for 118 h. The corrosion surfaces were studied by SEM after immersion for 1.5 h as well as for 118 h.

2.3 Hardness and Wear Tests

Hardness profiles on the top surface of the specimens were obtained with a 200-g load for a 10-s loading time by using a Vicker’s microhardness tester. The wear resistance of those ZK60 specimens was studied by using a friction and wear tester. Specimens were prepared by wire cut in 10 × 10 × 6 mm dimensions from four types of specimen. A load of 50 N was applied with a sliding speed of 50 mm/min for 20 min during the wear tests. In addition, all the microstructure examinations and surface property measurements were concentrated on the top surface of the specimens.

3 Results

3.1 Microstructure Evolution

Optical microstructures of four types of ZK60 plate are presented in Fig. 1. Coarse grains with an average grain size of ~ 68.2 μm were observed by the linear intercept method in the BM specimen. Figure 1b shows significant grain refinement in FSP ZK60 alloy. The small grains cannot be clearly observed in the optical photograph. Figure 1 indicates that the grain size of the 3-PASS specimen is larger than that of the 1-PASS and 2-PASS specimens, which would be verified by SEM images with high magnification. SEM images of four types of ZK60 plates are presented in Fig. 2. The average grain size of different specimens is measured based on more than seven SEM images, and the results are presented in Fig. 3. The average grain size in multipass FSP ZK60 alloy is 3.1 ~ 4.5 μm, which is significantly lower than that of the BM specimen (~ 68.2 μm). The observation confirms that the grains in multipass FSP of ZK60 plates are significantly refined compared with the BM.

Fig. 1
figure 1

Optical microstructure in different specimens: a BM specimen; b 1-PASS specimen; c 2-PASS specimen; d 3-PASS specimen

Full size image
Fig. 2
figure 2

SEM images showing microstructure in different specimens: a BM specimen; b 1-PASS specimen; c 2-PASS specimen; d 3-PASS specimen

Full size image
Fig. 3
figure 3

Average grain size in different ZK60 specimens

Full size image

The finest grains among the multipass FSP of ZK60 plates were in the 1-PASS specimen with an average grain size of 3.1 μm. The 2-PASS specimen had an average grain size of approximately 3.6 μm. The grains in the 3-PASS specimen were the coarsest, with the average grain size being approximately 4.5 μm. The results indicate that subsequent passes cannot further refine the grains for multipass FSP of ZK60 plates, but they cause an increase in the grain size in the stir zone. The observation is largely different from previous studies about multipass FSP of Mg–Al–Zn alloys. For example, Luo et al. [15] reported that subsequent passes of FSP caused the average grain size of AZ61 alloy to decrease from 13 to 7.8 μm. The reason is mainly the accumulative effect of dynamic recrystallization (DRX) during multipass FSP [15, 17, 18]. The unusual observation of grain size in ZK60 alloy would have a close relationship with the effects of precipitates on the DRX process in multipass FSP, which is discussed later in Sect. 4.

Precipitate particles can be observed in different ZK60 specimens, as presented in Fig. 2. Many precipitates with large sizes are segregated along grain boundaries in the BM specimen, which is similar to a previous study [23]. Previous studies show that precipitate particles in ZK60 alloy include Mg–Zn particles (MgZn2 or Mg4Zn7) and Zn–Zr containing particles (most likely Zr2Zn3) [22, 24, 25]. The former is the main reinforcing phase in ZK60 alloy, while the latter has significant effects on grain refinement [22, 26, 27]. Figure 2b–d show that many precipitates with small size are observed in three types of FSP ZK60 plate, and they are almost in a random distribution. In addition, little difference in precipitation is observed between the three types of FSP ZK60 plates. This indicates that subsequent passes have almost no effect on the size and distribution of precipitate particles in the multipass FSP of ZK60 plates.

3.2 Corrosion Resistance

Potentiodynamic polarization plots of different ZK60 specimens in a 3.5-wt% NaCl solution are shown in Fig. 4. The corrosion current (Icorr) and corrosion potential (Ecorr) values were measured considering the extensive linear region observed in the cathodic branch and a linear region in the anodic branch (Fig. 4). The results are presented in Table 1. The BM specimen has the worst corrosion resistance with a lower Ecorr of − 1383.4 mV and a higher Icorr of 7.3 × 10−4 A/cm2 than the three types of FSP ZK60 specimen. The results are consistent with those of previous studies reporting that FSP can enhance the corrosion resistance of Mg alloys [6, 23, 28, 29].

Fig. 4
figure 4

Polarization curves of different ZK60 specimens in 3.5-wt% NaCl solution

Full size image
Table 1 Electrochemical parameter for different ZK60 specimens in the potentiodynamic polarization tests
Full size table

Moreover, there are large differences in Ecorr and Icorr among the three types of FSP ZK60 specimen (see Table 1). The 1-PASS specimen presents the best corrosion resistance with a high Ecorr of − 1215.2 mV and a low Icorr of 4.3 × 10−5 A/cm2. The 3-PASS specimen shows the worst corrosion resistance with a low Ecorr of − 1339.7 mV and a high Icorr of 3.2 × 10−4 A/cm2 among three types of FSP ZK60 specimen. The comparative Ecorr and Icorr values of different specimens are observed in the following order. (a) Ecorr: 1-PASS specimen > 2-PASS specimen > 3-PASS specimen, and (b) Icorr: 1-PASS specimen < 2-PASS specimen < 3-PASS specimen. The results indicate that subsequent passes of FSP are not beneficial to the corrosion resistance of ZK60 plates.

Electrochemical alternating current (AC) impedance results of those ZK60 specimens in 3.5-wt% NaCl solution are presented in Fig. 5. All these samples exhibit one capacitive loop. The diameter of this capacitive loop is associated with the charge transfer resistance. The larger the diameter, the lower corrosion rate the electrode will have [8]. Figure 5 shows that the diameter of this capacitive loop is ranged in the following order: 1-PASS specimen > 2-PASS specimen > 3-PASS specimen > BM specimen. The results mean that 1-PASS specimen would have the best corrosion resistance. The 2-PASS specimen should have better corrosion resistance than the 3-PASS specimen, which is confirmed with the polarization measurements mentioned above.

Fig. 5
figure 5

Nyquist plots of different ZK60 specimens in 3.5-wt% NaCl solution

Full size image

Then hydrogen evolution of various ZK60 specimens immersed in 3.5-wt% NaCl solution is shown in Fig. 6a. After 118-h immersion, the emitted hydrogen volumes of four types of specimen are increased in the order: 1-PASS specimen ≈ 2-PASS specimen < 3-PASS specimen < BM specimen. The emitted hydrogen volume of the BM specimen is approximately 2.3 times that of the 1-PASS specimen. The 3-PASS specimen shows approximately 1.2 times that of the 1-PASS specimen. The corrosion rates of the four types of ZK60 specimen were evaluated based on the emitted hydrogen rate and are presented in Fig. 6b. They show that, initially, the corrosion rate fluctuated a little bit, and the corrosion rate of those specimens increased to the maximum at 4–6 h and then decreased. The corrosion rate of various specimens tends to be stable after immersion for 60 h, which increased in the order: 1-PASS specimen ≈ 2-PASS specimen < 3-PASS specimen < BM specimen. After 118-h immersion, the corrosion rates of the 1-PASS, 2-PASS, 3-PASS, and BM specimens were approximately 1:1:1.75:3.17.

Fig. 6
figure 6

Corrosion resistance of specimens immersed in 3.5-wt% NaCl solution: a hydrogen evolution; b corrosion rate curves

Full size image

The combined analysis of electrochemical AC impedance, potentiodynamic polarization tests and immersion experiments indicate that FSP can enhance the corrosion resistance of the cast and aged ZK60 alloy. However, subsequent passes of FSP are not beneficial, but rather are, harmful to the surface corrosion resistance of ZK60 plates. The reason has a close relationship with the microstructure evolution for multipass FSP of ZK60 specimens. As stated above, subsequent passes of FSP have little effect on the size and distribution of precipitate particles, but they cause an increase in the grain size for multipass FSP of ZK60 specimens (see Fig. 3), which leads to the decline of corrosion resistance of ZK60 alloy. The results are consistent with the previous conclusion that corrosion resistance would be reduced with the increase in the grain size for Mg alloys during FSW/FSP [30, 31].

The corrosion morphologies of the BM and 1-PASS specimens after immersion for 1.5 and 118 h are presented in Fig. 7. Intergranular corrosion occurs in the BM specimen because of the corrosion fissures along grain boundaries presented in Fig. 7a, which is similar to the results of a recent study [23]. It is regarded that the alloys with the precipitation of secondary phase at the grain boundaries are highly susceptible to intergranular corrosion [32]. Some studies suggested that true intergranular corrosion is not easy to occur in Mg alloys because the precipitates segregated along the grain boundaries are almost cathodic to the grains [33, 34]. However, a localized attack of Mg alloys at the grain boundary, being similar to the corrosion morphology in Fig. 7a, could be observed in the early stages of immersion. The corrosion behavior can be considered as intergranular corrosion [32, 35].

Fig. 7
figure 7

Surface morphologies of immersed specimens: a, c specimens with immersion 1.5 h; b, d specimens with immersion 118 h; a, b BM specimen; c, d 1-PASS specimen

Full size image

For the 1-PASS specimen, corrosion products are discretely distributed in the etched surface after immersion for 1.5 h. The corrosion products in Fig. 7c indicate that those regions were subjected to corrosion attack. Moreover, Fig. 7c shows that the corrosion products are cut into many irregular blocks. The size of those irregular blocks is measured to be 7–21 μm, which is much larger than the grain size in the 1-PASS specimen (~ 3.1 μm). It is inferred that the cracks on the corrosion products in Fig. 7c are not along the grain boundaries. Therefore, intergranular corrosion morphology is not found in the 1-PASS specimen. The reason may have a close relationship with the precipitates evolution in the ZK60 alloy during FSP. As stated above, the precipitates in ZK60 alloy were refined and changed in a random distribution after FSP (see Fig. 2). The effect of micro-galvanic corrosion between the precipitate and the matrix of α-Mg promotes localized corrosion around the small precipitates, which results in the corrosion products discretely deposited on the etched surface in the early stages of immersion, and no intergranular corrosion morphology is observed in the 1-PASS specimen (see Fig. 7c). Based on previous studies [28, 36], the corrosion products of Mg alloys might be Mg(OH)2, which has a low toughness and is very prone to cracking. Thus many irregular blocks of corrosion products are observed in Fig. 7c. As the immersion time increases, the number of corrosion products increases (see Fig. 7b, d). The corrosion products are deposited on the specimens, which forms a protective layer and reduces the corrosion rate for those specimens (see Fig. 6b). For 2-PASS and 3-PASS specimens, the corrosion mode and corrosion morphologies are similar to those of the 1-PASS specimen (no image provided).

3.3 Microhardness and Wear Resistance

Measured hardness values on the top surface of different ZK60 specimens are presented in Fig. 8. The 1-PASS specimen had the highest microhardness value of ~ 74.5 HV. The average hardness value of the BM specimen was ~ 65.6 HV. The result indicates that the hardness value of the ZK60 alloy can be improved by FSP. As reported in previous studies [23], grain refinement, as well as the broken up and dispersed precipitates, can result in an increase in the hardness for Mg alloys. Figure 8 shows that the average hardness values of the 2-PASS and 3-PASS specimens are ~ 70.2 HV and ~ 69.5 HV, respectively, which are lower than that of the 1-PASS specimen. The hardness change is well consistent with the grain size change in four types of ZK60 specimen in the present study. Previous studies [1, 37, 38] reported that the hardness value is proportional to the reciprocal of the square root of grain size (d, μm) for FSP/FSW Mg alloys. The Hall–Petch equation for the multipass FSP of ZK60 plates is expressed as HV = 64.2 + 11.3d−1/2. It should be stated that the intercept of the Hall–Petch equation in the present study is similar with that of the FSW AZ91D alloy reported in a previous study [39], in which the Hall–Petch equation was expressed as HV = 64 + 27 d−1/2. The lower slope of the Hall–Petch equation in the present study may be attributed to some reinforcing phases [22, 40], which weaken the effects of grain size on the hardness in ZK60 plates. The results indicate that subsequent passes of FSP can cause a slight decrease of hardness values for ZK60 alloy.

Fig. 8
figure 8

Mean hardness values in different ZK60 specimens

Full size image

Dry sliding wear tests were performed to examine the friction coefficient, and the results are presented in Fig. 9. The BM specimen had the maximum value of the average friction coefficient (~ 0.383). The minimum value of the friction coefficient was achieved by the 1-PASS specimen (~ 0.304). The average friction coefficients of the 2-PASS and 3-PASS specimens were ~ 0.369 and ~ 0.371, respectively. Previous studies reported that samples exhibiting a high friction coefficient would have a low wear resistance [4, 41]. Therefore, it is speculated that the 1-PASS specimen may have a high wear resistance, and the BM specimen would present a low wear resistance. A lower wear resistance may exist in the 2-PASS and 3-PASS specimens compared with the 1-PASS specimen. The results indicate that the surface wear resistance of ZK60 alloy can be enhanced by FSP. The subsequent passes of FSP have little positive effect on the improvement in the surface wear resistance for the cast and aged ZK60 alloy.

Fig. 9
figure 9

Variation of friction coefficient in different specimens: a BM specimen; b 1-PASS specimen; c 2-PASS specimen; d 3-PASS specimen

Full size image

4 Discussion

In previous studies, subsequent passes of FSP causing further grain refinement in the stir zone were widely observed in AZ31, AZ61, and AZ91 alloys [15, 17, 18, 42]. However, further refined grains cannot be obtained in multipass FSP of ZK60 alloy according to the experimental results in the current study. Processing parameters and tool geometries may have some effects on the grain refinement during FSP Mg alloys. Different DRX mechanisms between Mg–Al–Zn and Mg–Zn–Zr alloys during FSP mainly contribute to the interesting observation of grain size in ZK60 alloy. As is known, DRX proceeds by nucleation and nucleus growth. The nucleation of new grains in DRX occurs by bulging, subgrain rotation, and twinning for Mg alloys.

For Mg–Al–Zn alloys, Suhuddin et al. [2] deeply studied the DRX mechanisms and grain structure evolution of AZ31 alloy during FSW by a “stop-action technique”. It was found that twinning is activated in many original grains in the stir zone at the initial stage of FSW. Because of the geometrical requirements of strain, the original grains and twins tend to be geometrically reoriented in continuous deformation. The grain boundaries become wavy, and bulges appear along the grain boundaries, giving rise to fine equiaxed grains [2]. Moreover, twins tend to lose their original morphology and transform into irregular grains. Twin boundaries are converted into random high-angle grain boundaries. Finally, fine equiaxed grains are formed in stir zone after FSW/FSP [2]. Therefore, the DRX mechanisms and grain structure development for Mg–Al–Zn alloys have a close relationship with {10–12} twinning and limited discontinuous recrystallization.

For multipass FSP of Mg–Al–Zn alloys, the mechanisms of DRX and grain refinement in subsequent passes are similar with the first pass [16]. It should be stated that {10–12} twinning and bulging of grain boundaries would occur in subsequent passes. It would be more difficult than in the first pass because of the reduced grain size in the stir zone after the first pass of FSP. Therefore, subsequent passes can further refine the grains in the stir zone for Mg–Al–Zn alloys. The grain refinement effect would be significantly decreased compared with the first pass. The results reported by Luo et al. [15] confirm that single-pass FSP can refine the grains of cast AZ61 alloy from ~ 75 into ~ 12.5 μm, while it is further reduced to ~ 7.8 μm after multipass FSP.

To the authors’ knowledge, few studies have concentrated on the DRX mechanisms and grain structure evolution of ZK60 alloy during FSW/FSP. Mansoor et al. [11] suggested that basal slip, twinning and double twinning occur in FSW/FSP of ZK60 alloy. Galiyev et al. [43] found that the operating DRX mechanisms of ZK60 alloy are bulging of the original grain boundaries as well as subgrain growth at a nugget temperature of 573–723 K. Moreover, many precipitates exist in the matrix of Mg–Zn–Zr alloys, and these precipitates are much finer than Al12Mg17 particles in Mg–Al–Zn and Mg–Al–Mn alloys [23, 25]. Precipitates with a small size in ZK60 alloy have a pinning effect on the dislocation slip and grain boundary mobility [11, 44]. Concentrated strain in the vicinity of precipitates can increase the dislocation density and orientation gradient. Therefore, precipitates in ZK60 alloy can promote nucleation and hinder grain growth during recrystallization [45, 46]. This is the main reason for the significant grain refinement in the ZK60 alloy during the first pass of FSP compared with Mg–Al–Zn alloys.

Previous studies reported that precipitate particles in ZK60 alloy would be broken up and dispersed, with most of them being dissolved into the matrix, during FSW/FSP [25, 47]. No fewer than seven SEM images with high magnification were analyzed by software ImageJ to measure the volume ratio of particles in different ZK60 specimens roughly. The volume ratio of the particles was ~ 0.89% for the BM specimen, while it was ~ 0.33% for the 1-PASS specimen. The results indicate that significant dissolution of precipitates occurs during the first pass of FSP, which is consistent with previous studies [23, 25, 47]. It is measured that the volume ratio of particles in the 2-PASS specimen is ~ 0.35%, and that in the 3-PASS specimen is ~ 0.36%. The results further prove that there are no significant changes in precipitates among three types of FSP ZK60 specimen. Owing to the dissolution of precipitates after the first pass, the pinning effects of precipitates on the dislocation slip and grain boundary mobility would be significantly reduced during subsequent passes, which may promote grain growth during recrystallization. Therefore, the grains in ZK60 alloy are not further refined during subsequent passes, but an increase in the grain size is observed after multipass FSP. It is believed that the interesting observation of grain size evolution of ZK60 alloy has a close relationship with the effects of precipitates on the DRX process in multipass FSP.

Based on previous studies [4, 23], precipitates and grain size both have effects on the surface corrosion and wear resistance of Mg alloys. Precipitates have a more important effect on the surface properties compared with the grain size. The effect of galvanic corrosion between the precipitate and the matrix of α-Mg seriously reduced the corrosion resistance of Mg alloys [23]. Based on the experimental observation, compared with the first pass, almost no changes of precipitates in ZK60 plates were observed during the subsequent passes (see Fig. 2). As a result, the surface corrosion and wear resistance of the FSP ZK60 alloy was reduced by the subsequent passes. The declining number of surface properties was not particularly large. Finally, subsequent passes have little positive effect on the improvement in the surface corrosion and wear resistance for FSP ZK60 alloy.

5 Conclusions

  1. 1.

    The grains were significantly refined from ~ 68.2 μm in BM to 3.1–4.5 μm in multipass FSP of ZK60 plates. Subsequent passes increased the grain size in the stir zone, but they had almost no effect on the precipitate particles of FSP ZK60 alloy.

  2. 2.

    FSP improved the surface corrosion resistance of ZK60 alloy. However, subsequent passes were not beneficial, but rather they were harmful, to the corrosion resistance of FSP ZK60 alloy.

  3. 3.

    Subsequent passes of FSP resulted in a slight decrease of the hardness values of ZK60 plates. They had little positive effect on the improvement in the surface wear resistance of ZK60 alloy.