Keywords

Introduction

Presently, the automotive industry begins to implement magnesium sheets in series production. One reason for the limitations in plastic deformation of magnesium alloys is the crystallization of magnesium with a hexagonal close packed lattice structure. It is principally accepted that the comparably low formability of sheets is based on the resulting limitation in the number of active deformation mechanisms, at least in comparison to cubic metals. Such low formability [1, 2] of magnesium alloys also has a significant impact on the process ability during massive deformation. Magnesium alloys containing rare earth elements could partially compensate these disadvantages such as poor formability and corrosion properties [3].

Magnesium sheet materials with the addition of rare earths show weaker textures and finer-grained microstructures [4,5,6]. Furthermore, the addition of rare earths (RE) as alloying elements improves the ductility of these alloys. The alloy ME21 does not contain aluminum as an alloy element, which means there is no low-melting eutectic in contrast to the aluminum-containing Mg alloys. ME21 has the advantage of a good hot strength and creep resistance in comparison to other magnesium alloys [7]. The microstructure and texture development of magnesium during sheet rolling can be distinctly influenced by alloying magnesium with specific alloying elements. It has been shown that in the case of adding RE elements weaker textures result after rolling and annealing [8]. Hantzsche et al. also showed in binary alloys that the amount and kind of the rare earth elements has got an influence on the grain size and texture [9]. This circumstance is shown in Fig. 1.

Fig. 1
figure 1

Influence of different RE-elements and their amounts on grain size and basal texture in binary magnesium alloys [8]

Full size image

This work will report the sensitivity of the alloy ME21 to the amount and kind of the RE elements regarding microstructure and texture. The mechanical properties and the ductility are also investigated in tensile tests and Erichsen tests.

Experimental Procedure

In the first step of this work, 4 variations of the alloy ME21 were cast in cubic billets (120 mm × 70 mm × 270 mm) using a modified gravity casting process. The alloys ME20 and ME21 were alloyed with 0.5 and 1 wt% of Cerium based misch metal (Ce 50%, La 30%, Nd 16% and Pr 4%). In the other two alloys 0.5 and 1 wt% of pure Nd were added. The cast alloys and their alloy composition are listed in Table 1.

Table 1 Cast alloys and their alloy composition
Full size table

After casting the billets were heat treated to homogenize the microstructure and to reduce the amount of precipitates. For the rolling trials specimens of the dimensions 20 mm × 100 mm × 200 mm were machined and rolled down on the rolling mill at HZG, Danieli with roll dimensions Ø 360 mm × 500 mm, Fig. 2. The different alloys were rolled at a temperature of 450 °C and the rolling procedure consisted of two rolling schedules of 15 and 19 passes with different degrees of deformation between φ = 0.1 and φ = 0.2, leading to a final gauge of approximately 1.5 mm.

Fig. 2
figure 2

Rolling mill at HZG, Danieli Typ Ø 360 × 500

Full size image

φ is given as

$$ \phi = - \ln (h_{n + 1)}/h_{n}) $$

where n is the number of the pass and hn is the sample thickness after pass n. The rolling schedule is listed in Table 2.

Table 2 Parameters of the rolling trials
Full size table

In order to see how the different alloy compositions and rolling schedules influence the mechanical properties of the sheets, tensile tests according to ISO 6892 were performed. In order to homogenize the microstructure of the sheets, a heat treatment of 1 h at 450 °C was carried out after rolling. All samples were prepared in the sheet rolling and transverse direction after heat treatment. In order to investigate the deformation behavior under multi-axle load Erichsen tests (DIN EN ISO 20482/03) were performed.

After all processing steps, the microstructures and average grain sizes of the strips were analyzed and measured by using optical microscopy. Standard metallographic sample preparation techniques were employed and an etchant based on picric acid was used to reveal grains and grain boundaries [9]. Texture measurements were performed on the sheet mid-planes using a Panalytical X-ray diffractometer setup. The pole figures were measured up to a tilt angle of 70° which allowed recalculation of full pole figures based on a MTEX software routine [10]. The (0001) and (10-10) pole figures of the sheets in as rolled and heat treated condition, are used in this work to present the texture of the strips at midplane.

Casting Trials

Figures 3, 4, 5 and 6 display the microstructures of the cast billets in the as cast condition as well as after the heat treatment. Figure 3 shows the microstructures of the alloy ME20. The alloy ME20 exhibits a coarse grained dendritic microstructure (grain size app. 300 µm) with a very inhomogeneous grain size distribution. A small number of coarse Mn containing precipitates could also be observed located in the grains. Furthermore, finer RE containing precipitates are detected in the grains and at the grain boundaries. The heat treatment of 16 h at 450 °C reduces the number of precipitates at the grain boundaries, but increases the coarsens the grain structure. The microstructures of the alloys ME21 are displayed in Fig. 4. In comparison to the ME20 the microstructure of the ME21 (grain size app. 150 µm) is finer grained but also in a dendritic structure. There are significantly more precipitates in the ME21 than in the ME20. The composition of the precipitates is comparable to ME20, Mn and RE based. The RE containing precipitates in the ME21 are more located between the dendrite arms, but there are also coarse Mn containing precipitates in the grains. After the homogenization of 16 h at 450 °C there are no significant changes in the microstructure. However there are still coarse precipitates in the grains. Like in the ME20 the heat treatment leads to a significant grain growth, but compared to ME20 the grains in ME21 are much finer after heat treatment.

Fig. 3
figure 3

Microstructure of the alloy ME20 in as cast and heat treated condition

Full size image
Fig. 4
figure 4

Microstructure of the alloy ME21 in as cast and heat treated condition

Full size image
Fig. 5
figure 5

Microstructure of the alloy MN20 in as cast and heat treated condition

Full size image
Fig. 6
figure 6

Microstructure of the alloy MN21 in as cast and heat treated condition

Full size image

The microstructure of MN20 shows a dendritic grain structure in the as cast condition (Fig. 5). Generally, compared to the microstructure of the ME alloys the microstructure of MN20 is very coarse, partially the grains are larger than 2 mm. There is also a high amount of Mn and Nd containing precipitates in the microstructure with a homogeneous distribution in the grains. After the heat treatment at 450 °C for 16 h it could be seen that the precipitates are more concentrated in some grains (Fig. 5, right image). The grains are not significantly grown.

MN20 differs from the MN21 only by a small amount of additional Nd. The alloy MN21, in the as cast condition, shows a very coarse, dendritic microstructure (grain size larger than 2 mm) like the alloy MN20 (Fig. 6). A higher amount of Mn and Nd containing precipitates can be observed in the MN21 located in the grains. After the heat treatment at 450 °C for 16 h the dendritic microstructure disappeared and the precipitates are homogeneously distributed in the grains. Also a significant grain growth can be observed in the alloy MN21.

Rolling Trials

The heat treated slabs were rolled down to final gauge using the above mentioned rolling procedure (Table 2). All sheets were rolled at 450 °C successfully. In order to homogenize the microstructure, a heat treatment of 450 °C for 1 h was carried out after the rolling process. The microstructures of all alloys, which were produced according to the rolling schedule 1, are presented in Fig. 7 in the as rolled condition and after the heat treatment. The microstructures of all alloys of rolling schedule 1 are not fully recrystallized after the rolling process. Furthermore, in the alloys ME20 and ME21 some Mn containing precipitates are aligned in small lines, so called stringers. In the alloy ME21 a higher amount of these precipitates are observed. The alloys MN20 and MN21 show fine and homogeneous microstructure and the difference seen in the as cast microstructure of the alloys ME20 and ME21 disappear after rolling. After the heat treatment all alloys show a fully recrystallized microstructure, but the precipitates are still in the microstructure in the alloys ME20 and ME21 . The microstructures of all alloys rolled according to rolling schedule 2, in the as rolled and heat treated condition, are displayed in Fig. 8. The material behavior is similar to that seen in the material rolled by schedule 1. It can be clearly seen that the grains in all four alloys are more deformed in the as rolled condition. The higher deformation of the rolling schedule 2 leads to finer grained microstructure after the heat treatment. The smallest average grain sizes were archived in the microstructure of the sheets ME21 (11 ± 1.0 µm) and MN21 (11.3 ± 0.6 µm) which were produced by the rolling schedule 2. In general, the higher degree of deformation in the last rolling passes of rolling schedule 2 leads to a finer microstructure in the sheets after the heat treatment.

Fig. 7
figure 7

Microstructures of the all sheets of rolling schedule 1 in the as rolled (AR) and heat treated condition (450 °C/1 h), including the average grain size (average GS)

Full size image
Fig. 8
figure 8

Microstructures of the all sheets of rolling schedule 2 in the as rolled (AR) and heat treated condition (450 °C/1 h), including the average grain size (average GS)

Full size image

The basal and prismatic pole figures, after rolling and after heat treatment are presented in Figs. 9 and 10. The textures are best described by an alignment of basal planes close to the sheet plane but with a split peak component in the basal pole figure with tilt towards the rolling direction. The intensities of the basal pole distribution of the as-rolled sheets according to rolling schedule 2, vary between 7.4  and 8.1 m.r.d. Thus they are slightly higher compared to the intensities of the rolled sheets according to schedule 1, Imax between 6.2 and 7.2 m.r.d. This is consistent with the higher degree of deformation during rolling in rolling schedule 2. The pole figures of all four alloys after the heat treatment show distinctly lower intensities. For example, the basal pole figure intensity of alloy MN21 reduces from Imax = 7.9 m.r.d. to Imax = 4.5 m.r.d. Similar effects were observed in extruded ME21 profiles [11].

Fig. 9
figure 9

Textures of the all sheets of rolling schedule 1 in the as rolled (AR) and heat treated condition (450 °C/1 h)

Full size image
Fig. 10
figure 10

Textures of the all sheets of rolling schedule 2 in the as rolled (AR) and heat treated condition (450 °C/1 h)

Full size image

Tensile Tests

In order to see how the different alloying elements and rolling schedules influence the ductility of the sheets, tensile tests were performed. Tables 3 and 4 summarize the mechanical properties. The comparative stress-strain diagrams of all sheets are shown in Figs. 11 and 12. In general, it can be observed for all alloys that the tensile yield (TYS) and ultimate tensile stress (UTS) in transverse direction (TD) are slightly higher than in rolling direction (RD), but the elongation at fracture in TD are lower than in RD. The reason for the higher TYS and UTS is in the resulting textures, where the basal planes are more tilted in RD. In comparison to extruded ME21 sheet material the differences of the TYS and UTS are lower [12, 13]. The MN-based sheets exhibit higher elongations, especially in RD, compared to the ME-based alloys, corresponding to the absence of precipitates in the rolled and annealed condition. The differences in the mechanical properties of ME20 and ME21 are marginal, because the textures and the microstructures are quite similar.

Table 3 Results of the tensile tests of the sheets rolled according schedule 1
Full size table
Table 4 Results of the tensile tests of the sheets rolled according schedule 2
Full size table
Fig. 11
figure 11

Stress-strain curves of the ME- and MN-based alloys rolled according to schedule 1

Full size image
Fig. 12
figure 12

Stress-strain curves of the ME- and MN-based alloys rolled according to schedule 2

Full size image

The finer microstructure of ME21 after casting caused by the higher Cerium based misch metal amount does not result in better mechanical properties after rolling.

Quite the opposite can be observed, with a higher contain of Ce-misch-metal, the quantity of precipitates increases and leads to stringers which decrease the ductility. The MN21 in RD is significantly more ductile than the alloy MN20. It can be explained with the majority of the basal planes in the MN21 (25°) being more tilted out of the sheet plane than in MN20 (15°). This fact increases the Schmidt factor of MN21 in RD which results in a better ductility and is also responsible for the different TYS’s in RD and TD. The strength of all heat treated sheets rolled according to rolling schedule 2 is slightly higher compared to the sheets rolled according to rolling schedule 1 because of the finer microstructure. For all sheets after the heat treatment can also be observed that the tested rolling schedule does not influence the elongation significantly.

Conclusion

The results reveal that the alloy composition has got a significant influence on the resulting sheet properties. They are a first impression of resulting sheet properties of the investigated alloys. It is shown that the MN based sheets allow higher elongations compared to the ME based alloys. This corresponds with a higher solubility of Nd in the Mg–Mn-matrix compared to Ce. As a result, no precipitates form in the MN based sheet material. The high amount of Mn based precipitates in the ME based alloys leads to stringers in the microstructure what reduces the ductility significantly.

Furthermore, the results show that in ME-based alloys a Ce-misch-metal content higher than 1 w% does not result in better mechanical properties in comparison to the ME20 with less amounts of Cerium based misch metal, because of the higher content of precipitates and stringers in the microstructure. Aside from this results of the ME-based alloys the results of the MN-based demonstrate that in these alloys a Nd content of 1 w% results in significant better elongation at fracture in RD in comparison to the MN20 with less than 1 w% Nd. The texture measurements make clear that a heat treatment after rolling decreases the intensities of the textures substantially. Finally can be observed that a higher degree of deformation during the rolling process and a heat treatment after rolling lead in all alloys to higher strength without a decreasing ductility.