Abstract
The mechanical behaviour of AZ31 magnesium alloy sheets were evaluated at different directions tilting 0, 45 and 90° to rolling direction. A suitable critical point of pre-stretching has been investigated through the work hardening curve of the magnesium alloy sheets. The microstructure and mechanical properties were examined and compared. The results indicated that the average yield strength of the 0, 45 and 90° pre-hardened 4 specimens followed by annealing decreased 30 MPa and the average uniform elongation increased up to 38 compared with the as received sheet. Pre-hardening process is suggested as a potential technique to improve the ductility and anisotropy of magnesium alloy sheets.
Introduction
Magnesium alloy is considered as one of the most potential structural materials in the automotive and aircraft industries due to its series of advantages such as low density, high specific strength and excellent camping capacity. 1 Extrusion and rolling process are two of the most universal ways to obtain the semi-finished or finished industrial magnesium alloy sheets. However, it is commonly recognised that magnesium alloy sheets usually show poor stretch formability at room temperature because of its hexagonal–close-packed (hcp) crystalline structure, which extremely limits their applications.2,3
A lot of works have been done to improve the mechanical properties by randomising the texture of magnesium alloy sheets. Recently, Yang et al.4,5 suggested that the asymmetric extrusion process can effectively weaken the basal texture through asymmetric shear deformation during the extrusion process. Many other technologies, such as cross-roll rolling 6 and differential speed rolling, 7 which are more sophisticated, were also applied in the rolling process. Nevertheless, the above mentioned processes were not totally suitable for thin sheets, especially for which thickness is less than 3 mm, thus the formability of AZ31 sheets at room temperature still needed to be further improved. 8 Chino et al. 9 reported that microstructure played an important role during deformation and AZ31 alloy with larger grain size demonstrated higher stretch formability. Latterly, Zhang et al. 10 proposed that AZ31 sheet, which was pre-stretched up to the critical degree of deformation and annealed subsequently, exhibited highest Erichsen value (IE = 5·3). This was attributed to the formation of twins was easier in coarse grains during deformation and the deformation twins coordinated with the thickness strain. 9
In this present work, a new method of optimising pre-deformation process is put forward through analysing the work hardening behaviour of the magnesium alloy sheet, resulting in strong pointedness and good effect. Furthermore, different directions tilting of 0, 45 and 90° to rolling direction (RD) of specimens were analysed respectively to verify the reliability of the critical point of pre-hardening.
Experimental procedure
The as received hot rolling sheet of AZ31 (Mg–2·9Al–0·85Zn–0·3Mn, in wt-) with 1·12 mm in thickness, was cut into 100×46 mm (length×width) along the RD for pre-hardening (PRH) experiments. The specimens were pre-stretched by 2, 4 and 8 along RD, and then annealed at 300°C for 1 h respectively. It could be defined the samples with pre-hardening followed annealing as PHA for simplicity. PRH process was carried out on a CMT 6305-300KN testing machine with a speed of 1·5 mm min−1 at room temperature.
Dog bone tensile samples of 12 mm gauge length, 6 mm width and 1 mm thickness were machined from the as received sheets and PRH specimens with various directions tilting of 0, 45 and 90° to the RD respectively. The strain rate of uniaxial tensile test was 1·5 mm min−1. The microstructures were examined by optical microscopy (OM). The X-ray texture analysis of (0002) pole figure was carried out by Rigaku D/Max 2500. The electron backscatter diffraction data was collected using a diffractometer (EBSD; HKL Channel 5 System equipped with FEI Nova 400 FEG-SEM, Cambridge, MA).
Results and discussion
True stress–true strain curve of the as received sheet is examined in Fig. 1a. It can be seen the ductility behave a remarkably anisotropic and the uniform elongation of the 45° sample is relatively poor. Figure 1b shows the work hardening rate curve of three directions which is drew out through the above tensile test data of the as received AZ31 sheet. It can be seen that the work hardening rate is quite high after the yield stage and declines rapidly till the work hardening rate is going to be relatively stable, which consistent with the called ‘extended Voce model’. Point a, b and c are represent as directions tilting of 0, 45 and 90° respectively, can be defined as the critical point of low work hardening rate in the tensile test and their corresponding true strain are 3·69, 3·55 and 3·82. In order to study the effect of PRH values on mechanical properties, 2 (line A instead in Fig. 1), 4 (B) and 8 (C) as the value of PRH are chosen, based on the work hardening rate curve in Fig. 1b and considered of rapid linear decrease of the strain hardening rate stage (stage I, line A represents high work hardening rate), the critical point of low work hardening rate and the linear hardening change stage (stage II, line C represents far beyond the critical point). The related conclusions are presented.
a true stress–strain curves obtained from uniaxial tension tests and b Work hardening behaviour of as received AZ31 alloy under different tensile directions
Optical micrographs of the as-received and PHA specimens are revealed in Fig. 2. The as received specimen demonstrates fine equiaxed grains of ∼13 μm. There is no tensile twin appears in the optical micrographs after pre-stretched to 2, 4 and 8. However, great changes occur in the microstructure characteristics after annealed at 300°C for 1 h. For the PHA2 specimen, the average grain size increases to 58 μm. Besides, it is interesting that several fine grains are surrounded by major coarse grains and its size is almost identical with the fine previous equiaxed grains of the as-received AZ31sheet. As is shown in the EBSD map in Fig. 3b, the majority of the fine grains commonly have a strong basal texture orientation, such as the grain A and B. This suggests that the grains with weaker basal texture orientation may have priority to participate in plastic deformation, and few grains like A and B or others which did not participate in the plastic deformation may not become nucleation sites in the subsequent annealing process during the process of small tensile deformation. The phenomenon may in accordance with the previous study which weak texture favours the slip of dislocation. 11 Furthermore, the PHA4 sheet exhibits more uniform grains with reduced grain size of ∼37 μm, which is attributed to the rapid increasing recrystallisation nucleation sites after larger stretch deformation. Besides, the average grain size decreases to 25 μm when the PRH value up to 8. The evolution rule of metallographic organization at the 45 and 90° are consistent.
a as received alloy; b PHA2; c PHA4; d PHA8
a as received; b PHA2; c PHA4; d PHA8
The (0002) pole figures and EBSD orientation maps of the as received and PHA specimens are shown in Fig. 3. It can be seen that the basal texture of the as received sheet is divergent and its intensity value is only 7·7, since magnesium alloy usually possess a hcp crystalline structure, the basal slip dominates in the low strain deformation at room temperature.12,13 Previous studies suggested that it was difficult to activate the compression twin mechanism during tensile deformation at room temperature although the tensile load perpendicular to the c-axis of grains, which attributed to its high CRSS.14,15 This is the main reason why the twin deformation mechanism did not been activated during the process. The PHA4 sample, pre-stretching just beyond the critical point of high work hardening rate, exhibits more uniform grains and the lowest basal texture intensity, so it can be predicated that the combination of 4 pre-stretching and annealing at 300°C for 1 h could farthest diminish the work hardening effect of the as received AZ31 sheet. When the PRH value exceeds 4 to 8, the basal texture intensity rises to 6·4.
Table 1 shows the yield strength (YS), ultimate tensile strength (UTS), uniform elongation (Eu), strain hardening exponent value (n-value) and Lankford value (r-value) of the as received and PHA samples at three different directions. First, it can be seen that the YS decrease sharply after PHA process and have an increasing tendency with the rising of PRH value, which is related to the change rule of recrystallisation grain and the texture. Second, the UTS, however, demonstrate different evolution rule. For the 0° PHA sample, the ultimate tensile strength enhanced obviously after PHA, and the PHA4 sample shows a highest UTS value of 328·1 MPa. Besides, when the deformation exceeds the critical point, the UTS stop to add up. As for the 90° PHA sample, the UTS reduce evidently for the recrystallisation softening effect, while the PHA4 still remains the highest level compared with the PHA2 and PHA8 sample. Third, in addition, the Eu of the PHA samples at 0 and 45° improved effectively compared with the as received specimens. Meanwhile, the PHA4 sheets exhibit the highest ductility. It has been reported that the uniform elongation of magnesium alloy was strongly affected by hardening rate at room temperature. 16 However, in this study, little has changed in the n-value. Mohri et al. 17 suggested that the crack propagation at grain boundaries was easier activated with increasing grain size, which resulting a poor stretch formability. However, the PHA4 sample shows larger grain size but much higher ductility compared with the as received sheet. It should be noted that the weaker texture intensity of the PHA4 sample may result in a larger Schmid factor of <a> basal slip, which favours the improved formability as well as the uniform elongation. 18 In addition, the most attention is the reduction of the r-value, which indicates that the normal direction (ND) is relatively easier to deformation in tensile test. It may play a positive role in ductility in further deformation. However, the factors that affect uniform elongation are complex and required further investigation.
Yield strength (YS), ultimate tensile strength (UTS), uniform elongation (Eu), n-value and r-value of as received and PHA specimens at different directions tilting of 0, 45 and 90° to RD
Conclusions
The pre-hardened 4 and annealed samples represented best comprehensive mechanical properties. The average yield strength of the 0, 45 and 90° pre-hardened 4 specimens followed by annealing decreased 30 MPa and the average uniform elongation increased up to 38 compared with the as received sheet. Moreover, the ultimate tensile strength still remained a relatively high level. The stretch formability of the as rolled AZ31 sheet was improved due to the acquisition of the lowest Lankford value. It was proved that the detection of the critical point of pre-hardening was an efficient approach to improve the formability of AZ31 magnesium alloy sheets.
Acknowledgements
The authors are grateful for the financial supports from National Natural Science Foundation of China (51171212), Chongqing Science and Technology Commission (CSTC2012JJJQ50001, CSTC2013jcyjC60001, cstc2012ggB50003), and The National Science and Technology Program of China (2013DFA71070), and the Fundamental Research Funds for the Central Universities (CDJZR13138801).