Basic AZ31 alloy FSP trials

period and rotating rate, the tool rotates less cycles in a fixed distance when the processing speed increases. Therefore, ring spacing increases and the semicircular feature becomes rougher.

Figure 3-4 shows the cross-sectional views of the FSPed AZ31 alloy. It can be defined as the nugget zone, thermomechanically affected zone, and base metal. The feature of onion ring is contained in the nugget zone. The size of nugget zone is approximate equal to the size of pin. The pin tool with shoulder diameter, pin diameter, and pin length of 18, 6 and 6 mm, respectively, is used in present stage. The nugget zone is also called the dynamic recrystallized zone which experiences dynamic recrystallization. When the tool rotation, the material was stirred and forming a plastic material flow around the rotating pin and material is repeated in every rotation. Such material flow would lead to repetitive introduction of strain and strain rate into material, which probably produces an alternation of three-dimensional ellipsoidal surfaces, as shown in Fig. 1-10 (a). The alternation of the ellipsoids generates an onion ring structure in the nugget zone on the cross section perpendicular to the processing direction, as shown in Fig. 3-4. It also shows a region, named as TMAZ, which contains distorted grains between the nugget zone and base matrix in Figs.

3-4(c) and (d). The observed TMAZ region in Mg alloys is not as apparent as that in Al alloys.

Although the TMAZ underwent plastic deformation, recrystallization did not occur in this zone due to insufficient deformation strain. Typically, the recrystallization temperature for Mg alloys is around 0.5-0.7 Tm (or around 250-400 ^oC) which is lower than 0.6-0.8 Tm or around 300-450^oC for Al alloys. Therefore, Mg alloys can experience more complete dynamic recovery and dynamic recrystallization easily than Al alloys.

3.1.2 The microstructure of the modified AZ31 alloy made by FSP

The as-received AZ31 alloys contained coarse grains around 75 μm, as shown in Fig.

2-1. AZ31 alloy can be considered as solid-solution alloy without precipitates. Microstructure characterization in this study was focused on the dynamically recrystallized nugget zone. The as-received AZ31 alloy with a coarser grain size turned to posses equiaxed and smaller grains in the nugget zone (i.e. dynamically recrystallized zone) after FSP. Figs. 3-5 and 3-6 show the typical grain structures of AZ31 alloy specimens after one-pass FSP. Compared to the as-received AZ31 alloy, the grain sizes of FSP AZ31 alloy specimens are obvious smaller and can be refined drastic from 75 μm to less than 8 μm under all conditions in this study, as shown in Table 3-1. The grain shape observed from different cross-sectional planes is consistently of the equiaxed fully recrystallized type. The grain sizes near the bottom region of the dynamically recrystallized zone are typically smaller. For example, the grain sizes in top, middle and bottom regions of the FSP AZ31 alloy specimens at 800 and 1400 rpm are 3.7/3.6/2.0 μm and 4.7/3.8/2.6 μm, respectively. The larger grain sizes in the top regime are due to the more severe temperature rise caused by the friction heat generated from the tool shoulder. The recrystallized grains would gradually increase with increasing rotational rpm speed as shown in Figs. 3-5 and 3-7. The higher rotation rate causes severer rotation actions which could produce more heat energy. More generated heat brings a larger driving force for grain growth and results in larger grains. With regard to the influence of the advancing speed, the lower advancing speed would result in inputting more heat energy. Figures 3-6 and 3-8 show the variation of recrystallized grain size in the nugget zone for different advancing speeds. It could be seen that the parameter of 45 mm/min advancing speed result in the biggest grain than all others under the same rotational speed of 800 rpm. On the other hand, the parameter of 800 mm/min advancing speed leads to the finest grains. From the experimental results, it is suggested that heat input increases with decreasing advancing speed and/or increasing rotational speed. The growth of the nucleated grains would be affected easily by the heat generated during the process.

3.1.3 The temperature of the stirred zone of modified alloys

The typical temperature profiles measured by the inserted thermocouple into the AZ31 alloys are shown in the Fig. 3-9. The temperatures measured during FSP varied from 250^oC to 450^oC, depending on the FSP rotation speed. Higher rotation rate can generate more heat energy during FSP and result in larger residual grains in the nugget zone. The temperature rise duration is about 150 s. Such temperatures are compatible to those experienced during our previous hot extrusion research [24,52]. And the resulting grain sizes can also be compared. In addition, the temperature profiles also varied with advancing speed, as shown in Fig. 3-10. With a higher advancing speed, the measured peak temperature is lower and temperature rise duration is shorter. This yields the lower and shorter heat history during FSP and results in finer grain sizes. The summary of temperature measurements with different parameters is listed in Table 3-2. From the temperature measurements, it can be shown that the rotation rate and advancing speed are important parameters which mainly affect the heat history during FSP and final residual microstructure in the nugget zone. These temperature results are also consistent with the grain size observations in OM, and the measured temperature ranges in this study are also consistent with the typical recrystallization temperatures which are around 0.5-0.7 Tm (or around 250-400^oC) for Mg alloys.

3.1.4 Hardness measurements

The typical microhardness readings, Hv, in the central cross-sectional zones of the FSP specimens are depicted in Fig. 3-11. For the AZ31 alloy, after FSP, the hardness could increase from ~50 Hv up to ~90 Hv due to grain refinement via dynamic recrystallizaiton.

This hardness result is also consistent with the grain size observation in OM. After FSP, the

obvious increments in hardness of the nugget zone in AZ31 alloys were seen. It is unlike some precipitation hardening aluminum alloys which sometimes show drastic decreasing hardness values in the nugget zone after FSP. This is because that the precipitates of precipitation hardening alloys could dissolve into the matrix or form coarser precipitates during the thermal cycles of FSP, causing the decrement in hardness in the nugget zone. The typical microhardness readings, Hv, in the various zones of the FSP specimens are depicted in Fig. 3-12. Based on systematic hardness measurements of the base and FSP specimens, the Hall-Petch relationship is well followed (d, in μm), i.e.,

Hv = 40 + 72 d^-1/2. (4)

The average hardness values are also summarized in Table 3-3.

3.1.5 Grain orientations

With strong dynamic recrystallization occurred during FSP in the current AZ31 Mg, the equiaxed fine grains exhibit much lower texture intensities as compared with the as-extruded AZ31 specimens [24,170,171]. Figure 3-13(a) shows the computer simulated X-ray diffraction for completely random Mg. The (10 1 0) and (0002) peak heights are about half of the (1011) peak height, all present within the 2θ of 30-40^o. The XRD pattern for the transverse cross-sectional plane of the as-received billet is shown in the Fig. 3-13(b). The as-received billet basically exhibits nearly random grain orientation. The XRD diffraction patterns for the transverse cross-sectional plane of the as-received billet and its representative FSP specimens are shown in Fig. 3-14. The as-received billet exhibits basically nearly random grain orientation, reflecting its fully recrystallized coarse grained structure. As FSPed at a lower rotation speed of 800 rpm, the (0002) plane tends to lie on the transverse plane of

the FSP specimen (perpendicular to the pin travel direction and parallel to the onion back surface, as also observed by Park et al. [91]). With increasing rotation speed to 1400 or 1800 rpm, the (1011) peak height increases slightly, but never approaching to the random case.

3.1.6 Brief conclusions of basic AZ31 alloy FSP trials

The relationship between processing parameters and the resulting grain size and temperature and mechanical properties for FSP in AZ31 Mg is systemically examined. In the views of grain refinement and hardness enhancement, FSP is a useful process to refine grain size in Mg alloy and fine grain size also results in obvious increment of hardness value which follows well to the Hall-Petch relationship. Based on experimental results, the most important parameters are rotation rate and advancing speed, which are also related to the amount of generated heat energy. Therefore, in order to get finer grain size with increment of hardness, lower rotation rate and higher advancing speed are both considered to apply during the FSP.

In the above experiments, the finest recrystallized grain size can be refined from 75 μm to ~2 μm which is about 3% of the original billet grain size, and the hardness can be enhanced up to ~90 Hv which is 1.8 times of original base material (~50 Hv).

However, the means of only changing parameters in this step are not capable of producing extremely fine nano-scale microstructure with high hardness. The grain size can only be refined to micro-scale by the present step. Therefore, for the sake of achieving the goals of nano-scale grain size and triple of the hardness, more efforts and other specific approaches are needed. There are several improved ways which are considered to be carried on in the next steps. The flow chart is also shown in Fig. 2-1. The following steps are mainly classified into two ways. One is thought to reduce the generated heat or to accelerate heat conduction rate from the specimens. Another is trying to add reinforcements. With the help of

additional reinforcements combined with the base alloy, the hardness value can be further improved and the grain size can also be refined.