Effect of Y content and equal channel angular pressing on the microstructure, texture and mechanical property of extruded Mg-Y alloys

W.Yng, G.F.Qun, B.Ji, Y.F.Wn, H.Zhou, J.Zheng, D.D.Yin,?

a Key Laboratory of Advanced Technologies of Materials, Ministry of Education, School of Materials Science and Engineering, Southwest Jiaotong University, Chengdu, Sichuan 610031, China

b Nano and Heterogeneous Materials Center, School of Materials Science and Engineering, Nanjing University of Science and Technology, Nanjing 210094,China

c International Joint Laboratory for Light Alloys (Ministry of Education), College of Materials Science and Engineering, Shenyang National Laboratory for Materials Science, Chongqing University, Chongqing 400044, China

Abstract The microstructure, texture and mechanical property evolution of the extruded Mg-xY (x=1, 5wt.%) alloys during equal channel angular pressing (ECAP) were systematically investigated using an optical microscope, electron backscatter diffraction (EBSD) and uniaxial tensile test.The Mg-Y alloys exhibited a weakened basal texture before the ECAP, and the texture was further weakened with the max basal poles dispersed along ～45° between the extrusion direction and the transverse direction after the ECAP.The Mg-5Y alloys always exhibited a finer grain size comparing to that of Mg-1Y for the same ECAP process.With a proper ECAP process, both the strength and elongation of Mg-5Y alloy could be improved simultaneously after the ECAP, i.e., the yield strength (273.9±1.2MPa), ultimate strength (306.4±3.0MPa),and elongation (23.9±1.0%) were increased by 10%, 6%, and 72%, respectively, comparing to that before the ECAP.This was considered to be arose from the combined effects of grain refinement, significant improved microstructure homogeneity and solid solution hardening.In addition, it was found that Mg-Y alloy with better comprehensive properties could be obtained by the decreasing-temperature ECAP processes.The yield strength-grain size relationship could be well described by the Hall-Petch relation for all the ECAPed Mg-Y alloys,which was consistent with that the texture changes did not significantly affect the average Schmid factors of basal, prismatic and pyramidal slips for both Mg-Y alloys.

Keywords: Mg-RE alloy; Equal channel angular pressing; Texture; Mechanical property.

1.Introduction

Magnesium (Mg) alloys are widely used as lightweight materials due to the high strength/weight ratio, low density and good machinability [1-5].Mg alloys are typical hexagonal close-packed (HCP) crystal structure with limited slip systems and poor ductility at room temperature (RT) [6,7].The limited ductility and the relatively low strength hinder further applications of Mg alloys.Therefore, it is urgent to improve the strength and ductility of Mg alloys.

Severe plastic deformation (SPD) is considered as an effective technology to refine grains and improve ductility[8,9].Equal channel angular pressing (ECAP), as one of the SPD technologies, can accumulate large strains through repeated pressing without changing the size of the material, and consequently refine grains and improve ductility effectively[10-12].Some reports indicated that the ductility and strength of as-cast Mg alloys could be improved significantly after ECAP [13].However, when the initial billets for the ECAP were conventional wrought Mg alloys, which was the most common case, the increase of strength along extruded/rolled direction was limited or even decreased due to the initial strong basal texture.For instance, Kim et al.[11]reportedthat the elongation of AZ61 alloy could be significantly increased from 32% to 55%, while the yield strength decreased after 8 passes ECAP at 275°C.Xia et al.[14]also found that the superior ductility of AZ31 alloy could be achieved after ECAP due to the significant grain refinement.However, the strength decreased because of the effects of texture development.

It was well known that the effect of texture on mechanical properties of Mg alloys was essential [15].Besides, the original strong basal texture could be weakened after ECAP.Therefore, the effect of texture softening on mechanical properties should be paid more attention.The AZ system Mg alloys usually exhibit a typical strong basal texture after extrusion or rolling.For the extruded AZ61 alloy, Kim et al.[16]found that the initial basal texture was dispersed by ECAP.After 8 passes ECAP, the yield stress was reduced,indicating that the effect of texture softening was significant,which even exceeded the strengthening effect due to grain refinement.Masoudpanah et al.[17]also found similar phenomena in AZ31 alloy after ECAP.Therefore, texture softening was considered to be one of the main reasons for the limited increase or even a decrease of strength for Mg alloys after ECAP.

As is well known, the addition of rare earth elements to Mg, such as Yttrium (Y), Lanthanum (La), Cerium (Ce),Gadolinium (Gd), etc., can significantly modify the texture and produce the “rare earth” texture [18,19].The Y element was considered to be a potential benefit alloying element because it can weaken the basal texture, and improve strength and ductility of Mg alloys.Sandl?bes et al.[20]reported that the maximum intensity of the texture decreased from 17.16 multiples of random distribution (mrd) to 6.01 mrd after adding 3wt.% Y to Mg.Stanford et al.[21]found that basal texture presented a more dispersive and random distribution by adding more Y elements of 5wt.%.In our previous study [22], the as-extruded Mg-Y alloy exhibited a typical rare-earth texture, and both the pole distribution and maximum intensity of the basal pole were weakened significantly during the annealing process.Lu et al.[23]reported that the as-extruded Mg-1Y sheet exhibited a tensile elongation of 30.2%.Besides, they [23]found that the deformation behavior in extrusion direction (ED) and transverse direction(TD) of the Mg-5Y sheet were nearly isotropic at 300°C.Long et al.[24]found that Y alloying to Mg could significantly alter the deformation mechanisms, i.e., promote pyramidal＜c+a＞slip and depress twinning, during room temperature compression.Consistent results were also reported by Sandl?bes et al.[20]and Zhou et al.[25].Therefore, it is of great interest to investigate the microstructure, texture and mechanical properties evolutions of Mg-RE alloys, which initially exhibited a weaken basal texture,during ECAP.However, this kind of study is rather limited.

In this paper, the effect of Y content and ECAP processes on the microstructure, texture evolution and mechanical properties of the extruded Mg-Y alloys were systematically investigated.The correlations between microstructure and mechanical properties were discussed with special attention placed on the simultaneously improved strength and elongation with the proper ECAP process and the Y content.The present work can provide support for understanding and further application of high-performance Mg-RE based alloys by ECAP process.

Fig.1.Schematic diagram of the equal channel angular pressing (ECAP)mold.

2.Experimental materials and procedures

The Mg-1Y alloy and Mg-5Y alloy (wt.%, if there is no specific statement, compositions of the alloys in this paper are in weight percent) used in this paper were melted by high purity Mg(99.99%)and Mg-30Y(wt.%)master alloy.In the smelting process, the well resistance furnace was always filled with CO2and SF6.It was worth noting that the volume ratio was kept at 100:1.The as-cast ingots were cast into the cylinder-shaped mold.The as-cast ingots were homogenized at 530 °C for 8h by air cooling.For the next extrusion, the ingots were machined into billets.After preheated at 300 °C for 2h, the billets were extruded into bars with a diameter of 30mm.The extrusion ratio was 9 and the extrusion speed was 0.4 m·min?1.

The rectangular samples for ECAP were made from the center of the extrusion bars, and the geometry size are 20mm×20mm×70mm.In Fig.1, the curvature angle (ψ)and the internal angle(Φ)of the ECAP mold are 37°and 90°,respectively.After one pass of ECAP, the accumulated strain was about 1 [26].All the rectangular samples were processed by the BCroute, which was considered as the most effective route to obtain high angle grain boundaries and refine grains[27].The ECAP experiment conditions are listed in Table 1.Two ECAP methods were designed: the first one was carried out at constant-temperature for four passes, in which the processing temperature was 300 °C, 350 °C and 400 °C, respectively.It was known that the Mg alloys could be successfully ECAPed without any observable cracks, and there were obvious dynamic recrystallization (DRX) and grain growth at high processing temperature, which resulting in low finegrain strengthening.The grain growth would be suppressed at a low processing temperature, which helped improve the mechanical properties of the Mg alloys.Therefore, the other one was to reduce the processing temperature step by step.The decreasing-temperature ECAP process Ⅰwas conducted at 400 °C for the first pass, 350 °C for the second pass, and 300 °C for the last two passes.The decreasing-temperature ECAP process Ⅱwas conducted at 350 °C for the first pass and the second pass, and 300 °C for the last two passes.Before pressing, the rectangular samples needed to be preheated for 10min, and both the samples and the channel were lubricated by MoS2.The pressing speed (～20 mm·min?1) was controlled and kept constant during the experiments.The samples needed to be quenched in water immediately after ECAP.

Table 1ECAP experiments conditions for ECAP tests in the current study.

The microstructural observation and crystallographic orientation characterization were studied by optical microscope(OM, Zeiss Axio Lab A1) and electron backscatter diffraction(EBSD, Oxford Instrument Nordlys Nano).The observation plane for OM and EBSD of all the samples was the EDTD plane, as shown in Fig.1.The data processing of EBSD used a homemade MATLAB code based on the open-source toolbox MTEX [28].The average grain size was estimated by the linear intercept method (D=1.74×L, where L is the linear intercept grain size [22]).The volume fraction of the second phase was determined using a quantitative metallographic method by image analysis software (Image-Pro Plus 6.0) based on at least three optical micrographs.

Tensile tests were carried out by the MTS-CMT5105 universal testing machine at RT.The dog bone tensile sample was cut from the center of as-extruded and ECAPed Mg-Y alloys with gage dimensions of 2.5mm(width)×2mm(thickness)×10mm (length).Both the uniaxial tensile axis and the length of the sample were parallel to the ED, while the width of the sample was parallel to the TD (Fig.1).The strain rate was 1×10?3S?1.For repeatability and accuracy, each test condition was required to be repeated at least three times.

3.Results and discussion

3.1.Microstructural evolution

Fig.2 illustrates the optical micrographs of the Mg-1Y alloys before and after the constant-temperature ECAP for different temperatures and passes in the ED-TD plane.The optical micrographs are placed by row and column grid.The row represents the micrographs with different ECAP passes at the same temperature.From the bottom to the top, the temperature is 300 °C, 350 °C and 400 °C, respectively.While the column represents the micrographs with the same ECAP pass.The first column indicates the optical microstructures of the as-extruded Mg-1Y alloy, which exhibited abundant fine grains and many coarse grains, simultaneously.The coarse grains were elongated along with the ED.The average grain size of the as-extruded Mg-1Y alloy was 11.7±2.1μm,with a negligible volume fraction of second phase of about 3.8±0.3%.The size of the second phase was in the range of 0.2-1.4μm, and its average diameter was 0.6±0.4μm.At 300 °C, the decrease of grain size was not obvious in the first pass, while some coarse grains were substituted by fine grains after 2 passes, and the average grain size was 7.4±0.7μm.It was important to note that the sample exhibited obvious cracks during the third passes,as indicated by the arrows.At 350°C,there were still plenty of coarse grains and an undesirable grain size growth after 1 pass (11.7±2.1μm to 14.3±3.4μm).After 2 passes, the average grain size decreased, and the microstructure was got homogenized due to dynamic recrystallization (DRX) [29].However, the average grain size of the sample rose again after 3 passes.The similar grain size growth phenomenon was also found at 400 °C-1P(11.7±2.1μm to 23.5±1.6μm) with a more homogeneous microstructure.It was found that the grain grew quickly, and the grain size increased by 101%.So, the subsequent ECAP experiment at 400 °C was not continued.Our previous study[22]found that grain growth was temperature-sensitive.It was easily found that the microstructure became more homogeneous at a higher temperature with the same ECAP pass.Yoshida et al.[30]and Khani et al.[13]found the same phenomenon in ECAPed AZ31 and AZ91 alloys, respectively.

Fig.3 illustrates optical micrographs of the Mg-5Y alloys before and after the constant-temperature ECAP for different temperatures and passes.The average grain size of the asextruded Mg-5Y alloy was 9.1±1.1μm, which was smaller than that of Mg-1Y alloy.There were more coarse grains and the second phase (volume fraction was 6.1±1.8%) than as-extruded Mg-1Y alloy.The size of the second phase was in the range of 0.4-4.2μm, and its average diameter was 1.5±0.7μm.Fig.4 (a, b) illustrates the TEM bright field image and the corresponding EDS results of the second phase(area A)in the Mg-5Y sample after 300°C-1 pass ECAP.The TEM bright field image clearly showed it was cubic-shaped,and the EDS results showed this cubic-shaped particle was rich in Y element.Based on the above results and our previous studies [22, 23], the cubic-shaped particle could be concluded to be YH2phase with FCC crystal structure which was commonly observed in Mg-RE alloys.Fig.4 (c, d) illustrates the TEM bright field image and the corresponding selected area electron diffraction for the marked area (area B) in the Mg-5Y sample after 300 °C-1 pass ECAP.In Fig.4(d), there were only diffraction pattern ofα-Mg matrix without any extra diffraction points, indicated that there were no precipitates formed during the low temperature ECAP.This is expected as the maximum Y solute ability in Mg at 300 °C is approximate 5wt.%, and the solute ability increases significantly with increasing temperature [31].The direction of the incident electron beam was [110].At 300 °C, it was foundthat the average grain size decreased with further passes.The average grain size was too small to be estimated after 4 passes.Therefore, the average grain size was measured by the EBSD method instead of the linear intercept method as 1.24μm.The coarse grains were broken by mechanical shear and separated by obvious shear bands after 1 pass, which indicated the inhomogeneous microstructure.Besides, the homogeneity of the microstructure was still not improved after 4 passes.In 4 passes, long strip shape shear bands could still be observed.The microstructure evolution after ECAP at350 °C was similar to that at 300°C, and the smallest average grain size was obtained at 3 passes, which was 4.9±1.0μm.However, after ECAP at 400 °C, the grains were not refined.It was noteworthy that the homogeneity of the microstructure was greatly improved after the second passes, which was completely different from that at 300 °C and 350 °C.So, it was clearly found that the microstructure became more homogeneous at higher temperature with the same ECAP pass,which was also found in Mg-1Y alloy in Fig.2.

Fig.2.Optical micrographs of the Mg-1Y alloys before and after ECAP for different temperatures and passes.Only the 300 °C-3P sample exhibited obvious cracks, as indicated by the arrows.

Fig.3.Optical micrographs of the Mg-5Y alloys before and after ECAP for different temperatures and passes.All the samples presented here were successfully ECAPed without any observable cracks.

Fig.4.(a) TEM bright field image and (b) the corresponding EDS results of the cubic-shaped particle in the Mg-5Y sample after 300 °C-1 P ECAP; (c)TEM bright field image and (d) the corresponding selected area electron diffraction for the marked area in the Mg-5Y sample after 300 °C-1 P ECAP.No extra diffraction points except those for α-Mg matrix indicated that there were no precipitates formed during the low temperature ECAP.The direction of incident electron beam was [110].

Based on the results mentioned above and our previous studies [22,32], it was believed that DRX and grain growth could be promoted at 350 °C and 400 °C.However, no homogeneous microstructure was observed at any pass of Mg-5Y alloy after ECAP at 350 °C.It was speculated that the Y atoms might segregate to the grain boundaries [33],and it could suppress the grain boundary diffusion and DRX.The cubic-shaped particles in Mg-Y alloys contributed to the grain refinement during DRX by particle stimulated nucleation (PSN).Thermomechanical processing of the samples results in the formation of deformation zones around the precipitates.During the recrystallization treatment, nucleation of recrystallization occurs in the highly strained regions of the deformation zones, which called as PSN [34].PSN contributes to grain refinement by promoting nucleation of new grains.The role of large particles (＞1μm) in promoting recrystallization through PSN mechanism is well known, and it is also often inferred to have occurred in magnesium alloys[35].Xu et al.[36]reported that DRX only occurred near grain boundaries with discontinuousβ-Mg17Al12precipitates pinning at newly DRXed grain boundaries, when compressed at 300°C and 350°C.Kwak et al.[37]found that DRX grains formed firstly around the icosahedral phase, which indicated that icosahedral phase provided the preferred nucleation sites for DRXed grains through PSN.Robson et al.[35]found that the number of particles in the range 1-10μm diameter was increased by adding Mn element in Mg-Mn alloys,which provided additional PSN sites.In addition, the cubicshaped particles in this paper could have acted as pinning obstacles against the grain growth.Minárik et al.[38]reported that the distribution of small particles could effectively pin the grain boundaries, and significantly reduce the grain growth.Therefore, with the increase of Y content, the homogeneity of the microstructure did not improve significantly and grain size became smaller at the same ECAP process.Another noteworthy phenomenon was that all the samples presented in Fig.3 were successfully ECAPed without any observable cracks while the Mg-1Y alloy exhibited obvious cracks after 3 passes ECAP at 300 °C.It was speculated that it might be related to the high Y content improving the formability of Mg-Y alloy at low temperature.Many studies [19,20,22,25]had shown that an increase of Y content could weaken the basal texture and activate the non-basal slip, which might improve the formability at low temperature.

Fig.5.Optical micrographs of the Mg-Y alloys during the decreasing-temperature ECAP processes: (a) process Ⅰand (b) process Ⅱfor Mg-1Y, and (c) processⅠfor Mg-5Y.The process Ⅰis: 400 °C-1 pass following 350 °C-1 pass, and then 300 °C-2 passes.The process Ⅱis: 350 °C-2 passes following 300 °C-2 passes.The high-magnification micrographs for 4 passes samples are given at the right side.

From the above results, a fine-grain microstructure or homogeneous microstructure could be obtained by the constanttemperature ECAP processes.As is known,ECAP is one kind of SPD.It is hoped to get microstructure with finer grain and better homogeneity by another ECAP process.A nice way is to reduce the temperature step by step [39,40].Bry?a et al.[40]reported that decreasing-temperature ECAP process led to strong grain refinement about Mg-4Ag alloy.The grain size decreased from 350μm to 38μm after 1 pass ECAP at 370°C and 15μm after 1 pass ECAP at 330 °C.Therefore, the new decreasing-temperature ECAP processes were designed,as shown in Table 1.Fig.5 illustrates the optical micrographs of the Mg-Y alloys during the decreasing-temperature ECAP processes.Mg-1Y alloy could be ECAPed successfully without any observable cracks by the decreasing-temperature ECAP process,which was different from the ECAPed Mg-1Y alloy at 300°C in Fig.2.That was because the temperature of the first pass was set at 400 °C in the decreasing-temperature ECAP process Ⅰand 350 °C in the decreasing-temperature ECAP process Ⅱ, which improved the formability of the Mg-1Y alloy and overcame the cracking.It was easy to see that the grain size decreased with further passes.In Fig.5 (a, b),the difference between the two decreasing-temperature processes was that the rate of grain refinement was different,in which the decreasing-temperature ECAP process Ⅰwas higher.The smallest average grain size was 6.1±0.7μm at 4 passes of the decreasing-temperature ECAP process Ⅰand 7.1±0.5μm at 4 passes of the decreasing-temperature ECAP process Ⅱ,and both of them had homogeneous microstructure.Compared with Figs.2 and 5 (a, b), the microstructure of the Mg-1Y alloy with finer grains and higher homogeneity could be obtained through the decreasing-temperature ECAP process, which had achieved the design goal.For Mg-5Y alloy,only the decreasing-temperature ECAP process Ⅰwas carried out in this paper, as shown in Fig.5(c).It was found that with the increase of passes, the average grain size decreased and the homogeneity of microstructure increased.The smallest average grain size was 7.2±0.1μm at 4 passes.Compared with Figs.3 and 5(c), after the decreasing-temperature ECAP process Ⅰ, the average grain size of the sample was only smaller than that of the 4 passes ECAP at 400 °C.Although the degree of grain refinement was not as good as that of the constant-temperature ECAP at 300 °C and 350 °C,the homogeneity of the microstructure was obviously improved.As previously analyzed, it was believed that the 1 pass ECAP at 400 °C could improve the degree of DRX and significantly improve the homogeneity of the microstructure, a more homogeneous microstructure could be obtained under the decreasing-temperature ECAP process Ⅰ.Here is a reminder, the high-magnification micrographs for 4 passes samples were given at the right side in Fig.5.In these three optical micrographs, it was observed that the grain boundary was serrated borders, which was one of the typical characteristics after SPD [41].It was also found that the direction of the elongated grains was about 45° between ED and TD.

From Figs.2, 3, and 5, the evolution of grain size and microstructure homogeneity of Mg-1Y and Mg-5Y alloys during different ECAP processes were clear.To better understand the microstructure evolution, more factors need to be concerned.In general, there is some residual stress in the deformed alloy.Although the measurement of residual stress is relatively difficult, it is still very important to know.Through EBSD analysis, the kernel average misorientation (KAM) map in the grains can be obtained, which helps understand the resid-ual stress in the material.According to the analysis of optical microstructure, the relatively optimized ECAP processes of Mg-Y alloys were selected for further analysis.Fig.6 illustrates the KAM maps of Mg-Y alloys for relatively optimized ECAP processes.The residual stress is expressed in a different color: the blue indicates low residual stress, the green indicates medium residual stress, and the red indicates strong residual stress.Fig.6(a) and (d) indicate the KAM maps of Mg-1Y and Mg-5Y alloys before the ECAP, respectively.In both Fig.6(a) and (d), long strip shape grains were observed, which was the same as the optical metallographic microstructure of the above analysis.In these long strip shape grains, the KAM map appeared green and yellow, some of them even red, while the other recrystallized fine-grains were almost blue.It showed that there was high residual stress in the long strip shape grains,while there was almost no residual stress in the recrystallized fine-grains.Fig.6(b, c) indicates the KAM maps of the Mg-1Y alloy after the 350°C-3P ECAP and the decreasing-temperature ECAP process Ⅰ, respectively.Fig.6(e, f) indicates the KAM maps of the Mg-5Y alloy after the 300 °C-4P ECAP and 350 °C-3P ECAP, respectively.The KAM maps appeared green only in a few recrystallized finegrains and the distribution of green was more dispersed than that of Mg-Y alloys before ECAP, while the other fine-grains appeared blue.It indicated that there was a lower and more randomly located residual stress in the Mg-1Y and Mg-5Y alloys after ECAP.

Fig.6.Kernel average misorientation (KAM) maps of the Mg-Y alloys for different ECAP processes: (a) before the ECAP, (b) after the 350 °C-3P ECAP,and (c) after the decreasing-temperature process Ⅰfor Mg-1Y samples; (d) before the ECAP, (e) after the 300 °C-4P ECAP, and (f) after the 350 °C-3P ECAP for Mg-5Y samples.

Fig.7 illustrates the relative frequency of grain boundary misorientation angles of the Mg-Y alloys for the relatively optimized ECAP processes.Fig.7 (a-c) indicates the relative frequency of grain boundary misorientation angles of the Mg-1Y alloy before the ECAP, after the 350 °C-3P ECAP, and the decreasing-temperature ECAP process Ⅰ, respectively.The high-angle grain boundaries (HAGBs,θ≥15°) accounted for 82.2% of the Mg-1Y alloy before ECAP, which suggested the incomplete of DRX.The HAGBs frequency was 78.3% after 350 °C-3P ECAP, while the HAGBs frequency decreased to 66.5% when the processing temperature reduced in the decreasing-temperature ECAP process Ⅰ.From the above results, it was believed that DRX could be significantly affected by process temperature.Li et al.[42]also found that the temperature decreased would decrease the volume fraction of DRX, and the frequency of HAGBs was decreased correspondingly.In Fig.5, the grain boundaries were serrated,which indicated an incomplete continuous dynamic recrystallization (CDRX) after the decreasing-temperature ECAP process.Therefore, it was thought that a large amount of energy was reserved at the grain boundaries [33].According to literature reviews [29], continuous DRX (CDRX), discontinuous DRX (DDRX), and geometric DRX (GDRX) are described.It was also reported that the DRX process was very complex and affected by many factors, including original grain size, thermo-mechanical processing conditions and so on.In Fig.7(e, f), the same phenomenon was also found in the ECAPed Mg-5Y alloy.

3.2.Texture evolution

Fig.8 illustrates {0001} pole figures (PFs) and the inverse pole figures (IPFs) for ED of Mg-Y alloys with different ECAP processes.The color legend is shown in the upperright corner of Fig.8.It is shown that the blue color indicates minimum intensity, and the red color indicates maximum intensity for both PFs and IPFs.Fig.8(a-c)indicate the{0001}PF and IPF for ED of Mg-1Y alloy before ECAP, after the 350 °C-3P ECAP, and the decreasing-temperature ECAP process Ⅰ, respectively.The Mg-1Y alloy showed a weak basal texture before ECAP, and the basal pole was tilted to TD about 37° with a maximum intensity of 5.2 mrd, as shown in the PF.For the ED IPF, there was a main peak around＜100＞of the Mg-1Y alloy before ECAP.It was indicated that the＜100＞direction of most grains in the Mg-1Y alloy was almost parallel to the ED.However, the texture showed a trend of further weakening after ECAP.As shown in the PF, the basal pole was divided into several poles, while the maximum pole was tilted to about 40° along the ED with an intensity of around 5.2 mrd in Fig.8(b).In Fig.8(c), the maximum pole tilted about 42° to the opposite direction of ED, but the intensity increased to 8.0 mrd.For the ED IPF,there were also multiple poles, and the distribution of IPF was scattered with the invariant intensity of about 2.1 mrd in Fig.8(b).The added poles were＜101＞and＜111＞,of which the＜111＞was considered as the “RE texture”[18].In Fig.8(c), the main peak was around＜101＞and＜110＞.While its IPF intensity was only 1.7 mrd, which was considered to have no preferred orientation.

Fig.7.Relative frequency of grain boundary misorientation angles of the Mg-Y alloys for different ECAP processes: (a) before the ECAP, (b) after the 350 °C-3P ECAP, and (c) after the decreasing-temperature ECAP process Ⅰfor Mg-1Y samples; (d) before the ECAP, (e) after the 300 °C-4P ECAP, and (f)after the 350 °C-3P ECAP for Mg-5Y samples.

The Fig.8 (d-g) indicate the {0001} PF and IPF for ED of Mg-5Y alloy before ECAP, after the 300 °C-4P ECAP,350 °C-3P ECAP, and 400 °C-3P ECAP, respectively.The Mg-5Y alloy also exhibited a weak basal texture with several dispersed poles, and the maximum basal pole split along TD with tilted about 56° in the PF before ECAP.This was a typical RE-texture feature, and Sandl?bes et al.[20]also found that the strong basal texture changed into the bimodal texture by adding Y element to the pure Mg rolled bar.As shown in the IPF, the Mg-5Y alloy exhibited the main peak around＜100＞with an intensity of around 4.4 mrd before ECAP.Both pole distribution and maximum intensity of the basal texture had a significant weakening after ECAP.In Fig.8 (eg),it was found that a great quantity of secondary poles in PF and the maximum intensity almost unchanged or slightly reduced.However, the main peak of IPF was around＜110＞,and the intensity of IPF decreased to less than 2.0 mrd (the lowest was only 1.2 mrd), which was considered to have no obvious preferred orientation.In addition, it was found that the distribution of poles in the PF after ECAP tended to be about 45° between ED and TD.This was an interesting phenomenon, which was believed to lead to the soft orientation of basal slip and activate more basal slip [42].

3.3.Mechanical properties

Fig.9 illustrates the tensile engineering stress- engineering strain curves of the Mg-Y alloys before and after ECAP.The corresponding values of tensile yield strength (TYS), ultimate tensile strength (UTS) and elongation (EL) are shown in Table 2.For better understanding the effects of the ECAP processes on mechanical properties, a histogram is made, as shown in Fig.10.The relative changes after the ECAP are indicated in the figure.

Table 2Tensile mechanical properties of the Mg-Y alloys after different ECAP processes.

Fig.9(a) indicates the tensile engineering stress- engineering strain curves of the Mg-1Y alloy before ECAP,after 350 °C-3P ECAP, the decreasing-temperature ECAP process Ⅰ, and the decreasing-temperature ECAP processⅡ, respectively.There was no obvious strain hardening or strain softening in the plastic deformation stage except the 350 °C-3P ECAP.It was found that both the TYS and the UTS decreased while the EL increased after ECAP,and many researchers also found the same phenomenon[11,14,32].The EL of Mg-1Y alloy was greatly improved from 11.1±2.9% to 42.7±0.1% (increased by 285%) and the strength decreased significantly (TYS decreased by 55%from 237.7±6.8MPa to 106.2±4.1MPa and UTS decreased by 33% from 273.2±3.9MPa to 182.4±3.6MPa) after350 °C-3P ECAP.The loss of TYS and UTS might be due to the larger average grain size, weaker basal texture and lower residual stress.While the weaker basal texture could also cause an increase of EL.According to this result,by adjusting the ECAP process,the EL of the Mg-1Y alloy could be significantly improved at the expense of much strength.Wei et al.[32]also found that with the increase of passes, the strength of Mg-1Y alloy decreased while the EL increased from 1 pass to 4 passes at 400 °C.So, Wei et al.[32]got the best EL was 27.3% at 4 passes, which was significantly lower than 42.7%after 350 °C-3P ECAP in this paper.It could be concluded that the EL could be greatly improved by simply reducing the processing temperature in Mg-1Y alloy.In addition, it was found that the EL of Mg-1Y alloy increased by 109%from 11.1±2.9% to 23.2±1.2%, while the TYS decreased by only 11% from 237.7±6.8MPa to 211.2±6.1MPa and the UTS decreased by only 14% from 273.2±3.9MPa to 233.8±3.5MPa after the decreasing-temperature ECAP process Ⅰ.Compared with 350 °C-3P ECAP process, the higher strength of the decreasing-temperature ECAP process might be related to finer grain size and incomplete of DRX.Therefore,a large increase in EL of Mg-1Y alloy could be achieved at the expense of a small amount of strength,which was found after the decreasing-temperature ECAP process Ⅰ.

Fig.8.{0001} pole figures and inverse pole figures for ED of the Mg-Y alloys for different ECAP processes: (a) before the ECAP, (b) after the 350 °C-3P ECAP, and (c) after the decreasing-temperature ECAP process Ⅰfor Mg-1Y samples; (d) before the ECAP, (e) after the 300 °C-4P ECAP, (f) after the 350 °C-3P ECAP, and (g) after the 400 °C-3P ECAP for Mg-5Y samples.

Fig.9.Representative tensile engineering stress- engineering strain curves of the Mg-Y alloys before and after ECAP: (a) Mg-1Y and (b) Mg-5Y.

Fig.10.Tensile mechanical properties of the Mg-Y alloys before and after ECAP: (a) Mg-1Y and (b) Mg-5Y.The relative changes after the ECAP are also indicated in the figures.

Fig.9(b) indicates the tensile engineering stress- engineering strain curves of the Mg-5Y alloy before ECAP, after 300 °C-4P ECAP, 350 °C-3P ECAP, and 400 °C-3P ECAP,respectively.Different from the Mg-1Y alloy, the strain hardening phenomenon was more obvious to be observed for the Mg-5Y alloy.From Table 2,the TYS,UTS and EL of the Mg-5Y alloy before ECAP were 249.4±1.8MPa,290.2±1.4MPa and 13.9±1.9%.Compared with the Mg-1Y alloy before ECAP, the mechanical properties of the Mg-5Y alloy were all enhanced.This was due to the comprehensive effects of grain refinement, weaker texture and stronger solid solution hardening on the Mg-5Y alloy.The results of that strength reduction and EL increase were also found in the ECAPed Mg-5Y alloy.After 400 °C-3P ECAP, the EL significantly increased by 133% (increased from 13.9±1.9% to 32.4±0.6%), TYS and UTS decreased by 40% (decreased from 249.4±1.8MPa to150.6±7.6MPa) and 22% (decreased from 290.2±1.4MPa to 226.1±6.6MPa), respectively.Similarly, the EL increased by 58% from 13.9±1.9% to 21.9±0.4%, while the strength was only decreased by ～5% after 350 °C-3P ECAP.Compared with 400 °C-3P ECAP process, there were finer grain size and more long strip shape shear bands after 350 °C-3P ECAP, which resulting in higher strength.Therefore,strength and ductility could be adjusted as needed by adjusting the ECAP process.By comparing the Mg-1Y and the Mg-5Y alloys after the same process of 350 °C-3P ECAP,the higher strength could be obtained in the Mg-5Y alloy,in which the TYS increased by 122% from 106.2±4.1MPa to 236.2±5.8MPa, and the UTS increased by 49% from 182.4±3.6MPa to 271.4±1.4MPa, while the EL decreased from 42.7±0.1% to 21.9±0.4%.It was worth noting that both strength and EL of the Mg-5Y alloy could be enhanced simultaneously by the 300 °C-4P ECAP.After the 300 °C-4P ECAP, the TYS, UTS, and EL were increased by 10%(increase from 249.4±1.8MPa to 273.9±1.2MPa), 6% (increase from 290.2±1.4MPa to 306.4±3.0MPa), and 72%(increase from 13.9±1.9% to 23.9±1.0%), respectively.As shown above, the sample got a very weak basal texture and the average grain size was 1.24μm after 300 °C-4P ECAP.Before further analysis, it was necessary to understand what were the texture hardening and the texture softening.As is known , the slip system of Mg alloys was limited, and the basal slip was easiest to be activated at RT because critical resolved shear stress (CRSS) of the basal slip was the lowest in all the slip systems [3].Generally speaking, Mg alloys had strong basal texture after extrusion or rolling.This showed that the basal plane and ED were almost in the same direction [30](In this paper, the loading direction was also along the ED).The orientation of the loading direction and the basal plane led to the Schmid factor (SF) of basal slip closed to 0.So, the basal slip was seriously hindered, and the strength was greatly enhanced.This phenomenon was considered to be texture strengthening.However, after ECAP, the basal texture was obviously weakened and the strengthening effect was correspondingly weakened, which was called texture softening.It is known that the inclined basal texture can negatively affect the strength properties while improving the ductility of Mg alloys.For the AZ61 alloy, Kim et al.[16]found obvious texture weakening after ECAP, and the yield stress decreased after 8 ECAP passes, which indicated that the effect of texture softening on the strength was very significant.Sun et al.[43]found that the EL increased with the further weakening of the basal texture in AZ31 alloy.While the mechanical properties were significantly improved by rolling, which mainly due to the grain size refinement.While the weak basal texture of the Mg-5Y alloy in this paper was further weakened after ECAP, as shown in Fig.8.Besides, it was also found that the average SFs of basal, prismatic and 2ndpyramidal slips for Mg-5Y alloy had little change in Fig.12(b).Therefore, it was thought that the texture changes had a weak effect on SFs of basal, prismatic and 2ndpyramidal slips for Mg-5Y alloy.A more detailed discussion of the effects of texture will be provided later.It was known that theRe(Y,Gd) atoms tended to segregate to the grain boundaries [33].Hoseini-Athar et al.[44]and Xu et al.[45]had proposed solid solution as the main mechanism contributing to the yield stress in Mg-Gd alloys.Besides, the large cubic-shaped YH2particles (0.2-4.2μm) were observed after ECAP process.It is worth noting that no obvious changes of volume fraction for the cubic-shaped particles were observed during ECAP,and the amount of Y in the solid solution state was believed to be constant.Therefore, its effect on mechanical properties needed to be discussed.Hoseini-Athar et al.[44]reported that the second phase strengthening of Mg-2Gd-(1-3wt.%)Zn alloy was thought to can be neglected due to the small volume fraction (2.7%?5.4%) as well as large size (mostly larger than 1.5μm).Morozumi et al.[46]reported that YH2particles did not necessarily contribute to further strengthening of the Mg-Y alloys at RT.Therefore,the large cubic-shaped YH2particles (0.2-4.2μm) with small volume fraction (less than 6.1±1.8%) in Mg-Y alloys were considered to have no significant strengthening effect.Therefore, the critical factor for the increase of strength was significant grain size refinement,combined with solid solution hardening and significant improvement of microstructure homogeneity after the 300°C-4P ECAP.The significant DRX and improvement of microstructure homogeneity helped to improve EL.There were better comprehensive mechanical properties of the decreasingtemperature ECAP process for Mg-1Y alloy than that of the constant-temperature ECAP process.That was because the comprehensive properties of grain size and microstructure homogeneity of the decreasing-temperature ECAP process were better.

Fig.11 indicates the strain hardening curves of the Mg-Y alloys before and after ECAP.The strain hardening behavior of Mg-Y alloys is analyzed by strain hardening rateθ,calculated as [44]:

Whereσandεare true stress and true plastic strain, respectively.Fig.11(a)indicates the true plastic strain-strain hardening rate curve for Mg-1Y alloy before and after ECAP.In the earlier stage, the strain hardening rate of all the Mg-1Y alloy samples showed a similar trend of rapid linear decrease.After the rapid decrease, the curve declined much more slowly.In this stage, the strain hardening rate of Mg-1Y alloy after the decreasing-temperature ECAP process Ⅰand Ⅱwas similar to that before ECAP.However, the strain hardening rate increased significantly after 350°C-3P ECAP,which was consistent with the obvious strain hardening phenomenon shown in Fig.9(a).It was found that there were the largest average grain size and better homogeneity of Mg-1Y alloy after 350 °C-3P ECAP.For Mg-5Y alloy, the evolution trend of strain hardening rate was similar to that of Mg-1Y alloy.In addition, the strain hardening rate of Mg-5Y after 400 °C-3P ECAP was the highest in all the samples in the stage of slow decrease of strain hardening rate in Fig.11(b).It was also found that there were the largest average grain size and bet-ter homogeneity of Mg-5Y alloy after 400 °C-3P ECAP in Fig.3.

Fig.11.True plastic strain-strain hardening rate curves of (a) Mg-1Y and (b) Mg-5Y alloys before and after ECAP.

In addition, strain-hardening exponent (n) and hardening capacity (HC) are used to quantitatively analyze strain hardening behavior.The strain-hardening exponent is commonly expressed by the Hollomon relationship [44]and hardening capacity is defined in terms of UTS and TYS as [44]:

wherenis the strain-hardening exponent,Kis the strength coefficient,σis the true stress andεis the true plastic strain.The strain-hardening exponent and hardening capacity are listed in Table 2.For Mg-1Y alloy, it was found that the strain-hardening exponent (n=0.11) and hardening capacity(HC=0.72)after 350°C-3P ECAP were largest in all the samples, which indicated the optimal strain hardening and was consistent with the result of strain hardening rate curve in Fig.11(a).However, for Mg-1Y alloys before ECAP and after decreasing-temperature ECAP process Ⅰand Ⅱ, the strainhardening exponent and hardening capacity had little difference.For Mg-5Y alloy,the evolution trend of strain-hardening exponent and hardening capacity were similar to that of Mg-1Y alloy.After 400 °C-3P ECAP, the strain-hardening exponent (n=0.18) and hardening capacity (HC=0.50) were largest.

Fig.12.Grain size dependencies of yield strength for the Mg-Y alloys before and after ECAP.

How the ECAP process affected the strength and ductility will be discussed as follows.ECAP is one of the effective ways to refine grains, which usually leads to fine grain strengthening.Hall-Petch relation is used to understand the relationship between yield strength and grain size [16,17].Fig.12 illustrates grain size dependencies of yield strength for Mg-Y alloys before and after ECAP.The red line is the auxiliary line for Hall-Petch relation fitting except for Mg-Y alloys before ECAP.In Fig.12, yield strength-grain size relationship of all ECAPed Mg-Y alloys could be well fitted by Hall-Petch relation.However, different results that the yield strength-grain size relationship could not be well described by Hall-Petch relation were found in some AZ Mg alloys [8,16].It was believed that the texture modification during the ECAP process might cause the above results.Li et al.[42]got modified texture after ECAP, and the SF of basal slip for Mg-6.52Zn-0.95Y increased from 0.19 to 0.40, which was considered to be the cause of low TYS and high EL.Kim et al.[16]also calculated the SFs of basal slip and prismatic slip based on the most dominant texture.It was also found that there were obvious changes in SFs of basal slip and prismatic slip under different textures, which helped to understand the yield stress behavior.Therefore, the average SFs of basal, prismatic,and 2ndpyramidal slips for Mg-Y alloys before and after ECAP were calculated systematically, as shown in Fig.13.Different from the above papers, SFs of basal slip and pris-matic slip was slightly changed but not obvious, and the SF of 2ndpyramidal slip almost invariable for the Mg-1Y alloy in Fig.13(a).The SF of each slip mode for the Mg-5Y alloy had little change in Fig.13(b).Kim et al.[8]found that the AZ31 alloys showed a similar texture after ECAP.Besides,all ECAPed AZ31 alloys exhibited a standard Hall-Petch relation.Combined with the weak basal texture of the Mg-Y alloys before and after ECAP showed in Fig.8,it was thought that the texture changes did not significantly affect the average SFs of basal, prismatic and 2ndpyramidal slips for both Mg-Y alloys in this paper.In addition, it was easily seen that the Mg-1Y and Mg-5Y alloys before ECAP deviated far from the fitting line of the Hall-Petch relation.According to the above analysis, texture modification was not the cause of this result.Therefore, more consideration should be given to the microstructure of the samples.In Fig.6, there was high residual stress in the long strip shape grains, while there was almost no residual stress in the recrystallized fine-grains of the Mg-1Y and the Mg-5Y alloys before ECAP.Besides, all the ECAPed Mg-Y alloys exhibited a lower and more randomly located residual stress.Huang et al.[22]also found that the yield strength-grain size relationship of the as-extruded Mg-Y sheets was difficult to fit.The most likely causes were speculated to high residual stress and high dislocation density in the as-extruded Mg-Y sheets.Minárik et al.[47]found the same phenomenon in LAE442, which was proved to be attributed to high dislocation density.Therefore, the high residual stress in the Mg-Y alloys before ECAP might cause the failure of the fitting.

Fig.13.Average Schmid factor for various deformation modes versus different ECAP processes for the: (a) Mg-1Y and (b) Mg-5Y alloy.

4.Conclusion

The microstructure, texture and mechanical properties evolutions of the extruded Mg-xY (x=1, 5wt.%) alloys during ECAP were systematically investigated.The main conclusions of the present work are summarized below:

(1) Both the Mg-1Y and Mg-5Y alloys exhibited a weakened basal texture before ECAP.The texture was further weakened after the ECAP featured by the maximum basal poles dispersing along ～45° between ED and TD and the＜110＞direction nearly paralleling to ED for most grains.

(2) The finest grain (～1.24 μm) was obtained in Mg-5Y alloy after 300°C-4P ECAP.Besides,the grains became finer with the increasing Y content from 1.0% to 5.0%for the same ECAP process.

(3) With proper ECAP process and Y content, both the strength and elongation could be improved simultaneously which is difficult to be obtained in conventional Mg alloys with initial strong basal texture.After 300 °C-4P ECAP, the TYS (273.9±1.2MPa), UTS(306.4±3.0MPa), and EL (23.9±1.0%) of the Mg-5Y alloy increased by 10%, 6% and 72%, respectively.It was attributed to the grain refinement, significant improvement of microstructure homogeneity and solid solution hardening.In addition, better comprehensive properties of the Mg-Y alloys could be obtained by the decreasing-temperature ECAP processes.

(4) The yield strength-grain size relationship of all the ECAPed Mg-Y alloys could be well fitted by the Hall-Petch relation, which was consistent with that the texture changes did not significantly affect the average SFs of basal, prismatic and pyramidal slips for both Mg-Y alloys.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (Nos.51401172 and 51601003),Sichuan Science and Technology Program (2019YJ0238),Fundamental Research Funds for the Central Universities(2682020ZT114), and open funding of International Joint Laboratory for Light Alloys (MOE), Chongqing University.We would like to thank the Analytical and Testing Center of Southwest Jiaotong University for EBSD tests.