Kang Wei,Lirong Xiao,Bo Gao,Lei Li,Yi Liu,Zhigang Ding,Wei Liu,Hao Zhou,Yonghao Zhao
Nano and Heterogeneous Materials Center,School of Materials Science and Engineering,Nanjing University of Science and Technology,Nanjing 210094,China
Received 24 July 2019;received in revised form 16 September 2019;accepted 19 September 2019 Available online 30 June 2020
Abstract Due to the insufficient slip systems,Mg and its alloys exhibit poor ductility during plastic deformation at room temperature.To solve this problem,alloying is considered as a most effective method to improve the ductility of Mg alloys,which attracts wide attentions of industries.However,it is still a challenge to understand the ductilization mechanism,because of the complicated alloying elements and their interactions with Mg matrix.In this work,pure Mg and Mg-Y alloys were comparatively studied to investigate the effect of Y addition on microstructure evolution and mechanical properties.A huge increase of uniform elongation,from 5.3% to 20.7%,was achieved via only 3wt% addition of yttrium.TEM results revealed that the only activated slip system in pure Mg was basalslip,led to its poor ductility at room temperature.In contrast,a large number of stacking faults and non-basal dislocations with
Keywords:Magnesium alloys;Ductility;Stacking faults;Non-basal slip;Transmission electron microscopy.
As the lightest metallic materials,magnesium and its alloys have great potential for structural applications in many fields,such as automotive,aerospace,and electronic industries[1–3].However,the activable slip systems in hexagonal close-packed(hcp)structural materials are not enough for the uniform plastic deformation at room temperature,which limits the commercial applications of wrought magnesium alloys[4].The most popular strategy to solve this problem was to introduce more slip systems in Mg alloys.Especially,the introduction of pyramidal
However,the mechanism of ductilization is complicated,because it is difficult to obtain direct evidence of interaction between alloying elements and Mg matrix.Wang et al.[15]performed an in-situ tensile experiments on Mg-3Y alloy using three dimensional X-ray diffraction.They found that the addition of Y element promoted the activity of prismatic and pyramidal slip systems.Additionally,RE addition was proposed to reduce the basal plane stacking fault energy(SFE)of Mg,which was coincident with the results of density functional theory(DFT)simulations[16–18].Sandl?bes et al.[14,16]found that a large number of I1 stacking faults generated in a deformed Mg-RE alloys.They believed that the I1 stacking faults acted as heterogeneous sources for nucleation of pyramidal
Ingots of Mg-3Y(wt%)alloy were prepared from a high purity Mg(99.99%)metal and a Mg-25Y(wt%)master alloy in an electric-resistant furnace under a mixed protective gas of Ar and SF6that has a volume ratio of 100:1(The details of alloy preparation have been reported in Ref.[23]).The ascast ingots of pure Mg and Mg-3Y were homogenized in a vacuum furnace at 450°C for 20h and at 530°C for 12h,respectively,which were followed by water quenching to room temperature.Rolling deformation was performed on the two materials to eliminate the casting defects,which deformed the samples to a thickness reduction of 10%.The rolled samples were fully annealed for recrystallization at 325°C(pure Mg)and at 500°C(Mg-3Y)alloy,respectively.Dog-bone shaped specimens with a gauge length of 10mm and width of 2.5mm were cut from the annealed samples for tensile testing.The uniaxial tensile testing was performed on a walter+bai mechanical testing machine(LFM-20kN)with a strain rate of 2×10?3s?1at room temperature.The samples for TEM observations were obtained from the tensile specimens which interrupt at an engineering strain of~4%.The samples for optical microscopy observation were ground with sandpaper of 320,800,1200 and 2000 grits,and then polished by woolen cloth to a mirror finish.To observe microstructure under optical microscope,samples were etched by lab-prepared solution comprised of 95ml ethyl alcohol and 5ml nitric acid.
Cross-sectional TEM specimens were cut perpendicular to the tensile direction and gently polished to a thickness of~25μm.Perforation by ion milling(Gatan PIPS 691)was carried out on a cold stage(~?50°C)with low angle(5°)and low energy ion beam(4KeV).TEM observations were conducted on an FEI-Tecnai G2 20 LaB6 microscope operated at an accelerating voltage of 200kV,and the high-resolution TEM(HRTEM)observations were carried out on a Titan G2 60-300 at 300kV.
As shown in Fig.1a and b,the annealed samples of pure Mg and Mg-3Y alloy exhibit similar microstructure,which have average grain sizes of~196μm and~199μm,respectively.Equivalent densities of annealing twins exist in some grains of both specimens as marked by the white arrows.Twins are frequently observed in many Mg alloys,which is an important deformation mode to improve ductility similar to dislocations[24].However,the contribution of twinning to plasticity is very limited,because it only allows simple shear in one direction[25].Fig.1c shows the tensile stress–strain curves of the annealed pure Mg and Mg-Y alloy.The yield stress(YS,0.2% proof stress)and ultimate tensile strength(UTS)of the materials are very close to each other,which are~50MPa and~130MPa,respectively.Interestingly,the Mg-3Y alloy exhibits a four times higher uniform elongation(~21±1.2%)than that of pure Mg(~5±0.5%).Note that,pure Mg and Mg-3Y alloy specimens contain almost identical grain size and twinning density.Thus,it is reasonable to propose that such significant ductilization of Mg-Y alloy is mainly due to the addition of Y element.Fig.1d shows that the strain-hardening rate of the pure Mg decreased drastically after 5%of tensile strain.While,it is also found that the work hardening capability increased with addition of Y,which helps with retaining ductility of Mg.Previous researches believed that the ductilization of Mg-RE alloys were resulted from the weakening of basal plane texture,which was beneficial to activate non-basal slip systems in Mg-RE alloys[14,26].It indicated that dislocation type was the most fundamental issue for improving ductility of Mg alloys[13].Therefore,the types and configurations of dislocations in grain interior of pure Mg and Mg-Y alloy were investigated intensively.
To explore the underlying ductilization mechanism of Mg-Y alloy,detailed TEM characterizations(more than 30 grains for each sample)were performed on the pure Mg and Mg-Y samples subjected to a same strain of~4%.According to the“g?b”criterion,dislocations are invisible wheng?b=0,wheregandbrepresent diffraction vectors and Burgers vectors,respectively.In other words,dislocations withb=1/3<11-20>are invisible if the two beam condition is set asg=0002,and the
Fig.2a–c show the microstructures of a deformed pure Mg observed near the[2-1-10]zone axis.The two-beam diffraction conditions were set asg=0002 andg=0-110,as shown in Fig.2a and b,respectively.A twin marked by the white pentagram is selected as the label structure to determine the accurate positions of observation area.All the dislocations can only be observed in the condition ofg=0-110(Fig.2b and c),which indicates that these dislocations in pure Mg are mostly basaltype dislocations,as marked by yellow arrows.It is generally accepted that basal slip system is easier to be activated than other slip systems in pure Mg,because the resolved shear stress(CRSS)of basal slip is much lower than that of non-basal slips[29].Therefore,basal slip is dominated in the tensile deformation of pure Mg,resulting in poor ductility at room temperature.
Fig.1.Microstructures and mechanical properties of as-received samples:(a)and(b)Optical microscopy images of Pure Mg and Mg-3Y,respectively.The insets are grain size distribution charts;(c)Tensile curves of nominal stress vs.strain;and(d)True stress–strain and strain hardening curves from tensile tests.
Fig.2d–f show the TEM images of the Mg-Y alloy near the direction of[2-1-10]zone axis.A grain boundary marked by a white pentagram pattern is selected as the label structure.Besides basaldislocations shown in Fig.2e,numerous non-basal dislocations with
HRTEM observations were performed to investigate the detailed structure of stacking faults(SFs).As shown in Fig.3a and b,a high density of basal SFs(marked by the yellow arrows)are observed in the grain interior of Mg-3Y alloy.The spacing of SFs is in the range of 3–10nm.Meanwhile,some bright streaks(indicated by the white arrows)appear in the fast Fourier transform(FFT)pattern(Insert in Fig.3b),which is consistent with the diffraction shape effect of SFs on the basal plane.In order to identify the accurate types of SFs and their bounding dislocations,atomic-scaled observations were carried out on the aberration-corrected TEM operated at 300kV.Interestingly,all of the SFs observed in Mg-Y alloy are intrinsic faults,while no extrinsic faults are found in the alloy.Fig.3c and d show the atomic scaled TEM images of I1 faults,in which a Burgers circuit is drawn around the dislocation core.The start point D does not overlap with the finish point E,which indicates that vector DE(marked by the red arrow heads in Fig.3d)is the Burgers vector of Frank partial dislocation(b=1/6<20-23>).Similarly,the atomicscaled structures of I2 faults are also characterized,as shown in Fig.3e and f.The Burgers vector of its partial dislocation is1/3<10-10>,as illustrated by the Burgers circuit in Fig.3f.
Fig.2.TEM images in two-beam condition near the[2-1-10]zone axis:(a)and(b)Pure Mg viewed under g=0002 and g=0-110,respectively;(d)and(e)Mg-Y alloy viewed under g=0002 and g=0-110,respectively;(c)and(f)Enlarged images of selected area in(b)and(d),respectively;The white straight lines highlight the trace of(0001)basal plane.
As shown in Fig.4,the atomic models of I1 and I2 faults are established according to the experimental results.It is clear that I1 and I2 faults introduce a thin three-layer and four-layer of face-center cubic(fcc)stacking structure into the hcp matrix,respectively.The 2D atomic model in Fig.4b shows that the stacking sequence of I1 faults is ABABABCBCBCB.I1 intrinsic faults can be produced through the convergence of vacancies or interstitials on the basal plane,combined with the shear process of Shockley partial dislocations,thus resulting in I1 faults bounded by1/6<20-23>Frank partials[32].The corresponding dislocation reaction is described in Eq.(1).I1 stacking faults act as heterogeneous nucleation sources for perfect
Fig.3.Microstructures of a deformed grain in Mg-3Y alloy obtained from[2-1-10]axis:(a)and(b)the HRTEM images and corresponding fast Fourier transform(FFT)pattern of stacking faults;(c)and(d)the atomic scaled TEM images of I1 faults bounded by 1/6<20-23>Frank partials;(e)and(f)the atomic scaled TEM images of I2 faults bounded by 1/3<10-10>Shockley partials.Insets in(d)and(f)are the enlarged images of the corresponding square regions denoted by the dash lines,respectively,and red arrow heads highlight the Burgers vectors of the partials.(For interpretation of the references to color in this figure legend,the reader is referred to the web version of this article.)
Fig.4d reveals the 2D atomic structure of I2 faults yielding the sequence of ABABABCACACA.It is generally accepted that I2 faults are dissociated from the perfect basaldislocations.Its intrinsic dislocation reaction is represented in Eq.(2).Basically,I2 faults are bounded by Shockley partials(b=1/3<10-10>),which remain glissile on basal planes to accommodate the basal strain.Meanwhile,the basal partials provide Shockley shear for nucleation of perfect
Fig.4.Atomic models of stacking faults in Mg-Y alloy:(a)and(b)the 3D and 2D models of I1 faults,respectively;(c)and(d)the 3D and 2D models of I2 faults,respectively.
Based on the experimental results,the corresponding SFE ofγ,can be calculated by the following formula[16,35]:
WhereGrepresents the shear modulus,νthe Poisson’s ratio,βthe angle between the partial dislocations;bis the magnitude of the Burgers vector of the partials,anddis the measured width of the stacking faults.The values used for pure Mg(G=17GPa,ν=0.35,b=0.186nm for Shockley partials;b=0.320nm for Frank partials)were also adopted for the solid solution Mg-Y alloys[36].
The average I1 and I2 SFE of Mg-3Y determined via formula(3)amounts to 3.7±0.5 mJ m?2and 3.4±0.6 mJ m?2,respectively,which are distinctly lower than the basal SFEs(experimental measurements,by TEM analysis:>50 mJ m?2)of pure Mg[16].Recently,extensive DFT studies have been carried out to investigate the effect of alloying RE elements on the SFEs of binary Mg alloys[37–40].The calculated results further reveal that slight additions of Y into Mg can reduce the SFEs significantly(I1 SFE:from 18–21 to 4–9 mJ m?2,I2 SFE:from 30–38 to 14–23 mJ m?2),which are qualitatively coincide with the experimental results in our study.Physical features controlling the SFE with alloying Y element have been considered in previous works.Wu et al.[37]suggested that solute Y atoms led to charge redistribution of matrix Mg atoms,and the disturbance of the pseudo-atom bonds in the pristine lattice of Mg-Y caused the reduction of SFE.In addition,Zhang et al.[38]pointed out that changing of the SFE was expected to be related to the ionization energy and atomic radius of solute atoms.Substituting Mg atom by alloying Y atom with lower 1st ionization energy and bigger atomic radius tended to induce lattice expansion,which led to reducing of SFE.
The microstructure evolution and mechanical properties of pure Mg and Mg-Y have been comparatively studied.A huge enhancement of uniform elongation,from 5.3%to 20.7%,was achieved via only 3wt% of Y addition.The key findings are summarized as follows:
(1)Basal slip is dominated in the tensile deformation of pure Mg,resulting in poor ductility at room temperature.In contrast,besides basaldislocations,numerous SFs and non-basal dislocations with
(2)All of the SFs observed in Mg-Y alloy are I1 and I2 intrinsic faults,while no extrinsic faults are found in the alloy.The I1 and I2 faults are bounded by Frank partials(b=1/6<20-23>)and Shockley partials(b=1/3<10-10>),respectively.The significant decrease of SFE is proposed as the main reason for the formation of high density SFs.
Acknowledgments
This work was supported by the National Key R&D Program of China(2017YFA0204403),National Natural Science Foundation of China(51601003,51901103),and the Fundamental Research Funds for the Central Universities(30918011342).The authors wish to express their appreciation to the Jiangsu Key Laboratory of Advanced Micro&Nano Materials and Technology.TEM experiments were performed at the Materials Characterization and Research Center of Nanjing University of Science and Technology.
Journal of Magnesium and Alloys2020年4期