Wenchen Xu,Chun Yng,Hiping Yu,Xueze Jin,Guojing Yng,Debin Shn,*,Bin Guo

a School of Materials Science and Engineering & National Key Laboratory for Precision Hot Processing of Metals,Harbin Institute of Technology,Harbin 150001,China

b Institute of Machinery Manufacturing Technology,China Academy of Engineering Physics,Mianyang,Sichuan,621900,China

Abstract The continuous eddy current pulse treatment(ECPT)combined with heat treatment was employed to heal the microcracks in spin formed Mg alloy tubes and improve their mechanical properties in this study.The results show that all the microcracks in different tube specimens were almost healed after different continuous ECPT schemes up to 15 cycles.The schemes with less cooling intervals exhibited better healing effect and increased the strength and elongation of Mg alloy tubes more obviously.After aging treatment,the strength improvement of the specimens with ECPT was more remarkable than that of the specimens without ECPT,and the elongation decrease of the specimens with ECPT was less evident than that of specimens without ECPT due to the segregation of RE elements on the crack surface.Besides,after solution treatment,the strength reduction and ductility improvement of the specimens with ECPT were more pronounced than that of the specimens without ECPT owing to the notable decrease of dislocation density of the specimens with ECPT.Both narrowed cracks induced by ECPT and the segregation of precipitates in the vicinity of microcrack surface during aging treatment contributed to the maximum strength in the as-spun specimens with ECPT followed by aging treatment.? 2020 Chongqing University.Publishing services provided by Elsevier B.V.on behalf of KeAi Communications Co.Ltd.This is an open access article under the CC BY-NC-ND license(http://creativecommons.org/licenses/by-nc-nd/4.0/)Peer review under responsibility of Chongqing University

Keywords:Microcrack;Healing;Eddy current pulse treatment;Heat treatment.

1.Introduction

Magnesium alloys are of great interest for many potential applications including automotive,aircraft,aerospace,3C industries,and biomaterials and so on[1,2],especially rare element(RE)containing Mg alloys for its improved ductility and enhanced strength as compared to conventional Mg alloys[3,4].For example,Mg and its alloys differ from other biomaterials by presenting compatible mechanical and physical properties to human bone[5].The tubular component is one of most widely used metallic structural forms,and the universality of tubular component makes their plastic forming process attract considerable attention,such as spinning forming and extrusion.However,microcracks are easy to occur in the forming process of tube workpieces of difficult-to-orking metals,especially for magnesium alloys and titanium alloy.Although the crack propagation can be effectively arrested during hot forming because of the existence of a healing process under high temperature and pressure,some cracks still exist in the formed workpieces[6].The situation is usually more severe in cold-formed workpieces.These cracks existing in metal products decrease the mechanical performance and residual life of tubular components,which limits their widespread application and makes it hard to undertake further forming operations.Besides,the damage or degradation tends to take place due to the existence of cracks and then results in a failure of structural tube component,which often leads to huge losses in engineering applications[7].Therefore,it is necessary to repair or heal the crack to improve the mechanical properties or extend the lifetime of metallic tube components.

So far,many studies on crack healing have been found conducted for biological materials,polymers and ceramics.Examples include capsule-based self-healing materials[7,8],shape memory alloy(SMA)reinforced materials[9],plasma coating materials[10],etc.But it is more difficul to heal cracks in the metallic materials than in other materials because of high bond strength,small volumes and low diffusion rates of metallic atoms[11].Nevertheless,quite a few researches have been devoted to crack healing for metallic materials until now.For instance,the shape memory alloy wires were pre-added into off-eutectic metal matrix to heal the cracks by means of subsequent heating[12].Besides,the microcracks and voids in some aluminum alloys could be reduced significantl during supersaturated solid solution treatment[13].But these methods are not applicable to prefabricated materials because the SMA wires and precipitation elements should be added into the materials in advance[14].The hot plastic deformation method may repair cracks[15],but it could not be used for well-formed parts in that it would cause the deformation of the target parts.The heat treatment was also employed to heal submicron-scale voids embedded in a cold-rolled Al-Mg-Er alloy at a relatively low temperature of 453K[16],and the cracks in Mg-Gd alloys were also found to be healed during heating[17],whose disadvantage was that the atoms diffusion effect during heat treatment was not strong enough to heal large size cracks.Microcracks on metal surface can be repaired by plasma spraying,electroplating or gas carbonitriding.Zhou et al.took advantage of surface plasma spraying Ti3SiC2to repair microcracks,contributing to the improvement of fretting wear performance[18].Zhang R.et al.fabricated phytic acid conversion coatings on AZ31 magnesium alloy,the coatings change from amorphous magnesium phytate into crystalline Mg2P2O7when heat-treated at 300°C,and cracks on the surface of the coatings gradually healed up with the increase of the heat treatment temperature[19].Li et al.made use of gas carbonitriding technology to repair the surface cracks in 42CrMo steel bar,and the three-point bending test showed that the fracture strength of the repaired sample can be restored to 63.38% of the fracture strength of the crack free sample[20].Welding repair is also a common way of crack repair.Gunter et al.repaired the microcracks in 304L stainless steel by friction stir welding,and the repair width of microcracks could be from 0 to 2mm[21].Laser repairing technology is another novel way for crack repair in recent years.Cong et al.used laser remelting process to repair the thermal fatigue crack of annealed H13 steel die,and the results showed that the crack density and crack length are reduced,proving that it is an effective method to repair the crack and improve the thermal fatigue resistance[22].However,welding repair and laser repair are not suitable for inner cracks,which limit their application domain.

Besides the above methods for crack healing,the application of electric current pulse treatment(ECPT)has attracted increasing attentions in the recent years,which is suitable for almost all of the metals,such as silicon iron[23,24],titanium alloy[25,26]and stainless steel[14,27],which could work in a remarkably short time.However,nearly all of the electric current pulse treatments are used for the crack healing of plate specimens heretofore,which is not applicable to the complex-shaped parts,such as tube specimens,because electric sparking is easy to appear in the contact regions between the tube ends and electrodes.In this study,the microcracks in tube specimens of magnesium alloy were treated through employing pulse eddy current induced by high voltage pulse power supply,which could be applied in simple and safe conditions.In order to further improve mechanical properties of tubular specimens,subsequent heat treatment was also applied to the tube specimens.This study provided an effective way for healing crack and enhancing mechanical properties of non-ferrous metallic tube workpieces by coupling pulse eddy current and heat treatment.

2.Materials and methods

2.1.Materials

The as-cast ingot of magnesium alloy rich in rare earth(RE)was machined into the tubular blanks with the inner diameter of 72.3mm and wall thickness of 8.0mm,and the chemical composition of Mg-RE alloy was listed in Table 1.The power spinning process was employed to prepare the thin-walled Mg tubes by using RLE800 CNC spinning machine(DENN,Spain),as shown in Fig.1a.After fi e passes of power spinning at 450°C,the fina size of as-spun tubes were 1.7±0.10mm in thickness and more than 300mm in length,and the total thickness reduction of as-spun tubes reached 77%.The wire electrical discharge machining(WEDM,Ren Guang CNC Equipment Co Ltd,Suzhou,China)was employed to separate the as-spun tubes into several tube specimens with the length of 30mm for eddy current pulse treatment(ECPT),as shown in Fig.1a and b.

Table 1Chemical composition of Mg-RE alloy(wt.%).

Table 2Experiment scheme of ECPT and heat treatment of as-spun Mg-RE alloy specimens.

2.2.Eddy current pulse treatment

The cross section morphologies of tube specimen were recorded using Quanta 200FEG Scanning Electron Microscopy(SEM)with working distance of 10cm and accelerating voltage of 20KV.The observed zone on the cross section is shown in Fig.1b,and the SEM pictures of cross section morphologies are illustrated in Figs.3-5.And then the tubular specimens were treated by ECPT at room temperature.The designed schematic and equipment for ECPT was shown in Fig.1c and 1d,wherein the tubular specimens were fi ed upon the stainless steel shaft inside the copper coil.The copper coil was connected with a high voltage pulse power supply including a capacitor bank with the maximum discharging voltage of 15kV,and the capacitor could release the maximum energy of about 50kJ.When the capacitor discharge in pulse power supply took place,the damped attenuation wave appeared within the copper coils,as illustrated in Fig.1e.The tubular specimen was arranged outside of the stainless steel shaft in order to suppress radial contraction of tubular specimen,as shown in Fig.1c and d.

Fig.1.As-spun Mg alloy tube and design schematic of ECPT equipment(a)as-spun tube,(b)tube specimen cut from as-spun tube,(c)ECPT design schematic,(d)ECPT equipment,(e)output pulse current waveform,(f)the variation of maximum temperature on the cross section with time.

Table 2 presents the experiment scheme of ECPT and heat treatment of as-spun magnesium alloy specimens.The tube specimens A1-A4 taken from the same as-spun tube(tube I)were applied by different ECPT schemes,respectively.For comparison,the tube specimen A1 was not applied to ECPT,while the specimen A2-A4 underwent continuous ECPT up to 15 times at the discharging voltage of 7kV.The ECPT parameters,such as discharging voltage and treatment time,were selected to ensure the microcrack healing effect and avoid local melting.During the ECPT process,the tube specimen A2 was cooled down and the end surface was observed by SEM after 5,10 and 15 times of ECPT,respectively.The tube specimen A3 was cooled down and the end surface was observed after 7 and 15 times of ECPT,respectively,and the specimen A4 was cooled down and the end surface was observed only after 15 times of ECPT.During the continuous ECPT,since the charging process of the capacitor in pulse power supply only needed a couple of seconds(about～60 s),there was no long cooling intervals between two successive ECPT cycles in order to raise the efficien y of the ECPT,which was different from our previous study in which the tube specimen was cooled down to room temperature before the onset of the next ECPT cycle[28].In order to compare the heat accumulation during the continuous ECPT process,the FLIR 325 thermal infrared imager was employed to monitor the maximum temperature on the cross section of tube specimen,and the measured surface was indicated in Fig.1c.Both the thermal infrared imager and thermocouple were employed to measure the temperature of Mg-RE alloy specimen directly after ECPT at the same time,and then the infrared emissivity of the specifi Mg-RE alloy was adjusted to ensure the temperature detected by the thermal infrared imager be consistent with the thermocouple measured value.The maximum temperature variation with time of specimen A2 during the firs 5 ECPT cycles was shown in Fig.1f.

2.3.Microstructure evaluation

The group B and C of tube specimens were cut from another as-spun tube(tube II).Before heat treatment,the tube specimens in group B were subject to ECPT at the discharging voltage of 7kV,while the tube specimens in group C were not implemented to ECPT.The heat treatment for as-spun Mg alloy specimens was conducted in a box-type electric resistance furnace,which included solution treatment,aging treatment and solution/aging treatment.According to the study on heat treatment of Mg-6Gd-3Y-0.5Zn-0.3Zr alloy[29,30],the solution treatment was conducted at 500°C for 10h,while aging treatment was conducted at 210°C for 120h.The solution/aging treatment was the combination of solution treatment and subsequent aging treatment.The microstructures were observed before and after heat treatment through the Olympus optical microscope(OM),Quanta 200FEG scanning electron microscope(SEM)equipped with an energy dispersive spectroscope(EDS)and TecnaiG2F30 transmission electron microscope(TEM).The electron backscattered diffraction(EBSD)was used to map grain size distribution.Before the SEM and EBSD observation,the samples were polished by dry sandpapers of 200 mesh,400 mesh,600 mesh,800 mesh and 1000 mesh in turn in order to obtain smooth surfaces.Then the samples were electropolished to eliminate surface scratches.The electropolishing agent was the mixture of phosphoric acid and alcohol(5:3),and the polishing voltage and time were 5V and 10 min,respectively.The metallographic samples should be further treated by 5-10s metallographic etching,and the metallographic etching agent included picric acid(11g),alcohol(180ml),acetic acid(10ml)and distilled water(20ml).

Fig.2.Tensile specimen dimension in uniaxial tensile experiment.

2.4.Mechanical property test

After the ECPT,ten tensile specimens were cut from each tube specimen in each group,and the specimen dimension was illustrated in Fig.2.The uniaxial tensile tests of specimens were carried out on an Instron-5569 test machine with load cell of 30kN at a strain rate of 0.001s?1,and the tensile specimen dimension The Vickers hardness of specimens was measured in a Struers DuraScan-70G5 hardness test instrument with the load of 0.1Kg and the dwell time of 10s.

3.Results and discussion

3.1.Effect of continuous ECPT on crack healing and mechanical properties of as-spun tube

Fig.3 presents the evolution of the microcracks in the spin formed tubular workpiece.The microcracks on the cross section of tube specimens A2 after power spinning were observed,as shown in Fig.3a.When subject to ECPT for 5 times,the microcracks on the cross section of tube specimen A2 tended to close in area A,as shown in Fig.3b.The microcracks narrowed down evidently and even closed entirely in area A,and the microcracks in adjacent area B also manifested a strong tendency of crack healing.With the increase of ECPT times,more and more microcracks became basically closed or pronouncedly narrowed,and most of the microcracks were healed after 15 times of ECPT,as illustrated in Fig.3d.According to our previous study,the crack healing during ECPT was mainly attributed to the thermal compressive stress and material softening or melting in the vicinity of microcrack tips which was caused by the detour of eddy current around the microcrack tips,and the squeezing effect produced by Lorentz force also played an important role in crack healing[28].When the specimen A3 and A4 were treated by ECPT for 15 times,the similar phenomenon took place.As illustrated in Figs.4 and 5,the SEM images taken from area A of specimen A3 and A4 depicted the gradual crack closure in tube specimens,suggesting a notable healing effect of ECPT on the microcrack in tube specimens.Therefore,all the microcracks in various tube specimens were almost healed in different continuous ECPT schemes.Except the microcrack morphology,the microstructures of spin formed specimens did not change obviously because the relatively low temperature of tube specimen induced by the short duration of ECPT and rapid heat dissipation,and the white second phase also exhibited no evident morphological change in Figs.3-5.

Fig.3.SEM pictures of tube specimen A2 during ECPT(a)before ECPT,(b)after 5 times of ECPT,(c)after 10 times of ECPT,(d)after 15 times of ECPT.

Fig.4.SEM pictures of tube specimen A3 during ECPT(a)before ECPT,(b)after 8 times of ECPT,(c)after 15 times of ECPT.

The microcrack healing probably caused the change of mechanical properties of as-spun tube specimen,so the uniaxial tensile test of tube specimens was conducted in order to analyze the effect of continuous ECPT on the mechanical properties of different specimens.After the ECPT,the tensile experiment and hardness result was depicted in Table 3 and Fig.6.The average yield strength,average ultimate tensile strength,average elongation and average hardness(referred to as AYS,AUTS,AE and AH for simplification of as-spun specimen A1 was 209.5MPa,313.3MPa,10.2%and 91.3 HV,respectively.After 5-5-5 times of ECPT,the AYS,AUTS,AE and AH of specimen A2 reached 220.5MPa,330.74MPa,12.4% and 94.5 HV,respectively,increased by 5.2%,5.6%,20.9% and 3.5%,respectively.After 8-7 times of ECPT,the AYS,AUTS,AE and AH of specimen A3 reached 224.4MPa,336.2MPa,13.3% and 97.2 HV,respectively,increased by 7.1%,7.3%,30.2% and 6.4%,respectively.After 15 times of ECPT,the AYS,AUTS,AE and AH of specimen A4 reached 236.3MPa,343.2MPa,14.1% and 100.7 HV,respectively,increased by 12.8%,9.6%,37.5% and 17.0%,respectively.Obviously,the ECPT in all the experimental schemes contributed to the improvement of AYS,AUTS,AE and AH,which was a consequence of microcrack healing in Mg tube specimen.Clearly,the specimen A4 implemented to 15 times continuous ECPT exhibited the higher microcrack healing level and better mechanical performance than other specimens in group A.

Table 3Mechanical properties of specimens in group A,B and C.

As shown in Fig.1f,when the ECPT was continuously performed on the tube specimen for 5 times,the maximum temperature on the cross section increased dramatically after each ECPT.After the firs ECPT,the accumulated heat was transferred from the tube specimen to surrounding environment,especially the stainless steel shaft,so the temperature of the stainless steel shaft increased while the temperature of the tube specimen decreased obviously.Since the time interval between two successive ECPT cycles was very short,the temperature of the tube specimen was not cooled down to room temperature at the onset of the next ECPT,and then the temperature reached a higher value during the next ECPT than that during the firs ECPT.As a result,the temperature of tube specimen increased gradually as the ECPT times increased due to the thermal accumulation.According to the experimental results of tensile test of as-spun workpieces subject to continuous ECPT,it could be seen that the ECPT scheme with less cooling interval improved the mechanical properties of spin formed workpiece more evidently,indicating higher healing degree of microcracks.Therefore,the thermal accumulation was beneficia to some extent for the microcrack healing during ECPT.However,it should be noted that over high heat accumulation could lead to over high temperature rise within the tubular workpiece,which may lead to the fracture of Mg alloy workpiece during the ECPT process.Fig.7 presents the facture photo of as-spun Mg-RE alloy workpiece when the workpiece undertook 5 ECPT cycles under 8kV discharging voltage.Therefore,the continuous ECPT should be controlled in the certain process parameter range to facilitate the crack healing and thus enhance the mechanical properties of as-spun workpiece.

Fig.6.Uniaxial tensile experiment results of tensile specimens after different continuous ECPT(a)tensile experiment result,(b)hardness.

Fig.7.Fractured specimen after 5 continuous cycles of ECPT under 8kV discharging voltage.

3.2.Effect of heat treatment on mechanical properties of as-spun tube

In order to further improve the mechanical properties of Mg specimen,heat treatment was applied to the tube samples in group B after 5-5-5 times of ECPT and in group C without ECPT,respectively,for comparison.The tube B1 and C1 were not subject to heat treatment,while the tube B2 and C2,tube B3 and C3,tube B4 and C4 were subject to solution treatment,solution/aging treatment and aging treatment,respectively.After heat treatment,the tensile specimens were taken from the tube samples for uniaxial tensile experiment,and the experiment results were illustrated in Fig.8 and Table 3.The AYS,AUTS,AE and AH of specimens in group C1 was 209.1MPa,314.2MPa,10.2% and 91.3 HV,respectively,while the measurement values of specimens in group B1 increased to 227.2MPa,329.1MPa,12.5% and 94.5 HV respectively after ECPT.After solution treatment,the AYS,AUTS and AH of specimens in group B2 decreased to 151.5MPa,247.1MPa and 86.6 HV respectively,while the elongation increased to 20.0%.After solution/aging treatment,the AYS,AUTS and AH of specimens in group B3 reached 211.2MPa,335.0MPa and 92.3 HV,respectively,while the elongation decreased to 16.4%.After aging treatment,the AYS,AUTS and AH of specimens in group B4 evidently increased to 261.3MPa,365.2MPa and 126.2 HV,respectively,while the elongation decreased slightly to 11.1%.

In addition,the spin formed tube samples in group C without ECPT were also subject to heat treatment and their mechanical properties were compared with tube samples in group B.As shown in Fig.8,heat treatment influence the mechanical properties of the specimens without ECPT in a similar way to that of the specimens with ECPT,but the AYS,AUTS,AE and AH of the specimens in group C2,C3 and C4 were slightly lower than that of the specimens in group B2,B3 and B4,respectively.Besides,the enhancement of the mechanical performance due to heat treatment was also compared.After solution treatment,the decrease of AYS,AUTS and AH of specimen C2 were 70.2MPa(reduced by 33.5%,the same below),74.2MPa(23.6%)and 7.6 HV(8.3%),respectively,compared to specimen C1,while the absolute elongation increment of specimen C2 was 5.0%(relative increment was 5.1%).Meanwhile,the decrease of AYS,AUTS and AH of specimen B2 were 78.1MPa(34.4%),83.9MPa(25.5%)and 7.9 HV(8.7%),respectively,compared to specimen B1,while the absolute elongation increment of specimen B2 was 6.0%(relative increment was 7.5%).So the strength of specimen B2 with ECPT decreased more significantl than that of specimen C2 without ECPT after solution treatment,but the elongation of specimen B2 with ECPT increased more evidently than that of specimen C2 without ECPT.

Fig.8.Uniaxial tensile experiment results of specimen subject to ECPT and heat treatment(a)yield strength,(b)ultimate tensile strength,(c)elongation,(d)hardness.

After solution/aging treatment,the decrease of AYS,AUTS and AH of specimen C3 were 30.8MPa(14.7%),16.0MPa(5.1%)and 2.5 HV(2.7%),respectively,compared to specimen C1,while the absolute elongation increment of specimen C3 was 2.5%(relative increment was 24.5%).For specimen B3 with ECPT,the decrease of AYS,AUTS and AH were 18.4MPa(8.1%),4.0MPa(1.2%)and 2.2 HV(2.3%),respectively,compared to specimen B1,while the absolute elongation increment of specimen B3 was 3.3%(relative increment was 26.4%).It indicates that the strength of specimen B3 with ECPT decreased less obviously after solution/aging treatment than that of specimen C3 without ECPT,and the elongation of specimen B3 increased more obviously than that of specimen C3.The situation was similar to the specimens after aging treatment.After aging treatment,the increase of AYS,AUTS and AH of specimen C4 were 23.3MPa(11.1%),21.6MPa(6.9%)and 16.0 HV(17.5%),respectively,compared to specimen C1,while the absolute elongation increment of specimen C4 was 2.5%(relative increment was 24.5%).Under the same treatment process,the increase of AYS,AUTS and AH of specimen B4 with ECPT were 31.7MPa(14.0%),34.2MPa(10.4%)and 31.7 HV(33.5%)respectively,compared to specimen B1,while the absolute elongation value of specimen B4 was 1.8%(relative increment was 14.4%).Thus,the strength improvement of specimen B4 with ECPT was more remarkable than that of specimen C4 without ECPT after aging treatment,and the elongation decrease of specimen B4 with ECPT was less evident than that of specimen C4 without ECPT after aging treatment.

3.3.Microstructure evolution during ECPT and heat treatment

3.3.1.Spin formed microstructure

The microstructures of the specimens before and after heat treatment were observed through OM,SEM,EBSD and TEM to explore the cause of mechanical-property change.Fig.9 presents the microstructure of as-spun specimen B1 and Table 4 lists the average grain size of specimens in different groups.The optical microstructure and EBSD micrograph of specimen B1 in Fig.9a and b shows that the grain size distribution was relatively homogenous,and the average grain size was measured as 12.5μm.Obviously,the fin recrystallization structure existed in the Mg matrix of hot spinning tube workpiece.As depicted in the SEM pictures,amounts of white secondary phase were distributed continuously in the Mg matrix(see Fig.9c),which was a RE-rich phase with the actual chemical composition of Mg-4.15Y-3.10Gd-5.13Zn according to the EDS result(the point P1 in Fig.9d).The TEM bright fiel image in Fig.9f and diffraction pattern in Fig.9g further indicate that the continuous phase was 14H LPSO(long period stacking ordered)phase with the nominal composition of Mg12(Y,Gd,Zn)[31,32].

Fig.9.Microstructure and composition of specimen B1(a)metallographic picture,(b)EBSD picture,(c)SEM picture,(d)SEM picture of LPSO,(e)SEM picture of RE-riched particles and crushed RE phase,(f)TEM bright fiel picture of LPSO,(g)diffraction pattern of LPSO,(h)TEM picture of crushed RE phase,(i)diffraction pattern of crushed RE phase.

Table 4The average grain size and standard deviation of specimens in different groups.

In addition,some spherical RE phase particles were observed in spin-formed tube specimen(the point P4 shown in Fig.9e).Through the TEM analysis(in Fig.9h),the crushed RE phase was identifie as Mg5(Gd,Y,Zn)phase with face-centered cubic structure whose lattice constant wasa=2.223nm[33].Besides,many cuboid-shaped[33].Besides,many cuboid-shaped particles gathered around the continuous LPSO phase,while few of them were scattered over the Mg matrix(the point P2 shown in Fig.9e).The EDS result shows that the cuboid-shaped particles were rich in Y and Gd elements.Meanwhile,some spherical Zr-rich particles were located in the Mg matrix(P3 in Fig.9c).In fact,the Zr element was insoluble in Mg matrix and did not involve in the formation of LPSO phase,but it acted as a number of nucleation particles during the solidificatio process,leading to grain refinemen of Mg-RE alloy.It is noteworthy that the spin-formed tube specimen had a relatively low dislocation density owing to the occurrence of fully dynamic recrystallization[28],but the dislocation density would dramatically increase after several times of ECPT due to ECPT induced impact deformation in tube specimens,as shown in Fig.9h.

3.3.2.Solution treated microstructure

Fig.10 shows the microstructure of specimen B2 after solution treatment.The optical photographs in Fig 9a indicate that the grains grew up to the average size of 45.3μm after solution treatment(see Table 3).And more cuboid-shaped RE-rich particles appeared with a linear distribution in the rolling direction of the matrix after solution treatment(see Fig.10b and c).In addition,the spherical Zr particles were also found in the Mg matrix(see Fig.10d),while the crushed RE phase disappeared because of their dissolution into Mg matrix.After solution treatment,it is obvious that the dislocation density of specimens B2 with ECPT notably decreased(see Fig.10e)in contrast to the specimen B1(see Fig.9f).But the specimen C1 without ECPT had a relatively low dislocation density,and the change of dislocation density before and after solution treatment was not so obvious,leading to less evident softening effect than specimen B2.As a result,the strength decrease and ductility improvement of specimen B2 with ECPT were more evident than that of specimen C2 without ECPT.

Fig.10.Microstructure of specimen B2 after ECPT and solution(a)metallographic picture,(b)SEM picture,(c)RE-rich phase,(d)Zr phase,(e)TEM picture.

As is known to all,the solution treatment for as-cast Mg alloy could improve the ultimate tensile stress and elongation[34].This could be ascribed to the dissolution of solute atoms into Mg lattices,and a considerable amount of lattice distortion was generated because of the difference between elasticity modulus and atom radius of Mg atom and solute atom,resulting in the formation of the elastic strain fiel acting as strong obstacles to the dislocation motion.As a result,the solution treatment could increase the strength of Mg alloy,especially for Mg-RE alloy wherein the solid solubility of most RE elements in Mg matrix was in a relatively high degree.However,the grain growth may occur during solution treatment,and excessive solution temperature or too long holding time would cause grain coarsening and thus degrade the performance of materials.In this paper,the onset and peak temperatures of endothermic reaction of Mg-6Gd-3Y-0.5Zn-0.3Zr alloy were 521.5°C and 526.8°C,respectively,so the solution temperature was determined as 500°C.The fin grains of Mg alloy specimen caused by hot spinning grew up notably after solution treatment,and the grain coarsening dramatically decreased the strength of Mg alloy.Consequently,the combination effect of solution strengthening,grain growth and lower dislocation density during solution treatment decreased the strength and increased the elongation of as-spun Mg alloy in this study.

3.3.3.Solution/aging treated microstructure

Usually,the solution treatment,followed by aging treatment,could improve the strength of specimen but decrease the plasticity to some extent[34].As shown in Fig.11a,the grain size of specimen B3 after solution/aging treatment was similar to the specimen B2 after solution treatment(see Table 3).A large amount of RE-rich phase particles(P1 in the Fig.11c)scattered over the Mg matrix,as illustrated in Figs.11b and 10c,and some spherical and rod-shaped Zr particles(P2 in the Fig.11d)were also observed on the Mg matrix.

On the other hand,the solid solubility of RE elements in magnesium alloys decreased with the decrease of temperature,forming dispersive precipitates during aging treatment,as illustrated in Fig.11f.The TEM analysis shows that the main precipitated phase wasβ’phase[35],as shown in Fig.11e.The precipitated phase was a c-axis orthogonal structure(a=0.642nm,B=2.224nm,C=0.521nm),which pinned and hindered the dislocation movement,thus increasing the strength of magnesium alloy while reducing the elongation of the material.Combined with the effect of solution and aging treatment,the strength of magnesium alloy tube specimen B3 was slightly lower than that of the as-spun tube specimen B1,but the elongation increased slightly higher than that of the untreated specimen B1.

3.3.4.Aging treated microstructure

Actually,aging treatment can also be applied to magnesium alloys directly after hot deformation to improve their strength.Fig.12 shows the microstructure of specimen B4 after aging treatment.The grain size of specimen B4 did not change obviously compared with specimen B1(see Table 3).There were continuous LPSO phases(A,B regions)in the magnesium alloy matrix after aging treatment(Fig.12b).According to the local magnificatio picture,a large number of RE-rich precipitated phase particles were formed near the LPSO phase(Fig.12c).Large spherical Zr particles appeared in the B region of magnesium matrix(P1 point in Fig.12d).

Fig.11.Microstructure of specimen B3 after ECPT and solution/aging treatment(a)metallographic pictures,(b)SEM picture,(c)RE-rich phase,(d)Zr phase particle,(e)diffraction pattern of β’phase,(f)high resolution image of β’phase.

Without solution treatment,the crushed RE phase particles(P2 point in Fig.12c)were evenly distributed in the magnesium alloy matrix.The dispersion of crushed RE phase particles was also observed in TEM image of Fig.12f.Meanwhile,aging treatment for specimen B4 reduced the solid solubility of RE elements in magnesium alloys and led to the formation of the same dispersedβ’precipitates as in specimen B3(see Fig.12e),which could pin and hinder the dislocation movement,thereby increasing the strength and reducing the ductility of magnesium alloys.

3.4.Mechanism of different mechanical properties of as-spun tube with and without ECPT during heat treatment

Fig.13 shows the microstructures of the specimens in group C without EPCT after heat treatment.As illustrated in Fig.13a,the average grain size(47.8μm)of specimen C2 was similar to that of specimen B2 with ECPT(see Table 3).The SEM pictures in Figs.13b and 12c show that a large number of RE-rich phase particles scattered over the matrix of solution-treated specimen C2,and the RE-rich phase particles were approximately linearly distributed along the axial direction of tube workpiece.No crushed RE phase particles were observed in the matrix because the dissolution of crushed RE phase Mg5(Gd,Y,Zn)into Mg matrix occurred during solution treatment.The metallographic picture of specimen C3 treated by solution/aging treatment indicates that the grain size of specimen C3 without ECPT was similar to that of specimen B3 subject to ECPT,as listed in Table 3.The SEM images indicate that there were still a large number of RE-rich phase particles in the Mg matrix,which was similar to that of specimen B3 treated by ECPT.The metallographic image of aging-treated specimen C4 in Fig.13g shows that the grain size was close to that of specimen B4 with ECPT(see Table 3).As the SEM images depict in Fig.13h and i,the LPSO phase was continuously distributed in the matrix,and some RE-rich phase particles appeared near LPSO phase,which was also similar to that of specimen B4.Therefore,the microstructures of specimens in group C after heat treatment were similar to those of specimens in group B treated by ECPT,thus it was difficul to clarify why the strength improvement of specimens with ECPT was more significan than that of specimen without ECPT and the elongation decrease of specimens with ECPT was less evident than that of specimens without ECPT after aging treatment.

Fig.13.Microstructures of specimens in group C without ECPT after heat treatment(a-c)solution specimen C2,(d-f)solution/aging specimen C3,(g-i)aging specimen C4.

The SEM observation was carried out on the microcrack areas of the specimens after heat treatment.Besides,the distribution of Gd and Y elements around the microcracks was analyzed through EDS because the Gd and Y elements had a relatively high proportion than other RE elements in RE magnesium alloys.As illustrated in Fig.14,the distribution of Gd and Y elements in the specimen B1 after ECPT and B2 subject to both ECPT and solution treatment was fairly uniform.However,after solution/aging treatment,the content of RE elements was obviously greater at the crack edges than the neighbor areas of specimen B3.The accumulation of Gd and Y elements can also be observed on the crack surface of specimen B4 after aging treatment.The segregation of precipitates on the surface of cracks was beneficia to the fillin and closure of small-width microcracks[36].After multiple times of ECPT,the microcracks in magnesium alloy as-spun tubes were very narrow(as shown in Figs.2-4),so the fill ing of precipitates should promote the healing of microcracks more effectively for specimens in group B,which may explain the reason why the improvement of the mechanical property caused by aging treatment were more evident for specimen B3 and B4 with ECPT than that of specimen C3 and C4 without ECPT,respectively.

According to the above experiment results,the microstructure evolution process of as-spun specimens with and without ECPT during heat treatment could be described in Figs 14 and 15,respectively.As shown in Fig.15,the microstructure of as-spun specimen without ECPT possesses a low dislocation density due to sufficien dynamic recrystallization during hot spinning(Fig 14a).The grain size dramatically increased after solution treatment,resulting in the obvious decrease of strength(Fig.15b).During the subsequent aging treatment,the precipitated phase appears in the matrix and segregates in the vicinity of the crack surface(Fig.15c),which leads to the strengthening of tube specimen.As far as the ECPT specimen is concerned,the microcracks are narrowed and the dislocation density of specimen is increased after several times of ECPT(Fig.15d).After solution treatment,the grain size of ECPT specimen also dramatically increases,but the dislocation density decreases more evidently(Fig.15e),thus the strength drops more obviously than the as-spun specimen without ECPT.The subsequent aging process also leads to the segregation of precipitated phase around the microcracks,thus narrowed microcracks are easier to be fille by precipitated phase particles and probably be healed.As a result,more significan strength improvement appeared in ECPT specimen rather than as-spun specimen after solution/aging treatment.

Fig.14.The SEM and EDS analysis result of specimens after ECPT and heat treatment(a)SEM picture of B1,(b-c)Gd and Y distribution in B2,(d)SEM picture of B2,(e-f)Gd and Y distribution in B2,(g)SEM picture of B3,(h-i)Gd and Y distribution in B3,(j)SEM picture of B4,(k-m)Gd and Y distribution in B4.

Fig.15.The crack healing mechanism of solution/aging treatment on ECPT and as-spun specimen(a)As-spun specimen,(b)solution treatment on the specimen in(a),(c)aging treatment on the specimen in(b),d)ECPT specimen,(e)solution treatment on the specimen in(d),(f)aging treatment on specimen in(e).

Fig.16.The crack healing mechanism of the aging treatment on ECPT and as-spun specimen(a)as-spun specimen,(b)aging treatment on the specimen in(a),(c)ECPT specimen,(d)aging treatment on the specimen in(c).

When only the aging treatment is employed to treat the specimens,the precipitated phase segregation also occurred around the microcracks,but relatively low temperature during aging treatment is difficul to change the grain size and dislocation density,as shown in Fig.16.Due to narrower microcracks after the ECPT process(Fig.16b and d),the strength improvement of ECPT specimen is more obvious than that of as-spun specimen because of higher microcrack healing effect.Besides,both the specimens with and without ECPT showed better performance in strength than that of the specimens under solution/aging treatment,respectively.

4.Conclusion

The as-spun tube specimens of Mg-RE alloy with microcracks were treated by eddy current pulse,and the influenc of heat treatment was also analyzed in this study.The main conclusions can be summarized as follows:

(1)After different continuous ECPT schemes up to 15 cycles,all the microcracks in different tube specimens were almost healed.The schemes with less cooling intervals showed better healing effect and improved the strength and elongation of tube specimens more evidently,indicating that the thermal effect was beneficia for the crack healing during the ECPT process.

(2)The combination of solution strengthening and grain growth decreased the strength while increased the elongation of specimen.Aging treatment improved the strength while reduced the elongation.Solution/aging treatment was the combination effect of solution and treatment,leading to the strength decrease and slight elongation increase.

(3)After solution treatment,both the dislocation density of tube specimens with and without ECPT decreased remarkably.But the specimens without ECPT had relatively low dislocation density,and the dislocation density change before and after solution treatment was not so obvious,leading to less evident softening effect than the specimens with ECPT.As a result,the strength reduction and ductility improvement of specimen with ECPT were more evident than that of specimens without ECPT during solution treatment.

(4)The segregation of Gd and Y elements occurred in the vicinity of the crack surface after aging treatment.The segregation of precipitates on the surface of cracks was beneficia to the fillin and closure of microcracks.After multiple times of ECPT,the microcracks in as-spun magnesium alloy tubes were very narrow,so the fill ing of precipitates should promote the healing of microcracks more effectively,which was the reason why the mechanical property improvement under aging treatment were more evident for specimens with ECPT than that of specimens without ECPT.

Declaration of Competing Interest

The authors declare no conflic of interest.

Acknowledgment

The authors would like to thank the National Natural Science Foundation of China(Nos.51775137 and 51635005)for its support.

Author Contributions

Wenchen Xu and Chuan Yang conceived and designed the experiments;Wenchen Xu,Chuan Yang,Haiping Yu,Xueze Jin and Guojing Yang performed the experiments and analyzed the data;Bin Guo,Debin Shan provided guidance and all sorts of support during the work.