Comparative study on high temperature deformation behavior and processing maps of Mg-4Zn-1RE-0.5Zr alloy with and without in-situ sub-micron sized TiB2 reinforcement

S.K.Sahoo,S.K.Panigrahi,*

a Department of Mechanical Engineering,Indian Institute of Technology Madras,Chennai-600036,India

b Applied Magnesium Research Group,Center of Excellence in Materials and Manufacturing for Futuristic Mobility,Indian Institute of Technology Madras, Chennai-600036,India

Abstract Mg-4Zn-1RE-0.5Zr(ZE41)Mg alloy is extensively used in the aerospace and automobile industries.In order to improve the applicability and performance,this alloy was engineered with in-situ TiB2 reinforcement to form TiB2/ZE41 composite.The high temperature deformation behavior and manufacturability of the newly developed TiB2/ZE41 composite and the parent ZE41 Mg alloy were studied via establishing constitutive modeling of fl w stress,deformation activation energy and processing map over a temperature range of 250 °C-450 °C and strain rate range of 0.001 s-1-10 s-1.The predicted fl w stress behavior of both materials were found to be well consistent with the experimental values.A significan improvement in activation energy was found in TiB2/ZE41 composite(171.54 kJ/mol)as compared to the ZE41 alloy(148.15 kJ/mol)due to the dispersed strengthening of in-situ TiB2 particles.The processing maps were developed via dynamic material modeling.A wider workability domain and higher peak efficien y(45%)were observed in TiB2/ZE41 composite as compared to ZE41 alloy(41%).The Dynamic recrystallization is found to be the dominating deformation mechanism for both materials;however,particle stimulated nucleation was found to be an additional mode of deformation in TiB2/ZE41 composite.The twinning and stress induced cracks were observed in both the materials at low temperature and high strain rate.A narrow range of instability zone is found in the present TiB2/ZE41 composite among the existing published literature on Mg based composites.The detailed microstructural characterization was carried out in both workability and instability domains to establish the governing deformation mechanisms.? 2022 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:ZE41 Mg composite;In-situ TiB2 particles;ZE41 Mg alloy;Constitutive equations;Processing map;Dynamic recrystallization.

1.Introduction

Magnesium(Mg)matrix in-situ reinforced composites have attracted enormous interest in automobile and aerospace industries because of their inherent properties like high specifi strength,high specifi stiffness,good thermal stability and high wear resistance[1-4].Among various cast Mg alloys,the rare earth contained Mg alloys have proven as potential materials in several applications[5].ZE41(Mg-4Zn-1RE-0.5Zr;where RE=rare earth)is a commercial rare earth contained Mg alloy used for aircraft gearbox and generator housings,particularly in military helicopters[6].The performance,mechanical properties and tribological properties of ZE41 alloy can be improved significantl by reinforcing in-situ/ex-situ particles into ZE41 matrix.Among the conventional in-situ ceramic particles,Titanium diboride(TiB2)particles exhibit superior physical and mechanical properties for the development of Mg metal matrix composites(MMCs)because of high elastic modulus,high hardness,better thermal stability,and wear resistance[7].Apart from this,TiB2particles act as a grain refine for Mg metal matrix composites.In Mg matrix composites,researchers found that TiB2particles reinforced by the in-situ route exhibited superior mechanical properties than reinforced by the ex-situ process[8-10].Therefore insitu TiB2reinforced ZE41 Mg composite will be a prominent material for automobile and aerospace industries.

The addition of TiB2particles into soft magnesium alloy matrix often causes metallurgical defects during liquid state processing routes like casting.Generally,as cast particle reinforced Mg matrix composites have coarse grains and hardly homogeneous microstructure can be obtained.A secondary bulk processing is necessary to tackle these problems.The additional secondary bulk plastic forming,such as forging,rolling,and extrusion is required to refin microstructure in order to enhance the mechanical properties.However,limited number of slip systems due to hexagonal close packed(HCP)crystal structure of Mg and the addition of high modulus brittle ceramic reinforcements limits plastic forming ability of Mg composites at room temperatures.Therefore,to improve workability,magnesium alloys and their composites are usually deformed at high temperatures,by which the additional slip systems are activated[11].It is necessary to understand the high temperature deformation behavior of newly developed materials prior to the implementation of any kind of secondary thermo-mechanical/manufacturing processes.Many researchers have used constitutive models to describe the fl w behavior material with strain rate,stress,and deformation temperature[12-16].Processing map is used to evaluate the processing window for any kind of alloys and composites.The dynamic material model(DMM)is the most popular model to develop processing map.Based on the processing map,stable and unstable zones for a newly developed material can be identified

The approach of processing map has been used by many researchers to fin optimum processing windows for Mg based alloys and composites.The primary hot deformation mechanisms causing the workability and instability in the material are identifie based on the microstructural observation of deformed materials.Many researchers found the workability domains for the processing maps of Mg based composites at deformation temperatures greater than 300 °C.They observed continuous dynamic recrystallization(CDRX)and discontinuous dynamic recrystallization(DDRX)are the hot deformation mechanisms for workability domains based on the microstructural characterization[17,18].During the hot deformation of Mg based composites,the dislocation pile-up generally forms near the particles.This leads to the formation of particle deformation zone,which is favorable for the DRX nucleation.The hot deformation mechanism can be predicted through the activation energy(Q)and the stress exponent(n)through constitutive equations[20].The hot deformation activation energy of the SiCp/AZ91 composite[21]is found to be close to the value for grain boundary diffusion in pure Mg,whereas,with the addition of nano-sized in-situ TiB2particle,a much higher activation energy is found in TiB2/AZ91 composite as compared to that of the unreinforced AZ91 Mg alloy.The deformation mechanism based on the constitutive equation for Mg matrix composites is found to be the climb of dislocation[21,22].

Though several researchers have studied high temperature deformation behavior of Mg based composites via processing map,till now none of the researchers have established the processing map of ZE41 Mg based composite.In our previous research work,the manufacturing methodology to develop sub-micron sized in-situ TiB2reinforced ZE41 Mg MMCs(TiB2/ZE41)along with the scientifi knowhow was highlighted[23].The development of processing map and constitutive model via hot deformation for a novel in-situ sub-micron sized TiB2/ZE41 composite was carried out in this work for the firs time.Apart from these,none of the researchers have studied the comparison of processing map between any form of in-situ Mg composite vs their parent unreinforced counterpart so far.In the current work a firs attempt was carried out to compare the manufacturability via processing map of an in-situ Mg composite(TiB2/ZE41 composite)vs it’s unreinforced counterpart(ZE41).

The prime objectives of this present research work are(i)to investigate the high temperature deformation behavior of both ZE41 Mg alloy and in-situ sub-micron sized TiB2/ZE41 composite and study the influenc of in-situ TiB2particles on hot workability,(ii)to fin optimum hot working conditions for manufacturability via processing map,(iii)establishment of constitutive based models to describe the fl w behavior,(iv)to understand the deformation mechanisms via microstructural characterization in both stable and unstable zones of processing maps.

2.Experimental procedure

2.1.Fabrication of TiB2/ZE41 composite

A commercial ZE41 rare earth contained magnesium alloy was taken as a matrix material for the development of insitu composite.Titanium(Ti)and Boron(B)powders were used as the starting alloying elements with a weight fraction of 0.68:0.32 for the development of in-situ TiB2reinforcement.The initial metallic powder system and it’s weight fraction were identifie based on the thermodynamic calculation.The reaction temperature for the formation of TiB2phase was found by the differential thermal analysis(DTA)analysis as 900 °C.The Ti and B powers were mixed in a planetary ball milling and the milling was carried out for an optimized time period of 8 h at a milling speed of 350 rpm to form Ti+B clusters.The preheated Ti+B cluster powders were fille into the custom designed drilled holes of ZE41 Mg ingots slices and were placed inside the crucible of casting unit.The melt temperature was increased to 900 °C and held for 2 h after stirring.The Ti+B powder clusters were fragmented and diffused with each other to form in-situ TiB2in the molten ZE41 matrix.The pouring of the melt was done by a custom designed bottom pouring system in a preheated rectangular mold.The entire fabrication process was carried out under the argon gas environment to avoid any chances of fir and oxidation.The detailed fabrication process and in-situ formation mechanism of the developed TiB2/ZE41 composite were reported in the author’s earlier work[23].Similarly,ZE41 Mg alloy was casted to compare with TiB2/ZE41 composite.The weight fraction of TiB2reinforcement was 10 wt%(i.e.4.23 vol fraction)for the developed composite.The size of in-situ reinforcement was found in the range of 400 nm to 1.5μm.The average TiB2particle size was found to be 765 nm.To eliminate the casting defects like porosity,the developed base alloy and composite were partially warm rolled.The samples were heat-treated at 330 °C for 2 h before hot compression.

Fig.1.Schematic representation of the hot compression test.

2.2.Hot compression test

Hot compression tests of unreinforced ZE41 and in-situ TiB2reinforced ZE41 composite were carried out in a high precision INSTRON 3365,which has an integrated high temperature furnace.The hot compression cylindrical samples were cut from both ZE41 alloy and composite as per the ASTM-E9 standard.To ensure the repeatability of results,three samples in each condition were tested.The hot compression was performed at different deformation temperatures of 250,300,350,400 and 450 °C,and the strain rates of 0.001,0.01,0.1,1 and 10 s-1with a constant true strain of 0.4.The samples were heated to the deformation temperature and kept for 5 min before hot compression to get uniform distribution of temperature throughout the sample.The samples were quenched in cold water immediately after hot compression to preserve the deformed microstructure for further analysis.Fig.1 shows the schematic of hot compression process.

2.3.Microstructural characterization

The microstructural analysis of unreinforced ZE41 and TiB2/ZE41 composite were studied by using scanning electron microscopy(SEM),optical microscopy(OM)and transmission electron microscopy(TEM).The samples were cut parallel to the compressive loading axis and mounted before polishing.OM and SEM samples were etched by using picric acid after mechanical polishing with emery sheets of different grades followed by diamond polishing of 1 and 0.5μm.The samples for TEM characterization were mechanical polished till the thickness reach to 80μm with emery sheets of different grades and then ion milled.The TEM was carried out by using TECHNAI electron microscopy at 200 kV.

3.Constitutive equations and processing map

3.1.Constitutive equations

Generally,in hot working processes,constitutive equations based on mathematical models are significan to describe the relationship between processing.There are many constitutive mathematical models used by researchers for the prediction of fl w stress in hot deformation of a material.The Arrhenius model is one of the famous model to describe the thermo mechanical behavior of a material based on the activation energy[24,25].There are three types of power law relationship applied in hot deformation,such as(i)power law for low level stress,(ii)exponential law for high level stress and(iii)hyperbolic-sine law for a wide range of stress[26-29].The equations are as follows:

where,˙εis the strain rate(s-1);σis peak stress(MPa);Ris the universal gas constant(8.314 Jmol-1K-1);Qis the activation energy(kJ mol-1);Tis the temperature(K);nis the stress exponent.A1,A2,n1,β,αare the material constants.The termαis a stress multiplier and can be expressed asβ/n1[27].The use of above equation in hot deformation depends upon the stress level of the material.Generally,the power law relationship Eq.(1))is suitable when the stress level is low(ασ＜0.8),the exponential relationship(Eq.(2))is used in high stress level(ασ＞1.2),and the hyperbolic sine relationship(Eq.(3))is applied when there is a wide range of stress[25].Therefore,for metal matrix composites,a hyperbolic sine constitutive equation is used.To calculate the above constants,natural logarithm is taken on both sides of Eqs.(1)to(3)and expressed as follows:

Fig.2.Microstructures of(a,c)ZE41 alloy;(b,d)TiB2/ZE41 composite;(e)Mg7Zn3RE phase;(g)sub-micron sized in-situ TiB2 reinforcement,and(f,h)EDS results of corresponding selected areas.

From Eqs.(4)and(5),the valuesn1andβof and can be calculated by the linear regression ln˙εvs lnσand ln˙εvsσof respectively.The value ofαcan be get byβ/n1.The slope of in Eq.(6)is the stress exponentnat a specifi temperature.The expression fornis as follow:

The stress exponent(n)values represents the governing mechanism for hot deformation,n=2 for grain boundaries sliding,n=3 for the viscous glide of dislocation,n=5 for the climb of dislocation andn=8 for the cross-slip screw dislocation/constant substructure model[30-32].

The apparent activation energy(Q)required for deformation can be derived from Eq.(6)as follows:wheresrepresents the slope of ln[sinh(ασ)]and 1000/Tat a constant strain rate.

Besides this,the Zener-Hollomon equation can be used to combine the effect of deformation temperature and strain rate on the fl w behavior of a material[33].Zener-Hollomon parameter(Z)is the Arrhenius function and can be expressed as follows:

After the natural logarithm of Eq.(9)and combing with Eq.(6);the equation will be as follows:

From Eq.(10),lnAis the intercept of lnZvs.ln[sinh(ασ)].By substituting the values ofn,α,Q,andA,the constitutive model for a new novel material can be developed.

3.2.Principle of processing map

The processing map of a particular material is used to understand the workability at different hot deformation conditions in terms of microstructural mechanisms.The processing map has been used widely by many industries to optimize hot deformation parameters in order to avoid any defects during deformation.The processing map can be developed based on Dynamic material modeling(DMM).It is constructed by the superimpose of instability map over power dissipation efficien y map at various deformation temperatures and strain rates.The work piece dissipate power in terms of heat while material under goes hot deformation.This power dissipation in the material results in microstructural changes like dynamic recrystallization,phase transition,and superplastic fl w,which are reflecte by the power dissipation efficien y.

According to the DMM theory[34],the dissipated power(P)can be split into two complementary parts,such asGcontent andJco-content and can be expressed as follows:

The contentGexpresses the power dissipation due to plastic deformation and most of the heat is lost in the form of deformation temperature.In contrast,the co-contentJsignifie power dissipation due to microstructural transformations like dynamic recrystallization,dynamic recovery,phase transformation.

When deformation temperature is constant,the fl w stress of a material according to power law can be expressed as:

wherekandmrepresent material constant and strain rate sensitivity respectively.

The strain rate sensitivity(m)can be calculated in terms ofJandGas follows:

Eq.(12)can be used in Eq.(11)to obtained the co-contentJas:

For ideal linear dissipation,the value ofmis equal to one.So,the J co-content reaches to the maximum asJmax=σ˙ε/2.For non-linear dissipation,the power dissipation efficien y can be represented by a dimensionless parameter(η)and is the ratio of power dissipation to the maximum power dissipation via microstructural transformation.The expression for power dissipation efficien y as follows:

The higher the efficien y is the better the workability of a material.However,in some cases,there is deformation instability such as wedge cracking,cavitation,and fracture during hot working,which needs to be avoided.So,it is necessary to fin out the unstable domain in the processing map via instability criteria.Instability criteria are expressed based on extremum principles of thermodynamic irreversibility as applied to large plastic fl w continuum mechanics.Prasad et al.combine the Ziegler’s fl w theory with DMM to establish the fl w instability during the hot deformation process[35].According to Ziegler,the plastic fl w becomes unstable when:

whereDrepresents the dissipation function,which is the characteristic of the fl w behavior of the material,Prasad replacedDwith power dissipation co-contentJand the equation becomes as:

Which implies that

Fig.3.XRD analysis of unreinforced ZE41 Mg alloy and TiB2/ZE41 composite.

By substituting the value ofJfrom Eq.(14)into Eq.(18),the fl w instability criterion can be expressed as:

whereξ(˙ε)is a dimensionless fl w instability parameter developed by Prasad.Negative valuesξ(˙ε)show the f ow instability domain or unsafe domain in the processing map.

4.Results

4.1.Initial microstructure

The microstructures prior to the hot compression test for unreinforced ZE41 and TiB2/ZE41 composite are shown in Fig.2.Both the ZE41 alloy and composite show equiaxed grains with the distribution of ternary Mg7Zn3RE phases shown by the arrow mark.Fig.2a shows the OM image of ZE41 alloy havingα-Mg and ternary Mg7Zn3RE phases.Fig.2c shows the SEM image of the ZE41 alloy.As compared to the ZE41 alloy,the ternary phases in composite material are broken(Fig.2b,14c).The average grain size for ZE41 and TiB2/ZE41 are 41±2.3μm and 26±1.9μm respectively.The presence of TiB2particle cluster in TiB2/ZE41 composite is shown in Fig.2d.The grain size of TiB2/ZE41 composite material is smaller compared to the unreinforced ZE41 alloy.The high magnifie SEM image of Mg7Zn3RE phase is shown in Fig.2e with corresponding EDS result in Fig.2f.The EDS result confirm the presence of cerium(Ce)and neodymium(Nd)as rare earth elements.Fig.2g shows the high magnifica tion SEM image of in-situ TiB2particle and Fig.2h shows the corresponding EDS result.The EDS analysis result confirm the presence of Ti and B in-situ reinforcement as elemental content.The TiB2particle is mostly hexagonal in shape.Fig.3 shows the XRD results of ZE41 Mg alloy and in-situ TiB2/ZE41 composite.The XRD peaks of ZE41 show mainlyα-Mg and the presence of ternary phase i.e.Mg7Zn3RE.The XRD spectra of TiB2/ZE41 composite reveals the presence of TiB2phase includingα-Mg,Mg7Zn3RE and MgO.

4.2.Flow behavior of ZE41 and TiB2/ZE41

The hot compression test was carried to understand the fl w behavior of both ZE41 alloy and in-situ TiB2/ZE41 composite at different temperatures and strain rates.Fig.4 and Fig.5 show the true stress-true strain curve of ZE41 alloy and TiB2/ZE41 composite at various temperatures(250-450 °C)and at different strain rates(0.001-10 s-1)respectively.The fl w stress significantl depends on hot deformation parameters(mainly temperature and strain rate).The fl w stresses of both ZE41 alloy and in-situ TiB2/ZE41 composite increased with increasing in strain rate and decrease with increase in deformation temperature.It could be seen from Figs.4 and 5 that,during initial stage of compression i.e.up to a very low true strain,the fl w stress of both alloy and composite increase steadily irrespective of the deformation temperature.This rapid increase in fl w stress is the characteristics of work hardening of the alloy and composite at the initial stage of deformation.The generation and accumulation of dislocation at the earlier stage results in increase of fl w stress in the material.From the fl w behavior of ZE41 and TiB2/ZE41,it can also be observed that,the rate of increase in fl w stress gradually decreases with progress in hot compression i.e.increase in true strain.The ZE41 Mg alloy deformed at 350 °C with strain rate of 0.001 s-1and for TiB2/ZE41 composite deformed at 300 °C with strain rate of 1 s-1show dynamic softening type fl w behavior after work hardening(Fig.4c,Fig.5b).The fl w stress in the composite is found to be higher than the unreinforced ZE41 alloy(Fig.5).This is due to the presence of fin TiB2particles.These fin particles(i)enhance strengthening effect(ii)provide better thermal stability to the material and(iii)also oppose the dislocation movement during deformation.In addition to this,the average grain size of TiB2/ZE41 composite is decreased by 37% than that of unreinforced ZE41 Mg alloy.This is due to the pinning action of TiB2particles,which restricts the grain growth in composite.According to the Hall-Petch relationship,the strength of material increases with decrease in grain size[11].Hence,the increase in fl w stress of TiB2/ZE41 is also due to the decrease in grain size.

The 3D contour map of true stress vs.temperature vs.strain rate(log scale)is shown in Fig.6 for both ZE41 alloy and TiB2/ZE41 composite.With the increase in strain rate and decrease in deformation temperature,fl w stress increases in both the materials.This shows that,strain rate and deformation temperature have significan impact on the true stress of the material.Similar fl w behavior has also been observed by many researchers for Mg alloys and it’s composites[17,28].

4.3.Constitutive analysis

Fig.4.True stress-true strain curves of ZE41 alloy at different temperatures(250 to 450 °C)and various strain rate(0.001 to 10 s-1).

Fig.5.True stress-true strain curves of TiB2/ZE41 composite at different temperatures(250 to 450 °C)and various strain rate(0.001 to 10 s-1).

The quantitative relationship between hot deformation conditions and fl w stress can be predicted for both ZE41 Mg alloy and TiB2/ZE41 composite by using the constitutive equations.The compression testing was carried out up to a true strain of 0.4 for both materials and the fl w stresses are selected at a true strain of 0.25 in order to establish the constitutive equations.The constitutive constantsn,β,andαcan be calculated from the above explained constitutive equations.From Eq.(4),n1has been calculated from the linear regression of ln˙εvs.lnσas shown in Fig.7.The value ofn1for ZE41 and TiB2/ZE41 are found to be 8.48 and 11.19 respectively.Similarly,the value ofβis obtained from the slope of ln˙εvs.σas shown in Fig.8 and are found to be 0.065 and 0.074 for ZE41 and TiB2/ZE41 respectively.Then,αfor both alloy and composite are found to be 0.0076 and 0.0067 respectively.From the linear fittin between ln˙εvs.ln[sinh(ασ)](Fig.9)the average stress exponent(n)values for both ZE41 and TiB2/ZE41 are calculated and found to be 5.96 and 8.09 respectively.So,the dominating deformation mechanism for hot compression of ZE41 alloy and TiB2/ZE41 composite are climb of dislocation and cross-slip of screw dislocation/constant substructure model respectively.The average values ofsare calculated from the linear fittin of 1000/Tvs ln[sinh(ασ)](Fig.10).The activation energy of both ZE41 and TiB2/ZE41 are calculated and the quantitative values are represented in Table 1.

Fig.6.3D contour maps of true stress vs.strain rate(log scale)and temperature for(a)ZE41 alloy and(b)TiB2/ZE41 composite.

Fig.7.Relationship between ln˙εand lnσfor(a)ZE41 alloy and(b)TiB2/ZE41 composite.

The activation energy of ZE41 is found to be 148.15 kJ/mol,whereas for TiB2/ZE41 the value is 171.54 kJ/mol.The lattice self-diffusion activation energy of pure magnesium is 135 kJ/mol[36].The ZE41 alloy has higher activation energy of than pure Mg due to the presence of a stable Mg7Zn3RE ternary phase.These ternary phases obstruct the dislocation movement by giving back stress.The activation energy of the composite is higher than the pure Mg and ZE41 alloy.These fin thermally stable in-situ TiB2particles obstruct the dislocation movement during hot deformation,which improves the activation energy of deformation.Apart from this,there is a miss match in the coeffi cient of thermal expansion(CTE)between in-situ TiB2particles and ZE41 matrix,resulting in the generation of dislocations around particles[37]and increase in activation energy.Many researchers also reported the higher activation energy of Mg matrix composites during hot deformation[22,38].

The lnAvalue was calculated from lnZand ln[sinh(ασ)]and shown in Fig.11.The lnAfor both ZE41 and TiB2/ZE41 are 26.37 and 31.07 respectively.Based on the above constitutive analysis,the constitutive parameters(n,β,andα)are calculated and represented in Table 1.

Fig.8.Relationship between ln˙εandσfor(a)ZE41 alloy and(b)TiB2/ZE41 composite.

Fig.9.Relationship between ln˙εand ln[sinh(ασ)]for(a)ZE41 alloy and(b)TiB2/ZE41 composite.

Table 1Calculated values of constitutive equation constants for both ZE41 and TiB2/ZE41 composite.

So,the constitutive equation during hot deformation of ZE41 Mg alloy can be expressed as follows:

Similarly,the constitutive equation of sub-micron sized insitu TiB2reinforced ZE41 composite during hot deformation is expressed as:

Fig.10.Relationship between 1000/T and ln[sinh(ασ)]for(a)ZE41 alloy and(b)TiB2/ZE41 composite.

Fig.11.Relationship between lnZ and ln[sinh(ασ)]for both ZE41 alloy and TiB2/ZE41 composite.

The fl w stress values can be predicted based on the developed Arrhenius based constitutive equations(Eq.(20)and Eq.(21))for both ZE41 Mg alloy and TiB2/ZE41 composite.Fig.12 represents the predicted fl w stress values and the experimental stress values.Most of the stress values for both ZE41 Mg alloy and TiB2/ZE41 composite are near the fittin line of the plot.The R2values for ZE41 and TiB2/ZE41 are 0.967 and 0.959 respectively.These high values of R2show a good correlation of experimental values with the predicted values.This indicates the proposed Arrhenius based constitutive model equations for both materials can well signify the fl w behavior.

4.4.Processing maps

The processing map for hot deformation can be developed by superimpose of instability map on power dissipation map.Fig.13 shows the processing maps of unreinforced ZE41 and TiB2/ZE41 composite at a true strain of 0.25.The contour lines represent the iso power dissipation efficien y and the higher the efficien y is,the better the workability of the material will be.It has been reported by many researchers for Mg alloys that,the efficien y greater than 30-35% is suitable for workability of materials[39].The higher values of power dissipation efficien y enhance the nucleation and growth of DRX.So,the workability domains are chosen by considering power dissipation efficien y greater than 33%.The peak effi ciency zones for both ZE41 alloy and TiB2/ZE41 composite are marked with dotted rectangles.The instability or unsafe zones are shown by shaded areas,where the fl w instability parameter is negative(ξ(˙ε)＜0)(Fig.13).

The efficien y values and processing parameters in different zones of processing map for both ZE41 alloy and TiB2/ZE41 composite are shown in Table 2.In the ZE41 alloy,the power dissipation efficien y value decreases with increase in strain rate at low deformation temperature(250-300 °C).Whereas,at moderate and high temperatures with an increase in strain rate,the efficien y value increases.There are three workability domains(Zone-I,II and III)found in processing map of ZE41 alloy.In Zone-I,the efficien y values were found to be 0.33 to 0.41 at a low strain rate(0.001-0.0055 s-1)and a temperature range of 300 °C-375 °C.The efficien y value in Zone-II is found to be 0.33 at a strain rate of 0.1 to 3 s-1(408-434 °C)and in Zone-III the peak efficien y of 0.33-0.39 is observed at a strain rate range of 0.1 to 3 s-1(350-400 °C)(Table 2).The Zone-A represents the instability domain in the processing map of ZE41 alloy(Fig.13a).The instability/unsafe domain is observed at a strain rate ranging from 0.025 to 10 s-1with a temperature range of 250-350 °C(Table.2).Fig.13b shows the processing map for in-situ TiB2/ZE41 composite.Similar to ZE41 alloy,the efficien y value decreases with an increase in strain rate at low temperatures.Two peak efficien y zones(Zone-I and Zone-II)are observed in the case of composite material.In Zone-I,the efficien y value lies in the range of 0.33-0.45 at a stain rate range of 0.001-0.008 s-1(292-384 °C)and in Zone-II,the values range is 0.33-0.39 at strain rate range of 0.0058 to 1 s-1(384-447 °C)(Table.2).These zones in composite material are considered to be the safe workability domain.The instability domains(Zone-A and Zone-B)in the composite are smaller than that of ZE41 alloy.The instability Zone-A is observed at a strain rate of 0.0675-10 s-1(250-300 °C),whereas in Zone B,the instability is found at a high strain rate(6.5-10 s-1)and high temperature(385-450 °C).

Fig.12.Comparative analysis of the experimental and predicted fl w stress data from the proposed Arrhenius based constitutive model equation for(a)ZE41 alloy and(b)TiB2/ZE41 composite.

Table 2Different zones in processing maps of both ZE41 and TiB2/ZE41 composite.

5.Discussion

The present research work emphasized on the manufacturability of novel ZE41 with and without addition of sub-micron sized in-situ TiB2particles.The hot deformation behavior was studied by developing processing map.In the processing map,each domain is related to different microstructural evolutions and deformation mechanisms.Therefore,for each workability domain and instability domain,microstructural characterization was analyzed in depth for both ZE41 and TiB2/ZE41 composite to understand the hot deformation mechanism discussed in the following section.

5.1.Workability domain

The maximum power dissipation efficien y zones in the processing map represent the workability domain for a particular material with certain hot deformation parameters.There are three workability domains(Zone-I,Zone-II and Zone-III)found for ZE41 and two workability domains(Zone-I and Zone-II)in case of TiB2/ZE41 composite(Table 2).Each zone of workability domains is discussed as follows:

5.1.1.Zone-I

The Zone-I workability domain possess at medium deformation temperature range(290-385 °C)and at slow strain rate(0.001 s-1)for both the materials(Table.2).However,the Zone-I in composite is wider(almost 1.5 times)than un-reinforced ZE41 alloy.At this workability zone,both base and composite exhibit peak power dissipation efficiencies So,the microstructures at this zone(350 °C and 0.001 s-1)have equiaxed DRX grain structure for both base and composite(Fig.14a,b).Theα-Mg of ZE41 alloy is significantl refine(around 80%)with equiaxed fin grains and average grain size of 8μm(Fig.14c).The ternary Mg7Zn3RE phases are fragmented(Fig.14e).The microstructure of composite also reveals equiaxed fin grains with well-define grain boundaries.The composite microstructure shows the distribution of Mg7Zn3RE phases with some TiB2particle clusters(Fig.14f).The EDS results gives the evidence of presence of phases and particles(Fig.14g,h).These particle clusters are present mainly around the grain boundaries.As compared to the grain size of deformed ZE41 in Zone-I,the average grain size of composite found to be smaller(～6μm)(Fig.14d).The decrease in grain size is mainly due to the in-situ particle pinning effect on grain boundaries which limit the grain growth.The significan refinemen ofα-Mg grains in Zone-I hot deformation condition is primarily due to the fully dynamic recrystallization of material.The DRX is a thermally activated process[41].The extent of DRX depends on temperature and time.In zone-I deformation condition(300-384 °C;0.001-0.008 s-1),the combined action of slow strain rate and moderate temperature helps the material fully recrystallize.For Mg based alloys and composites,the DRX is often predominant due to the activation of dislocation in non-basal slip systems and grain boundary migration at lower strain rate(＜0.01 s-1)and relatively higher temperature(＞320 °C)[41,42].The fin grain structure with random orientation in zone-I is mainly due to dislocation clime and cross-slip,which provides nuclei for continuous dynamic recrystallization(CDRX).In addition to this,the value of stress component(n)for ZE41 and TiB2/ZE41 are 5.9 and 8.09(Table.1),which indicates the hot deformation is mainly controlled by the climb of dislocation for ZE41 and cross-slip of screw dislocation for TiB2/ZE41.The CDRX is the primary hot deformation mechanism for the zone-I deformation condition.

Fig.13.Processing maps for(a)ZE41 alloy and(b)TiB2/ZE41 composite.

The TEM image of ZE41 alloy deformed at Zone-I(350 °C,0.001 s-1)reveals the presence of DRX grains(Fig.15a).The segregation of Ce,Nd rare earth content in term of T-Mg7Zn3RE phases effectively reduce the DRX grains size by solute drag effect during hot deformation[43].Apart from this,Ce,Nd content rare earth phases(Mg7Zn3RE)have better thermal stability and creep properties at this deformation temperature range.Meanwhile,the activation energy during hot deformation has also increased to 148 kJ/mol which is higher than the self-diffusion activation energy of pure Mg(135 kJ/mol).This is due to the presence of rare earth content T phase present in the matrix,which hinder the dislocation motion.The TEM observation shows the dislocation accumulation around the Mg7Zn3RE phase(Fig.16a).It also observed that,there is pile-up of dislocations around this rare earth phase and these phases act as barrier to the dislocation movement(Fig.16a),which helps to promote DRX in ZE41.However,in composite,along with Mg7Zn3RE phases,the presence of sub-micron sized in-situ TiB2particles helps to generate and accumulate more dislocations than ZE41 alloy.Due to the mismatch in the CTE between in-situ TiB2and matrix alloy,a strain fiel is created and hence dislocation clusters are formed around the particles.In addition to this,these in-situ particles act as barrier to dislocation movement,which results in additional pile-up of dislocations in TiB2/ZE41(Fig.16b).The insert image in Fig.16b shows the selected area electron diffraction(SAED)pattern of TiB2particle along zone axis of[0-111].This results in the formation of high dislocation density at the vicinity of particles.These regions are known as particle deformation zone(PDZ)[44]and is significan when the particle size is in the sub-micron or micron level.The DRX and nucleation of new grains are promoted with these high energy stored in PDZ and this mechanism is called as particle stimulated nucleation(PSN)[44].The microstructure reveals the presence of fin DRX grains surrounded by the in-situ TiB2particle(Fig.15b).Under similar deformation conditions(350 °C;0.001s-1)Zhou et al.[45]have observed fin DRX grains for AZ91-5SiC-0.5CNTs composite.They have also found the growth of these fin DRX grains were restricted due to the pinning action of particles.Several researchers have also reported that with the addition of fin reinforcements in Mg metal matrix composite,the nucleation of DRX is also promoted[19,46].

Fig.14.OM image,grain size distribution plot,SEM image and EDS result of samples deformed at 350 °C;0.001 s-1 for;(a,c,e,g)ZE41 alloy;and(b,d,f,h)TiB2/ZE41 composite.

Fig.15.TEM microstructures of samples deformed at 350 °C;0.001s-1(a)ZE41 alloy and(b)TiB2/ZE41 composite.

Fig.16.TEM images of dislocations arrested by(a)Mg7Zn3RE in ZE41 alloy and(b)in-situ TiB2 particles in TiB2/ZE41 composite;(c)TEM images for fragmentation of phase due to particle.

5.1.2.Zone-II

Fig.17.OM image,grain size distribution plots,SEM image and EDS results of(a,c,e,g)ZE41 alloy at 425 °C,1 s-1 and(b,d,f,h)TiB2/ZE41 composite at;425 °C,0.1 s-1.

Another workability domain i.e.Zone-II was found for hot deformation of both ZE41 alloy and TiB2/ZE41 composite at high temperature range(385-447 °C)and medium strain rate(Table.2).As compared to the workability zone-I,the power dissipation efficien y is lower for both materials.The power dissipation efficiencie are 0.33 and 0.33-0.39 for base and composite respectively.But,the zone-II of composite is much wider(almost twice)than base alloy.This is due to the existence of thermally stable in-situ TiB2particles in the material.The microstructure of ZE41 deformed at 425 °C and 1 s-1reveals equiaxed grain and the Mg7Zn3RE phases are mostly present in grain boundaries(Fig.17a,e).However,the average grain size for ZE41 at this zone is 31μm(Fig.17c),which is much higher than that of zone-I,where as in composite it is 20μm(Fig.17d).The grain size is lower in composite because of the presence of in-situ particle.Some particle cluster also present in the composite material(Fig.17b),which is shown in high magnificatio SEM image(Fig.17f).The corresponding EDS results revels the presence of phases and particles(Fig.17g,h).DRX for Mg alloy and its composite is a thermally activated process and the extent of DRX depends on time and temperature.As the deformation progress with higher temperature,the DRX grains grow up faster.The grain size of deformed ZE41 at Zone-II is much larger than that of the sample at Zone-I.At high deformation temperature(＞400°C),the DRX grains completely grow with a homogeneous microstructure.The size of grains in Zone-II is increased with an increase in deformation temperature due to the enhancement of mobility of grain boundaries.The main mechanism of DRX is grain boundary migration because the energy supply is high at higher deformation temperature.Similar kind of grain growth characteristics are also observed at deformation temperature range of 400-450 °C and stain rate range of 0.01-1 s-1for different Mg based alloys and composites like AZ31,Mg-5Zn and SiCp/Mg-5Zn composite[46,47].

5.1.3.Zone-III

Apart from the two workability zones in both ZE41 and TiB2/ZE41 materials,another peak power dissipation efficien y was found in case of un-reinforced ZE41 alloy(Fig.13a).In comparison to the other two workability domains,this Zone-III workability domain is smaller.The Zone-III workability domain have power dissipation efficien y of 0.33-0.35 and this domain is observed at deformation temperature of 350-400 °C with strain rate range of 5 to 10 s-1(Table 2).Fig.18 shows the typical optical microstructure of ZE41 sample deformed at Zone-III workability domain(375 °C and strain rate of 10 s-1).The ZE41 deformed at this condition was characterized with a mixture of fin dynamic recrystallized and coarse grains.These fin DRX grains are formed as necklace structure(Fig.18).As per the microstructural observation and grain size distribution plot(Fig.18b),a duplex nature of microstructural variation(bimodal grain size)is found in Zone-III i.e.(i)fin necklace type DRX grains(＜5μm)with a number fraction of 65% along the grain boundaries and(ii)coarse grain structure.It is also clearly observed that the volume fraction of course grains(＞20μm)is much higher than the fin grains.The necklace structure DRX grains are observed as nucleated around the bulging grains.This types of DRX grains are mainly discontinuous dynamic recrystallization(DDRX)grains.Mostly,DDRX are found in the hot deformation of low stacking fault energy materials like Mg[36].The fin DDRX grains were also observed at the vicinity of the rare earth phases(T-Mg7Zn3RE).This is due to more fraction of dislocation accumulation around the T-phases.These rare earth phases are mostly found along the grain boundaries.During hot deformation,the nucleation of new grains starts the areas having high store energy.At high strain rate and high temperature,Mg7Zn3RE had high stored energy than theα-Mg.So,the new grains start nucleating at the vicinity of rare earth phases.With the constant rise in deformation temperature in hot working,more new fin grains are nucleated and form necklace structure.A similar observation was also found for hot deformation of Mg alloys at deformation temperatures greater than 300 °C with a higher strain rate range(＞1 s-1)[28,33,48].

Based on the above study,it is found that,for both ZE41 alloy and TiB2/ZE41 composite,DRX is the dominating mechanism for hot deformation.The optimum workability domain is suggested with high power dissipation efficien y.All the workability domains are at temperature ranges from medium to high domain(300 to 450 °C).In the case of ZE41 alloy,the workability domains are existing at low,medium and high strain rates,whereas in composite,the workability domains are only limited to low and medium strain rates.As compared to the unreinforced ZE41 alloy,the workability domains in the composite are wider.This shows,incorporation of sub-micron sized in-situ TiB2reinforcements expands the workability of ZE41.The DRX phenomenon is a thermally activated process and it happens as long as the activation energy of DRX is satisfied During hot deformation,the material’s internal energy has key role to provide activation energy for DRX.The calculated activation energy of composite is higher than that of ZE41 alloy(Table 1).Apart from the activation energy,the presence TiB2particles promotes the dislocation generation and pile up.This provides additional internal energy required for DRX.Hence TiB2/ZE41 composite possesses better workability domain than unreinforced ZE41 alloy.

Based on the above microstructural observations,the governing mechanisms of each workability zone for both ZE41 Mg alloy and TiB2/ZE41 Mg composite are identified The continuous dynamic recrystallization(CDRX)is the deformation mechanism for zone-I workability domain.The CDRX is mainly due to the climb of dislocations for ZE41 Mg alloy and cross-slip of screw dislocation for TiB2/ZE41 composite.The PSN effect in composite also helps to promote the DRX in TiB2/ZE41 Mg composite.The hot deformation mechanism in zone-II workability for both materials is mainly due to the grain growth kinetics of DRX grains at high deformation temperature domain(400-450 °C).The zone-III workability of ZE41 Mg alloy forms a typical necklace kind of fin grain structure and this is due to discontinuous dynamic recrystallization(DDRX).

5.2.Instability domain

The instability in processing map represents the instantaneous workability of a material.The shaded area in the processing map shows the instability domain of the material.The processing map of un-reinforced ZE41 alloy has one instability domains(Zone-A),however,in case of TiB2/ZE41 composite,there are two instability domains(Zone-A and Zone-B)(Fig.13).The instability of both alloy and composite have occurred mainly at low deformation temperature domain(250 °C-300 °C)and high strain rate(Table.2).The instability area in composite is smaller than that of base alloy.Generally,fl w localization and cracking are the phenomenon for instability in a Mg alloy.Typical optical microstructure in the instability zone for ZE41 sample deformed at 250 °C and 10 s-1is shown in Fig.19a.The microstructure is characterized with the number of mechanical twins,intergranular cracks.The mechanical twins are shown by red arrows while the intergranular cracks with yellow arrows.Fig.19c shows the corresponding SEM image.This shows the presence of micro crack in ZE41 alloy deformed at instability domain.These cracks are generated near to the TMg7Zn3RE phases of the ZE41 alloy.The dislocation movement in Mg alloy can be achieved when the applied stress will be greater than the critical resolved shear stress(CRSS).The required stress for hot deformation is temperature and time-dependent.For Mg alloy,at low temperature,only basal mode of slip systems is activated since the CRSS of basal slip system is lower compare to other slip systems.The additional twins get activated due to lack of slip systems in low-temperature deformation.But,with increase in temperature,the other slip system gets activated,which suppresses the extent of twin formation.So,in low deformation temperature twin mode of deformation is dominated.Apart from this,deformation at a high strain rate increases stress concentration,resulting in the nucleation of voids and cracks.This is because at high strain rate,there is not sufficien time for the material to dissipate heat during hot deformation[49].

In the case of in-situ TiB2/ZE41 composite,the instability domains are smaller than those of the ZE41 alloy.Fig.19b shows the optical microstructure of TiB2/ZE41 deformed at 250 °C and 1 s-1.The microstructure shows presence of intergranular cracks and mechanical twins in the material.However,intergranular cracks are present like the ZE41 alloy and these cracks are mostly initiated from the particle cluster,which is marked as white ellipse(Fig.19b).Fig.19d shows the SEM image of TiB2/ZE41 sample deformed at instability domain(250 °C,1 s-1).The presence of void and crack around the in-situ TiB2particle clusters are clearly visible.This is due to the dislocation generation and accumulation because of the mismatch in CTE and elastic modulus between the TiB2particle and alloy matrix.During deformation,the generated dislocation may concentrate around the in-situ particles,which further leads the stress concentration at the interface.As the stress concentration exceeds the interface bond strength,intergranular cracking happens.Due to this stress concentration,the interface debonding occurs between the in-situ TiB2particle and the matrix.Fig.19e shows the debonding of a sub-micron size in-situ TiB2particle with the matrix.The corresponding EDS result proofs the evidence of TiB2(Fig.19f).The debondings are the initiation of cracks and voids near the particles.During the deformation of SiCp/Mg composite at deformation temperature of 270 °C and strain rate of 1s-1,similar cracks are observed mainly at the interface between Mg matrix and SiCp[27].The TEM image shows the presence of twins along with high dislocation density present in TiB2/ZE41 composite deformed at instability domain(250 °C and 1s-1)(Fig.20).During deformation at low temperature and high strain rate,twining plays a key role in accommodation of dislocations due to limited number of slip systems.Meanwhile increase in dislocation fraction in composite material leads to an increase in the difficult of plastic deformation and results in fl w instability.So,mechanical twinning is the main cause of fl w instability for the ZE41 Mg alloy.Whereas stress concentration and debonding between TiB2particle and ZE41 matrix interface are causes the fl w instability in composite material.

Fig.19.Microstructure of sample deformed at 250 °C,10 s-1(a,c)ZE41 alloy;(b,d)TiB2/ZE41 composite and(e,f)debonding between TiB2 and matrix with corresponding EDS result.

In order to compare our result of instability domains for TiB2/ZE41 composite with the published literature of other Mg matrix composites[18,22,26,27,40,46,50,51],a thorough literature survey was carried out and the results are shown in Fig.21.Most of the researchers conducted hot compression up to a maximum strain rate limit of 1s-1,so the comparison is made within the strain rate range of 0.001s-1to 1s-1.In most of the cases,the instability domains are found in low to medium deformation temperature range(250 to 350 °C).The instability zone for TiB2/ZE41 composite is also found in low temperature and high strain rate as like most of the composites.The fl w instability in most of the composites is due to the micro cracks,voids,wedge cracking and mechanical twinning.They found the cracks and voids at the vicinity of the particles due to the stress concentration.The wider instability domain was found in case of high-volume fraction of reinforcement for the processing map of Mg composites[27,50].Apart from this,micron sized Mg composites exhibit high instability than that of sub-micron/nano size composites.But in few case studies,it was observed that,even though addition of nano reinforcement to Mg alloy matrix,there is increase in fl w instability domain.It was found that,all these composites were developed by ex-situ processing route.So,the size of the reinforcement,volume fraction and processing route play important roles in hot deformation of Mg composites.In this present work,the instability domain of TiB2/ZE41 has become narrow in comparison with other published literature.Because of the novel in-situ processing route and sub-micron sized TiB2particles(average particle size of 765 nm),the developed composite shows wide range of workability domain with narrow instability zone.These in-situ particles provide strong interfacial bonding betweenα-Mg and TiB2.Apart from this,in-situ TiB2particles give better thermal stability than that of ex-situ Mg composite[3].During high temperature deformation of Mg composites,the matrix alloy is getting thermal softening faster specificall at high strain rate than the reinforcements.The matrix alloy undergoes plastic deformation and accumulate stress surrounding the particles.This generates the stress concentration and hence exceeds the critical stress value.If the interface is not strong as in case of in-situ particles,there will be thermal residual stress at the interface.This leads to fl w instability and cracking near the particles.Apart from this,the nano/submicron size particle reinforced Mg composites developed via ex-situ route have more particle clusters than the in-situ Mg composite.This is due to the agglomeration and surface oxidation of starting fin powders,which results in poor wettability with the matrix.During hot deformation of ex-situ nano/sub-micron composites,these particle clusters increase the stress concentration which results cracks and voids.So,the developed sub-micron sized in-situ TiB2/ZE41 composite gives a better choice for manufacturability.

Fig.20.TEM image of TiB2/ZE41 composite at instability domain.

Fig.21.Comparison of instability domains of different Mg composites with TiB2/ZE41.

6.Conclusions

In this present research work,the hot deformation behavior of both ZE41 and in-situ TiB2/ZE41 composite are studied at different temperatures(250 °C-450 °C)and strain rates(0.001 s-1-10 s-1)based on constitutive analysis and processing maps.The detailed analysis of deformation mechanisms forboth workability and instability are discussed based on the microstructural characterization.The following important conclusions are drawn:

1)Compared with monolithic ZE41 alloy,the addition of in-situ sub-micron sized TiB2particles effectively enhanced the fl w stress of TiB2/ZE41 composite by hindering the strain-induced dislocations.

2)The apparent activation energy of TiB2/ZE41 composite is increased from 148.15 kJ/mol to 171.54 kJ/mol due to the effective dispersion strengthening action of in-situ TiB2particles.

3)The developed constitutive model equations for both ZE41 Mg alloy and TiB2/ZE41 composite are as follows:

The predicted fl w stress values from the constitutive equations of both materials are correlated well with the experimental results.

4)According to the developed processing maps,the workability domains at peak efficiencie are found to be at 300-375 °C,0.001-0.0055 s-1for ZE41 Mg alloy and at 292-384 °C,0.001-0.008 s-1for in-situ TiB2/ZE41 composite.The DDRX is the primary governing mechanism in the workability domain of both the materials.The fl w instability for both the materials occurs at a low temperature with a high strain rate,and stressinduced cracking,twinning are the main characteristics of fl w instability.

5)The peak efficien y of TiB2/ZE41 composite(45%)is higher than that of ZE41 alloy(41%).The processing map of TiB2/ZE41 composite results in wider workability domains and higher peak efficien y(45%)as compared to ZE41 Mg alloy.The presence of these fin in-situ TiB2particles promotes the DRX of the composite due to the additional PSN effect in addition to DDRX.Hence,the deformation of TiB2/ZE41 composite is more stable and homogenous with a wide range of manufacturability.

6)The presently developed TiB2/ZE41 composite showed a smaller range of instability domain as compared to that of existing Mg based composites published in the literature till date.The higher interfacial bonding between ZE41 matrix and in-situ TiB2reinforcement due to in-situ reaction prevent earlier debonding during high temperature deformation.

Acknowledgment

The authors show gratitude to Department of Science and Technology,India[grant number of DST/TDT/AMT/2017/211(G)(MEE/18-19/412/DSTX/SUSH)for the financia support and FIST grant,Department of Science and Technology,India[grant number SR/FST/ET11-059/2012(G)]for funding electron microscope facility.This research work is a part of Center of Excellence(CoE)in Applied Magnesium Research(A Vertical of Center for Materials and Manufacturing for Futuristic Mobility),IIT Madras.The authors acknowledge the Ministry of Human Resource and Development for funding this CoE through grant number-SB20210992MEMHRD008517.