Hot deformation mechanisms, microstructure and texture evolution in extruded AZ31–nano-alumina composite

Author's Accepted Manuscript

Hot deformation mechanisms, microstructure and texture evolution in extruded AZ31-Nano Alumina composite T. Zhong, K.P. Rao, Y.V.R.K. Prasad, F. Zhao, M. Gupta

www.elsevier.com/locate/msea

PII: DOI: Reference:

S0921-5093(13)01047-2 http://dx.doi.org/10.1016/j.msea.2013.09.062 MSA30315

To appear in:

Materials Science & Engineering A

Received date: 3 April 2013 Revised date: 16 September 2013 Accepted date: 18 September 2013 Cite this article as: T. Zhong, K.P. Rao, Y.V.R.K. Prasad, F. Zhao, M. Gupta, Hot deformation mechanisms, microstructure and texture evolution in extruded AZ31-Nano Alumina composite, Materials Science & Engineering A, http://dx.doi. org/10.1016/j.msea.2013.09.062 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Hot deformation mechanisms, microstructure and texture evolution in extruded AZ31-Nano Alumina composite T. Zhonga, K.P. Raoa*, Y.V.R.K. Prasadb, F. Zhaoc, M. Guptad a

Department of Mechanical and Biomedical Engineering, City University of Hong Kong, 83 Tat Chee Avenue, Kowloon, Hong Kong SAR, China b

Consultant, processingmaps.com, No. 2/B, Vinayaka Nagar, Hebbal, Bengaluru 560024, India (formerly with City University of Hong Kong)

c

The Peac Institute of Multiscale Sciences, Jiang'an Campus, Sichuan University, Chengdu, Sichuan 610207, China (formerly with City University of Hong Kong) d

Department of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, Singapore 117575, Singapore

Abstract Processing map for hot working of hot extruded AZ31-1.5 vol.% nano-alumina magnesium composite (AZ31-NAL) prepared by disintegrated metal deposition (DMD) technique has been developed in the temperature range 250–500 oC and strain rate range 0.0003–10 s-1. The starting composite microstructure is fine grained and is much less textured compared with the base AZ31 material prepared by similar technique (AZ31-DMD). The processing map for the composite exhibits three domains in the temperature and strain rate ranges: (1) 250–350 o

C/0.0003-0.01 s-1; (2) 375–500 oC/0.0003–0.01 s-1; (3) 300–400 oC/1–10 s-1, which are all

similar to those exhibited by the base alloy. In Domains #1 and #3, dynamic recrystallization occurs and results in fine grained microstructure, while in Domain #2, grain boundary sliding leading to wedge cracking in compression and intercrystalline cracking in tension have been identified. In comparison with the base alloy AZ31-DMD, the efficiency in the first domain has decreased, the third domain moved to lower temperatures and the effect on second domain is marginal. These differences are attributed to the lower intensity of starting texture in the nano-composite compared with the base alloy. A study of texture changes in the deformed specimens revealed that nano-alumina additions are helpful in reducing the preferred orientation in AZ31 alloy. Keywords: AZ31 magnesium alloy, Nano alumina composite, Hot deformation, Processing map, Microstructure, Texture * Corresponding author. Tel.: +852 3442 8409; Fax: +852 3442 0172 E-mail address: [email protected] (K.P. Rao)

Page 1

1. Introduction With the advantages of low density, superior damping capacity, good electromagnetic shielding characteristics and high specific strength, magnesium alloys are favored for various components in auto and aerospace industries. Mg-3Al-1Zn (AZ31) is one of the most favored wrought magnesium alloys, and has been investigated extensively. However, the alloys have limited workability at room temperature due to insufficient slip systems. Thus, bulk forming, such as extrusion, forging and rolling, is usually done at elevated temperature, where additional slip systems would be activated [1-8]. Microstructural refinement is an effective way to improve strength and ductility of the alloys, but thermo-mechanically refined microstructures are susceptible to coarsening and instability at high temperature [9, 10]. The addition of nano-particle dispersoids is being considered [11-13] for achieving microstructural refinement and strengthening. In the raw material preparation stage of this research, a new material synthesis method, named disintegrated melt deposition (DMD), was applied [14]. This technique would help not only in refining the microstructure of the billet but also in distributing nano-dispersoids more uniformly in the microstructure of the composite. Workability of the Mg-nano dispersion composites is an important aspect since the strengthening by the dispersoid particles would narrow the safe workability window. Earlier, the hot working characteristics of a Mg-1 vol. % nano alumina composite have been evaluated [15, 16]. The composite may be hot worked at higher strain rates similar to the base Mg material but at lower strain rates, the apparent activation energy for hot working is much higher than in the matrix material. The nano alumina particles have a tendency to migrate preferentially to the prior particle boundaries of the composite and stabilize them from migration. The aim of this research is to establish a safe and optimum workability window for AZ31-1.5 vol. % nano alumina composite (AZ31-NAL) and also to investigate the development of

Page 2

microstructure and texture during hot compression as well as the deformation mechanisms during hot working of this composite. With a view to bring out the effect of nano-dispersion on the behavior of the composite, it is proposed to compare with the behavior of the composite with that of the base alloy AZ31 produced by a similar process, namely, DMD [17]. The principles and basis for processing maps are briefly introduced here. Processing map is an explicit representation of the response of a material, in terms of microstructure mechanisms, to the imposed processing parameters and consists of a superimposition of a power dissipation map and an instability map. These are developed on the basis of the Dynamic Materials Model, which is essentially a continuum model using the concepts of systems engineering, extremum principles of irreversible thermodynamics with application to continuum mechanics of large plastics flow and those describing the stability and self organization of chaotic systems [18,19]. The efficiency of power dissipation occurring through microstructural changes during deformation as a function of temperature and strain rate is given by η=

2m m +1

(1)

In the Eq. (1), m is strain rate sensitivity. The efficiency variation of power dissipation with temperature and strain rate represents a power dissipation map, which is generally viewed as an iso-efficiency contour map. This efficiency η represents the relative rate of internal entropy production during hot deformation and characterizes the dissipative microstructure change under different temperature and strain rate. In addition, the extremum principles of irreversible thermodynamics as applied to continuum mechanics of large plastic flow [20] are explored to define a criterion for the onset of flow instability given by the equation for the instability parameter

ξ (ε ) =

∂ ln[m /( m + 1)] +m≤0 ∂ ln ε

(2)

Page 3

In the Eq. (2), m is strain rate sensitivity and ε is strain rate. This formula could be physically interpreted in simple term as follows: if the material system does not produce entropy constitutively at a rate that at least matches the rate of entropy input through imposed parameters, the flow becomes localized and causes flow instability, where ξ (ε ) ≤ 0 [18]. Combining an instability map with a corresponding power dissipation map could make a processing map, which helps to identify temperature-strain domains where workability of a material is optimal, such as dynamic recrystallization (DRX) and also to avoid regimes of flow instabilities, such as flow localization.

2. Material and experimental procedure AZ31–1.5 vol. % nano alumina (AZ31-NAL) composite was prepared by Disintegrated Melt Deposition (DMD) technique [14]. Briefly, the procedure consists of the following steps: AZ31 magnesium alloy ingots (with chemical composition of 2.94 wt. % Al, 0.87 wt. % Zn, 0.57 wt. % Mn, 0.0027 wt. % Fe, 0.00112 wt. % Si, 0.0008 wt. % Cu, 0.0005 wt. % Ni and balance Mg) were cut into rectangular pieces. To obtain reasonable distribution of nanoalumina particulates (with a diameter of 50 nm) into the AZ31 matrix, holes (12 mm diameter and 30 mm deep) were drilled into these pieces and required amount of 1.5 vol. % of the particulates was filled in these holes. In primary processing, synthesis of these materials involved heating the magnesium alloy pieces to 750 oC and stirring in an inert Ar gas atmosphere in a graphite crucible using a heating resistance furnace. The composite melt slurry was subsequently deposited onto a metallic substrate. Billets of 40mm diameter were obtained following the deposition stage. In the secondary processing, the deposited billets were soaked for 1 hour at a temperature of 400 oC before extrusion and then hot extruded using a die preheated to 350 oC using an extrusion ratio of 12.96:1 to produce rods of 10 mm diameter. During this process, colloidal graphite was used as lubricant.

Page 4

Cylindrical specimens of 9.8 mm diameter and 15 mm height were machined from the extruded rods for isothermal uniaxial compression along the direction of extrusion. For inserting a thermocouple to measure the instantaneous specimen temperature during deformation, a hole of 0.8 mm diameter and 4.9 mm depth was machined at mid height to reach the center of the specimen. Conditions for the isothermal uniaxial compression were as follows: Compression direction is along the extrusion direction, strain rates in a range of 0.0003-10 s-1 and temperatures in a range of 250-500 oC. An exponential decay of actuator speed in the servo hydraulic machine was applied to achieve a constant true strain rates during the compression. Graphite powder mixed with grease was used as the lubricant in all the experiments. The specimens were deformed up to a true strain of about 1 and then quenched in water. The load-stroke data were converted into true stress-true strain curves using the following equations [18]: True stress, σ = P/Ao (1 - e)

(3)

True strain, ε = - ln (ho/h)

(4)

where P is load, Ao is the original area of cross-section of the specimen, e is engineering strain [(ho - h)/ho], ho is the original height of the specimen and h is its instantaneous height. The flow stress values were corrected for the adiabatic temperature rise which was measured during each compression test and the actual temperature was obtained. The corrected flow stress was computed by fitting smooth curves for flow stress as a function of temperature at all the relevant strain rates within the experimental points. This correction was significant at lower temperatures and higher strain rates. The deformed specimens were sectioned in the center parallel to the compression axis and the cut surface was mounted, polished and etched for metallographic examination. The etching solutions for microstructure investigation contained: 4.2 g picric acid, 10 ml acetic acid, 10 ml distilled water, 70 ml ethanol. EBSD

Page 5

analysis was carried out using a JEOL 5600 scanning electron microscopy (SEM) equipped with a HKL Channel 5 software under an acceleration voltage of 20KV, together with a working distance between 10-20 mm and a sample tilt angle of 70o. The texture was recorded on a plane parallel to the compression direction and represented in the form of inverse pole figures. Transmission electron microscopy (TEM) was carried out using Philips CM 20 model operating at 20 KV. Specimens for TEM were mechanically polished to a thickness of 80-100 µm and then thinned further using a twin jet electro-polishing. Electro-polishing was performed using methanol, n-butyl alcohol and perchloric acid electrolyte solution.

3. Results and discussion 3.1. Starting Microstructure and texture The starting microstructure of the hot extruded AZ31-NAL is shown in Fig. 1(a). The vertical direction of the micrograph is parallel to the extrusion direction of the starting rod. The microstructure exhibits equiaxed fine grains along with stringers of nano-alumina particles aligned with the extrusion direction. Some undissolved hard intermetallic particles containing Mn have also been observed which are not covered in Fig. 1(a). The average grain size in the extruded material is 7.2 µm. For the purpose of comparison, the microstructure of the AZ31 DMD base material is shown in Fig. 1(b). It may be noted that the grain structure of the base alloy is non-uniform with a mixture of larger grains and finer grains and the average grain size 9.2 µm. The microstructural refinement in AZ31-NAL may be attributed to nano alumina particles which act as nucleation sites to cause more uniform recrystallization. The inverse pole figures obtained on the extruded rod on a plane parallel to the extrusion direction are shown in Fig. 2. Here, the inverse pole figure for Y0 represents that for the extrusion direction while X0 and Z0 represent the ones for the two orthogonal directions. The inverse pole figures reveal that the texture of the extruded AZ31-NAL material is about 2.3 times the random,

Page 6

which may be termed as weak compared with that recorded on the base alloy AZ31-DMD (about 7 times random) [17]. It may be inferred that nano-particles have an important effect on reducing the texture which may be due to their restricting influence on the slip processes. This is further supported by the presence of high dislocation density in the regions surrounding the nano-particles as revealed by the TEM micrograph of AZ31-NAL starting material which is shown in Fig. 3. However, as can also be seen in Fig. 1(a), it may be noted that the nano-particles got agglomerated at many places in the microstructure.

3.2. Stress-strain curves The true strain-stress curves obtained on the composite at 300 and 400 oC for different strain rates are shown in Fig. 4(a) and (b), respectively. The flow curves obtained at strain rates of 10 and 1 s-1 exhibit significant work softening which is less for lower strain rates. It is commonly believed that DRX is one of the mechanisms responsible for the flow softening behavior although flow softening may also result from flow instability and localization. In comparison with the base alloy AZ31-DMD [17], this nano-composite had higher strength only when the strain rates are higher than 1 s-1 and the flow stresses are nearly the same at lower strain rates.

3.3. Processing map The processing map developed on AZ31-NAL at strains of 0.5 on the basis of the Eqs. (1) and (2) is shown in Fig. 5. The map does not reveal any regimes of flow instability in the temperature and strain rate ranges covered in this study. It exhibits two domains in lower strain rate range: #1 in the temperature range of 250–350 oC and strain rate range of 0.00030.01 s-1 with a peak efficiency of about 30% occurring at 300 oC/0.0003 s-1, and #2 in the temperature range of 375–500 oC and strain rate range of 0.0003-0.01 s-1 with a peak

Page 7

efficiency of about 43% occurring at 450 oC/0.0003 s-1. In addition, a third domain is found at the temperature range of 300-400 oC and strain rate range of 1-10 s-1 with a peak efficiency of around 32% occurring at 350 oC/10 s-1. The microstructure obtained on specimen deformed at 300 oC/0.001 s-1 near the peak efficiency in Domain #1, is shown in Fig. 6. This exhibits fine dynamically recrystallized grain structure suggesting that DRX is occurring in this domain. The variation of grain size with temperature is shown in Fig. 7. The microstructure obtained near the peak efficiency condition in Domain #2 is shown in Fig. 8. The grain structure looks recrystallized and is analyzed further in terms of the angles of grain boundaries with respect to the compression axis. This is done because the grain structure resembles the so-called “diamond” configuration [21], where many of the boundaries are oriented at 45o with respect to the compression axis. Such a configuration will promote sliding of the boundaries since a shear stress is resolved along them. The statistical distribution of grains in different ranges of angles is shown in Fig. 9, which reveals that a high population of grains occur in the range of 40-50o, confirming that the configuration promotes sliding of grain boundaries. Since grain boundary sliding is a viscous process, it will exhibit higher strain rate sensitivity or higher efficiency of power dissipation (44% in this case). Grain boundary sliding may result in wedge cracking or superplasticity if the wedge cracks are repaired by accommodation of stress concentration by diffusional flow or plastic deformation [22]. With a view to differentiate between these two processes, tensile test has been conducted on a cylindrical specimen with 6 mm diameter and 24 mm gage length at the corresponding strain rate. The measured total elongation is only 42% and hence superplasticity is not occurring in this domain. On the other hand, SEM image of the region near the fracture end, shown in Fig. 10(a), reveals the presence of small cracks at grain

Page 8

boundary triple junctions which ultimately cause intercrystalline fracture in tension (Fig. 10(b)). Thus Domain #2 represents wedge cracking caused by grain boundary sliding. The microstructure of the specimen deformed under peak efficiency conditions in Domain #3 is shown in Fig. 11, which reveals fine grained recrystallized microstructure suggesting that this is also a DRX domain. The grain size variation with temperature in this domain is shown in Fig. 6 and the variation is similar to that occurring in Domain #1.

3.4. Texture evolution The inverse pole figures obtained on specimens deformed at 300oC/0.0003 s-1 (Domain #1) is shown in Fig. 12. Compared with the texture in the starting material, the texture intensity almost stayed the same but some of the crystallites are oriented such that <10-10> direction is along the compression axis. Such a change would occur if DRX takes place with basal+prismatic slip and climb controlled by diffusion through the lattice. In the second domain, the inverse pole figures in Fig. 13 indicate marginal changes for the intensity of texture with some of the crystallites with <10-10> direction rotating away from the compression axis. This domain occurs at higher temperatures and is associated with the slip largely on second order pyramidal systems. When it is restricted by the presence of nanoparticles, slip on basal+prismatic systems continues to occur. The inverse pole figure for specimen deformed in Domain #3 is shown in Fig. 14. The texture intensity is nearly unchanged and also associated with crystallite rotation similar to that in Domain #2. In view of temperatures being lower than 400 oC, slip occurs mainly on basal+prismatic systems, which is responsible for the rotation.

Page 9

3.5 Comparison with the processing map of the base alloy AZ31-DMD The processing map obtained on the composite, shown in Fig. 4, may be compared with that for the base alloy AZ31-DMD [17]. The features of the map are essentially similar although the following differences may be noticed: (1) In Domain #1, the peak efficiency has decreased from 43% in the base alloy to 30% in the nano-composite, which may be attributed to the effect of weaker initial texture in the composite. Stronger basal texture would promote increased prismatic slip and increase the efficiency of power dissipation as in the case of base alloy. (2) Domain #2 has moved to slightly higher temperatures (25 oC) and the efficiency remained nearly the same. This suggests that the nano dispersion has only a marginal effect on the deformation behavior in this domain. (3) Domain #3 has moved to lower temperatures by about 50 oC with slightly higher efficiency which may be attributed to the grain refinement that promotes grain boundary diffusion controlling the deformation in this domain.

3.6. Kinetic analysis The standard kinetic rate equation relating the flow stress (σ) to strain rate ( ε ) and temperature (T) is given by [23]:

ε =Aσn exp[

−Q ] RT

(5)

where A is constant, n is stress exponent, Q is activation energy and R is gas constant. The rate-controlling mechanisms are identified on the basis of the activation parameters n and Q. This kinetic rate equation is obeyed within the domain and can be applied to evaluate the apparent activation energy.

Page 10

The variation of ln (flow stress normalized with the shear modulus, μ) with ln (strain rate) in Domain #1 has shown linear fits for different temperatures and the estimated strain rate sensitivity value is 0.17 (n = 5.88). Fig. 15(a) is an Arrhenius plot showing the variation of ln(flow stress normalized with the shear modulus, μ) with ln(inverse of absolute temperature). The variation is linear and the apparent activation energy is estimated to be 116 kJ/mol. This apparent activation energy value is lower than that for lattice self-diffusion (135 kJ/mol) for magnesium [24], but was suggested to be the rate controlling mechanism for a similar domain in AZ31 [17]. Domain #3 is in high strain rate range, and the variation of ln (flow stress normalized with the shear modulus, μ) with ln (strain rate) in the domain has yielded a strain rate sensitivity value of 0.19 (n = 5.26). The variation of ln(flow stress normalized with the shear modulus, μ) with ln(inverse of absolute temperature) is given in Fig. 15(b) and the estimated values of apparent activation energy is 119 kJ/mole, which is also lower than that for lattice self-diffusion but higher than that for grain boundary self-diffusion [24], which was identified to be the rate controlling mechanism in a similar high strain rate domain in AZ31 [25, 26].

3.7. Workability and microstructural control in processing With a view to further evaluate Domains #1 and #3 from workability view point, tensile tests were conducted under conditions of peak efficiency in these two domains and the fracture surfaces have been observed in SEM. The fractographs are shown in Fig. 16(a) and (b) corresponding to the test conditions of 300 oC/0.0003 s-1 (Domain #1) and 350 oC/10 s-1 (Domain #3), respectively. It can be observed that both these exhibit dimple structure suggesting ductile fracture. Thus, it may be concluded that Domains #1 and #3 are safe for processing AZ31-NAL, which would result in fine grained microstructure without defects.

Page 11

Processing in Domain #3, however, may be desirable for industrial applications since the speed is high and processing will be economically viable. Within the DRX domains identified in the processing map (Domain #1 and Domain #3), the grain size variations may be correlated using the temperature compensated strain rate parameter, Z, defined as [23]: Z = ε exp[Q / RT ]

(6)

The average grain sizes measured on all the compressed samples are given in Table 1. The values of grain sizes (d) obtained within the domain are plotted against the calculated Z values in Fig. 17 on a log-log scale. A linear relationship is observed which may be expressed in the form of the following equations: Domain #1 :

log(d) = 1.7662-0.1077log(Z)

(7)

Domain #3 :

log(d) = 2.7563-0.1611log(Z)

(8)

The variation of grain size with strain rate at a temperature of 400 and 450 oC is shown in Fig. 18. It may be noted that abnormal grain growth has occurred at a strain rate of 0.1 s-1 for specimens compressed at 350, 400, 450 and 500 oC. In the processing map (Fig. 4), a changeover among Domains #1, #2 and #3 occurs in the vicinity of strain rate of 0.1 s-1 and this region is akin to a bifurcation in dynamical systems terminology [20, 27]. Similar bifurcation has also been observed during hot deformation of rolled AZ31 [25] where the two relevant dynamic recrystallization domains are controlled by lattice self diffusion and grain boundary self diffusion. The situation at the bifurcation may be explained in simple terms where the dissipative energy is highest in the bifurcation region and lowest in the domains on either side and therefore the bifurcation has a saddle-point configuration. Any fluctuations in the applied parameters in the bifurcation region can cause abnormal changes in the dissipative energy state, which in the present case is represented by the microstructure. Thus, the abnormal grain

Page 12

growth observed in the vicinity of strain rate of 0.1 s-1 may be attributed to the occurrence of the bifurcations in the processing maps. Microstructural control is difficult if the material is hot worked in these regions where probability plays an important role [20, 27]. This behavior is similar to that reported on the base alloy AZ31-DMD [17].

4. Conclusions On the basis of processing maps, kinetic analysis, microstructural features and fractographies, the following conclusion are drawn on the hot working characteristics of AZ31-NAL material produced by disintegrated metal deposition (DMD) technique: (a)

The starting composite microstructure is fine grained but is much less textured compared with the base alloy AZ31 prepared by similar technique (AZ31-DMD).

(b)

The processing map exhibits three domains in the temperature and strain rate ranges: Domain #1: 250-350 oC / 0.0003-0.01 s-1, Domain #2: 375-500 oC / 0.0003-0.01 s-1, and Domain #3: 300-400 oC / 1-10 s-1, similar to those observed in the map for the base alloy.

(c)

In comparison with the base alloy AZ31-DMD, Domain #1 occurs under similar conditions but the peak efficiency is decreased to 30% from 43%, which was attributed to weaker texture in the starting composite.

(d)

Domain #3 has moved to lower temperatures by about 50oC, which may be attributed to the grain refinement that promotes grain boundary diffusion controlling the deformation in this domain.

(e)

In Domain #2, which moves to slightly higher temperatures compared with the base alloy, grain boundary sliding occurs leading to wedge cracking in compression and intercrystalline cracking in tension.

Page 13

(f)

Comparing to the starting nano-composite, crystallites of the deformed specimens in Domains #2 and #3 rotate away from <10-10> direction; whereas, crystallites of the deformed specimens in Domains #1 rotate along <10-10> direction, although the max intensities for the specimens in those three domains are almost as weak as that for the starting composite.

Acknowledgement The work presented in this paper has been supported by a strategic research grant (Project #7008098) from City University of Hong Kong and General Research Fund (Project #114811) of the Research Grants Council (RGC) of the University Grants Council (UGC) of Hong Kong SAR.

References [1] K.B. Muller, in: H.J Kaplan (Ed.), Magnesium Technology 2002, TMS, Warrendale, PA, 2002, pp. 187-192. [2]

Baoping Zhang, Yin Wang, Lin Geng, Chunxiang Lu, Mater. Sci. Eng., A 539 (2012) 56-60.

[3]

Y. Uematsu, K. Tokaji, M. Kamakura, K. Uchida, H. Shibata, N. Bekku, Mater. Sci. Eng., A 434 (2006) 131-140.

[4]

W. Yuan, R.S. Mishra, Mater. Sci. Eng., A 558 (2012) 716-724.

[5]

R.Ye.Lapovok, M.R. Barnett, C.H.J. Davies, J. Mater. Proc. Technol. 146 (2004) 408414.

[6]

H. Watanabe, T. Mukai, K. Ishikawa, J. Mater. Sci. 39 (2004) 1477-1480.

[7]

D.L. Atwell, M.R. Barnett, W.B. Hutchinson, Mater. Sci. Eng., A 549 (2012) 1-6.

Page 14

[8]

T. Sakai, H. Utsunomiya, S. Minaminguchi, H. Koh, in: M. O. pekguleryuz, L. W. F. Machenzie (Eds.), Metallurgy and Petroleum, Canadian Institute of Mining, Montreal, Canada, 2006, pp 205-215.

[9]

Z.M. Li, A.A. Luo, Q.G. Wang, L.M. Peng, P.H. Fu, G.H. Wu, Mater. Sci. Eng., A 564 (2013) 450-460.

[10] H.K. Kim, J. Mater. Sci. Lett. 39(2004) 7101-7103. [11] W.L.E. Wang, M. Gupta, Adv. Eng. Mater. 9 (2007) 902-909. [12] S.F. Hasan, M.J. Tan, M. Gupta, Mater. Sci. Eng. A 486 (2008) 56-62. [13] P. Lukac, Z. Trojanova, Int. J. Mater. Product Tech. 23(2005) 121-137. [14] S.F. Hasan, M. Gupta, J. All. Compds. 429 (2007)176-183. [15] Y.V.R.K. Prasad, K.P. Rao, M. Gupta, J. Comp. Mater. 44 (2010) 181-194. [16] Y.V.R.K. Prasad, K.P. Rao, M. Gupta, Comp. Sci.Technol. 69 (2009) 1070-1076. [17] T. Zhong, K.P. Rao, Y.V.R.K. Prasad, M. Gupta, Mater. Sci. Eng. A 559 (2013) 773781. [18] Y.V.R.K. Prasad, S. Sasidhara, Hot working Guide: A compendium of Processing Maps, ASM International, Materials Park, OH, 1997, pp 3-24. [19] Y.V.R.K. Prasad, T. Seshacharyulu, Inter, Mater. Rev. 43 (1998) 243-258. [20] I. Prigogine, Science 201 (1978) 777-787. [21] V. Singh, P.R. Rao, G.J. Cocks, D.M.R. Taplin, J. Mater. Sci. 12 (1977) 373-383. [22] G. Rai, N.J. Grant, Metall. Mater. Trans. A 14 (1983) 1451-1458. [23] J.J. Jonas, C.M. Sellars, W.J. McG. Tegart, Metall. Rev. 14 (1969) 1-24.

Page 15

[24] H.J. Frost, M.F. Ashby, Deformation-mechanism maps, Pergamon Press, Oxford, 1982, pp 44. [25] Y.V.R.K. Prasad, K.P. Rao, Mater. Sci. Eng. A 487 (2008) 316–327. [26] Y.V.R.K. Prasad, K.P. Rao, Mater. Des. 30 (2009) 3723–3730. [27] R.C. Hilborn, Chaos and Non-Linear Dynamics, Oxford University Press, New York and Oxford, 1994. Table 1: Average grain sizes in µm measured on AZ31-NAL specimens deformed in hot compression at different temperatures and strain rates. Strain rate, s-1

Temperature, oC 250

300

350

400

450

500

10

6.7

9.5

10.2

13.8

15.8

18.9

1

8.1

10.4

11.1

20.1

25.6

33.3

0.1

7.2

11.2

18.5

31.2

71.5

98.7

0.01

6.1

6.9

9.8

10.2

14.5

22.8

0.001

6.9

7.2

10.7

12.8

19.5

25.6

0.0003

7.9

8.9

13.3

20.8

33.3

41.6

Page 16

Fig. 1: Microstructure of the extruded (a) AZ31-NAL in the starting condition and, (b) AZ31DMD base material.

Page 17

=20 µm; BC+E1-3; Step=0.8571 µm; Grid119x149

Fig. 2: Inverse pole figures obtained on the starting extruded rod of the AZ31-NAL on a plane parallel to the extrusion direction (Y0). X0 and Z0 are the two orthogonal directions on this plane. In the inverse pole figures, red color represents maximum intensity. The extrusion direction is vertical in the microstructure.

Page 18

Fig. 3. TEM micrograph showing high dislocation density around the alumina particles.

Page 19

180

(a)

o

AZ31-NAL: 300 C

160

True stress, MPa

140 -1

Strain rate, s

120

10

100

1 80

0.1

60

0.01

40

0.001 0.0003

20 0 0.0

0.2

0.4

0.6

0.8

1.0

1.2

True strain 100

(b)

o

AZ31-NAL: 400 C

True stress, MPa

80

-1

Strain rate, s 10

60

1 0.1

40

0.01 0.001 0.0003

20

0 0.0

0.2

0.4

0.6

0.8

1.0

1.2

True strain

Fig. 4: Stress-strain curves of AZ31-NAL at temperatures of (a) 300 oC and (b) 400 oC obtained in uniaxial compression at different strain rates.

Page 20

Fig. 5: Processing map of AZ31-NAL at a compression strain of 0.5. The numbers associated with the contours represent efficiency of power dissipation in percentage.

Fig. 6: Microstructure obtained on the specimen deformed at 300 oC/ 0.001 s-1.

Page 21

22

AZ31-NAL -1 Domain #1, Strain rate of 0.1 s -1 Domain #1, Strain rate of 0.01 s -1 Domain #1, Strain rate of 0.003 s -1 Domain #3, Strain rate of 10 s -1 Domain #3, Strain rate of 1 s

20

Grain size, μm

18 16 14 12 10 8 6 240

260

280

300

320

340

360

380

400

420

o

Temperature, C Fig. 7: The variation of grain size with temperature within Domain #1 and Domain #3.

Fig. 8: Microstructure obtained on the specimen deformed at 500 oC/ 0.0003 s-1.

Page 22

70

Grain boundaries

60

50

40

30

20

10

0 0-10

10-20

20-30

30-40

40-50

50-60

60-70

70-80

Angle with respect to compression axis,

80-90

o

Fig. 9: Distribution of number of grains having boundaries at different angles with respect to the compression axis in the specimen deformed at 500 oC/0.0003 s-1 (Domain #2).

Page 23

Fig. 10: Microstructural and fractural features of specimens deformed at a strain rate of 0.0003 s-1 in Domain #2 ⎯ (a) Wedge cracks occurring at the grain boundary triple junctions in the region close to the fracture end of the tensile specimen deformed at 450 oC, and (b) Surface of the fractured specimen tested in uniaxial tension at 500 oC.

Page 24

Fig. 11: Microstructure of the specimen deformed at 350 oC/10 s-1 (Domain #3).

Page 25

=20 µm; BC+E1-3; Step=0.4 µm; Grid184x94

Fig. 12: Inverse pole figures obtained on the AZ31-NAL compressed at the condition of 300 o

C, 0.0003 s-1. The compression direction is parallel to (Y0). X0 and Z0 are the two orthogonal

directions on this plane. In the pole figures, red color represents intensity max. The compression direction is vertical in the microstructure.

Page 26

=50 µm; BC+E1-3; Step=0.8 µm; Grid213x116

Fig. 13: Inverse pole figures obtained on the AZ31-NAL compressed at the condition of 450 oC, 0.0003 s-1. The compression direction is parallel to (Y0). X0 and Z0 are the two orthogonal directions on this plane. In the pole figures, red color represents intensity max. The compression direction is vertical in the microstructure.

Page 27

=20 µm; BC+E1-3; Step=0.4 µm; Grid181x102

Fig. 14: Inverse pole figures obtained on the AZ31-NAL compressed at the condition of 350 oC, 10 s-1. The compression direction is parallel to (Y0). X0 and Z0 are the two orthogonal directions on this plane. In the pole figures, red color represents intensity max. The compression direction is vertical in the microstructure.

Page 28

-5.2 -5.4

Ln (Flow stress/μ)

-5.6

AZ31-NAL -1 0.01 s -1 0.001 s -1 0.0003 s

-5.8 -6.0 -6.2 -6.4

n = 5.88 Q = 116 kJ/mol

-6.6 -6.8 1.55

1.60

1.65

1.70

1.75

1.80

1.85

1.90

1.95

-1

1000/T, K

(a) -4.4

Ln (Flow stress/μ)

-4.6

AZ31-NAL -1 10 s -1 1s

-4.8

-5.0

-5.2

-5.4

n = 5.26 Q = 119 kJ/mol

-5.6

-5.8 1.45

1.50

1.55

1.60

1.65

1.70

1.75

1.80

-1

1000/T, K

(b) Fig. 15: Arrhenius plot showing variation of flow stress normalized with shear modulus with inverse temperature at a strain of 0.5 in (a) Domain #1 and (b) Domain #3.

Page 29

Fig. 16: Fracture surfaces of the specimens tested in uniaxial tension at the conditions of (a) 300 oC/0.0003 s-1 (Domain #1) and (b) 350 oC/10 s-1 (Domain #3).

Page 30

1.35

Domain #1 Domain #3

1.30 1.25

Log (d, μm)

1.20 1.15 1.10 1.05 1.00 0.95 0.90 0.85 0.80 0.75 5.5

6.0

6.5

7.0

7.5

8.0

8.5

9.0

9.5 10.0 10.5 11.0 11.5 12.0 12.5 -1

Log (Z, s )

Fig. 17: Variation of grain size with Zener-Hollomon parameter for Domains #1 and #3. 80 o

400 C o 450 C

70

Grain size, μm

60 50 40 30 20 10 0 -4

-3

-2

-1

0

1

2

-1

Log (strain rate, s )

Fig. 18: Variation of grain size at temperatures of 400 oC and 450 oC. The change-over among domains #1, #2 and #3 occurs at about 0.1 s-1 (corresponds to -1 on X-axis).

Page 31