Increasing elastic modulus, strength and CTE of AZ91 by reinforcing pure magnesium with elemental copper

(1)

Increasing elastic modulus, strength and CTE of AZ91 by reinforcing pure magnesium with elemental copper

S.F. Hassan, K.F. Ho, M. Gupta*

Department of Mechanical Engineering, National University of Singapore, 9 Engineering Drive 1, Singapore 117576, Singapore Received 15 August 2003; received in revised form 6 January 2004; accepted 7 January 2004

Available online 13 February 2004

Abstract

Heat-treatable AZ91 and 3.9 vol.% copper particulate reinforced magnesium composite was synthesized using an innovative disintegrated melt deposition (DMD) technique followed by hot extrusion. Microstructural characterization of the composite material revealed retention and uniform distribution of reinforcement with defect free interface with the matrix. Physical properties characterization revealed improved dimensional stability of composite when compared to AZ91. Mechanical properties characterization revealed an increase in average values of modulus, 0.2% yield strength and ultimate tensile strength of un-heat-treated composite when compared to T6 heat-treated AZ91 while the ductility was adversely affected. An attempt is made in the present study to compare the microstructural, physical and mechanical properties of Mg/CuPcomposite with that of the commercially used AZ91 alloy.

Keywords: AZ91; Composite materials; Microstructure; Thermal expansion; Mechanical properties

1. Introduction

The ability of magnesium-based materials to exhibit high specific mechanical properties has been instrumental in attracting the attention of manufacturers for their possible use in automobile, aerospace, space, electronics and sports industries [1 – 3]. The major limiting factors in using magnesium include its low elastic modulus, rapid loss of strength with an increase in service temperature and poor creep resistance. Conventional addition of alloying elements and, more recently, reinforcement with stronger and stiffer materials are well-known practices to circumvent these limita- tions. Cast alloy, such as AZ91, is one of the most widely used existing magnesium-based engineering material, which exhibits excellent mechanical properties over pure magnesium when it is heat-treated. Heat treatment of a product adds extra cost and may also add non-uniformity in microstructural and mechanical properties in the case of composite formulations. Again, the use of SiC particulates reinforcement, for example, increases the stiffness, specific strength at

room and elevated temperature, dimensional stability, damp- ing capability and creep properties of magnesium[4 – 6]with limited success, due to the high brittleness of ceramic – Mg formulations. The selection of metal – metal composite formulations was prompted since the wettability of high melting point solid metal (such as copper) by low melting point liquid metal (such as magnesium), especially those that are mutually soluble, is excellent[7]. Accordingly, the primary aim of the present study was to develop high-performance magnesium material incorporated with elemental copper that does not need heat treatment to improve its properties. The synthesis was accomplished using an innovative disintegrated melt deposition (DMD) [8] technique followed by hot extrusion. Particular emphasis was placed to compare the microstructure, physical and mechanical properties of AZ91 and Mg/Cu material synthesized in the present study.

2. Experimental procedures

DMD technique [8] was used to synthesize magnesium (99.9% pure) reinforced with 17.2 wt.% of copper particulates (99% purity with 8 – 11 Am size range). The proce-

doi:10.1016/j.matlet.2004.01.011

* Corresponding author. Tel.: +65-874-6358; fax: +65-779-1459.

E-mail address: mpegm@nus.edu.sg (M. Gupta).

www.elsevier.com/locate/matlet Materials Letters 58 (2004) 2143 – 2146

(2)

dure involved melting and superheating the magnesium turnings with reinforcement particulates (placed in multi- layer sandwich form) to 750 jC under inert Ar gas atmosphere in a graphite crucible. Superheated slurry was stirred at 460 rpm for 5 min using a zirtex (86.0%

ZrO2, 8.8% Y2O3, 3.6% SiO2, 1.2% K2O, Na2O and 0.3%

trace organics, by weight) coated twin blade (pitch 45j) to facilitate the incorporation and uniform distribution of reinforcement particulates in the metallic melt. The melt was then released through an orifice at the base of the crucible and disintegrated by two jets of argon gas at a preselected flight distance, orientated normal to the melt stream, and subsequently deposited onto a metallic sub- strate. The synthesis of AZ91 was carried out using similar steps except that no reinforcement particulates were added.

The resulting ingots were subsequently hot extruded at 350 jC employing an extrusion ratio of 20.25:1. Subsequently, the AZ91 samples were subjected to T6 heat treatment by solutionizing at 413 jC for 1 h with subsequent ageing at 168 jC for 9 h [9].

Density of polished extruded materials was determined using Archimedes’ principle [6,8]. Distilled water was used as immersion fluid.

Retained metallic reinforcement in extruded composite was determined by inductively plasma atomic emission method, using three randomly selected samples. This method involved: (a) dissolving a known amount of sample in the nitric acid, (b) atomizing the solution into plasma and (c) analyzing the plasma in the inductively coupled plasma spectrometer, which detects the wavelength of Cu in the composite samples.

Microstructural characterization studies were conducted on metallographically polished extruded composite samples to investigate reinforcement distribution, interfacial integrity between the matrix and reinforcement, and the presence of porosity. JEOL JSM-5800 LV Scanning Elec- tron Microscope (SEM) equipped with Energy Dispersive Spectroscopy (EDS) was used.

Extruded AZ91 and composite samples were exposed to CuKaradiation (k = 1.5418A˚ ) with a scan speed of 2j/min on an automated diffractometer. The Bragg angles and the interplanar spacing, d, obtained were subsequently matched with standard values[10]of expected phases.

The coefficients of thermal expansion (CTE) of the extruded AZ91 and composite samples were determined using an automated SETARAM 92-16/18 thermo-mechanical analyzer with 5 jC/min heating rate and 1.2 l/min Ar gas flow. Displacement of the AZ91 and composite sam-

ples (each 15 mm long) as a function of temperature (50 – 400 jC) was measured using an alumina probe and was subsequently used to determine the CTE.

The smooth bar tensile properties of the extruded AZ91 and composites samples were determined in accordance with ASTM test method E8M-96 using Instron 8516 machine with a crosshead speed set at 0.254 mm/min on round tension test specimens of 5 mm diameter and 25 mm gauge length. Instron 2630-100 Series Clip-On type extensometer was used for strain recording. Elastic modulus was measured from the tensile test data corresponding to the elastic regime of the stress – strain curve.

3. Results

The result of macrostructural characterization conducted on the as deposited AZ91 and Mg/CuP samples did not reveal any presence of macropores or shrinkage cavity.

Following extrusion, there was also no evidence of any macrostructural defects.

Table 1

Density, quantimet studies and CTE measurement result for AZ91 and Mg/CuP

Materials Reinforcement Density (g/cm³) CTE ( 10 ⁶/jC) (wt.%) (vol.%)

AZ91 – – 1.820 F 0.026 30.7 F 0.5

Mg/3.9CuP 17.18 3.9 2.076 F 0.051 27.0 F 0.4

Fig. 1. Representative SEM micrographs showing: (a) microstructural characteristics and (b) CuP/Mg interfacial characteristics exhibited by Mg/

CuPcomposite samples.

S.F. Hassan et al. / Materials Letters 58 (2004) 2143–2146 2144

(3)

The result of the density measurements is given in Table 1. The results of inductively coupled plasma spectrometer showed successful retention of copper in the extruded composite matrix (see Table 1).

Scanning electron microscopy conducted on the extruded composite specimens showed reasonably uniform presence of secondary phases (including copper and its reaction product) in the magnesium matrix (see Fig.

1(a)), good CuP– Mg interfacial integrity (see Fig. 1(b)) and the presence of minimal porosity. It may be noted that good interfacial integrity indicates the absence of voids and/or debonded areas. The X-ray diffraction results revealed the presence of Mg and Mg2Cu phases in Mg/

CuP composite, and Mg, Al12Mg17 and h-AlMg phases in AZ91.

The results of CTE measurements obtained from extruded AZ91 and Mg/CuP samples are listed in Table 1.

The results show enhanced dimensional stability of Mg/

CuP samples when compared to that of AZ91.

The results of ambient temperature tensile tests of Mg/

CuP (see Table 2) revealed higher average values of modulus, 0.2% YS, and UTS and reduced ductility of Mg/CuP samples when compared with T6 heat-treated AZ91 samples.

4. Discussion

4.1. Synthesis of AZ91 and Mg/CuP

Synthesis of AZ91 and Mg/CuP materials was success- fully accomplished by DMD process followed by hot extrusion. Observations made on deposited ingot revealed:

(a) minimal oxidation of magnesium, (b) absence of macropores and blowholes and (c) no detectable reaction between graphite crucible and melts (AZ91 melt and Mg/

CuP composite slurry). The inert atmospheric condition used during melt processing, dispersion, deposition and solidification was instrumental in the prevention of reaction between air/oxygen and Mg melts. The absence of macro-pores, blowholes and segregation or agglomeration of reinforcement particulates indicates the suitability of stirring parameters and solidification conditions during deposition. The absence of reaction between AZ91 melt/

composites slurry with the graphite crucible can be attributed primarily to the inability of magnesium to form stable carbides [5].

4.2. Microstructure

The results of the microstructural characterization studies conducted on the extruded composite samples are discussed in terms of: (a) microstructural uniformity, (b) reinforcement – matrix interfacial characteristics and (c) the presence of porosity. The fairly uniform distribution of secondary phases (including copper and its reaction product Mg2Cu) can be attributed to: (i) limited agglomeration of reinforcement during melting of matrix due to thin layered arrangement of raw materials in crucible for melting, (ii) minimal gravity-associated segregation due to judicious selection of stirring parameters, (iii) good wetting of reinforcement by the matrix melt [7] and (iv) disintegration of the composite slurry by argon jets and its subsequent deposition in metallic mold. The microstructural characterization results revealed the reduction in the average size of copper particulates to f 1 Am (as received size range was 8 – 11 Am) following processing and this can be attributed to the severe reaction of Cu with Mg leading to the formation of Mg2Cu intermetallics [11] as also supported by XRD results. XRD was not able to detect Cu due to its low volume fraction and small size [12]. The results of microstructural characterization of composite material also revealed a near defect free interface formed between reinforcements and matrix (see Fig.

1(b)), which was assessed in terms of interfacial debonding and microvoids at the particulate – matrix interface. The presence of minimal porosity in composite materials (see Fig. 1(a) and (b)) can be attributed to: (i) good compati- bility between metal – metal system [6] and (ii) the use of an appropriate extrusion ratio [5,8].

4.3. Coefficient of thermal expansion

CTE measurements in the temperature range of 50 – 400 jC revealed that the presence of copper as reinforcement resulted in a noticeable increase in dimensional stability of Mg matrix (seeTable 1) when compared with AZ91. This can be attributed to the lower CTE of copper when compared to magnesium (27.0 10 ⁶/jC and 17.4 10 ⁶/jC for Mg and Cu[13], respectively). It may be noted that the effect on CTE of the matrix can also be attributed to the presence of Mg2Cu phase. Such a correlation, however, is not attempted in this study due to the unavailability of CTE of Mg2Cu in the open literature.

4.4. Tensile behavior

Addition of copper as reinforcement marginally in- creased elastic modulus of magnesium matrix (see Table 2) over AZ91 and this can be attributed to: (a) the high modulus of reinforcement (i.e., 129.8 GPa for Cu[13]) and (b) uniform distribution of reinforcement with good interfacial integrity. It may be noted that uniform distribution of reinforcement coupled with good matrix – reinforcement

Table 2

Density, CTE and room temperature tensile properties of AZ91 and Mg/CuP

Materials Modulus

(GPa)

0.2% YS (MPa)

UTS (MPa)

Ductility (%)

AZ91 45 F 2 272 F 3 353 F 0 3.7 F 0.5

Mg/3.9CuP 47 F 1 355 F 8 358 F 7 2.2 F 0.9

Mg/9.3SiCP[6] 44 F 2 120 F 5 181 F 6 4.7 F 1.3

Mg/10SiCP[18] 45 120 160 2

(4)

interfacial integrity leads to a significant increase in internal stress between reinforcement and matrix resulting in the enhancement of elastic modulus [4]. It may very interestingly be noted that Mg/CuPformulation can realize a theoretical modulus value, 46.5GPa (using 129.8 GPa [13] and 43 GPa [14] as the modulus for Cu and Mg, respectively), predicted by well-known rule-of-mixture [15] assuming absence of reaction product Mg2Cu in the matrix. A higher average value of modulus observed experimentally in the case of Mg/CuPformulation confirms that Mg2Cu contributes positively. The relative contribu- tion of Mg2Cu could not be determined due to the absence of mechanical properties of Mg2Cu in the open literature.

Significant increase in 0.2% YS and marginal improve- ment in UTS (seeTable 2) of Mg/CuPsamples can primarily be attributed to: (a) the presence of uniformly distributed elemental Cu particulates coupled with the strengthening effect of Mg2Cu [16] and (b) effective transfer of applied tensile load to the uniformly distributed and well-bonded Cu and Mg2Cu phases. Poor ductility of Mg/CuP samples can be attributed to the presence of brittle intermetallic phase (Mg2Cu) in matrix, which serves as crack nucleation sites leading to the reduction in ductility under tensile loading conditions[17]. It may be noted that the overall combina- tion of mechanical properties of the Mg/CuP composite developed in this study remained superior to AZ91 and Mg/SiCPformulations with even higher volume percentage SiC particulates (see Table 2). The results of the present study thus show tremendous promise for developing entirely new class of cost-effective metal-reinforced material for specific strength-critical applications that do not need any heat treatment.

5. Conclusions

(1) Disintegrated melt deposition technique coupled with hot extrusion can be used to synthesize AZ91 and elemental copper reinforced magnesium composites.

(2) The reasonably uniform distribution of reinforcement particulates, strong reinforcement – matrix interfacial integrity and the presence of minimal porosity in the composite microstructure indicate the suitability of

primary processing and secondary processing parameters used in the present study.

(3) Results of coefficient of thermal expansion measurement indicate that the un-heat-treated Mg/CuP formu- lation investigated in this study exhibits enhanced dimensional stability when compared to T6 heat-treated AZ91 alloy.

(4) The results of tensile characterization revealed that Mg/

CuPformulation developed in this study exhibits higher modulus, 0.2% YS, and UTS when compared to heat- treated AZ91 alloys, while ductility gets adversely affected.

References

[1] D. Eliezer, E. Aghion, S. Froes, Synthesis of Lightweight Metals, Part 3, TMS, Warrendale, PA, USA, 1999, p. 139.

[2] R. Edgar, Proc. of the 56th Annual Meeting of the Int. Magnesium Association, IMA, Rome, Italy, 1999, p. 21.

[3] I.J. Polmear, Metallurgy of Light Alloys, Halsted Press, New York, 1996, pp. 1 – 4.

[4] A. Luo, Metall. Trans., A, Phys. Metall. Mater. Sci. 26 (1995) 2445.

[5] D.J. Lloyd, Int. Mater. Rev. 39 (1994) 1.

[6] M. Gupta, M.O. Lai, D. Saravanaranganathan, J. Mater. Sci. 35 (2000) 2155.

[7] N. Eustathopoulos, M.G. Nicholas, B. Drevet, Wettability at High Temperatures, vol. 3, Elsevier, UK, 1999, pp. 1 – 53, 175 – 197.

[8] L.M. Tham, M. Gupta, L. Cheng, Mater. Sci. Technol. 15 (1999) 1139.

[9] ASM Handbook: Heat Treating, vol. 4, ASM International, USA, 1991, p. 899.

[10] Powder Diffraction File, International Center for Diffraction Data, PA, USA, 1991.

[11] A.A. Nayeb-Hashemi, J.B. Clark, Phase Diagram of Binary Magne- sium Alloys, ASM International Publication, USA, 1988, p. 94.

[12] C. Suryanarayana, Prog. Mater. Sci. 46 (2001) 1.

[13] E.A. Brandes, G.B. Brook, Smithells Metals Reference Book, 7th ed., Butterworth-Heinemann, London, 1998, pp. 14 – 3, pp. 14 – 19, pp. 15 – 2.

[14] S.F. Hassan, M. Gupta, Mater. Res. Bul. 37 (2002) 377.

[15] T.W. Clyne, P.J. Withers, An Introduction to Metal Matrix Composite, Cambridge Univ. Press, NY, USA, 1993, p. 14.

[16] J.H. Westbrook, Intermetallic Compounds, Wiley, NY, USA, 1967, p. 11, p. 471.

[17] M. Gupta, Alum. Trans. 1 (1999) 33.

[18] M.R. Krishnadev, R. Angers, C.G.K. Nair, G. Huard, JOM 45 (1993) 52.