Materials Science and Engineering A xxx (21) xxx xxx Contents lists available at ScienceDirect Materials Science and Engineering A journal homepage: www.elsevier.com/locate/msea Recycling of high grade die casting AM series magnesium scrap with the melt conditioned high pressure die casting (MC-HPDC) process S. Tzamtzis, H. Zhang, M. Xia, N. Hari Babu, Z. Fan BCAST (Brunel Centre for Advanced Solidification Technologies), Brunel University, Kingston Lane, Uxbridge, Middlesex UB8 3PH, UK article info abstract Article history: Received 2 November 21 Received in revised form 3 November 21 Accepted 1 December 21 Available online xxx Keywords: Magnesium alloys Casting Mechanical characterization Intensive shearing Recycling Strict government regulations and environmental concerns are the driving forces behind the increased use of magnesium alloys aimed at weight reductions. This however inevitably leads to increased magnesium alloy scrap and calls for effective recycling processes. In this paper, the melt conditioned high pressure die casting (MC-HPDC) process has been investigated as a physical approach for the recycling of AM series magnesium alloy die casting scrap. Process optimization was required to eliminate hot cracking phenomena. The experimental results showed that intensive melt shearing alters the size and morphology of MgO present in the scrap melt, leading to an effective grain refinement that was reflected in the mechanical properties of the recycled alloy. The MC-HPDC process showed excellent potential as a physical recycling approach for Mg alloy scrap, producing casting with properties comparable to those of fresh Mg alloys. 21 Elsevier B.V. All rights reserved. 1. Introduction The existing tight government regulations for sustainable economic development and the continuously increasing concerns for environmental protection have been the main driving forces for the recent attempts made by the automotive industry to improve fuel consumption and reduce greenhouse gas emissions. The general consensus is that for every 1% reduction in vehicle weight there will approximately be a corresponding 6 8% decrease in fuel consumption, which would also lead to reduced CO 2 emissions [1]. Magnesium alloys, being lightest of all structural metallic materials, are finding increased applications by the automotive industry, triggered by the demand for vehicle weight reduction. A recent study of the overall magnesium consumption in 27 in the United States has shown that 32% was for die-casting applications whilst the global demand has been expected to increase at a 7.3% average annual growth rate until 212 [2]. This increasing usage of magnesium alloys will inevitably lead to a fast increase in Mg-alloy scrap from both manufacturing sources (new scrap) and end-of-life vehicles (old scrap). The European Union has launched a directive on end-of-life vehicles, which states that 85% of the vehicles weight will have to be recycled by 215 [3]. According to a German magnesium inventory analysis up to Corresponding author. Fax: +44 1895269758. E-mail addresses: spyridon.tzamtzis@brunel.ac.uk, spyrostzamtzis@gmail.com (S. Tzamtzis). 22 [4], the major source of magnesium scrap is end-of-life vehicles. The general picture of magnesium inventory is expected to be similar in the rest of the developed countries [1]. In addition, the source for new scrap is also growing fast. In typical magnesium diecasting operations only around 5% of the material input ends up as finished products, and the remaining 5% of magnesium alloys is accumulated as scrap and handled by the foundry or an external recycler [5]. Both new and old magnesium alloys scrap contain substantial amount of inclusions and impurity elements which cause severe loss of strength and ductility and significantly reduce the corrosion resistance. Thus, a major barrier to overcome in magnesium recycling is dealing with the increased inclusions and impurity elements. Several chemical approaches have been followed in order to reduce the amount of such inclusions and impurities [6]. However, these recycling technologies remain complex, energy demanding and exhibit low productivity. Given that an effective recycling technology for scrap magnesium alloys would require less than 1% of the energy required for primary production [7], the application of recycled magnesium can lead to a substantial reduction of greenhouse gas emissions and help the automotive industries conform to the tight government regulations for sustainable economic development. Therefore, recycling magnesium alloys is becoming a major technical, economical and environmental challenge. The melt conditioned high pressure die casting (MC-HPDC) process has been developed at Brunel Centre for Advanced Solidification Technologies (BCAST) and has been used for the production of high quality magnesium cast components [8,9]. In the 921-593/$ see front matter 21 Elsevier B.V. All rights reserved. doi:1.116/j.msea.21.12.1 Please cite this article in press as: S. Tzamtzis, et al., Mater. Sci. Eng. A (211), doi:1.116/j.msea.21.12.1
2 S. Tzamtzis et al. / Materials Science and Engineering A xxx (21) xxx xxx 3 25 25 MC-HPDC UTS (MPa) 2 15 1 Elongation (%) 2 15 1 5 5 1 2 3 4 5 6 7 8 Sample ID MC-HPDC 9 1 11 12 13 1 2 3 4 5 6 7 8 Sample ID 9 1 11 12 13 Fig. 1. The variation of mechanical properties of MC-HPDC recycled AM series scrap prior to process optimization. MC-HPDC process intensive shearing is directly imposed on the alloy melt prior to die-filling. The implementation of the shearing is carried out by the melt conditioning by advanced shear technology (MCAST) process, involving a specially designed twin screw machine which can be directly attached to a standard HPDC machine. The twin screw mechanism is used to impose the high shearing dispersive mixing action to the melt, so that the melt is treated in such a way that its uniformity in chemistry and temperature is improved. Inside the barrel of the twin screw machine, there is a pair of specially designed screws, co-rotating, fully intermeshing and self-wiping, creating an environment of high shear rate and high intensity of turbulence for the alloy melt. The sheared melt is then cast by the conventional HPDC process and is expected to offer unique solidification behaviour, and an improved fluidity and die-filling during the subsequent HPDC process. In this study, we investigate the potential of the MC-HPDC process as a physical approach for the recycling of high-grade AM series magnesium die casting scrap. 2. Experimental A mixture of AM5A and AM6B die casting high-grade scrap supplied by Meridian Technologies United Kingdom, consisting of biscuits and gates, runners, overflows and dross. A top loaded electrical resistance furnace was used to melt the scrap alloy in a steel crucible, under a protective atmosphere of pure N 2 gas containing.5 vol.% SF 6, and the melt was then subjected to the intensive shearing prior to casting. Detailed description of the shearing mechanism of the MCAST process can be found elsewhere [1 12]. The rotation speed of the twin screws of the melt conditioner was controlled to be 5 and 8 rpm and the temperature ranged between 615 C and 64 C. After the scrap alloy was sheared for 45 s, it was directly transferred to a standard 28-ton cold chamber HPDC machine (LK Machinery Co. Ltd., Hong Kong) which was used to produce the standard tensile test specimen, 6.4 mm in gauge diameter and 25 mm in gauge length. Specimens for optical microscopy (OM) were cut from the top, middle and bottom sections of the gauge length of the tensile test components. A pressurised melt filtration technique was used to collect inclusions and oxides from the AM series scrap alloy melt, so that their local concentration was enough to facilitate SEM examination. The scrap alloy melt, either taken directly from the furnace or after being sheared in the MCAST unit, was transferred into a pre-heated crucible positioned in the pressure chamber of the filtration unit. Argon was introduced to the pressure chamber to force the liquid metal to flow through a steel filter attached at the bottom of the crucible. Solid inclusions and oxides were collected above the porous filter in the crucible together with the remaining melt, which were allowed to solidify in the crucible outside the pressure chamber under the same protective atmosphere for melting. The residue material close to the filter concentrated with the intermetallic particles was sectioned and prepared for microscopic examinations. The metallographic specimen for OM were prepared by grinding with SiC abrasive paper and polishing with an Al 2 O 3 suspension solution, followed by etching in a solution of 5 vol.% concentrated HNO 3 and 95 vol.% ethanol. A Carl Zeiss Axioskop 2MAT optical imaging system equipped with image analysis software was used for the OM observation. In order to obtain further microstructural information, color etching in a solution of 7 ml ethanol, 1 ml water, 2 ml acetic acid and 4.2 g picric acid was used, with the colored orientation contrast being created under polarized light of the Zeiss optical microscope. Scanning electron microscopy (SEM) was carried out using a Zeiss Supra 35 machine with a field emission gun, equipped with an energy dispersive spectroscopy (EDS) facility. Finally, the mechanical properties of the tensile test specimen were measured at room temperature by a universal materials tensile testing machine (Instron 5569) at a crosshead speed of 1 mm/min (strain rate:.66 1 3 s 1 ). 3. Results 3.1. Optimization of the MC-HPDC process Mechanical testing of standard tensile test components cast with the MC-HPDC process produced highly scattered and inconsistent results. These are provided in Fig. 1. Minor changes in the processing parameters resulted in a large variation of the mechanical properties of the component, where even samples cast under identical parameters exhibited inconsistent mechanical behaviour, with their properties varying from sample to sample. The origin of the large variations in the mechanical properties of cast components was identified after careful examinations of their macro- and microstructure. Failure of the samples that exhibited unexpected mechanical performance was located outside the gauge length of the tensile sample and always occurred at the shoulder of the sample. Inspection of the castings prior to mechanical testing identified the presence of dark in contrast lines on the surface of the majority of the samples that would fail unexpectedly. Metallographic sections of samples where these dark in contrast lines were found revealed the presence of cracks in the microstructure, starting from the surface and extending well into the sample body. A characteristic micrograph is presented in Fig. 2(a). Optical microscopy examinations in higher magnification, as seen in Fig. 2(b), exposed the detailed structure of these shoulder cracks, which consisted of porosity and solidified eutectic liquid. Please cite this article in press as: S. Tzamtzis, et al., Mater. Sci. Eng. A (211), doi:1.116/j.msea.21.12.1
S. Tzamtzis et al. / Materials Science and Engineering A xxx (21) xxx xxx 3 Fig. 2. (a) Shoulder cracks, the major defect of MC-HPDC recycled AM series scrap tensile samples prior to optimization and (b) the detailed structure of shoulder crack. According to the relationship between magnesium die-casting defects and casting parameters, as presented by Avedasian and Baker [13], the die-casting parameters that could optimized in order to eliminate these defects were the die temperature, metal temperature and cavity fill time. The die temperature was controlled directly from the die-casting machine control panel and the metal temperature could be altered by varying the shearing temperature of the magnesium scrap alloy melt in the MCAST unit. Finally, to control the cavity fill time we determined that that the die-casting parameter that had to be altered was the position where the intensifying pressure is applied during the casting cycle. The first parameter under investigation for the optimization of the MC-HPDC process for magnesium scrap was the position of the plunger at which the intensifying pressure was applied, termed here intensifier position. A lower intensifier position value indicates that the pressure was applied earlier in the casting cycle resulting in faster filling of the die cavity. The ratio of castings that exhibited the shoulder cracks to the overall number of castings is defined as defective rate. Fig. 3(a) provides the defective rate as a function of the intensifier position. It was found that in order to produce sound castings with no shoulder cracks, the die filling time had to be short. Thus, the intensifier position was reduced to 18 mm, where the defective rate was only 6.25%. The second parameter that was studied in order to optimize the MC-HPDC process and reduce the occurrence of shoulder cracks was the die temperature of the die-casting unit. It was found that as the die temperature decreased, the quality of the samples improved. There was a lower limit to which the temperature of the die could be reduced, before the shoulder crack phenomena reappeared. The effect of the die temperature on the defective rate of the castings is presented in Fig. 3(b). A very useful observation was that although the visual examination gave a satisfactory indication for the probability of a shoulder crack in relation to the die temperature, as the die temperature decreased this indication was less accurate and microstructural examination of the samples was necessary. As the die temperature decreased the shoulder crack did not initiate as close to the sample surface, and was not visible under visual examination. However, the castings exhibited a clear trend to improved performance with decreasing die temperature. The optimized die temperature for the MC-HPDC process was selected to be 18 C, as this was the temperature that consistently produced the highest quality of castings. The final parameter to be optimized therefore was the melt temperature, determined by the shearing temperature in the MCAST unit prior to casting. Since the liquidus temperature of the AM series alloy scrap was found to range between 62 and 625 C, the shearing temperature was determined in relation to the liquidus. The shearing temperatures studied were equal to T L 5 C, T L +5 C and T L +15 C. The results of the optimization process are presented in Fig. 3(c), where the sample defective rate percentage is plotted as a function of the shearing temperature. It was found that when the melt was sheared with a higher superheat (+15 C), the defec- a Defective rate (%) b Defective rate (%) c Defective rate (%) 1 9 8 7 6 5 4 3 2 1 1 9 8 7 6 5 4 3 2 1 1 9 8 7 6 5 4 3 2 1 Visual examination 18 19 2 215 Intensifier position (mm) Visual Examination Microstructural examination 16 17 18 2 22 Die Temperature (ºC) Visual Examination Microstructural examination 24 T L - 5 C T L +5 C T L +15 C Shearing Temperature (ºC) Fig. 3. The casting defective rate as a function of (a) the intensifier position, with the die temperature ranging between 16 and 24 C and the processing temperature ranging between T L 5 and T L +15 C; (b) the die temperature, with the optimized intensifier position fixed at 18 mm and the processing temperature ranging between T L 5 and T L +15 C; and (c) the processing temperature, with the optimized intensifier position fixed at 18 mm and the optimized die temperature fixed at 18 C. Please cite this article in press as: S. Tzamtzis, et al., Mater. Sci. Eng. A (211), doi:1.116/j.msea.21.12.1
4 S. Tzamtzis et al. / Materials Science and Engineering A xxx (21) xxx xxx Table 1 Optimized parameters for the MC-HPDC process of high grade magnesium scrap. Shearing time Shearing temperature Die temperature Intensifier position 45 s T L +SC 18 C 18 mm tive rate was about 3% based on visual inspection of the samples and increased to 5% when the samples were sectioned and their microstructure was examined. When the melt was sheared at a temperature close to, but below the liquidus temperature ( 5 C), the defective rate decreased to 1% with visual examination but was still as high as 35% when the microstructure was examined. However, when the shearing temperature was just above the liquidus of the alloy (+5 C), no defective samples were found either by visual or microstructural examination. In summary, the optimized parameters of the MC-HPDC process for AM series magnesium alloy scrap are given in Table 1. 3.2. The effect of intensive shearing on inclusion and oxides As received magnesium scrap was melt and transferred to the pre-heated crucible of the Prefil unit at 63 C for casting. Al 8 Mn 5 intermetallic particles that form during solidification of manganese-containing magnesium alloys were found in the microstructure, with a large number of them being needle shaped and plate-like particles, apart from the common faceted Al 8 Mn 5 intermetallics. Image analysis carried out on the microstructures of the non-sheared magnesium scrap calculated the average size of all the Al 8 Mn 5 intermetallic particles equal to approximately 7.5 m. To investigate the effect of intensive shearing on the intermetallic particles and the oxide found in magnesium scrap, the alloy melt was intensively sheared in the MCAST unit at 63 C and 5 rpm for 45 s and then transferred to the pre-heated crucible of the Prefil unit for casting. It was found that intensive shearing of the magnesium alloy scrap prior to solidification changed the morphology of intermetallic particles, as no needle shaped Al 8 Mn 5 particles were found in the microstructure. Also, intensive shearing resulted in the formation of Al 8 Mn 5 particles with a finer size of approximately 3.5 m compared to 7.5 m of the non-sheared alloy. There was also a narrower size distribution of the intermetallic particles, as seen in Fig. 4. Higher magnification backscattered electron SEM examinations of the filtered magnesium scrap using revealed the presence of a very large number of oxide inclusions. Fig. 5 shows MgO inclusions in the non-sheared AM series scrap alloy, which was found to have two different morphologies. The first morphology type of MgO can be described as oxide particle clusters, seen in Fig. 5(a), which were developed during melt handling and pouring. They consisted of extremely fine MgO particles (1 2 nm) dispersed Frequency (%) 5 4 3 2 1 5 1 15 Average particle size (μm) Non-sheared Sheared Fig. 4. The Al 8Mn 5 intermetallic particle size distributions of the non-sheared and sheared AM series alloy scrap. in the alpha-mg matrix. The second MgO morphology type can be described as ingot skin, seen in Fig. 5(b), which was carried in from the solid ingots. These are usually straight segments with a thickness of 1 15 m and a length that can extend to more than 1 m. These oxide skins also consisted of MgO particles dispersed in alpha-mg matrix. Intermetallic Al 8 Mn 5 particles can also be seen in the microstructures, both in their faceted and needleshape morphologies. On the other hand, after the implementation of intensive shearing, both the MgO particle clusters and the ingot skin were eliminated from the microstructure. SEM examination of the sheared sample at high magnifications showed that both the young oxide films and the oxide skins were broken up and dispersed as extremely fine individual particles with a size of 1 2 nm, as presented in Fig. 6. 3.3. Microstructure and mechanical properties Fig. 7 shows the microstructure of a tensile test specimen produced after the optimization of the MC-HPDC process. The casting exhibited a significant improvement in both microstructural uniformity and refinement compared to conventional HPDC castings normally characterized by a non-uniform dendritic structure. As a result of the intensive shearing of melt prior to casting, the primary magnesium particles had a fine spherical morphology. No dendrite fragments or dendrites as the result of the secondary solidification inside the die were found in the microstructure. The optimization of the processing parameters for the MC-HPDC process was confirmed not only by the zero defective rates of the produced castings, seen in Fig. 3, or the refined and uniform microstructure seen in Fig. 7.Itwas also reflected in the uniformity of the mechanical properties of the 2 Fig. 5. High magnification backscattered electron SEM micrographs of the nonsheared AM series scrap showing (a) the large in size MgO particles clusters and (b) the MgO ingot skins, with the element peaks from the EDS analysis shown in the insets. Please cite this article in press as: S. Tzamtzis, et al., Mater. Sci. Eng. A (211), doi:1.116/j.msea.21.12.1
S. Tzamtzis et al. / Materials Science and Engineering A xxx (21) xxx xxx 5 Table 2 Tensile strength and elongation of AM series Mg-alloy scrap recycled with the MC-HPDC process and properties of fresh AM5A and AM6B alloy castings reported in the literature. Alloy Processing UTS (MPa) Elongation (%) Reference AM5A AM6B HPDC HPDC 21 8.5 [14] 21 1 [13] 225.5 ± 33.5 6.9 ± 2.1 [15] 241 ± 7 12.9 ± 2.1 [16] 225 8 [13] 214 ± 43.3 7.5 ± 4.2 [17] 25.8 ± 4.4 9.2 ± 5.5 [18] AM series scrap HPDC 23.9 ± 17.2 12.4 ± 3.4 Current study MC-HPDC 231.5 ± 11.1 14.1 ± 1.9 Current study 3 4 Fig. 6. High magnification backscattered electron SEM micrograph of the MgO particles in the sheared AM series scrap alloy, revealing the very fine size dispersion. samples that were subjected to tensile tests. Fig. 8 shows the UTS and elongation of a series of casting produced with the MC-HPDC process after the processing parameters were optimized. In comparison to Fig. 1, the properties exhibit a high level of consistency. No failures of the tensile test components were observed at unexpectedly low strengths or elongations. Failure was always located at the middle of the gauge length confirming the elimination of the shoulder crack defect. Table 2 summarizes the tensile properties of commercial purity AM5A and AM6B fresh alloy casting reported in the literature. The MC-HPDC process was shown to have excellent potential as a physical recycling technology high grade Mg-alloy scrap, producing casting with tensile properties UTS and elongation comparable to these properties reported in the literature for fresh AM5A and AM6B alloy castings. Fig. 7. Polarized optical micrograph showing the detailed solidification microstructure of AM-series recycled alloy scrap processed at 63 C, 8 rpm for 45 s with the MC-HPDC process. UTS (MPa) 25 2 15 1 5 2 4 6 8 1 12 Sample ID 14 UTS Elongation Fig. 8. Consistency of the mechanical properties after the process optimization, showing the reliability and reproducibility of the MC-HPDC process. 4. Discussion 4.1. Elimination of hot cracking phenomena in recycled AM series alloy scrap Hot cracking, also known as hot tearing or solidification cracking is a major defect in alloy castings and has been comprehensively reviewed by Eskin et al. [19]. Hot cracks nucleate and grow in the late stages of solidification, close to the solidus temperature when the metal is in a mushy zone with 85 95% solid fraction. At that stage, the growing dendritic grains in the conventional HPDC process start to interlock and liquid is isolated resulting in pore and crack formation with further contraction of the solidifying metal [19]. Cao and Kou [2] studied the hot cracking phenomena in binary Mg Al castings containing.25 8 wt. % aluminium concluding that the tendency for hot crack formation is increased at lower solid fractions. In terms of the die-casting cycle in the MC-HPDC process, this means that the nucleation and growth of hot cracks in the casting will start faster once the molten metal is injected into the die. Hence the intensifying pressure needs to be applied earlier to ensure fast and adequate feeding of the regions between the grains and prevent the hot cracking phenomena. This is in agreement with the results presented in Fig. 3(a), where smaller intensifier position leads to elimination of the hot cracks. The optimization of the MC-HPDC process in terms of die temperature and melt processing temperature can be explained according to the grain size refinement and morphology change. A lower die temperature is expected to offer a higher cooling rate and result in a fine grain structure. At the same time, intensive shearing close to the liquidus temperature has been found 16 18 3 2 1 Elongation (%) Please cite this article in press as: S. Tzamtzis, et al., Mater. Sci. Eng. A (211), doi:1.116/j.msea.21.12.1
6 S. Tzamtzis et al. / Materials Science and Engineering A xxx (21) xxx xxx to produce the smallest grain size for magnesium alloys [21]. In addition, it has been shown that this fine grain size produced with the MC-HPDC process is very uniform in the microstructure of the castings [9]. This is also confirmed by the micrograph of the MC-HPDC AM-series alloy scrap seen in Fig. 7. The fine grain structure produced by intensive shearing at a temperature equal to T L +5 C and promoted by the high cooling rates achieved at lower die temperatures, as well as the near spherical and uniform grain structure, will prevent grain interlocking and allow easier feeding of the regions between the solidifying grains at any time during the casting cycle. As a result, less hot cracking susceptibility is expected at those processing conditions. This is experimentally confirmed by the defective rates presented in Fig. 3(b) and (c) as a function of the die temperature and the melt processing temperature. 4.2. Microstructural refinement and mechanical properties of MC-HPDC processed alloys The solidification of magnesium alloys can be described in terms of heterogeneous nucleation by solid particles that exist in the alloy melt. For effective heterogeneous nucleation and grain refinement to take place, the number and size distribution of the potent nucleating particles are critical [22]. According to the free growth theory developed by Greer et al. [23], the undercooling T fg required for achieving the state of free growth is given by: T fg = 4 sl (1) S v d pi where S v is the volumetric entropy of fusion and d pi is the diameter of the potent inoculants particles. Hence, the undercooling for free growth is inversely proportional to inoculants particle size, where the larger particles are more potent for heterogeneous nucleation and are activated first. Under typical solidification conditions, only a small fraction of the particles become active before recalesence occurs. In order to promote the efficiency of inoculation, more particles need to contribute to heterogeneous nucleation; hence a narrow size distribution is desirable. It has recently been shown that MgO particles have the potential to act as nucleation sites for alpha-mg grains [24]. As seen in Section 3.2 melt shearing breaks down the MgO clusters and skins and disperses individual MgO particles with a fine size. This is in agreement with our previous work on intensive shearing of liquid metal, which has demonstrated that solid particle agglomerates added in a metal alloy can be broken down and dispersed uniformly by intensive shearing [25]. In addition, a very recent study by Men and co-workers [26] has both modelled and experimentally confirmed that intensive melt shearing effectively disperses MgO particles and increases the density of active nucleating particles, resulting in a much finer magnesium grain size after solidification. The strength of an alloy can be correlated to the grain size according to the Hall Petch equation [27,28]: y = + k y d 1/2 g where y is the yield strength, is the resistance of the lattice to dislocation motion (or the yield strength of a single crystal), k y is the strengthening coefficient and d g is the grain diameter, and improved mechanical properties are predicted for alloys with fine grain sizes. This relationship between grain size and strength has been confirmed for numerous metals and recent work by Ono et al. [29] confirmed the validity of applying this relationship on polycrystalline magnesium. Thus, it can be used to justify the improved (2) mechanical properties of castings produced with the MC-HPDC process, showing a simultaneous increase in both strength and elongation. 5. Conclusions The potential of the MC-HPDC process to be employed as a physical recycling process for high grade magnesium alloy scrap was investigated in this work, using a mixture of AM5A and AM6B die casting scrap. A successful optimization process was followed by altering the die-casting conditions. It was found that intensive shearing should take place at 5 C above the melt liquidus temperature, the die temperature kept to 18 C and short die filling times should be used by reducing the intensifier position to 18 mm in order to produce defect free castings that exhibit consistent properties. Intensive shearing was found to significantly alter the size and morphology of MgO particle clusters and oxide skins, reducing their size and dispersing them in the matrix. The MC-HPDC process was shown to have excellent potential as a physical recycling technology high grade Mg-alloy scrap, producing casting with consistent tensile properties UTS and elongation which compare very well to the properties reported in the literature for fresh AM5A and AM6B alloy castings. Acknowledgements The authors are grateful to the EPSRC (UK) for financial support. We are also grateful to Dr. Y. Wang for his help with the SEM work. References [1] G. Liu, Y. Wang, Z. Fan, Mater. Sci. Eng. A 472 (28) 251 257. [2] United States Geological Survey, 27 Minerals Yearbook Vol. I Metals & Minerals, United States Government Printing Office, Washington, 28. [3] D. Fechner, N. Hort, C. Blawert, K.U. Kainer, in: S. Beals, A.A. Luo, N.R. Neelameggham, M.O. Pekguleryuz (Eds.) Magnesium Technol. (27) 287 291. [4] C. Scharf, C. Blawert, A. Ditze, in: K.U. Kainer (Ed.), Magnesium, Wiley-VCH, Weinheim, 24. [5] H. Antrekowitsch, G. Hanko, P. Ebner, in: H.I. Kaplan (Ed.), Magnesium Technol. (22) 43 49. [6] A. Javaid, E. Essadiqi, S. Bell, B. Davis, in: A.A. Luo, N.R. Neelameggham, R.S. Beals (Eds.) Magnesium Technol. (26) 7 12. [7] C. Scharf, in: K.U. Kainer (Ed.), Magnesium Alloys and Technologies, Wiley VCH, Weinheim, 23. [8] S. Tzamtzis, H. Zhang, N. Hari Babu, Z. Fan, in: E.A. Nyberg, S.R. Agnew, N.R. Neelameggham, M.O. Pekguleryuz (Eds.) Magnesium Technol. (29) 91 96. [9] S. Tzamtzis, H. Zhang, N. Hari Babu, Z. Fan, Mater. Sci. Eng. A 527 (21) 2929 2934. [1] Z. Fan, G. Liu, Acta Mater. 53 (25) 4345 4357. [11] Z. Fan, G. Liu, Y. Wang, J. Mater. Sci. 41 (26) 3631 3644. [12] Z. Fan, X. Fang, S. Ji, Mater. Sci. Eng. A 412 (25) 298 36. [13] M.M. Avedasian, H. Baker, ASM Specialty Handbook Magnesium and Magnesium Alloys, ASM International, Materials Park, OH, 1999. [14] T. Aune, H. Westengen, in: B.L. Mordike, F. Hechmann (Eds.) Magnesium alloys and their applications (1992) 221 228. [15] J. Song, S.M. Xiong, M. Li, J. Alloy. Compd. 477 (28) 863 869. [16] S. Ji, Z. Zhen, Z. Fan, Mater. Sci. Technol. 21 (25) 119 124. [17] C.D. Lee, Mater. Sci. Eng. A 454 455 (27) 575 58. [18] C.D. Lee, J. Mater. Sci. 42 (27) 132 139. [19] G.D. Eskin, L. Suyitno, Katgerman, Prog. Mater. Sci. 49 (24) 629 711. [2] G. Cao, S. Kou, Mater. Sci. Eng. A (26) 23 238. [21] M. Xia, Y. Wang, H. Li, S. Arumuganathar, Y. Zuo, G.M. Scamans, Z. Fan, in: E.A. Nyberg, S.R. Agnew, N.R. Neelameggham, M.O. Pekguleryuz (Eds.) Magnesium Technol. (29) 135 14. [22] T.E. Quested, A.L. Greer, Acta Mater. 52 (24) 3859 3868. [23] A.L. Greer, A.M. Bunn, A. Tronche, P.V. Evans, D.J. Bristow, Acta Mater. 48 (2) 2823 2835. [24] Z. Fan, Y. Wang, M. Xia, S. Arumuganathar, Acta Mater. 57 (29) 4891 491. [25] S. Tzamtzis, N.S. Barekar, J.B. Patel, N. Hari Babu, Z. Fan, Compos. A: Appl. Sci. Manuf. 4 (29) 144 151. [26] H. Men, B. Jiang, Z. Fan, Acta Mater. 58 (21) 6526 6534. [27] E.O. Hall, Proc. Phys. Soc. B 64 (1951) 747 753. [28] N.J. Petch, J. Iron Steel Inst. 174 (1953) 25 28. [29] N. Ono, R. Nowak, S. Miura, Mater. Lett. 58 (24) 29 43. Please cite this article in press as: S. Tzamtzis, et al., Mater. Sci. Eng. A (211), doi:1.116/j.msea.21.12.1