MA TE RI A L S CH A R A CT ER IZ A TI O N 59 ( 20 0 8 ) 1 7 81 8 7

An investigation on microstructural and mechanical properties

of solid mould investment casting of AZ91D magnesium alloy
S. Lun Sin, D. Dub, R. Tremblay
Department of Mining, Metallurgical and Materials Engineering, Universit Laval, Quebec, Qc, Canada G1K 7P4

AR TIC LE D ATA

ABSTR ACT

Article history:

In this work, AZ91D magnesium alloy was cast in solid plaster mould using vacuum

Received 1 February 2007

assistance. The influence of process parameters and wall thickness on the microstructure

Accepted 24 April 2007

and tensile properties of cast specimens was studied. Within the range of experimental
parameters, it was found that casting and mould preheating temperatures have minor

Keywords:

influence on mechanical properties. However, gating design proved to affect the

Investment casting

microstructure and tensile properties of cast specimens. Compared to bottom filling, top

AZ91D alloy

filling led to casting defects that seriously impaired the mechanical properties. Generally,

Microstructure

tensile properties increased with a decrease of section thickness, this improvement being

Tensile properties

explained by a reduction of dendrite arms spacing and grain size. Finally, the mechanical

Casting defects

properties of solid mould investment cast specimens were compared with those obtained by
foamed plaster casting, sand casting, permanent mould casting and die-casting.
2007 Elsevier Inc. All rights reserved.

1.

Introduction

Investment casting methods are used for the production of

short series of premium quality components having complex
shapes and thin walls [13]. They are also useful for the
fabrication of prototypes, particularly in high-pressure diecasting of magnesium alloys where smooth surface finish and
near-net-shape are essential [4,5].
It is recognised that the mechanical properties of cast
alloys depend on their microstructure: grain size and texture,
secondary phases and porosity. However, there is little
information about the influence of casting parameters on
microstructure and mechanical properties of investment cast
magnesium alloys [69].
Idris and Clegg [8,9] have studied the influence of investment casting process variables on mechanical properties of
magnesium alloys cast into an alumino-silicate mould.
However, no clear relationship between solidification conditions, microstructure and strength was established. Kim et al.
[10] observed a decrease of the grain size and an increase in
the hardness and the ultimate tensile strength with a decrease

of the mould temperature. They also found that any casting

temperature variation on the interval of temperature considered (650710 C) produced little variation on the hardness.
In the case of conventional solid mould investment casting,
Davenport and Orton [11] have studied the mechanical
properties of investment cast AZ63 and AZ92 magnesium
alloys, in the as-cast state. Their properties were similar to
those obtained by sand casting, for specimens having
comparable solidification time. A decrease in the casting
temperature led to better tensile properties. Herrick [6] and
Pellegrini [12] studied the mechanical properties of MgAl- and
MgZr-based alloys respectively. In particular, Herrick [6]
reported that mould temperature (from 20 to 340 C) and
casting temperature (from 670 to 740 C) as well as section
thickness (from 1.6 to 12.7 mm) had minor influence on the
average grain size. However, increasing the investment mould
temperature from 20 to 340 C or the pouring temperature
from 670 to 740 C proved to have a deteriorating influence on
the average mechanical properties of all alloys and tempers
investigated. The negative influence of mould and pouring
temperatures was attributed to the presence in varying

Corresponding author. Tel.: +1 418 656 3533; fax: +1 418 656 5343.
E-mail address: dominique.dube@gmn.ulaval.ca (D. Dub).
1044-5803/$ see front matter 2007 Elsevier Inc. All rights reserved.
doi:10.1016/j.matchar.2007.04.026

M A TE RI A L S CH A RACT ER IZ A TI O N 59 ( 20 0 8 ) 1 7 8 1 8 7

179

quantity of undissolved phases after heat treatment. In the

case of MgAl-based alloys, Herrick [6] found that the tensile
properties were comparable to those obtained by sand casting
methods. Among all the alloys studied, the combination of
strength, ductility and castability was found to be best for
AZ91C alloy [6,13].
Previous papers have presented the results of a study about
the influence of process and mould parameters on the fluidity,
surface finish and reactivity of magnesium alloys produced by
using vacuum-assisted plaster mould investment casting
[14,15]. In the present work, the influence of casting conditions
on microstructure and mechanical properties of AZ91D
magnesium alloy cast was examined. The following experimental parameters were studied: gating design, thickness of
cast specimen, casting temperature and mould preheating
temperature.

2.

Experimental Procedures

2.1.

Production of Cast Specimens

Tensile test specimens were produced using solid mould

investment casting method with vacuum assistance, a
process described in a previous paper [14]. AZ91D ingots
were melted in a 430 stainless steel crucible under CO20.5%
SF6 protective gas mixture. Moulds were also flushed with the
CO20.5%SF6 gas mixture prior to and during pouring of the
molten alloy in order to help reducing mouldmetal reactions
[9,16].
The pouring temperature and mould preheating temperature influence the surface finish of the cast specimens and
mouldmetal reactivity [9,15]. Unless specifically stated otherwise, pouring temperature and mould preheating temperature
were set to 750 and 350 C respectively.

2.2.

Experimental Parameters

2.2.1.

Gating Design

The design of the cluster and its gating is critical with thin
components as filling must be completed before solidification
significantly reduces melt flow. In investment casting, it is
usual to pay relatively little attention to the design of the
filling system of a casting [17]. The top-filled system is widely
adopted, the gating system being usually short and wax
patterns directly connected to the sprue [1719]. In this way,
mould cavities are filled rapidly at the lowest cost. However,
previous works showed that bottom filling reduces melt
turbulence, leading to better mechanical properties [17,19
26]. It must be noted however that the degree of improvement
in mechanical properties depends on the nature of the alloy
studied. Significant improvement was observed in Al-based
alloys, while this effect was less clear for Fe- or Ni-based alloys
[19,26].
In the case of magnesium alloys, such study has not yet
been reported in the literature so far. The first objective of this
study is thus to assess the influence of gating design on
microstructure and tensile properties of AZ91D magnesium
alloy. In this perspective, two types of gating system were used
as illustrated in Fig. 1a. Some specimens were cast using

Fig. 1 Top and bottom filling configurations.

bottom filling (Fig. 1b), for which the molten magnesium alloy
enters the mould at its base, rises up and enters the specimen
cavity more quietly. Other specimens were cast by direct
filling (Fig. 1c), the molten alloy entering the specimen cavity
from the top under the influence of gravity. The thickness of
test specimens was set to 4.3 mm.

2.2.2.

Pouring and Mould Preheating Temperatures

Both pouring temperature and mould preheating temperature

influence mouldmetal reactivity and cooling rate during
solidification [15]. In order to study the influence of pouring
temperature on microstructure and mechanical properties,
2.2 mm thick specimens were bottom filled at a mould
preheating temperature of 350 C and at casting temperatures
of 700, 720, 740, 750 and 760 C. This range of experimental
pouring temperatures was selected in order to prevent
excessive contamination of the melt and ensure complete
filling. The effect of mould preheating temperature was
studied by using 4.3 mm thick specimens, bottom filled at
750 C. Mould preheating temperature was set to 250, 300 and
350 C. It was found with experience that preheating at

180

MA TE RI A L S CH A R A CT ER IZ A TI O N 59 ( 20 0 8 ) 1 7 81 8 7

temperature higher than 350 C was detrimental to surface

finish [15].

2.2.3.

Thickness of Cast Specimens

During solidification, the cooling rate of cast specimens and

their microstructure are influenced by wall thickness [27]. In
this series of experiments, the influence of section thickness
on microstructure and mechanical properties of cast specimens was investigated using both bottom and top filling. The
thickness of specimens was set to 1.0, 1.3, 1.6, 2.2, 3.2 and
4.3 mm for top filling and to 1.6, 2.2, 3.2 and 4.3 mm for bottom
filling. It was not possible to completely fill tensile specimens
thinner than 1.3 mm using bottom filling, within the range of
experimental casting parameters tested during this study.
However, the surface filled was sufficiently large to be able to
observe the microstructure and make porosity measurements
on these specimens.

2.3.

were evaluated from the stressstrain data obtained. The real

elongation at break (ef) was calculated by subtracting the
elastic strain (100 fracture stress / elastic modulus) from the
total elongation, according to ASTM B557M [33]. The tensile
yield strength at 0.2% offset (YS) was calculated using a
nominal elastic modulus (E) value of 45 GPa, since AZ91
magnesium alloy does not exhibit a definite elastic region [34
38].
The fracture surfaces of selected tensile test specimens
were examined using a scanning electron microscope,
equipped with an energy dispersive X-ray analyser (EDS). All
specimens were coated with a thin layer of AuPd to prevent
charging.

3.

Results and Discussion

3.1.

General Microstructure of Cast Specimens

Microstructural Studies

Selected specimens were mechanically polished through

successively finer grits of silicon carbide papers (down to grit
size 1200) and diamond polishing suspension (particle size
down to 0.1 m). Polished specimens were etched with acetic
glycol to reveal the general microstructure [28]. A different
etchant was used to reveal the granular structure [29].
Specimens were examined by optical (OM) and scanning
electron microscopy (SEM). Grain size was measured according to ASTM E112 [30].
Solidification time was estimated by measuring the
secondary dendrite arm spacing (SDAS), using the following
relation obtained for AZ91D magnesium alloy solidified under
various conditions [31]:
k 35:5s0:31

Fig. 2 shows the microstructure of a 1 mm thick specimen cast

at 750 C in a mould preheated at 300 C, revealing
microconstituents typically found in specimens during this
study. Primary equiaxed dendrites of -magnesium are visible
(Fig. 2a). At higher magnification (Fig. 2b), a lamellar structure
consisting of alternate -phase (Mg17Al12) and -magnesium
solid solution is detected along with large intermetallic -

1
1

where (m) is the secondary arm spacing and (K s ) is the

cooling rate during solidification. Measurements of SDAS were
made at a minimum of ten different locations on dendrites
revealing at least four secondary arms. Solidification time was
calculated considering a liquidus of 595 C and a nonequilibrium solidus of 425 C, i.e., a liquidussolidus interval
of 170 C [32].
The porosity of tensile specimens was evaluated by using
density measurements in distilled water (Archimede's
method) and calculated using the following relation:
% porosity 100 

Dt  Da
Da

where Da is the actual density of the specimen and Dt is the

theoretical density of the alloy, taken as 1.81 g cm 3 for AZ91D
magnesium alloys [32]. A minimum of three tensile specimens
were used for these measurements.

2.4.

Evaluation of Tensile Properties and Fractography

Tensile strength was evaluated with as-cast sub-size specimens of rectangular cross-sections having a width of 6.0 mm
and a 25.0 mm gage length. Tensile tests were carried out at a
deformation rate of 0.003 s 1, according to ASTM B557M [33].
The ultimate tensile strength (UTS) and the total elongation

Fig. 2 (a) Optical micrograph of solid mould cast AZ91D alloy;

(b) detailed view of the lamellar structure (1) and Mg17Al12
phase (2).

181

M A TE RI A L S CH A RACT ER IZ A TI O N 59 ( 20 0 8 ) 1 7 8 1 8 7

Table 1 Coefficients of variation of ultimate tensile

strength (UTS), elongation (ef) and yield strength (YS) for
top and bottom filling

Top filling
Bottom filling

CV of UTS (%)

CV of ef (%)

CV of YS (%)

19
3

60
19

3
2

where s is the standard deviation and x is the mean value of

the considered property. Table 1 gives the coefficient of
variation for ultimate tensile strength (UTS), yield strength
(YS) and elongation (ef). It can be seen that the coefficient of
variation is lower for ultimate tensile strength and elongation
in the case of bottom filling, revealing a reduced scatter in
these properties. The comparable coefficients of variation for

yield strength confirm the weak influence of gating design on

this property, in accordance with Campbell [17] and Cceres
et al. [40].
Previous studies on aluminium alloys containing magnesium, which have a relatively high propensity to form oxide
films, produced similar results [19,24,25,41,42]. It was reported
that, with top-filling running systems, the disruption and the
entrainment of oxide films caused by turbulent flow leads to
wide scattering of tensile strength and hence low mechanical
reliability [19,24,25,41,42]. In this work, the detrimental
influence of top filling on ultimate tensile strength, ductility
and elastic modulus is also attributed to casting defects.
Fig. 5 shows the visual aspect of cast tensile test specimens.
It reveals that top-filled specimens exhibit flow marks on their
surface (Fig. 5a), while these features were not visible in the
case of bottom-filled specimens (Fig. 5b). Flow marks are
surface defects which are usually associated with turbulent
flow [43]. With the top-filling gating system, the molten
magnesium alloy falls through the mould in a turbulent,
uncontrolled way, explaining the occurrence of flow marks.
Fig. 6 shows the cross-section of top- and bottom-filled
specimens, which were observed with scanning electron
microscope. It reveals the presence of porosity, inclusions
and oxide films in all cast parts which were top-filled (Fig. 6a),
while these defects were generally not observed in the case of
bottom-filled specimens (Fig. 6b). Quantitative porosity measurements and fractographs confirm the presence of casting
defects in top-filled specimens. As shown in Fig. 7, bottomfilled specimens are less porous than top-filled specimens.
The higher porosity content in top-filled specimens is ascribed
to the turbulent flow which entrains gases during the filling of
mould. Finally, Fig. 8 shows the SEM fractographs of selected
specimens, which revealed features of brittle fracture, with
cleavage and quasi-cleavage as the principal fracture modes
(Fig. 8a and b). For top-filled specimens, the presence of
relatively large round pores was also observed (Fig. 8c), along

Fig. 4 Influence of gating design on ultimate tensile strength

(UTS), elongation (ef) and yield strength (YS) at 0.2% offset of
4.3 mm thick specimens. An error bar corresponding to two
standard deviations is also shown for each measured
property.

Fig. 5 Influence of gating design on occurrence of flow

marks, (a) top-filled specimen showing flow marks (see
arrows) at its surface, (b) bottom-filled specimen showing no
flow mark at its surface.

Fig. 3 Typical stressstrain curves obtained for top- and

bottom-filled specimens.

phase. Cross-section of as-polished specimens also reveals

the presence of intermetallic AlMn based paricles, most of
them being located at the surface of cast specimens [39].

3.2.

Influence of Gating Design

Fig. 3 shows typical stressstrain curves obtained for top- and

bottom-filled specimens. It can be noted that bottom-filled
specimens exhibit higher tensile strength, effective elastic
modulus and elongation at break. As shown in Fig. 4, the
average tensile strength of bottom-filled specimens is 36%
superior to that of top-filled specimens. Bottom filling also
proved to have a remarkable effect on ductility, which was
found to be three times higher as compared with top filling.
However, gating design seems to have only a slight influence
on yield strength.
In order to evaluate the relative dispersion of mechanical
properties, the coefficient of variation (CV) was calculated
using the following relation:
CV

s
x

182

MA TE RI A L S CH A R A CT ER IZ A TI O N 59 ( 20 0 8 ) 1 7 81 8 7

to the slightly reduced cooling rate during solidification. The

influence of mould preheating temperature on grain size,
however, is not conclusive. No significant influence of mould
temperature on ultimate tensile strength, yield strength or
ductility was observed, as shown in Fig. 10b, despite the
moderate but systematic variation of secondary dendrite arm
spacing.

3.4.

Fig. 6 SEM micrograph of (a) top-filled specimen showing

microshrinkage, inclusions and oxide film and
(b) bottom-filled specimen.

with regions containing intact (protruding) dendrites characteristic of shrinkage cavities [44] (Fig. 8d). Porosity was found
mostly along interdendritic arms. For bottom-filled specimens, some casting pores were also observed, but intact
dendrites were much less visible.

3.3.
Influence of Casting and Mould Preheating
Temperatures
Fig. 9a shows the influence of casting temperature on grain
size and secondary dendrite arm spacing in cast specimens. It
reveals that, in the range of temperatures studied, SDAS does
not vary with casting temperature. A slight increase of grain
size with casting temperature was also observed, but it does
not significantly influence the ultimate tensile strength,
elongation or yield strength of cast parts, as shown in Fig.
9b. These results are similar to those obtained by Siaminwe
and Clegg [45] on AlSiMg alloy. These authors also observed
that casting temperature did not influence the microstructure
and had little effect on tensile properties.
Fig. 10a shows the influence of mould preheating temperature on grain size and secondary dendrite arm spacing. In the
range of temperature studied, a slight increase (20%) of
secondary dendrite arm spacing was found as mould preheating temperature was raised, which can be ascribed essentially

Thickness of Cast Specimen

The tensile properties of top- and bottom-filled specimens are

presented in Fig. 11. For section thicknesses above 1.6 mm,
mechanical properties of top- and bottom-filled specimens
follow a similar tendency. It is observed that ultimate tensile
strength and elongation increase as the section thickness
decreases, while yield strength remains relatively constant.
Below 1.6 mm, for top-filled specimens, it is observed that
these properties decrease with a continuing decrease of
section thickness although it would be expected that the
mechanical properties would further increase with a decrease
of section thickness [46,47]. However, a similar trend was
observed in the work of Easton et al. [48] and Abbott et al. [49]
on die-cast AZ91 for UTS and elongation. The variation in
ductility and UTS with section thickness was ascribed to the
presence of various defects in the cast parts. The yield
strength, in their work, tended to be higher for the thinnest
sections and this variation was attributed to the grain size of
the specimen.
In this study, the variation of ultimate tensile strength and
elongation can also be explained using microstructural
considerations. First, for bottom-filled specimens, Fig. 12
shows that grain size and secondary dendrite arm spacing
increases with section thickness. This increase of secondary
dendrite arm spacing is explained by the reduction of
solidification rate, the cooling rate being inversely related to
section thickness [27]. Besides, the size of grains closely
follows the variation of dendrite arm spacing, longer freezing
times promoting coarser microstructure, as shown in Fig. 12.
In Fig. 7, it is shown that the degree of porosity depends on
the section thickness of specimens. Within the margin of
experimental error, porosity does not vary much between 1.6
and 4.2 mm, but significantly increases when the section
thickness decreases to 1.0 mm, this increase being more
important for top-filled specimens. Fractographs (Fig. 8) also

Fig. 7 Influence of section thickness and gating design on

porosity level in investment cast specimens. Error bars
correspond to two standard deviations.

M A TE RI A L S CH A RACT ER IZ A TI O N 59 ( 20 0 8 ) 1 7 8 1 8 7

183

Fig. 8 SEM micrographs of the fracture surface of a tensile test specimens showing (a) and (b) cleavage features (top and bottom
filling), (c) a large spherical pore on the fracture surface (top and bottom filling) (d) exposed intact dendrite (top filling).

revealed that shrinkage porosity, with intact dendrites clearly

visible, is all the more present as the section thickness of cast
parts decreases. The higher porosity level in thinner sections
is ascribed to a more difficult filling.
Finally, as illustrated in Fig. 13a, it was observed that for
section thicknesses above 1.6 mm, the microstructure of
bottom- and top-filled specimens consists of equiaxed grains,
homogeneously distributed within specimens. For section
thicknesses below 1.6 mm, bottom-filled specimens still
present a uniform microstructure. However, in top-filled
specimens, areas with abnormally large grains along with
areas containing much smaller grains were observed (Fig. 13b
and c). These abnormally large grains are approximately
500 m in diameter, much larger than those found in the
thickest specimens cast in this study. Their presence in the
microstructure is detrimental to tensile properties, in agreement with previous studies [50,51]. Their origin is not clear but
is likely related to the presence of oxide films which are
unfavourable nucleation sites.
As for the yield strength, since the grain size and secondary
dendrite arm spacing are simultaneously reduced with the
thickness of as-cast specimens, it is expected that it should be
improved even more efficiently with a reduction in thickness
of specimens, in agreement with a previous study of Couture
and Meier [46]. However, it was observed that it does not vary

much, except at the lowest section thicknesses in top-filled

specimens. The reverse trend observed with thinner specimens in the present work is ascribed to the presence of
microshrinkage, to fine particles or to surface defects, which
have relatively more importance in thin specimens and
reduce the grain size contribution.

3.5.
Comparison of Mechanical Properties of Investment
Cast Parts with Other Casting Methods
From the different results obtained, it can be seen, in the case
of bottom filling, that tensile properties exceed the minimum
requirements of ASTM B403 for AZ91C-F magnesium alloy
(UTS: 124 MPa, YS (0.2%): 69 MPa, elongation not specified) [52].
For comparison purposes, the typical tensile properties of
AZ91D magnesium alloy in as-cast condition for different
casting processes are listed in Table 2. Moreover, Fig. 14
compares the yield strength () of as-cast AZ91 from different
sources [46,49,5359] as a function of the inverse square root of
grain size (d 1/2). A least square regression was calculated
using a model based on the HallPetch relationship [60]:
r r0 kdd1=2

where 0 and k depend on alloy composition. Values of 379

and 64 were obtained for 0 and k respectively with a R2 of

184

MA TE RI A L S CH A R A CT ER IZ A TI O N 59 ( 20 0 8 ) 1 7 81 8 7

Fig. 9 Influence of casting temperature (a) on grain size and

secondary dendrite arm spacing and (b) on mechanical
properties of bottom-filled specimens. Error bars correspond
to two standard deviations.

Fig. 11 Influence of section thickness on mechanical

properties (a) bottom filling, (b) top filling. Error bars
correspond to two standard deviations.

0.77, based on properties measured in MPa for yield strength

and m for grain size respectively. Grain size was comprised
between 10 and 1100 m.
As expected, solidification time is shorter in the case of diecasting, resulting in a finer microstructure and higher
mechanical properties [49,55,57,58]. In the case of permanent
mould casting, the results display significant differences [53
55,58,59]. The higher values of yield strength obtained by
Maltais et al. [53] can be explained by considering the
secondary dendrite arm spacing. In this work, the SDAS was
maintained constant to 16 m, while it was maintained to
40 m in the work of Sasaki et al. [54]. The reduction of the
effective size of primary phase should even more increase the
yield strength. Comparatively to die-casting and permanent
mould casting, the solidification time in plaster mould casting

Fig. 10 Influence of mould preheating temperature (a) on

grain size and secondary dendrite arm spacing and (b) on
mechanical properties of bottom-filled specimens. Error bars
correspond to two standard deviations.

Fig. 12 Influence of section thickness on grain size and

secondary dendrite arm spacing (bottom filling). Error bars
correspond to two standard deviations.

M A TE RI A L S CH A RACT ER IZ A TI O N 59 ( 20 0 8 ) 1 7 8 1 8 7

185

Fig. 14 Influence of grain size on tensile yield strength from

various sources.

provided by a lower mould temperature. However, they are

superior to those obtained by Fantetti et al. [61] in the case of
foamed plaster casting (without vacuum assistance). This can
be explained by the fact that the section thickness used was
higher and that foamed plaster is more thermally insulating
than dense plaster.

4.

Fig. 13 Optical micrographs of (a) a bottom-filled specimen

4.3 mm thick, (b) a top-filled specimen 1.6 mm thick and (b) a
top-filled specimen 1.3 mm thick. The arrow shows
microporosity observed in cast part.

[56] and sand casting [46,49] is higher, which leads to a coarser

microstructure and lower mechanical properties.
In the present work, the results obtained reveal that the
mechanical properties of investment cast AZ91D magnesium
alloy are slightly inferior or comparable to those found for
sand casting and permanent mould casting. The results are
lower than those published by Herrick [6] who found a value of
UTS of 198 MPa and an elongation of 2.5% in the as-cast
condition in the case of AZ91C alloy cast at 670 C in a mould at
20 C. This is probably due to a higher solidification rate

Table 2 Typical tensile properties of as-cast specimens

of AZ91 magnesium alloy prepared using different casting
methods
UTS (MPa) YS (MPa)
Plaster (foamed)
casting
Sand casting
Permanent mould
casting
Die-casting

ef (%)

Reference

103

91

1.0

[61]

153
145
100112
170
233

104
87
8895
93
163

2.5
2.6
0.50.9
1.4
6.0

[38]
[55]
[53]
[59]
[55]

Conclusions

In this work, the influence of process parameters on microstructure and mechanical properties of vacuum-assisted solid
investment cast AZ91D magnesium alloy was investigated.
In the range of temperature investigated, it was found that
casting and mould preheating temperatures do not significantly influence the microstructure and tensile properties of
as-cast specimens.
However, the influence of top and bottom filling systems
on the properties of AZ91D castings were compared. Topfilling systems promote the formation of microporosity and
other casting defects, which appear to adversely affect
ultimate tensile strength and elongation of cast AZ91D alloy.
Ductility was particularly affected, since an elongation three
times lower was obtained, as compared to bottom filling.
However, it was also observed that it was not possible to fill
section thicknesses inferior to 1.6 mm with bottom filling.
An improvement of tensile strength and elongation with a
decrease of section thickness was observed in bottom-filled
specimens. It was essentially ascribed to a reduction of the
grain size and secondary arm spacing. The thinnest top-filled
specimens display decreasing properties. This was ascribed to
the presence of porosity and abnormally large grains.
Finally, the ultimate tensile strength and yield strength
obtained were higher than the minimum requirements of
ASTM B403 for AZ91C-F magnesium alloy. The results were
slightly lower than those obtained by sand casting and
permanent mould casting, but higher than those obtained by
foamed plaster casting.

Acknowledgements
The authors are grateful to the National Research Council
Canada (NSERC) for financial support during this project. The

186

MA TE RI A L S CH A R A CT ER IZ A TI O N 59 ( 20 0 8 ) 1 7 81 8 7

assistance of Mrs. M. Larouche, MM. G. Bureau, D. Marcotte, M.

Choquette and A. Ferland is kindly acknowledged.

REFERENCES
[1] Casting. In: Boyer HE, Gall TL, editors. Metals handbook, desk
edition. Metals Park, Ohio: American Society for Metals; 1985.
p. 23.123.64.
[2] Beeley PR, Smart RF. Investment casting. London: Institute of
Materials; 1995. 486 pp.
[3] Clegg AJ. Precision casting processes. Oxford: Pergamon Press;
1991. 293 pp.
[4] Defouloy D, Monnier L. Prototype parts in magnesium alloys by
express casting (Pices Prototypes d'Alliages de Magnsium en
Fonderie Express). Hommes et Fonderie, 318; 2001. p. 1222.
[5] Ruden T. Prototype alternatives for magnesium die castings.
Die Cast Manage 1991;9:269.
[6] Herrick K. Magnesium-base alloys investment cast properties.
AFS Trans 1961;69:17988.
[7] Campbell JB. Magnesium investment castings save weight.
Mater Methods 1955;41:945.
[8] Idris MH, Yong MS, Clegg AJ. Precision casting of a
magnesium-base alloy. Foundryman 1997;90:1404.
[9] Idris MH, Clegg AJ. Processing and evaluation of investment
cast magnesium-base alloy. AFS Trans 1996;104:23744.
[10] Kim S, Kim M, Hong T, Kim H, Kim Y. Investment casting of
AZ91HP magnesium alloy. Met Mater Int 2000;6:2759.
[11] Davenport WF, Orton GW. Investment casting magnesium.
Iron Age 1950;165:946.
[12] Pellegrini CJ. Investment cast zirconium-bearing
magnesium-base alloys. AFS Trans 1962;70:122934.
[13] Anonymous. Magnesium investment casting. Prec Metal
Molding 1965;43:389.
[14] Lun Sin S, Dub D. Influence of process parameters on fluidity
of investment-cast AZ91D magnesium alloy. Mater Sci Eng A
Struct Mater Prop Microstruct Process 2004;386:3442.
[15] Lun Sin S, Dub D, Tremblay R. Interfacial reactions between
AZ91D magnesium alloy and plaster mould material during
investment casting. Mater Sci Technol 2006;22:145663.
[16] Neelameggham R, Wheeler OK and Weed GB. Magnesium
alloy casting in plaster mold, US Patent No. 4,579,166. 1 Apr.
1986.
[17] Campbell J. Castings. Oxford: ButterworthHeinemann; 1993.
288 pp.
[18] Vainer MI, Lerner YS. Vacuum-assisted investment casting of
AlNibronze. AFS Trans 1999;107:3542.
[19] Cox M, Wickins M, Kuang JP, Harding RA, Campbell J. Effect of
top and bottom filling on reliability of investment castings in
Al, Fe, and Ni based alloys. Mater Sci Technol
2000;16:144552.
[20] Elliott HE, Mezoff JG. Effect of gating design on metal flow
conditions in the casting of magnesium alloys. AFS Trans
1948;56:22345.
[21] Cox M, Harding RA, Campbell J. Optimised running system
design for bottom filled aluminium alloy 2L99 investment
castings. Mater Sci Technol 2003;19:61325.
[22] Dai X, Yang X, Campbell J, Wood J. Effects of runner system
design on the mechanical strength of Al7SiMg alloy
castings. Mater Sci Eng A Struct Mater Prop Microstruct
Process 2003;354:31525.
[23] Dai X, Yang X, Campbell J, Wood J. Influence of oxide film
defects generated in filling on mechanical strength of
aluminium alloy castings. Mater Sci Technol 2004;20:50513.
[24] Green NR, Campbell J. Influence of oxide film filling defects on
the strength of Al7SiMg alloy castings. AFS Trans
1995:3417.

[25] Jolly M, Cox M, Harding R, Griffiths B, Campbell J. Quiescent

filling applied to investment castings. Mod Cast 2002;92:368.
[26] Li DZ, Campbell J, Li YY. Filling system for investment cast Nibase turbine blades. J Mater Process Technol 2004;148:3106.
[27] Flemings MC. Solidification processing. New York: McGrawHill; 1974. 364 pp.
[28] Metals Handbook. Metallography and microstructures, 9th
ed., vol. 9. Metals Park, OH: American Society for Metals; 1988.
p. 42534.
[29] Maltais A, Dub D, Fiset M, Laroche G, Turgeon S.
Improvements in the metallography of as-cast AZ91 alloy.
Mater Charact 2004;52:10319.
[30] ASTM E112. Standard test methods for determining average
grain size. Annual book of ASTM standards, vol. 03.01. West
Conshohocken, PA: ASTM International; 2001. p. 24366.
[31] Dub D, Couture A, Carbonneau Y, Fiset M, Angers R,
Tremblay R. Secondary dendrite arm spacings in magnesium
alloy AZ91D: from plaster moulding to laser remelting. Int
J Cast Met Res 1998;11:13944.
[32] Busk RS. Magnesium products design. New York: Marcel
Dekker; 1987. 544 pp.
[33] ASTM B557M. Standard test methods of tension testing
wrought and cast aluminum-and magnesium-alloy products
[Metric]. Annual book of ASTM standards, vol. 02.02. West
Conshohocken, PA: ASTM International; 2001. p. 41929.
[34] Aune TK, Westengen H. Property update on magnesium die
casting alloys. Magnesium in vehicle design. SAE, Inc.; 1995.
p. 2131.
[35] Avedesian MM, Baker H. ASM specialty handbook: magnesium
and magnesium alloys. Materials Park, OH: ASM International;
1999. 314 pp.
[36] Caceres CH. Material properties and process assessment in
MgAl casting alloys using strengthductility charts. Int J Cast
Met Res 2001;14:18597.
[37] Zhang Z, Couture A. Investigation of the properties of
MgZnAl alloys. Scripta Mater 1998;39:4553.
[38] Carbonneau Y, Lepage M, Renaud J, Dub D. Elastic modulus,
yield strength and elongation of AZ91 alloy. Magnesium 2000.
Proceedings of the Second Israeli International Conference on
Magnesium Science and Technology. Dead Sea, Israel;
February 2224 2001.
[39] Lun Sin S, Dub D and Tremblay R. Characterization of AlMn
particles in AZ91D investment castings. Mater Charact in
press; doi:10.1016/j.matchar.2006.10.010.
[40] Caceres CH, Davidson CJ, Griffiths JR, Newton CL. Effects of
solidification rate and ageing on the microstructure and
mechanical properties of AZ91 Alloy. Mater Sci Eng A Struct
Mater Prop Microstruct Process 2002;325:34455.
[41] Nyahumwa C, Green NR, Campbell J. Effect of mold-filling
turbulence on fatigue properties of cast aluminum alloys. AFS
Trans 1998;106:21523.
[42] Runyoro J, Boutorabi SMA, Campbell J. Critical gate velocities
for film-forming casting alloys: a basis for process
specification. AFS Trans 1992;100:22534.
[43] Hnon G, Mascr C, Blanc G. Recherche de la Qualit des
Pices de Fonderie. Paris: Editions techniques des industries
de la fonderie; 1971. 394 pp.
[44] Lee SG, Patel GR, Gokhale AM, Sreeranganathan A,
Horstemeyer MF. Variability in the tensile ductility of
high-pressure die-cast AM50 Mg-alloy. Scripta Mater
2005;53:8516.
[45] Siaminwe L, Clegg AJ. Effect of processing variables on
structure and tensile properties of investment cast AlSiMg
casting alloy. Mater Sci Technol 1999;15:81220.
[46] Couture A, Meier JW. The effect of wall thickness on tensile
properties of MgAlZn alloy castings. AFS Trans
1966;74:16473.
[47] Sequeira WP, Murray MT, Dunlop GL, StJohn DH. Effect of
section thickness and gate velocity on the microstructure and

M A TE RI A L S CH A RACT ER IZ A TI O N 59 ( 20 0 8 ) 1 7 8 1 8 7

[48]

[49]

[50]

[51]

[52]

[53]

[54]

mechanical properties of high pressure die cast magnesium

alloy AZ91D. In: Automotive Alloys, Das SK, editors.
Proceedings of the 1997 TMS annual meeting, The Minerals,
Metals and Materials Society (TMS). Warrendale, PA; 1997.
p. 16983.
Easton M, Abbott T, Cceres C. The effect of microstructural
features and defects on the ductility of high-pressure die
cast AS21, AM60 and AZ91. Materials Science Forum
2003;419422:14752.
Abbott T, Easton M, Song W. Mechanical behaviour of
cast magnesium alloys. Materials Science Forum
2003;419422:1416.
Kurzydlowski KJ, Bucki JJ. Flow stress dependence on the
distribution of grain size in polycrystals. Acta Metall Mater
1993;41:31416.
Bowles AL, Griffiths JR, Davidson CJ. Ductility and the skin
effect in high pressure die cast MgAl alloys. 2001 TMS
Annual Meeting, New Orleans, LA; February 1115 2001.
ASTM B403. Standard specification for magnesium-alloy
investment castings. Annual book of ASTM standards, vol.
02.02. West Conshohocken, PA: ASTM International; 2001.
p. 34952.
Maltais A, Fiset M, Dub D. Grain refinement of magnesium
alloy AZ91D cast in permanent mold using mechanical
vibrations. Materials Science Forum 2003;426432:52732.
Sasaki H, Adachi M, Sakamoto T. Effect of solidified structure
on mechanical properties of AZ91 alloy. Proceedings of IMA
53: Magnesium- A Material Advancing to the 21st Century;

[55]

[56]

[57]

[58]

[59]

[60]
[61]

187

June 24 1996, vol. 02.02. Ube City, Yamaguchi, Japan:

International Magnesium Association (IMA); 1996. p. 8692.
Zhang Z, Couture A, Tremblay R, Dub D. Microstructure and
mechanical properties of permanent mold and die casting
AZ91 magnesium alloy. International Symposium on Light
Metals as held at the 38th Annual Conference of Metallurgists
of CIM; August 2226 1999. Quebec City, Canada.
Fantetti N, Thorvaldsen A, Couture A. Properties of magnesium
plaster castings. Proceedings of SAE International Congress
and Exposition; February 25March 1 1991. Detroit, MI.
Technical Paper 910413.
Luo A, Hu H, Lo SHJ. Microstructure and mechanical properties
of squeeze cast AZ91D magnesium alloy. Light Met August
2529 1996;1996 [Montreal, Canada].
Lee CD, Shin KS. Effects of precipitate and dendrite arm
spacing on tensile properties and fracture behavior of as-cast
magnesiumaluminum alloys. Met Mater 2003;9:217.
Choi BH, You BS, Yim CD, Park WW, Park IM. Microstructure
and mechanical properties of Ca containing AZ91 magnesium
alloys. Materials Science Forum 2005;475479:247780.
Dieter GE. Mechanical metallurgy. 3rd edition. New York:
McGraw-Hill; 1986. 751 pp.
Fantetti N, Pekguleryuz MO, Avedesian MM. Magnesium
plaster-cast prototypes vs. diecastingsa comparative
evaluation of properties. Magnesium at the Threshold? 48th
Annual World Magnesium Conference; June1821 1991.
Quebec City, Canada.