ASM Materials Solutions 2002 Conference & Exposition,

2nd International Aluminium Casting Technology Symposium,

Columbus, Ohio, USA, October 2002.

Effect of Sr, Na, Ca & P on the Castability of Foundry Alloy A356.2

T. N. Ware
Comalco Research & Technical Support, Melbourne Australia
A. K. Dahle
Division of Materials Engineering, The University of Queensland, Brisbane Australia
S. Charles
M. J. Couper
Comalco Research & Technical Support, Melbourne Australia

Abstract
The effect of small variations in composition on castability has been characterised. Specifically, the effects of strontium (Sr),
calcium (Ca), sodium (Na) and phosphorous (P) were investigated in alloy A356.2. These elements are normally present as an
additive or trace element. A preliminary experimental program was conducted with 19 combinations of elemental levels out of a
maximum possible 96 (4x4x3x2 matrix). The castability was evaluated using Tatur moulds, fluidity spirals, an elbow casting and
a cylindrical casting. Assessment included the amount, size and shape of porosity, grain structure, microstructural phases and
surface appearance. The distribution of porosity was found to depend strongly on Sr content. Combined additions of Na and Sr
did not enhance modification but had a marked effect on porosity. Ca-rich intermetallic particles were observed at high levels of
Ca. An effect of P on porosity was also noted.

Introduction understand the implications of small variations in Sr, Na, Ca

and P on the ability to make a sound casting.
The castability of an alloy incorporates a number of aspects
but essentially refers to the ability of an alloy to form a sound Background
casting. The soundness of a casting is usually evaluated in
terms of a number of mechanical properties as well as its Modification of Al-Si alloys with trace amounts of Sr or Na is
conformance to dimensional design requirements. The an accepted means of enhancing the material properties. It has
mechanical properties not only depend on the material long been observed that this practice is often accompanied by
properties derived from microstructure, but also the continuity increased tendency for castings to contain microshrinkage
of the material in terms of the presence of void defects such as porosity. Recent studies suggest that the development of
porosity. Porosity may be caused by air entrapment, localised microshrinkage porosity is related to the formation of eutectic
shrinkage, hydrogen evolution during solidification, or a between primary dendrites in the final stages of solidification
combination of all. [1]. A eutectic solidification interface extends in the growth
direction creating a eutectic mushy zone. The nature of the
Two important aspects of castability that control the primary dendrites and the mode of eutectic nucleation and
susceptibility of an alloy to form shrinkage defects are the growth affects the feeding to this zone and hence porosity
fluidity and feedability of an alloy. Fluidity refers to the formation.
ability of the alloy to fill a mould. Poor fluidity is usually
associated with macroshrinkage defects at the surface of The eutectic has been shown to nucleate in three ways in
castings. Feedability refers to the ability of molten metal to be hypoeutectic Al-Si alloys [2]: on primary aluminium
fed to the solid / liquid interface to accommodate for dendrites, at the mould wall and at independent heterogeneous
contraction during solidification. Porosity stems from sites in interdendritic spaces. The eutectic solidification mode
microshrinkage due to difficulties in interdendritic feeding. has been shown to be strongly dependent on Sr levels. It was
Poor feedability results in distributed porosity in castings. found that the eutectic grew from primary dendrites in
unmodified alloys, whilst the eutectic grew independently
Alloy composition is important for the control of castability as between primary dendrites in Sr modified alloys (levels of 70
small variations have been shown to not only affect the final and 110ppm). Similar to unmodified alloys, the eutectic grew
microstructure of a casting, but also its tendency for porosity from the primary dendrites in alloys containing very high
formation. This work characterises the castability of A356.2 amounts of Sr (500ppm).
in terms of microstructure, feedability and fluidity to

Ware, 1 of 10
Nucleation of the eutectic on primary aluminium dendrites to promote a random distribution of porosity throughout a
(unmodified or overmodified with Sr) is likely to promote a casting without surface slumping.
concentration of porosity in the hot spots of the casting. The
eutectic would grow with the primary dendrite solidification For the combined effect of Ca in Sr modified alloys, Velasco
front causing the bulk of the shrinkage to be accommodated in et al. reported that Ca enhances the modification effect of Sr
the last regions of the casting to solidify. However, primary (for Ca contents between 0 and 130 ppm and Sr between 0 and
dendrite coherency may limit interdendritic feeding during 312ppm) [10]. However, Abdollahi and Gruzleski found that
eutectic solidification resulting in some distribution of the Ca and Sr (50 ppm) interact to reduce the effectiveness of each
porosity. other [11]. Nakae et al. also reported an interaction between
Ca and Sr and suggested that they react with each other to
Independent heterogeneous nucleation of eutectic in form AlSrSiCa-silicide particles in the liquid [12].
interdendritic spaces (for intermediate Sr levels) may increase
the tendency for surface slumping and shrinkage cavities. P additions to hypoeutectic alloys are regarded as deleterious,
Since eutectic nucleation may occur anywhere between resulting in acicular eutectic structures [13]. AlP particles act
dendrites, feeding between dendrites may be directed from as Si nucleation sites, reducing the modifying effect of Sr or
mould walls. This type of growth mode would produce a Na. It has been reported that increased Sr or Na modifier
random distribution of porosity along eutectic grain additions can act to neutralise the AlP particles to balance
boundaries. their deleterious effect [14, 15]. P has also been reported as
detrimental to fluidity [13], particularly at high levels.
Argo et al. have also observed changes in the nature of
shrinkage porosity with increased Sr additions to A356 alloys. It is impossible to evaluate castability through one measure
A dispersive effect on porosity was observed in sand cast bars alone. A range of tests for an alloy is required to characterise
with 30-120ppm Sr [3]. Sample densities measured at a all aspects that affect castability, including feedability, fluidity
specific point increased with Sr addition up to 120 ppm. For and material properties. In the present study, the castability of
samples containing 120-360ppm Sr the porosity distribution alloy A356.2 was evaluated by using a suite of castings
changed and the density of samples declined with increasing summarised in Table 1. Tatur castings have previously been
Sr. The change in porosity was attributed to either used to quantify the difference between unmodified and Sr
overmodification effects or the entrainment of non-metallics modified alloys in terms of microporosity and macroshrinkage
acting as nucleants for porosity. [3]. In the current study this type of casting was used
qualitatively to observe the characteristics of porosity
The use of Na to modify A356.2 alloys has been seen to formation. The elbow mould was originally developed to
produce equivalent microstructural properties in the eutectic model the spoke-rim junction in a wheel casting, an area
and a similar trend for increased porosity formation as Sr. It where shrinkage defects commonly occur.
has been found that less Na is required to achieve the same
optimum level of eutectic modification compared to Sr Table 1. Summary of mould types used to evaluate
(200ppm Sr compared to 100ppm Na) [4]. Despite similar castability of alloys.
microstructures, Na modified alloys have been observed to Mould Mould Material Purpose
have an increased propensity to microporosity formation Type
compared to Sr modified alloys [5]. Tatur Stainless Steel Assess feedability -
observe porosity and
Nucleation of eutectic from mould walls has been found in Na cavity characteristics
modified alloys [6] and tends to prevent surface slumping of a Elbow Stainless Steel Assess feedability -
casting as a solid reinforcing shell is developed early during observe porosity and
solidification. Formation of porosity would tend to occur cavity characteristics
away from the mould walls. Cylindrical Graphite Metallography sample-
observe microstructure
The effect of Na in Sr modified alloys has not been well Spiral Sand Quantify fluidity
quantified. Griffiths et al. reported no deleterious reaction
between Na and Sr, but rather that a combination is beneficial
in giving prolonged modification (less fading)[7]. Wang and Experimental
Gruzleski also reported no deleterious effect in combining Na
and Sr [8]. A356.2 ingot from Comalco - Bell Bay Smelter was used as
the base alloy for the experiments. About 80 kg of the base
For 356 alloys containing up to 90ppm Ca with low Sr alloy ingot was melted in an electrical resistance furnace and
(<10ppm), the eutectic grains have been found to nucleate held at 740oC. No filtering or degassing was performed. Table
independently in the interdendritic liquid but growth 2 shows the composition of the base alloy.
predominantly occurs from the mould walls to the centre of
the casting [9]. This type of eutectic solidification is expected

Ware, 2 of 10
Table 2. Chemical analysis of the base alloy Table 3. Targeted and actual compositions for the
(Wt%) sample taken after initially remelting the ingot Tatur, Elbow and metallographic castings.
(unmodified). Actual Composition
Variant (ppm)
Si Mg Fe Ti B Ca Na Target Condition (ppm)
#
7.09 0.34 0.13 0.12 0.0009 0.0002 0.0001 Sr Na P Ca
1 Unmodified 11 1 6 2
Sr V Mn P Cu 2 100Sr 105 1 5 3
0.0011 0.011 0.0011 0.0006 0.0017 3 300Sr 364 1 6 4
4 300Sr + 25Ca 397 1 6 32
The following compositional levels (ppm) for Sr, Ca, Na and
5 300Sr + 50Ca 429 2 8 58
P in A356.2 alloy were selected for investigation:
6 300Sr + 100Ca 334 2 6 113
Sr - 0, 100, 200, 300
Ca - 0, 25, 50, 100 7 50Na 21 39 6 13
Na - 0, 20, 50 8 100Sr + 20Na 89 23 6 11
P - 0, 10 9 100Sr + 50Na 74 62 5 12
10 200Sr + 20Na 154 38 6 10
As a preliminary investigation into possible effects from these 11 50Na + 50Ca 17 49 6 58
elements, 19 of the possible 96 combinations of selected levels
12 200Sr + 20Na + 50Ca 168 26 6 50
were tested.
13 200Sr + 50Na + 10P + 25Ca 233 53 14 39
Table 3 and Table 4 detail the compositions targeted and 14 200Sr + 50Na 156 57 6 10
actual compositions achieved for the experiments. Alloying 15 300Sr + 20Na 263 27 6 9
additions were made successively to the melt where 16 300Sr + 50Na 253 42 6 9
appropriate. Chemical analysis samples were cast and 17 200Sr 192 4 6 3
analysed by optical emission spectroscopy. In some cases the 18 200Sr + 10P 173 3 10 2
composition of the melt was assured prior to pouring. 19 200Sr + 10P + 25Ca 207 3 14 41

Sr was added as AlSr10 waffles, Ca as AlCa10 waffles, Na as

Nalok tablets, and P as Al-15Fe-5P tablets. Na levels were Table 4. Targeted and actual compositions for the
varied by controlling the holding time after addition. Spiral castings.
Sometimes compensation for melt losses of some elements Actual Composition
Variant
Target Condition (ppm)
was required to maintain the targeted levels (eg. Sr). The melt #
Sr Na P Ca
was stirred well on alloying and prior to pouring, and the
20 Unmodified - - - -
surface of the melt was skimmed well prior to pouring.
Composition targets were not easy to achieve, especially for 21 200Sr 188 1 6 1
Na where the exact composition of the master alloy and the 22 200Sr + 50Na 169 39 6 2
recovery of the alloy under these test conditions was 23 200Sr + 20Na + 10P + 25Ca 200 16 14 30
unknown. For the levels requiring no specific addition (ie. 24 200Sr + 50Na + 10P + 25Ca 180 37 12 24
Zero levels), the residual amounts in the melt were below
10ppm. The samples for metallographic examination were sectioned
1cm from the bottom and prepared by standard metallographic
Two Tatur castings, two Elbow mould castings, one techniques for examination in an optical microscope. Selected
metallography sample, and two Ransley hydrogen samples samples were analysed using SEM and WDS for composition
were produced for each melt composition. The Tatur mould of particles.
was preheated to a temperature of 400oC, while the Elbow
mould was preheated to 300oC. The temperature of the Elbow The Tatur and Elbow mould samples were sectioned through
mould was verified with a contact thermocouple prior to the centre and carefully machined to give a clean surface. A
pouring. A sample for metallographic investigation was cast in schematic of the Elbow and Tatur sections showing features
a cylindrical graphite mould preheated to the same can be seen in Figure 1. Elbow sections were photographed
temperature as the melt. It was placed on an insulating and Tatur sections were scanned and manipulated with an
fiberfrax felt with a fiberfrax lid to ensure radial heat flow outline function to demonstrate porosity distributions.
with a cooling rate of about 1 K/s.
Selected Elbow sections were etched with Poultons Reagant
Sand moulds were used for the spiral testing. Three spirals for 1 minute to reveal the grain structure and solidification
were cast for each composition. The pouring temperature was patterns of castings for subsequent relation to porosity
720oC1 and was measured prior to pouring each spiral. characteristics.

Ware, 3 of 10
 central
region / The porosity distributions in Tatur sections, Figure 2, showed
hot spot that increasing amounts of Sr caused more distributed porosity
hot with a reversion to more concentrated porosity occurring for
spot the higher Sr sample (364 ppm). The 192ppm Sr sample
rim displayed a completely random distribution of porosity, while
the 105ppm Sr sample still showed some concentration of the
porosity towards the centre of the casting.
base
casting Sr additions were seen to increase the number and size of
length top porosity in Elbow sections (refer to Figure 3). No porosity
junction was observed close to mould walls. A large amount of
interconnected shrinkage porosity was observed in the hot spot
Figure 1. Schematic of Elbow and Tatur sections
region of the Elbow section containing 105 ppm Sr (Figure
showing characteristic casting regions.
3b). The size and amount of porosity decreased with Sr
addition. There was some suggestion that the porosity
Results and Discussion increased for the highest Sr containing sample. One 192 ppm
Sr casting had small external tears at top and base of the
I. The Effect of Sr modifier in A356.2 Alloys junction.
The effects of Sr were observed in samples containing 105,
192 and 364 ppm Sr (variants 2, 17 and 3 in Table 3) and were
compared to unmodified samples (variant 1).

Microstructural examination showed that a Sr level of 105ppm

was sufficient for good modification of the eutectic.
Increasing the Sr level further resulted in more dark particles,
probably Sr-containing, on the eutectic boundaries. The
samples with the highest level of Sr (364ppm) displayed some
coarsening of the eutectic. a) b)

c) d)
Figure 3. Photographs of Elbow sections showing
porosity distributions a) unmodified b) 105ppm Sr c)
192ppm Sr d) 364ppm Sr.
a) b)
Optimum mechanical properties have previously been
reported between 50-150ppm Sr with a decline in properties
above this level [16]. The reversion for porosity reported at
around 120ppm Sr [3] and the formation of an intermetallic
(Al4SrSi2) above 300ppm Sr [16] may contribute to the
reported change in mechanical properties. This is consistent
with findings in the present study where a reversion to higher
levels of porosity in a more concentrated section of the casting
is observed at the high Sr level (364 ppm). It is also consistent
with the increase in small dark particles observed with
increased Sr additions.
c) d) The Tatur castings at the low Sr level (105 ppm) still exhibit
Figure 2. Scanned images of Tatur sections showing some concentration of porosity at the centre of the casting.
porosity distributions a) unmodified b) 105ppm Sr c) This could be explained by there still being relatively good
192ppm Sr d) 364ppm Sr. feeding of the eutectic mushy zone extending from the mould
wall for lower Sr levels under these casting conditions. The

Ware, 4 of 10
higher Sr level (192 ppm) displays completely distributed
porosity suggesting reduced feeding of the eutectic mushy
zone, probably from increased nucleation events of eutectic
colonies blocking feeding channels. The reversion back to
some concentration towards the centre of the Tatur castings
for the high Sr level (364 ppm) could be explained by a
change in the eutectic nucleation mode (from independent
interdendritic nucleation to nucleation on primary aluminium
dendrites). This mode may provide better feeding to the a) b)
mushy zone growing from the mould wall leading to a Figure 5. Photographs of Elbow sections showing
concentration of porosity towards the hot spot of the casting. porosity distributions a) 39ppm Na b) 49ppm Na+58ppm
These trends in the relative porosity distribution are consistent Ca.
with observations in the Elbow castings.
III. The Effect of Ca in Na modified Alloys
II. Na modification of A356.2 The effect of Ca in Na modified alloys was evaluated by
Samples containing 39ppm Na (variant 7) were examined and comparing variants 7 and 11 (Table 3).
compared to unmodified samples (variant 1).
The addition of Ca to Na samples did not produce any
An addition of 39ppm Na was observed to give a very well significant observable changes in microstructure compared to
modified eutectic microstructure equivalent to Sr modified the Na modified alloy. No Ca-containing particles were found
samples. using SEM. The porosity appeared slightly less and smaller in
Tatur sections (refer to Figure 4). Smaller and finer porosity
The porosity distribution in Tatur sections revealed that Na concentrated closer to the junction was observed in Elbow
modification results in an increased amount and complete sections (refer to Figure 5).
distribution of porosity compared to the unmodified alloy
(shown in Figure 4). Na modification would appear to give an Figure 6 shows the effect of Na and Ca on the surface
equivalent volume of porosity to that observed in the Sr appearance of Tatur castings. The casting containing Na only
modified samples (compare Figure 2 with Figure 4 a)). shows a small amount of dull grey interdendritic pattern
(Figure 6a)). Slightly more discolouration is observed in the
Na modification was observed to increase the number and size Na + Ca sample (Figure 6b)). Discolouration in an
of porosity in Elbow sections (compare Figure 3a) with Figure interdendritic pattern is indicative of eutectic clustering at the
5 a)). An external tear was observed at top of the junction in mould wall and would suggest the propensity of the eutectic to
one casting with Na. There also appeared to be some nucleate and grow from the mould wall. This suggests that
concentration of porosity towards the top region of the clustering of eutectic at the surface occurs to a greater extent
junction. in castings containing combined Na and Ca addition compared
to Na modified only.
The tendency for Na to promote eutectic growth from the
mould walls may explain the concentration of porosity It would appear that in these castings, Ca additions to Na
towards the top region of the junction (hot spot - last region to modified alloys provide some positive effect on porosity. If
solidify). This may also explain the appearance of the external Ca further promotes eutectic nucleation and growth from the
tear at the top of the junction caused by the premature mould walls, then this may explain the observed surface
solidification and shrinkage of the spoke region before the appearance of Taturs and the further concentration of porosity
junction. towards the junction in Elbow sections.

a) b) c)
Figure 6. Photographs of surface appearance for
lower external section of various Tatur castings for a)
39ppm Na b) 49ppm Na+58ppm Ca c) 429ppm Sr +
a) 58ppm Ca
b)
Figure 4. Scanned images of Tatur sections showing
porosity distributions a) 39ppm Na b)49ppm Na+50ppm
Ca.

Ware, 5 of 10
 IV. The Effect of Na in Sr Modified A356.2
The trends in samples containing base levels of approximately
100, 200 and 300ppm Sr with the addition of increasing
amounts of Na were evaluated (variants 2, 8 & 9; 17, 10 & 14;
3, 15 & 16 in Table 3).

Combining Na with the lower Sr levels (~100ppm Sr)

appeared to give some refinement of the eutectic
microstructure. Na otherwise did not appear to cause extra a) b)
modification of the eutectic in Sr-containing alloys. Figure 8. Photographs of Elbow sections showing
porosity distributions a) 154ppm Sr+38ppm Na b)
Porosity distributions in Tatur sections showed that combining 156ppm Sr+ 57ppm Na
Na with Sr gave an increase in porosity (Figure 7a) and b)).
Very large pores were observed for the high Na variants The morphology of pores observed in Elbow sections also
(Figure 7c)). There was a shift from concentrated porosity in appeared to be affected by Na additions. Distinct differences
the central region to completely distributed porosity with Na in pore shape across the spoke sections of the Elbow castings
addition in the ~100Sr and ~300Sr base case castings. In the were observed. Pores either appeared distinctly directional or
~200Sr base case, porosity remains distributed with and random in shape (refer to schematic in Figure 9). Pores
without Na. Castings containing the higher levels of Na appeared random in shape for samples containing Sr only,
showed instances of interconnected porosity / tearing in the whilst pores appeared fine and directional in shape for samples
central region (shown in Figure 7b) and c)). containing a combination of Na and Sr (refer to Figure 9).
The appearance of fine, directional porosity started at the top
of the junction in Elbow sections and continued across the
spoke section (shown Figure 9). Samples containing Na only
showed some indication of directionality in porosity, however
it was not as distinct as for the Sr and Na combination.

a) b)

d)
c)
Figure 7. Scanned images of Tatur sections showing
porosity distributions a) 192ppm Sr b) 154ppm Sr+38ppm
Na c) 156ppm Sr and 57ppm Na d)168ppm Sr+26ppm a) b)
Na+ 50ppm Ca. Figure 9. Photographs of etched Elbow sections
with schematics to explain porosity morphology a) 364ppm
Small additions of Na to Sr containing alloys caused finer and Sr (random porosity) b) 263ppmSr + 27ppm Na
more distributed porosity along the spoke region of Elbow (directional porosity). Sections etched with Poultons
sections (up to 38ppm Na) (compare Figure 3 c) to Figure 8). Reagant for 1 minute.
Higher Na levels caused reversion back to more concentrated
porosity towards the top of the junction (hot spot region). The grain structure of Elbow sections revealed by etching
This was the case for base levels of ~100, ~200 and ~300ppm showed that the sample containing Na and Sr had a
Sr. significantly more refined grain size than the sample
containing Sr only (refer to Figure 9). One of the following

Ware, 6 of 10
mechanisms may therefore explain why porosity appears in solidification. The tendency for clustering appears higher
distinctly directional in Elbow castings containing Na: for higher Na additions.

- Nucleation and growth of the eutectic occurs from the The porosity distributions in cast sections also suggest that Na
mould walls early in solidification. As the solidification changes the nature of solidification when combined with a Sr
front moves from the walls inwards, feeding of the modified alloy. Not only is the size and number of pores
eutectic channels occurs perpendicular to the spoke along affected by Na levels, but the morphology of the pores also
the growing primary dendrites. Therefore, any blockage appears to change in the Elbow sections. These trends in
to the eutectic feeding channels results in porosity with a casting characteristics appear consistent despite no significant
biased shape in the perpendicular direction. enhancement to the modification of the eutectic
microstructure.
- Na additions may promote grain refinement. As the
dendritic structure develops and the eutectic solidifies The observed changes suggest that Na additions to a Sr
along the length of the spoke region, the pressure head modified alloy alters the contributing mechanisms to eutectic
that develops in the casting causes a series of collapses solidification in the castings. Increased Na additions appears
in the dendritic network (burst feeding mechanism). to cause a progressive dulling of Tatur surfaces in an
Interdendritic feeding is restricted at the boundaries of interdendritic pattern and concentration of porosity in Elbow
these collapsed regions resulting in small shrinkage castings. These trends could be explained by the progressive
cavities perpendicular to the spoke. effect of Na to promote eutectic nucleation and growth from
the mould wall in Sr modified alloys.
The surface appearance of Tatur castings also appeared to be
affected by Na additions (shown in Figure 10). All Tatur V. The Effect of Ca in Sr Modified A356.2
castings containing amounts of Na (23-62 ppm) had a dull The effect of Ca on Sr modified A356.2 was observed in
grey discolouration in an interdendritic pattern. Sr modified samples containing a base level of approximately 300ppm Sr
alloys without Na only have a small amount of the dull grey and 32-113ppm Ca (variants 3, 4, 5 and 6 in Table 3).
pattern. Figure 10 shows the progression of surface
discolouration from none in the unmodified alloy to a small Microstructural evaluation of metallographic samples showed
amount in the Sr modified alloy to increasing amounts of that Ca addition to Sr modified alloys did not yield extra
discolouration with increasing Na additions. The castings modification of the eutectic. In the high Ca containing sample
containing high Na levels were almost completely dull grey. (113ppm), large needle-shaped Ca-Si-Mg particles were
This was found to be the case for all the base levels of Sr identified (example shown in Figure 11). Blocky Ca-Si-Mg
tested. particles were also observed.

Small Ca-rich needle-like particles appeared at levels as low

as 32ppm Ca (shown in Figure 12). Clusters of blocky Ca-Si
particles and Ca-Si-Mg script particles 25-30 m in size were
also present.

Initial Ca additions to the Sr alloy produced no obvious

changes to the porosity in Tatur sections (distributed with
some concentration towards the hot spot). There was some
a) b) evidence of increased porosity for the very high Ca sample.
Elbow sections showed some evidence of finer and more well
distributed porosity (particularly in the very high Ca sample).
Ca additions showed no discolouration to Tatur surfaces but
did show a characteristic wrinkling in surface texture (refer to
Figure 6 c).

c) d)
Figure 10. Photographs of surface appearance for
lower external section of various Tatur castings for a)
unmodified b) 192ppm Sr modified c) 154ppm Sr +38ppm
Na d) 156ppm Sr + 57ppm Na

The surface appearance of Tatur castings suggests that Na

additions promote eutectic clustering at the mould wall early

Ware, 7 of 10
 effect of Sr and Ca on the eutectic solidification mode is not
reported. The distribution of porosity in these castings could
be consistent with a combined mode of eutectic solidification
with poor feeding to the mushy zone extending from the
mould wall resulting in a distribution in the porosity.

To further understand porosity formation in castings

containing Ca and to further investigate the appearance of
finer more distributed porosity with increased Ca additions,
two Elbow sections containing two different Ca levels were
etched to reveal the grain structure (refer to Figure 13). Ca
additions are seen to result in a finer grain structure than
castings containing Sr only (compare to Figure 9a)).
However, the castings with Ca and Sr are not as well refined
as the casting containing Na and Sr (compare to Figure 9b)).
Further increasing Ca in the castings does not appear to affect
grain size. The difference in porosity therefore can be
Si explained by one of the following mechanisms:

- Different Ca levels promote different modes of eutectic

Mg nucleation and therefore result in a difference in porosity
Ca distribution.

- Ca containing particles that form in castings with higher

Ca levels promote porosity formation and hence change
the relative distribution of porosity compared to castings
Figure 11. SEM micrograph of 334ppm Sr + that contain less Ca-containing particles.
113ppm Ca containing sample showing large Ca-Si-Mg
needle-particle, corresponding WDS spectrum shown for
particle.

Ca-Si-Mg
particle

Figure 13. Photographs of etched Elbow sections a)

alloy containing 397ppm Sr +32ppm Ca, b) alloy
containing 334 ppm Sr + 113 ppm Ca. Sections etched
with Poultons Reagant for 1 minute

VI. The Effect of Ca in Sr Modified Alloys Containing Na

The effect of Ca on a Sr modified alloy containing Na
Figure 12. SEM micrograph of 397ppm Sr + 32ppm (variants 10 and 12) was evaluated and compared to the effect
Ca containing sample showing small Ca-Si-Mg needle- of Ca on a Na modified alloy (variant 11).
particle.
An increased number of larger dark particles appeared in the
The appearance of needle-like Ca-containing particles is microstructure when Ca was added to an Sr-Na alloy. Plate-
expected to be deleterious to mechanical properties. Ca in Al- like particles were seen to be similar to those observed in Sr-
Si alloys forms CaSi2 intermetallic, which is thought to cause Ca samples. The addition of Ca would appear to have a
reduced foundry and mechanical properties. This is consistent dispersive effect on the porosity distribution of Sr-Na alloys in
with the formation of plate-like and blocky particles observed Tatur sections (refer to Figure 7). Elbow sections show an
at levels above 32ppm Ca. equivalent amount of distributed porosity for the Sr-Na and
Sr-Na-Ca combinations.
It is reported for alloys containing Ca (up to 90ppm) with low
Sr (<10ppm), the eutectic grains nucleate independently in the The Na-Ca combination was found to have significantly less
interdendritic liquid but growth predominantly occurs from the porosity in Tatur sections than the Sr-Na-Ca combination.
mould walls to the centre of the casting [9]. The combined Elbow sections showed porosity tending to be concentrated

Ware, 8 of 10
towards the junction in Na-Ca samples compared to a VIII. Spiral Testing
complete distribution along the spoke in the Sr-Na-Ca sample. A summary of the cast lengths for the three cast spirals for
Ca appears to increase the feedability in Na modified alloys each alloy variant is plotted in Figure 6. Note that the alloys
but the same effect is not clear in Sr modified alloys. The were obtained by successive additions to the same melt.
slight dispersive effect of Ca in Sr modified alloys containing
Na is likely to be a result of increased nucleation sites for 106
porosity formation. 104
102
VII. The Effect of P in Sr Modified A356.2

Length (cm)
100
The effect of P on Sr modified A356.2 was observed for
98
samples containing ~200ppm Sr and ~10ppm P (variants 17
and 18). The combined effect of P, Ca and Na in a Sr modified 96
A356.2 was also observed (variants 13 and 19) 94
92
P additions did not appear to significantly affect the 90
modification level or eutectic structure. Tatur sections showed
that the addition of P to a Sr containing alloy caused an d ified pm S
r
pp m
Na
pm C
a
pm C
a
u nm o 1 88 p Sr+39 P +30 p P +24 p
p m 4p pm 2p pm
increase in porosity and pore size (refer to Figure 14), but a 1 69 p Na+1 Na+1
subsequent Ca addition seemed to reduce porosity with pores r +16 pp m r +37 pp m
pm S pm S
remaining large. A further P addition to a Sr-Na-Ca modified 2 00 p 1 80 p
alloy produced a poor casting.
Figure 15. Results from spiral testing (average of
three tests for each alloy). Range of spiral lengths
measured at each composition shown as lines (3per
condition).

For these castings, modification seems to give a significant

improvement in fluidity. The addition of Na, P and Ca did not
cause a large difference to the flow length of the metal. For
the castings in the current study there is some suggestion that
further trace amounts of Na, Ca or P are detrimental to the
fluidity of Sr modified alloys, however more data is required.
a) b)
Figure 14. Scanned images of Tatur sections showing
porosity distributions a) 192ppm Sr with 6ppm P b) Conclusions
180ppm Sr and 14ppm P.
The following conclusions were established for trace element
Elbow sections show some evidence that P improves effects on castability in A356.2 alloys:
distribution of porosity up to wall edges (from Sr modified Sr:
only). The extra addition of Ca further distributes pores along - For Sr modified alloys (no P, Ca, Na additions), a level of
the spoke region (possibly slightly finer). An extra Na ~100ppm Sr gives good eutectic modification, but a
addition showed evidence of larger pores concentrated complete distribution of porosity in experimental castings
towards the top of the junction. was not achieved at this level. Distributed porosity only
occurred in the Sr modified samples at an intermediate Sr
Although it has been reported that P additions result in level (192ppm). Further concentration of porosity in
acicular eutectic structures in hypoeutectic alloys [13], the castings occurs at high Sr levels (364ppm). The extent of
addition of P did not appear to affect the level of eutectic eutectic refinement by modification was the same at all Sr
modification. However, the nature of the porosity in the levels. Changes to the porosity distribution at different Sr
castings appeared to change. Any effect on the level of levels could be explained by the mode of eutectic
modification was probably not apparent, as the amount of solidification and the relative feeding to the mushy zone
modifier present in the alloys was sufficient to eliminate its during solidification.
effect while maintaining the level of modification. This is
consistent with reports that show any deleterious effects of P Na:
on the modifying effect of Sr and Na can be solved with - Na appears to give equivalent modification to Sr modified
higher levels of modifier addition [14, 15]. alloys for the levels tested, and an equivalent amount of
porosity to the Sr modified castings tested.
- Na changes the nature of porosity when combined with a
Sr modified alloy, despite no significant enhancement to

Ware, 9 of 10
 eutectic modification. Not only are the size and number
of pores affected by Na levels, but the morphology of the [2] A. K. Dahle, K. Nogita, J. W. Zindel and L. M.
pores would also appear to change in the elbow sections. Hogan, Metallurgical and Materials Transactions A,
The appearance of a dark grey interdendritic pattern on vol 32A, 2001, pp. 949-960.
the surface of all Tatur castings containing Na is [3] D. Argo and J. E. Gruzleski, AFS Transactions, vol
indicative of eutectic colonies at the mould wall. 98, 1988, pp. 65-74.
Observations suggest that Na promotes a mechanism of [4] F. Paray and J. E. Gruzleski, AFS Transactions,
increased eutectic nucleation and growth from the mould vol.102, 1994, pp. 833-842.
wall early during solidification. The mechanism would [5] R. Fuoco, H. Goldenstein and J. E. Gruzleski, AFS
appear to be more dominant in Na-Sr alloy combinations Transactions, vol.102, 1994, pp297-306.
than Na modified alloys only. The finer grain size in Sr [6] K. Nogita, S.D. McDonald, J. W. Zindel and A. K.
modified Elbow castings containing Na explains the Dahle, Materials Transactions, vol. 42 (9), 2001, pp.
refinement in porosity. 1981-1986.
- Ca additions to Na modified alloys would appear to have [7] W.D. Griffiths, MR Jolly, W Kattlitz, R Kendrick
some positive effect with a refinement in porosity in and AM Niblett, Light Metals 1991, TMS, pp. 1047-
castings. Ca in Na modified alloys may promote eutectic 1055.
nucleation and growth from the mould wall (evidenced by [8] W. Wang and J.E. Gruzleski, Mat. Sci. Tech., vol. 5,
an increase in a dull grey interdendritic surface 1989, pp. 471-475.
appearance on Tatur castings). No Ca containing [9] A. Knuutinen, "Effects of Ba, Ca, Yb and Y on the
particles could be found in the Na-Ca combination sample Eutectic Solidification and Porosity Formation in an
Al-7%Si Alloy", Master of Science and Engineering
Ca: Thesis, University of Queensland, 2001.
- Ca in combination with Sr produced no extra modification [10] E Velasco et al, Light Metals 1998, TMS, 993-1001
in samples. Si containing particles become more [11] A. Abdollahi and J.E. Gruzleski, Int. J. Cast Metals
needle/plate-like for Sr-Ca combinations in comparison to Res., vol 11, 1998, pp. 145-155
samples containing Sr only. Small needle-like particles [12] H. Nakae, K. Song and H. Kanamori, Proc Third
appeared at levels as low as 32ppm Ca, which may be Asian Foundry Congress, Ed. Z.H. Lee, C.P. Hong
deleterious to mechanical properties. Samples containing and M.H. Kim,Kyongju South Korea, 1995, pp. 94-
high Ca levels (113ppm) can result in very large needle- 99.
particles. Most Ca-containing particles in all samples [13] C. R. Loper and J. I. Cho, "The Influence of Trace
appear blocky in shape and often occur in clusters. Amounts of Phosphorous in Aluminium Castings - A
- There was some evidence of increased porosity in the Review of the Literature", AFS Transactions, 2000
Tatur castings and finer porosity in the Elbow casting for [14] M. Garat and R. Scalliet, AFS Transactions, vol.86,
a high Ca level (113ppm), although there was no 1978, pp. 549-562.
observable change in grain size with increased amounts of [15] G. Nagel and R. Portalier, Int. Cast Metals Research
Ca. J., vol.5, 1980, pp. 2-6.
- Larger, intermetallic Ca-Mg-Si particles were observed in [16] B. Closset and J. E. Gruzleski, Metallurgical
the microstructure when Ca was added to a Sr modified Transactions A, vol. 13A 1982, pp. 945-951.
alloy containing Na. Ca appeared to have a dispersive
effect on porosity in Tatur castings. This effect may be
explained by increased nucleation sites for porosity.
Small Ca and P additions to a Na + Sr combination appear
to slightly improve fluidity.

P:
- The addition of P did not appear to affect the level of
eutectic modification at the levels studied, although the
nature of the porosity in the castings appeared to change.
Pores were larger and more numerous in castings
containing Sr, Na, Ca and P combinations.

References

[1] A.K. Dahle, J.A. Taylor and D.A. Graham,

Aluminum Transactions, vol. 3, 2000, pp. 17-30.

Ware, 10 of 10