Thermodynamics Calculation of Extra Mn Addition in the Recycling
of Al-Si-Cu Aluminium Alloys
Shouxun Ji, Feng Gao, Zhongyun Fan
Brunel Centre for Advanced Solidification Technology (BCAST)
Brunel University London, Uxbridge, Middlesex UB8 3PH, United Kingdom
Email: [email protected]
Keywords: Aluminium alloys, thermodynamics, phase formation, materials recycling, iron
removal
Abstract. Iron contamination from scrapped materials is always a problem in producing high
quality secondary aluminium alloys. Consequently, the iron removal during recycling of
aluminium alloys is essential and important in industrial practice. This work aims to study the
effect of extra Mn addition on the effectiveness and efficiency of iron removal during
recycling. The thermodynamics assessment was carried out for Al-Si-Cu alloys to find out the
variation of balanced iron and manganese in the liquid melt and in the sediment solid Fe-rich
intermetallics with different levels of extra Mn addition. The effect of alloy composition and
processing temperatures was investigated. The findings help to understand the capability and
fundamentals of iron removal in aluminium alloys.
1.
Introduction
Aluminium is the leading metal in a number of industrial sectors such as aeronautics,
automotive, packaging, construction & energy. Compared to other high volume materials,
aluminium production has one of the largest energy differences between primary and
secondary productions: 186 MJ/kg for primary compared to 10-20 MJ/kg for secondary,
which means that only 5% of the energy is needed for recycling of aluminium alloys [1].
Therefore, recycling is a major sustainable approach to continue the use of aluminium
because of the enormous benefits in energy resilience manufacturing and the cost saving of
products. As a result, the application of secondary materials is increasingly attractive for
manufacturing such as transport because light-weighting is an important consideration.
Currently, more than a third of all the aluminium produced globally originates from traded
materials and scraped materials. In addition to the obvious economic dimension, growing
environmental concerns and heightened social responsibility for the last ten years have served
to boost recycling activity of aluminium alloys in order to conserve resources and to avoid
littering. However, although the recycled aluminium can be utilised for almost all aluminium
applications, the accumulation of impurities in the recycled materials provides a significant
and long-term compositional barrier to maintain the aluminium alloys within required range.
The problematic impurities are different for various aluminium alloys, but it is always
associated with iron because of the easy picking up during manufacturing [2]. Therefore, the
key problem associated with the accumulation of unwanted elements in the recycled material
is to reduce or diminish iron levels in aluminium alloys.
There is a variety of solutions to deal with the iron removal from aluminium melt during
recycling; each presents a trade-off between cost and improvement in scrap utilisation (or
recycling) potential [ 3 ]. A common solution used in industry today is to form related
intermetallics with subsequent sediment in alloy melt [3]. Therefore, metal recycling is
essentially a metallurgical process and it is a compositionally determined cap to recycling
rates. This is governed by the laws of thermodynamics. However, aluminium presents a high
degree of difficulty in the removal of tramp elements due to thermodynamic barriers. In order
to mitigate the detrimental effects of element accumulation, the effectiveness and efficiency
are important in operational strategies. Therefore, it is necessary to have a thermodynamic
assessment in order to identify the effective strategies throughout the production process to
mitigate the elemental accumulation within contaminated secondary materials.
This study intends to identify more precisely for the expected ranges of compositions of
the recycled metal before and after recycling. The study was carried out to explore the effect
of adding extra manganese into melt to promote the formation of Fe-rich intermetallics, which
can be eventually separated from the melt during recycling. The balanced Fe and Mn
concentration in the alloy melt after the formation of intermetallics is assessed in order to
obtain the required concentration in the recycled materials.
2.
Methodology
The investigation was focused on the thermodynamics evaluation using the PanAl database
in Pandat software [4] to understand the solidification behaviour and compound formation at
different contents of Mn and Fe in the Al-Si-Cu alloy. The COST507 thermodynamic
database [5] was used for constituent alloy systems and the α-AlFeMnSi and β-AlFeSi phases
were treated as a stoichiometric phase during the modelling. The composition of the LM 24
is given in Table 1. The minor and low levels of elements were not considered in the
calculation. All the compositions and solid fractions were given in weight percentage (wt-%)
unless otherwise stated.
Table 1 Composition of the Al-Si-Cu alloys used for calculation (wt-%).
Element Si
Range
9.0
Cu
4.0
Zn
3.0
Fe
1.1
Mg
0.3
Mn
0.5
Pb
0.2
Sn
0.2
Ti
0.2
Al
Bal.
3. Phase formation with increased Mn content
The calculated equilibrium phase diagram on the cross section of Al-9Si-4Cu-3Zn with
varied Mn and Fe contents is shown in Fig. 1. It was seen that the calculated equilibrium
phase diagram could be divided into several regions with different levels of Fe contents.
When the Mn addition was at 0.5wt-%, the phase formation during solidification could be
described as:
(a) L→ α-AlFeMnSi+ α -Al+ θ-AlCu at Fe<1.1 wt-%
(b) L→α-AlFeMnSi+ β-AlFeSi+ α-Al+ θ-AlCu at Fe>1.1 wt-%.
In both cases, α-AlFeMnSi phase was the prior phase, although the subsequent transformation
would form different types of phases.
When the Mn content was increased to a higher level at 1.0 wt-%, the solidification
process showed no change in terms of phase formation. It was seen that the α-AlFeMnSi
phase was still prior phase, but its formation range was enlarged to 1.7wt-%Fe, representing
an increase of 55% in comparison with the alloy with 0.5wt-%Mn. This was further
confirmed by the increase of Mn content to 2.0 wt-% and 3.0wt-%, the solidification could
also be divided into two areas. The prior α-AlFeMnSi phase was formed when Fe was below
2.6wt-% in Fig. 1c. It is clear that the increase of Mn content in the alloy was able to promote
the formation of α-AlFeMnSi phase with increased Fe content. This revealed that the
formation of β-AlFeSi phase in the as-cast microstructure was also significantly affected by
the Mn content, which started from 1.1 wt-%Fe when Mn was at 0.5 wt-% and 2.6 wt-% Fe
when Mn was at 2.0 wt-% in the experimental alloys.
700
700
(a)
(b)
L
L
L+αAlFeMnSi+αAl+Si
L+αAlFeMnSi+βAlFeSi
L+ αAlFeMnSi
L+αAlFeMnSi+βAlFeSi
L+αAlFeMnSi+βAlFeSi+αAl+Si
500
L+ αAlFeMnSi
600
Temperature (oC)
Temperature (oC)
600
αAlFeMnSi+αAl+Si+θAlCu
L+αAlFeMnSi
+αAl+Si
500
αAlFeMnSi+βAlFeSi+αAl+Si+θAlCu
αAlFeMnSi+αAl+
Si+θAlCu
αAlFeMnSi+βAlFeSi+αAl+Si+θAlCu
400
0
1
2
400
3
L+αAlFeMnSi+βAlFeSi+αAl+Si
0
1
Fe (wt.%)
700
2
3
Fe (wt.%)
700
(c)
(d)
L
600
L+ αAlFeMnSi
Temperature (oC)
Temperature (oC)
L
L+αAlFeMnSi+βAlFeSi+αAl
L+αAlFeMnSi+αAl+Si
500
L+ αAlFeMnSi
600
L+αAlFeMnSi+αAl+Si
500
αAlFeMnSi+αAl+θAlCu+Si
αAlFeMnSi+αAl+θAlCu+Si
αAlFeMnSi+βAlFeSi+αAl+θAlCu+Si
400
0
1
2
400
3
0
1
2
3
Fe (wt.%)
Fe (wt.%)
Fig 1 The equilibrium phase diagram of Al-9wt-%Si-4wt-%Cu-3wt-%Zn alloy with different
Fe and Mn contents, (a) 0.5wt-%, (b) 1.0wt-%Mn, (c) 2.0wt-%Mn and (d) 3.0wt-%Mn.
Fe (wt.%)
4
In Fig. 1, it was seen that Mn increased
L+αAlFeMnSi+βAlFeSi
the area of forming α-AlFeMnSi
intermetallic compound in the alloy. The
3
liquidus temperature of forming αAlFeMnSi phase was moved to higher
L+βAlFeSi
temperatures for the alloy with increased
2
Mn content, which revealed that the
addition of Mn increased the liquidus
L+αAlFeMnSi
temperature of the alloy. It was also seen
1
that the Fe content to form β-AlFeSi phase
L
was raised to higher values with increased
0
Mn content in the alloy. Therefore, the
3
4
0
2
1
processing window to form α-AlFeMnSi
Mn (wt.%)
phase was significantly enlarged with the
Fig. 2 Calculated phase formation at 600oC in
increase of Mn content in the alloy. In the
the Al-Si-Cu alloy.
meantime, it was seen that the addition of
Mn reduced the equilibrium concentration
of Fe in the liquid melt of aluminium alloy. This was important for industrial application as it
revealed that the Fe content in the melt could be controlled through the addition of Mn into
the alloy melt. These provided the fundamentals for the iron removal in aluminium alloys.
The relations between Fe content and Mn content in the Al-Si-Cu alloy can be seen from Fig.
2. In order to form α-AlFeMnSi phase, the Mn and Fe should satisfy a ratio in the alloy. With
different Fe contents, the Mn content was required to maintain a certain level to form αAlFeMnSi phase in the alloy.
The solidification process can be understood from Fig. 3 for the variation of phase fraction
during cooling. Clearly, Mn increased the precipitation temperature and the volume fraction
of prior α-AlFeMnSi phase, but it did not change the eutectic temperature as the temperature
and volume fraction was shown at the same level for Si precipitation. A significant change
was the disappearance of β-AlFeSi phase in Fig.3, which was transferred to α-AlFeMnSi
phase due to the increase of Mn content in the alloy.
1.0
1.0
(a)
0.8
Phase fraction
Phase fraction
0.8
(b)
Liquid
αAlFeMnSi
βAlFeSi
αAl
Si
θAlCu
0.6
0.4
0.2
0.4
0.2
0
400
Liquid
αAlFeMnSi
αAl
Si
θAlCu
0.6
600
500
700
0
400
o
Temperature ( C)
600
500
700
o
Temperature ( C)
Fig. 3 The variation of different phases during solidification for the Al-Si-Cu alloy with (a)
0.5wt-%Mn and 2wt-%Fe and (b) 2wt-%Mn and 2wt-%Fe.
4. The balanced Fe and Mn concentration in the melt
The concentrations of Fe and Mn in the alloy melt are critical for determining the
effectiveness and efficiency of iron removal capability. When the melt was cooled down
below its liquidus temperature, the precipitation of α-AlFeMnSi phase could consume Al, Fe,
Mn and Si in the melt. This process could continue until the temperature was able to form αAl phase. Therefore, there existed a thermodynamic balance for each element between the
precipitated phase and the remnant liquid phase. The variation of Fe and Mn contents during
solidification could be understood more clearly by calculating the equilibrium concentration
in the liquid with different Mn/Fe ratios. The results are shown in Fig. 4. As the variation of
Si was negligible, it was therefore not given here. The results showed that the balanced
concentrations of Fe and Mn in the liquid phase were significantly reduced with the decrease
of the temperature in the interval of liquidus and solidus. Moreover, the balanced
concentrations of Fe and Mn in the liquid phase also varied with the different levels of Mn
content in the alloy. The representative data can be seen in Table 1 for the alloy at 600 oC and
590oC with different Fe/Mn ratios. When Mn/Fe=0.5, the equilibrium concentration of Fe was
decreased from the initial 1.1 wt-% to 0.78 wt-% at 600 oC in the liquid phase. In the
meantime, the initial Mn concentration of 0.5 wt-% was decreased to 0.19 wt-% at 600 oC in
the liquid phase, showing a significant reduction of 30% for Fe and 60% for Mn. The
balanced concentrations of Fe and Mn were further reduced at 590oC, which was 0.71wt-%
for Fe and 0.14wt-% for Mn. The similar situation could also be observed when Mn/Fe was at
1:1. The equilibrium Fe concentration was decreased from 1.1wt-% to 0.5 wt-% at 600oC and
0.42wt at 590oC. The equilibrium Mn concentration was decreased from 1.0 wt-% at liquidus
temperature to 0.29 wt-% at 600 oC and 0.21wt% at 590oC. These showed an increase of Mn
concentration and a decrease of Fe concentration in the melt in comparison with the results
showed in Mn/Fe=0.5, indicating a better efficiency of iron removal when adding more Mn in
the alloy. A further improvement could be seen when Mn/Fe was at 2:1. From these
calculations, it was confirmed that the increase of Mn/Fe ratios resulted in a significant
decrease of Fe content in the liquid phase of the alloy when it was maintained at a
temperature between the liquidus and solidus. Therefore, the Fe content in the liquid could be
controlled by adjusting the Mn/Fe ratios and temperatures in practical operation. It should be
emphasised that the addition of Mn into the melt should be mainly controlled by the remnant
content in the melt because of the defined limitation of alloy specification. If the Mn content
was over the limitation, it would cause secondary contamination and other problems even the
Fe was within the requirement after processing. In the meantime, it should be carefully
controlled the different variables for iron removal as it was a complex and problematic
process.
0.8
0.4
1.2
(b)
Fe@liquid
Mn@liquid
Al concentration in liquid (wt.%)
Al concentration in liquid (wt.%)
1.2
(a)
Fe@liquid
Mn@liquid
0
Fe@liquid
Mn@liquid
0.8
0.8
0.4
0.4
0
500
650
550
600
Temperature (oC)
700
(c)
Al concentration in liquid (wt.%)
1.2
0
500
650
550
600
Temperature (oC)
70
0
500
550
650
600
Temperature (oC)
700
Fig. 4 The equilibrium concentration of Fe and Mn in the liquid phase of the Al-Si-Cu alloy
during solidification, calculated from the equilibrium phase diagram, (a) Mn/Fe=0.5, (b)
Mn/Fe=1.0, and (c) Mn/Fe=2.
Table 1 The Fe and Mn concentration in the melt after forming α-AlFeMnSi intermetallics at
different temperatures.
Mn content
Element in melt
Melt at 600oC
Melt at 590oC
0.5wt-%Mn
Fe
Mn
0.78
0.19
0.71
0.14
1.0wt-%Mn
Fe
Mn
0.50
0.29
0.42
0.21
2.0wt-%Mn
Fe
Mn
0.41
0.26
0.32
0.21
Among the Fe-rich intermetallics, α-AlFeMnSi is always referred as quaternary
intermetallic Al15(Fe,Mn)3Si2 compound and β-AlFeSi is always referred as Al5FeSi phase.
Because of the difference in constituent, the Fe content is 17wt-% in the Al15(Fe,Mn)3Si2
phase and 27wt-% in the Al5FeSi phase. Therefore, the formation of β-AlFeSi can consume
more Fe in the melt and provide better efficiency. However, the balanced Fe concentration
after forming β-AlFeSi is much higher than that of forming α-AlFeMnSi phase in the melt.
This means that the remained Fe in the melt will be at a higher level if forming β-AlFeSi.
Therefore, it is not applicable if the low Fe content in the melt is essentially required in the
recycled alloys. However, this can be practically useful if the final Fe content is required at
relatively high level.
When Mn is added into the alloy as a naturalisation element to form α-AlFeMnSi
intermetallics, the actual Fe content in the alloy should be considered because the excessive
Mn addition in the Al alloy will promote the formation of Al-Mn intermetallics, for example
Al6Mn and AlMnSi [ 6 ]. These intermetallics are not able to consume Fe in the alloy.
Therefore it needs to be avoided. In general, the amount of extra Mn addition should be
selected to form Al15(Fe,Mn)3Si2 compound, rather than Al-Mn and other Fe-free
intermetallics.
Conclusions
The phase diagrams, phase fractions and balanced concentration in the melt of an Al-Si-Cu
system were thermodynamically analysed using the CALPHAD method in order to
understand the effectiveness and efficiency of iron removal mechanism. The main
conclusions are:
The CALPHAD calculation suggests that extra Mn addition can effectively form αAlFeMnSi intermetallics with increased Fe content in the alloy, which is capable of
maintaining a relatively low level of Fe content in the melt. Therefore, extra Mn addition is an
active approach to remove Fe in the alloy.
The remained Fe content in the alloy melt can be controlled by the Mn content and the
processing temperature of forming α-AlFeMnSi intermetallics. The higher the Mn content and
the lower the holding temperature, the lower the Fe content in the remained melt. However,
the remained Mn in the melt should be controlled within the requirement of defined
specification. Meanwhile, the formation of Al6Mn and AlMnSi intermetallics should be
avoided after adding Mn because they consume Mn and Si without decreasing Fe in the alloy.
Acknowledgement
The financial support from EPSRC and TSB (UK) is gratefully acknowledged.
References
[ 1 ] International Aluminium Institute, Global Aluminium Recycling, A Cornerstone of
Sustainable Development, 2009.
[2] S. Ji, W. Yang, F. Gao, D. Watson, Z. Fan, ‘Effect of iron on the microstructure and
mechanical property of Al-Mg-Si-Mn and Al-Mg-Si diecast alloys’, Mater Sci Eng A
2013(564) 130-139.
[ 3 ] L. Zhang, J. Gao, L. Nana, W. Damoa, D. G. Robertson, ‘Removal of iron from
aluminum: A review’, Mineral Processing & Extractive Metall Rev, 2012(33): 99–157.
[4] S.L. Chen, S. Daniel, F. Zhang, Y.A. Chang, X.Y. Yan, F.Y. Xie, R. Schmid-Fetzef, W.A.
Oatesd, Calphad, 2002, 26, 175.
[5] I. Ansara, A.T. Dinsdale, M. H. Rand (eds.), COST 507-final report, Thermodynamic
Database for Light Metal Al-alloys, vol.2, European Communities, Brussels 1998.
[6] M. Warmuzek, A. Gazda, and J. Sieniawski, ‘Influence of retrogression and reaging
(RRA) heat treatment on microstructure, mechanical and chemical behaviour of an Al-ZnMg alloy’ Adv Mater Sci 2003(1) 33-41.