The aim of the project is to determine the effect of alloying additions on the microstructure and properties of aluminium-silicon (Al-Si) alloys for piston applications. A number of common analysis methods including optical microscopy, energy dispersive X-ray (EDX) spectroscopy, X-ray diffraction (XRD) and thermodynamic calculations will be carried out to identify the phases present in each of the three selected Al-Si alloys. The results will be combined with data from high temperature, high frequency fatigue testing to compare the selected piston alloys performance, with that of the current commercial alloys.


The Piston

The automotive piston is responsible for utilizing energy from the combustion process. Due to the harsh conditions that are experienced by the piston, the material used must have properties that allow the piston to provide optimal performance. This includes high thermal conductivity and a low thermal expansion, to be able to cope with the high operating temperatures of near 300C, and good wear resistance to accommodate for the abrasion action between the piston and piston rings. In early low speed engines, the chosen material for piston applications was cast iron, which although provided good performance, is too heavy to be used in modern engines. As engines have become more powerful and energy efficient, higher strength-to-weight materials such as aluminium are desirable to save weight where possible. The chosen piston materials for modern commercial engines are aluminium (Al) alloys, which are approximately one-third the mass of the equivalent cast iron piston. The most commonly used alloy is based on the aluminium-silicon (Al-Si) system, due to excellent castability, strength and wear resistance. As the content of silicon (Si) increases up to the eutectic composition, the fluidity increases, resulting in faster and more uniform casting, hence the preferred use of near-eutectic alloys for automotive components [1]. However, hypereutectic alloys are generally only produced up to a composition of 25 % Si (all % are by weight unless otherwise stated), because the alloy becomes more prone to cracking, due to the brittle nature of the Si particles.

The Al-Si system

The binary Al-Si system (figure 1) comprises of face centred cubic Al solid solution and essentially pure diamond cubic Si. The Al-Si alloys used for automotive applications generally contain between 11 and 13% Si, giving them their near eutectic composition. The following information can be obtained from the phase diagram:

  • The eutectic is located at a composition of 12.2% Si and a temperature of 577C.
  • The maximum solid solubility of Si in Al is 1.5% at the eutectic temperature, which decreases to 0.05% at a temperature of 300C.
  • The maximum solubility of Al in Si is at a composition of 0.016% Al at 1190C [2].

Upon slow cooling through the eutectic, large quantities of Si plates are distributed throughout the microstructure. No intermetallic phases will form at the eutectic composition. Due to the brittle nature of Si below temperatures of 626C, the boundaries between these large plates and Al grains are prone to cracking, as little deformation takes place [3]. A faster cooling rate results in the production of a fibrous structure of primary Si, due to the link between cooling rate and particle size; because the solidification is spread over a longer period of time, slow cooling allows the particles to grow bigger, and vice versa [4, 5]. Another theory behind this, is the twin plane re-entrant (TPRE) mechanism. The presence of the re-entrant groves between twinned Si plates act as sites for Si atoms to adhere. It has been shown by Ho & Cantor [6] that the addition of sodium (Na) contributes to the non-preferential growth of the Si plates parallel to the twin boundaries. The main theory for this, is that Na inhibits/poisons the groves, stopping growth in the <112> direction, forcing growth in other crystallographic directions.

Alloying Additions

Although solidification has a large influence on the microstructure produced and hence the mechanical properties, the alloys can be tailored by the use of alloying additions in the casting process. These additions are split into two categories: major alloying additions and minor / grain refining alloying additions.

The main effect major alloying additions have on the alloy, are increases in strength and hardness, but usually result in a decrease in ductility. As shown by Ammar [8], Magnesium (Mg) is a common addition that improves the strength of the alloy by precipitation of the hard phase, Mg2Si. Usually added in compositions of between 0.40 and 0.70%, the addition of Mg can nearly double the proof stress of the alloy after precipitation hardening. The addition of copper (Cu) is usually small, because although it increases the machinability of the alloy, the hot tear resistance and castability is decreased. A correct balance of nickel (Ni) with the Cu reinforces the increase in high temperature strength, and reduces the disadvantageous effects. Another major addition is iron (Fe), which although increases the hot tear resistance, reduces the ductility of the alloy. Fe forms insoluble phases in the alloy, such as FeAl3 and FeMnAl6, the fraction of which are increased with increasing Fe content. However the content must be controlled, as larger amounts result in the formation of sludging phases (containing other elements), which can adversely reduce the castability.

Minor alloying additions are used in the process of grain refinement, in order to achieve a uniformly fine, equiaxed microstructure. This is achieved by varying the grain refiner content added, which has an effect of the grain size. Other advantages of grain refining include a reduction in porosity and hot tearing, as well as improving the casting surface finish and an increase in mechanical properties. Grain refiners are added in the form of master alloys, which are usually between 0.5-2kg t-1 in weight. These master alloys produce inoculate particles which act as nucleation sites, on which a-Al grains can grow [9]. Common master alloys are based on the Al-Ti and Al-Ti-B systems. The number of grains that nucleate around the particles, is determined by the Titanium (Ti) content, with context to the initiation of the peritectic reaction (the reaction between a liquid and solid phase to create a different solid phase) [4]. The main action of Boron (B) is to allow for a reduction in the content of Ti required to satisfy the conditions of the peretectic reaction, achieved by altering the liquidus line [10]. As stated by Quested [9], the inoculate particles produced by a master alloy based on the Al-Ti-B system (AL3Ti, TiB2 and AlB2), offer superior grain refining effects than master alloys based on the Al-Ti or AL-B systems alone. Although still under study, other master alloys, such as those based on the Al-Zr system, operate using the same principals as the more common ones [9].

Heterogeneous Nucleation

Because of the large activation energy required for the development of an a-Al grain (a spherical cap) on a flat substrate (e.g. a mould wall), it is harder to achieve crystal formation by the means of homogeneous nucleation. Heterogeneous nucleation involves the transfer of an atom from the undercooled liquid to a nucleation site, where the spherical cap is formed. This action is aided by the inoculate particles. As shown in figure 2, the contact (wetting) angle, ?, between the a-Al phase and substrate depends on the amount of undercooling. A different rate of undercooling will result in different interfacial energies, s, of the nL (nucleant substrate-liquid), nS (nucleant substrate-solid) and LS (liquid-solid) interfaces; these must be balanced in order for nucleation to occur (figure 2). The contact angle is represented by the equation: . The rate of nucleation increases as the interfacial energies decrease, hence a smaller undercooling is required to form a stable solid. As stated by Quested [9] TiB2 and Al3Ti particles are strong influences, as they can incur small contact angles as a result of undercoolings of less than 1K. An increasing contact angle towards 180 leads to homogeneous nucleation, which occurs at a much slower rate. The change in free energy associated with the formation of the a-Al grain, is given as:

Three as-cast pistons were selected due to their differences in composition; the full compositions of which can be seen in table 1. Highlighted in red are the main areas of interest, comparing high* and low* additions between the alloys.

Sample Preparation

To obtain samples consisting of material from the equiaxed zone, the pistons were sectioned down the central axis using a band saw. Each piston was then further sectioned to achieve samples for optical microscopy and X-ray diffraction, using an abrasive cutting wheel. The samples cut for optical microscopy were mounted in conducting Bakelite for ease of use and compatibility with the scanning electron microscope (SEM). When mounted, the samples were ground in progression from 240 grit, through to 1200 grit paper and then polished to a 1m finish on diamond cloth. To obtain the best possible finish for viewing in the (SEM), the samples were polished using colloidal silica solution. The samples were immediately placed under running water and swabbed with cotton wool, followed by a dry-off using methanol. This ensured that no silica particulate was left on the sample surfaces. The rectangular sections cut for XRD analysis were ground to a 1200 grit finish on one side for two purposes; to achieve a relatively smooth surface (remove any prominent scratches) and to ensure the two surfaces were parallel.

Optical Microscopy

Each of the three mounted samples were viewed in the un-etched condition using a Reichert metallurgical-type microscope. Each was viewed in turn at magnifications of 100, 200 and 500 times, and digital photographs recorded using a high resolution Q-imaging camera. A calibrated measurement scale slide was also photographed using each of the three lenses, in order to produce an accurate representation of 100m at the different magnifications.

X-Ray Diffraction (XRD)

XRD was performed on the three alloys using a Bruker D8 AXS X-ray diffractometer, with Cu Ka radiation of wavelength 1.54 at 40kV and 40mA. Testing was performed using a 2? range of 20-100 with a step size of 0.02 and an 8 second dwell time per step. The traces were interpreted using XRD analysis software, adding ICDD values to identify the phases present.

Thermodynamic Calculations

Thermodynamic calculations were carried out for the three alloys using the MTDATA software.

Equilibrium: For the Al-Si samples, the Al alloys database was selected. The compositions for each alloy was entered into the software to display the phases that may be present; initially all possible phases were considered, and eliminated after their absence from the thermodynamic trace. A temperature range of 300-700C, with a step size of 5C was used.

Scheil-Gulliver: The compositions of each alloy were entered into the Scheil-Gulliver module, along with a temperature range of 520-793K (the range in which solidification occurs). The program was run in 1K steps, in order to identify the phases present at the given compositions.

Results and Discussion

Optical Microscopy

Optical analysis of the samples was performed to observe the general microstructure of the alloys and compare the size, form and distribution of the main phases present. The main observations made were concerning the Si particle size and distribution, which can be seen to have a slate-grey colour. All of the alloys were near-eutectic compositions, so had Si particles distributed throughout the microstructure (between 12.9 and 13.71% Si). As shown in figure 3, two main types of Si were observed; primary Si present in a block form ?and plates of secondary Si .

Alloy 1: The microstructure presented a uniform distribution of secondary Si, however, the primary Si was observed to be in clusters. This would result in the alloy having concentrated brittle areas, which may influence cracking under stressed conditions. The main intermetallic phase (light grey?) was uniformly distributed and consisted of plates measuring 50-80m in width. In an area near the edge of the alloy, a dendritic aluminium structure was observed, with fibrous secondary Si present around the dendrite arms. At a high magnification, widely distributed black intermetallic phases existed; these measured less than 30m in width.

Alloy 2: The main observation made was the reduced amount of primary Si (as expected, due to the lower Si content) and smaller particle size compared with alloy 1. This suggests a faster cooling rate was experienced. The secondary Si had a uniform distribution, but was observed to be associated with the primary Si. At high magnification the black intermatallic phase was present, however, another intermatallic phase (light brown ) was observed to be associated with the secondary Si (figure 4). Little dendritic structure was observed.

Alloy 6: The microstructure contained primary Si of a similar size to alloy 2, however a larger amount was present, which was mostly organised into clusters. Alike the other two alloys, the dominant intermetallic phase ? was uniformly distributed throughout, but was present in branched clusters. A dendritic structure surrounded by secondary Si similar to that of alloy 1 was also present in some regions. Alike the other alloys, the black intermetallic phase was observed at a higher magnification.


Using the data produced from the XRD analysis of the three alloys, traces were generated of intensity versus 2, in order to identify the prominent peaks and their equivalent intensities. From the analysis of previous work, ICDD values for phases expected to be present were obtained and added into the XRD analysis programme. In most cases, the phases were easily identifiable, with the intensity ICDD lines cutting through the centre of the peaks, or slightly off-centre. However, when viewing the lower intensity intermetallic peaks, a large amount of interference makes them hard to identify; this is due to the restriction of the 2resolution. This outlines the limitations of solely using XRD and justifies the need to use more than one method to identify the phases present. A typical method that could be used is Differential Scanning Calorimetry (DSC), which although is not sensitive to d-spacings, is much more sensitive to phases than XRD.

As shown in figure 5, the traces from the three alloys have been positioned so that the prominent peaks can be compared and any unique phases identified. By firstly applying the ICDD value for Al, it was observed that the relative peaks of alloy 6 were close to random, as expected of a fine microstructure alloy. On the other hand, alloys 1 and 2 were far from random, with the {200} peaks being higher than the expected intensities; again this is expected, due to the coarse microstructure of both alloys.

Alloy 2 was chosen to be analysed further, as the high intensity peaks mean that the diffractions present would be above the noise. Figure 6 shows the trace generated from the XRD data, with the prominent Al and Si peaks identified. The intensity of the {200} peak was observed to be double that of the corresponding ICDD.

Table 2 shows the phases identified through equilibrium thermodynamic calculations; most of these phases were expected to be present. The trace generated of alloy 2 (figure 8) shows that solidification initiates at a temperature of 873K, with the first phase to form being AlSiFeMn (ALPHA). By then carrying out Scheil-Gulliver calculations, it was possible to analyse the solidus and liquidus (figure 9) in 1K steps until 99.9% solidification, as it assumes there is no solid-state diffusion. Further the phases identified using equilibrium calculations, the following additional phases were present: Al8FeMg3Si6, Al7Cu4Ni, AlFeSi, Al7Cu2Fe, Al2Cu. The following phases were also identified during XRD after a brief analysis of the prominent peaks: Al, Si, Al3Ni, Al9FeNi, Al7Cu4, AlFeSi.


From the work carried out to date, the following conclusions can be made:

  1. Two main types of Si were observed in the three alloys, however, all have different distributions and particle sizes. Smaller Si particles were observed in alloys 2 and 6, suggesting a faster cooling rate was experienced.
  2. A light grey intermetallic phase was observed to dominate in the alloys, being uniformly distributed throughout the microstructure in plate form.
  3. XRD analysis of the three alloys showed that alloy 6 was close to random, whereas alloys 1 and 2 were far from random; this was expected of the fine and coarse microstructures respectively.
  4. Further analysis and identification of the Al and Si peaks of alloy 2, showed that the intensity of the {200} peak was double that of the corresponding ICDD.
  5. Thermodynamic calculations carried out met expectations of the phases that were present in the alloys.

Further Work

The work carried out to date has given an initial view of the phases present in the three chosen Al-Si alloys. To gain a deeper understanding of the phases present and the effects they have on the alloy, the following work will be carried out:

  • EDX and Electron Back Scatter Diffraction (EBSD) will be used to analyse the samples, to identify a full list of phases present and compare the results to the optical micrographs.
  • Identification of all of the prominent peaks in the XRD traces, to achieve a phase list that is comparable with SEM results.
  • Differential Scanning Calorimetry (DSC) will be used to observe phase transitions in each of the three alloys.
  • High temperature, high frequency fatigue testing will be carried out to compare the mechanical performance between each of the alloys, and against current commercial alloys.


  1. Heisler, H. Vehicle and Engine Technology. Vehicle and Engine Technology. London: Edward Arnold, 1985.
  2. J.L. Murray, A.J. McAlister. The Al-Si (Aluminium -Silicon) System. 1984, Bulletin of Alloy Phase Diagrams volume 5, p. 74-77.
  3. L.F. Mondolfo. Aluminim Alloys, Structure and Properties. London: Butter Worths, 1976, p. 369.
  4. I.J. Polmear. Light Alloys: Metallurgy of Light Metals. London: Arnold, 1995, p. 167-181
  5. Edwards, W. Microstructural and mechanical property modelling for the processing of Al-Si alloys. 2002, Loughborough PhD Thesis, p. 128.
  6. C.R.Ho, B. Cantor. Modification of Hypoeutectic Al-Si alloys. 1995, Journal of Materials Science volume 30, p. 1912.
  7. J.E Hatch. Aluminium: Properties and Physical Metallurgy. USA: American Society for Metals, 1984.
  8. H.R. Ammar, C. Moreau, A.M. Samuel, F.H. Samuel, H.W. Doty. Influences of alloying elements, solution treatment time and quenching. 2008, Materials Science and Engineering A489, p. 426.
  9. Quested, T. E. Understanding mechanisms of grain refinement. 2004, Materials Science and Technology volume 20, p. 1357-1362.
  10. M.N. Binney, D.H. St John, A.K. Dahle, J.A. Taylor, E.C. Burhop, P.S. Cooper. Grain Refinement of Secondary Aluminium-Silicon Casting Alloys. 2003, Light Metals, p. 1-2.
  11. ASM International. Casting. 1988, ASM Handbook volume 15, p. 212.
  12. N. Saunders, The Application of Calculated Phase Equilibria to Multi-Component Aluminium Alloys, 2001, JILM Vol 51, p145

Please be aware that the free essay that you were just reading was not written by us. This essay, and all of the others available to view on the website, were provided to us by students in exchange for services that we offer. This relationship helps our students to get an even better deal while also contributing to the biggest free essay resource in the UK!