1 30 Int. J. Materials and Product Technology, Vol. 39, Nos. 1/2, 2010 On the ability of producing FGMs with an AlSi12 aluminium alloy by using centrifugal casting G. Chirita Mechanical Engineering Department, School of Engineering, University of Minho, Campus de Azurém, Guimarães, Portugal I. Stefanescu Faculty of Mechanical Engineering, Dunarea de Jos University, Galati, Romania D.F. Soares and F.S. Silva* Mechanical Engineering Department, School of Engineering, University of Minho, Campus de Azurém, Guimarães, Portugal *Corresponding author Abstract: This paper deals with the study of the centrifugal effect on aluminum castings under high G values. In this study it will be shown that aluminum alloys, without reinforcement, and where there is almost no difference of density or low solubility of different phases, may also give rise to a Functionally Graded Material-FGM. It seems that the main variable that affects the mechanical and metallurgical properties is the fluid dynamics of the process. Keywords: FGMs; functionally graded materials; centrifugal casting; aluminium alloy. Reference to this paper should be made as follows: Chirita, G., Stefanescu, I., Soares, D.F. and Silva, F.S. (2010) On the ability of producing FGMs with an AlSi12 aluminium alloy by using centrifugal casting, Int. J. Materials and Product Technology, Vol. 39, Nos. 1/2, pp Biographical notes: George Chirita is a PhD student at the Department of Mechanical Engineering of Minho University, Guimaraes, Portugal. Since February 2002, he is a teaching assistant at Mechanical Faculty, Low Danube University, Galati, Romania. His research interests are material science, machine elements, tribology. He is also a member of ART (Romanian Association of Tribology) and RoAMET (Member of Romanian Association of Mechanical Transmissions). Copyright 2010 Inderscience Enterprises Ltd.
2 On the ability of producing FGMs with an AlSi12 aluminium alloy 31 Ioan Stefanescu received his PhD from Low Danube University, Galati, Romania in In 1968, he joined the academic staff from Faculty of Mechanics, Galati and since 1992 as full Professor of Mechanical Engineering. He is the Director of The Research Center of Mechanical and Tribology of Superficial Layer and president of the Galati branch of the Romanian Association of Tribology. His current research includes tribology, study of: the mechanism of damaging of the superficial layer by contact fatigue, the influence of the thermal treatments on the rolling contacts, the tribological properties of vegetable oils. Delfim F. Soares received his MSc in 1993, and PhD in 2008, both at Minho University, and is Professor at the University of Minho on the Materials area. He works at the Center for Mechanical and Materials Technologies CT2M, at Minho University, and his research interests include processing technologies and phase equilibria for FGMs-Functionally Graded Materials. Filipe S. Silva received his MSc in 1996, and PhD in 2002, both at Minho University, and is Professor at the University of Minho on the Mechanics of Materials area. He works at the Center for Mechanical and Materials Technologies CT2M, at Minho University, and his research interests include processing technologies for FGMs-Functionally Graded Materials and fatigue and fracture of FGMs. 1 Introduction Most of studies in the domain of Functionally Graded Materials (FGMs), obtained by centrifugal casting deal with functionally graded composites where an alloy is reinforced with a solid phase as for example silicon particles. In these cases the gradation as well as the improvement of mechanical properties is attributed to the reinforcement. Moreover in some cases it is possible to obtain FGMs with metallic materials where there is a high difference of density and low solubility of different phases or different materials of the same alloy. In this last case the phases with higher density move in the radial direction to the outer surface of the casting due to the centrifugal force (Janco, 1998). However, in this study, an aluminium alloy, without reinforcement, and with very close densities between phases, is used. It will be shown that the casting may be very influenced by the centrifugal effect and functionally graded castings may be obtained. In order to understand why a homogeneous alloy, with similar densities between phases, may give rise to an FGM, by using the centrifugal effect, an analysis of the most important effects of the centrifugal casting process on metallurgical features is necessary. As a fact the properties of a specific alloy (hypoeutectic, eutectic or hypereutectic) can be attributed to the individual physical properties of its main phase components (α-aluminium solid solution and silicon crystals) and to the volume fraction and morphology of these components (ASM International, 2004). Perhaps the most characteristic property of Al-Si alloys is a high tensile strength in relation to density compared with that of other cast alloys, such as ductile cast iron or cast steel. The high specific strength of aluminium alloys is very strongly influenced by their polyphase microstructure (ASM International, 2004; Hengcheng et al., 2002).
3 32 G. Chirita et al. As a fact, in order to understand the relation between microstructure and mechanical properties investigations have been carried out to clarify the relation between microstructural features such as secondary dendrite arm spacing SDAS (Wang, 2003; Guang et al., 2006), the eutectic particles (Si and Fe intermetalics) size and shape (ASM International, 2004; Liao et al., 2002; Numan et al., 2004; Nikanorov et al., 2005; Saigal and Fuller, 2001) the alloy composition (Shabestari and Moemeni, 2004) and mechanical properties. Wang (2003) discussed the relationship between the Secondary Dendrite Arm Spacing (SDAS) and tensile properties of an A356 aluminium alloy. Their results show an increase of the ultimate tensile strength and also the ductility of the A356 alloy with the increasing of the SDAS. Studies performed on a sand cast A356-T6 obtained by hot isostatic pressing process also revealed an increase of ultimate tensile strength and ductility of both A356 and A357 alloys with SDAS (Guang et al., 2006). There are different solutions to control these micro structural features, for example by adding some alloying elements (Liao et al., 2002; Shabestari and Moemeni, 2004) to refine the grain. However perhaps the most common way to improve mechanical properties of cast Al-Si alloys is by changing the casting technology (ASM International, 1997). Each technology has particular aspects that interfere on microstructure and consequently on mechanical properties. Regarding the effect of the centrifugal technique on vertical centrifugal castings there are mainly three aspects that may strongly affect the microstructure of the components thus obtained, namely: fluid dynamics; vibration (inherent to the system); and centrifugal force (Janco, 1998). In this paper, a solidification characterisation on different points along the mould is made, in order to have an accurate idea of both the fluid dynamics inside the mould during the casting and the solidification behaviour in the different parts of the component. This analysis will be related to some metallurgical features (phase distribution; SDAS; eutectic silicon content and shape) along the component and mainly along the direction of the centrifugal pressure. For comparison purposes castings obtained by centrifugal casting technique will be compared with traditional gravity casting technique, as a reference. Finally, a model for rupture strength and strain is tested in order to correlate rupture strength and strain with metallurgical features namely silicon lamellas thickness and length, SDAS and volume fraction of the eutectic phase. 2 Experimental methods The material used for obtaining the castings was an AlSi12 commercial alloy (Chirita et al., 2006). The material was melt with an induction vacuum furnace at a temperature of 670 C. Before pouring, the permanent mould was preheated at 130 C for all castings. On centrifugal casting the mould was rotating with an acceleration (G) of 24.5 g around the central axis of the casting machine and the molten aluminium was poured into the mould cavity (Figure 1(a)) (Chirita et al., 2006). For gravity castings the same induction vacuum melting equipment and the same melting temperatures used in
4 On the ability of producing FGMs with an AlSi12 aluminium alloy 33 centrifugal casting were used, but in this case the melt was manually poured into the mould (Figure 1(b)). Figure 1 Schematic representation of (a) centrifugal casting and (b) gravity casting (see online version for colours) (a) (b) Three ingots for each casting technique were produced (Figure 2). Each ingot, from the centrifugal castings and from the gravity castings (Chirita et al., 2006) were cut to obtain three specimens in order to compare the properties of the aluminium alloys in different places of the ingot and in particular along the radial direction of centrifugal casting (Figures 1 and 2) along the first 15 mm. Globally, 18 specimens were tested, nine of each casting technique. Figure 2 Position of the specimens in the casting: (a) centrifugal and (b) gravity (see online version for colours) (a) (b) Tensile tests were performed in a Dartec tensile testing machine at room temperature. Samples were analysed by optical and scanning electron microscopy SEM/EDS. The amounts of the constituents were statistically quantified by image analysis. Temperature readings (at 1.5 mm from internal mould surface), during castings, were performed with a data acquisition system (reading of 8 points/s), in three different points of the mould during the solidification (Figure 3). Three reading positions were selected: mould bottom to determine the pouring time; side and front part with, respectively, low and high mould wall mass.
5 34 G. Chirita et al. Figure 3 Schematic views of temperature reading positions in: (a) walls of the mould and (b) casting: 1: down; 2: front; 3: side (see online version for colours) (a) (b) 3 Results and discussions Figures 4 8 show different thermal cycles during the castings. Figures 4 and 5 provide the initial temperature evolution just after the pouring. Figures 6 8 provide comparative analysis between gravity and centrifugal castings of the thermal cycles after pouring. Figure 4 Temperature profile with time, for gravity casting, in the different positions of the mould (see online version for colours) Figure 5 Temperature profile with time, for centrifugal casting, in the different positions of the mould (see online version for colours)
6 On the ability of producing FGMs with an AlSi12 aluminium alloy 35 Figure 6 Temperature profile with time, for down position, and for both gravity and centrifugal castings (see online version for colours) Figure 7 Temperature profile with time, for side position, and for both gravity and centrifugal castings (see online version for colours) Figure 8 Temperature profile with time, for front position, and for both gravity and centrifugal castings (see online version for colours) From Figures 4 and 5 it can be seen that the melt, in centrifugal casting, reaches the side and front position, (at half height of the mould) almost at the same time as the down position (the first point where the melt touches at pouring). In gravity technique the melt takes about 2 s to reach the same previous positions. This means that the mould is filled
7 36 G. Chirita et al. much faster and/or the first melt that touches the down position, due to the fluid dynamics, splits in different directions touching the inside mould walls in side and front positions. Probably both situations, faster mould filling and more turbulent fluid dynamics, occur simultaneously for the centrifugal casting, as compared to the gravity casting process. In Figures 6 8 it is possible to compare the heat transfer between the casting processes, for the different positions. In Figure 6 it is observed that the heat accumulation in down position is higher for gravity casting then for centrifugal casting (higher maximum temperature and higher heat transfer rate as inferred by the slope of the temperature curve). This means that the first melt that reached the down position remained there and the material that was continuously poured to the mould touched that position first and kept transferring heat to that position. In centrifugal casting, as the melt that first touched the down position, split to other directions, was not in touch anymore with the down position but with the other mould walls. As a fact, in Figures 7 and 8 it is clear that the side and front walls were affected by the melt quicker in the centrifugal casting then in the gravity casting. Thus, and based on Figures 4 8, it can be concluded that the main variable that distinguishes both casting processes is the fluid dynamics. In the centrifugal casting process the mould is filled with a more turbulent flow, with the melt splitting in different directions touching the inside mould walls, in side and front positions, then the gravity casting process. Accepting the previous explanation as valid, a substantially different solidification behaviour would occur for the different techniques with substantial differences in the phase morphology in different regions of the casting. However, prior to the microstructure analysis, an analysis on the chemical composition was performed. It can be seen if Figure 9 that there is no significant difference in both the aluminium and silicon contents between the casting processes and along the radial direction of the centrifugal casting. This means that there was no effect of different phase densities due to the centrifugal force, as occurs in alloys where there is a high difference of density and/or low solubility of different phases or different materials of the same alloy. In this case this effect is not observed. Otherwise a variation in chemical composition would be observed (Janco, 1998). Figure 9 Chemical composition in different position of ingots (see online version for colours)
8 On the ability of producing FGMs with an AlSi12 aluminium alloy 37 Further to the previous effect of the radial force, the radial centrifugal pressure is not expected to have any interference in the solidification behaviour. For the specific case used in this study the pressure level is about 2 4 MPa and it is known that pressure only influences the solidification diagram for pressures higher than 50 MPa (Ghomashch and Vikhrov, 2000). Regarding the microstructure effects due to the solidification behaviour, Figures 10 and 11 show that there is a difference between casting techniques (Figure 10) and also a substantial difference between positions, along the radial direction, in centrifugal casting. The microstructure is thinner in centrifugal casting and, in this process is also thinner in the outer position (position 1 see Figure 2). Figure 10 Microstructure for both: (a) gravity and (b) centrifugal castings, in position 1 of the ingots (500 ) (a) (b) Figure 11 Microstructure of Al alloy in the three positions of a vertical centrifugal casting ingot (500 ) Figures provide a quantification of the previous evidence. Silicon lamellas are thinner (Figure 12) and shorter (Figure 13) for centrifugal casting and decrease from the inside part of the cast (position 3) to the outside part of the mould (position 1). In Figure 14 is also observed the same behaviour for the SDAS, e.g., bigger for gravity in position 3. In Figure 15 is also observed that the phase distribution differs between casting techniques and positions. The amount of the eutectic constituent is higher for the centrifugal process and increases from position 3 to position 1.
9 38 G. Chirita et al. Figure 12 Thickness of silicon eutectic lamellas (see online version for colours) Figure 13 Length of silicon eutectic lamellas (see online version for colours) Figure 14 Secondary Dendrites Arm Spacing (SDAS) (see online version for colours)
10 On the ability of producing FGMs with an AlSi12 aluminium alloy 39 Figure 15 Volume fraction of phases in different positions of the ingots for both gravity (G) and centrifugal (C) Al-alloy casting (see online version for colours) From Figures it can be concluded that in centrifugal casting the solidification was quicker and/or there were no conditions for phases to grow. As a fact, a more turbulent flow inside the mould would promote a thinner microstructure. The first melt that reaches the down position, in contact with the cold mould, formed some germens of solidification. These germens were then split all over the mould promoting a faster solidification in all regions of the cast and, due to the increased number of solidification germens, there were no chance for the phases (dendrites), to substantially grow, because they would get in touch with the neighbouring phases. Thus, in centrifugal casting process, it is observed a thinner microstructure all over the casting. This effect is more accentuated in the down position because this side of the mould remains colder (in the solidification stage first seconds after pouring) than in the gravity casting process. This reason may be able to explain the more accentuated gradient that is observed in centrifugal castings as compared to gravity castings (Figure 16(b)) These results are in accordance with the thermal cycles observed in Figures 4 8 where it is clear a more uniform heat transfer distribution in the whole mould, for centrifugal casting, then in gravity casting where the heat transfer was mainly concentrated in the down position. Besides the fluid dynamics effect, and the centrifugal force that was already previously shown not to affect the microstructure, the variable that are supposed to influence the microstructure of the components is, according to (Janco, 1998) the inherent vibration of the process. However most of the existing studies report that its effect is an improvement of the mechanical properties of the materials (Moffat et al., 2005; Fisher, 1973; Ivanov and Krushenka, 1992; Kadir, 2006) through a global faster solidification. The mechanism supposed to be responsible for this improvement is a faster solidification behaviour (liquid phase is cooling from a lower maximum temperature: C in Figures 6 8) due to an internal movement of the melt that promotes a quicker heat transfer inside the melt and to the mould walls. This effect is not supposed to be responsible for the gradient observed between position 1 and 3 but could be, along with the centrifugal effect, responsible for a global improvement of the rupture strength.
11 40 G. Chirita et al. Thus, it may be concluded that the gradient (between positions 1 3) (Figure 16) that is more pronounced in the centrifugal casting process is due to the fluid dynamics of the centrifugal process and consequent heat transfer in the walls of the mould. Further to this it is clear that an FGM is possible to be obtained, not through a chemical composition variation along the volume but through a change in the solidification behaviour. Figure 16 Experimental results of ultimate tensile: (a) strength and (b) strain, for gravity and centrifugal casting (see online version for colours) (a) It is observed that the rupture strength and strain can be adequately described by equation (1) (Figure 17), that may be derived for both casting techniques: 1 feut 2Stress orstrain (b) σ = k + V k (1) where, σ is the ultimate tensile strength; k 1 is an empirical factor that introduces the influence of several metallurgical features (such as SDAS; intermetallics amount and distribution); k 2 represents phase morphology and distribution on the eutectic constituent; and V f eut is the eutectic s volume fraction.
12 On the ability of producing FGMs with an AlSi12 aluminium alloy 41 Figure 17 Experimental and calculated values of (a) stress and (b) strain, for centrifugal and gravity castings in positions 1 3 of the ingot (see online version for colours) (a) Table 1 presents the obtained values of k 2 for the two casting processes. (b) Table 1 Correlation factor obtained for the casting processes, between eutectics volume fraction and rupture stress and rupture strain Alloy (12% Si) Casting process k 2 stress k 2 strain Gravity Centrifugal 4 Conclusions The main conclusions of this work are: The centrifugal process, as compared to the gravity casting: does not influence the chemical composition along the casting strongly influences the phase morphology and distribution along the casting
13 42 G. Chirita et al. these effects are related to the solidification process, namely the nucleation process and location fluid dynamics is the main process variable that is responsible for the gradient in properties along the radial direction giving origin to an FGM. Acknowledgements The research presented here was carried out in Materials Testing Laboratory of the Mechanical Engineering Department of University of Minho, and was supported by Fundação para a Ciência e Tecnologia (Portugal) through the PhD grant with the reference SFRH/BD/19618/2004. References ASM International (1997) Metal Handbook, 9th ed., Casting, Vol. 15, pp ASM International (2004) Aluminium-Silicon Casting Alloy: Atlas of Microfractografs, pp.1 9. Chirita, G., Stefanescu, I., Soares, D. and Silva, F.S. (2006) Centrifugal versus gravity casting techniques over mechanical properties, Proceedings of XXIII Encuentro de Grupo Español de Fractura, Albarracin, Marzo, Vol. I, pp Fisher, T.P. (1973) Effects of vibrational energy on the solidification of aluminium alloys, Br. Foundryman, Vol. 66, No. 3, pp Ghomashch, M.R. and Vikhrov, A. (2000) Squeeze casting: an overview, Journal of Material Processing and Technology, Vol. 101, pp.1 9. Guang, R., Zhou, J.G. and Wang, Q.G. (2006) The effect of hot isostatic pressing on the microstructure and tensile properties of an unmodified A356-T6 cast aluminum alloy, Journal of Alloys and Compounds, Vol. 421, Nos. 1 2, pp Ivanov, A.A. and Krushenka, G.G. (1992) Preparation of Al Si alloying composition by means of vibration, Liteinoe Proizvod, Vol. 3, pp.7 8, (Russian); Met. Abs., Janco, N. (1998) Centrifugal Casting, American Foundrymen Society, Inc., ISBN Kadir, K. (2006) Effect of low frequency vibration on porosity of LM25 and LM6 alloys, Materials and Design, pp.1 9. Liao, H.C., Sun, Y. and Sun, G. (2002) Correlation between mechanical properties and amount of dendritic α-al phase in as-cast near-eutectic Al 11.6% Si alloys modified with strontium, Materials Science and Engineering, Vol. A335, pp Moffat, A.J., Barnes, S., Mellor, B.G. and Reed, P.A.S. (2005) The effect of silicon content on long crack fatigue behaviour of aluminium silicon piston alloys at elevated temperature, International Journal of Fatigue, Vol. 27, pp Nikanorov, S.P., Volkov, M.P., Gurin, V.N., Yu, A., Burenkov, L.I., Derkachenko, B.K., Kardashev, L.L., Regel, W.R. and Wilcox, W.R. (2005) Structural and mechanical properties of Al Si alloys obtained by fast cooling of a levitated melt, Materials Science and Engineering, Vol. A390, pp Numan, A-D., Khraisheh, M., Saito, K. and Male, A. (2004) Silicon morphology modification in the eutectic Al-Si alloy using mechanical mold vibration, Materials Science and Engineering, Vol. A393, pp
14 On the ability of producing FGMs with an AlSi12 aluminium alloy 43 Saigal, A. and Fuller Jr., E.R. (2001) Analysis of stresses in aluminium alloys, Computational Materials Science, Vol. 21, pp Shabestari, S.G. and Moemeni, H. (2004) Effect of copper and solidification conditions on the microstructure and mechanical properties of Al-Si-Mg alloys, Journal of Materials Processing Technology, pp.153, 154, Wang, Q.G. (2003) Micro structural effect on the tensile and fracture behaviour of aluminium casting alloy A 356/357, Metallurgical and Materials Transaction, Vol. 34A, December, pp