Aluminum-4 mass%copper/alumina Composites Produced from Aluminum Copper and Rice Husk Ash Silica Powders by Powder Forging

Transcription

1 Materials Transactions, Vol. 51, No. 4 (2010) pp. 756 to 761 #2010 The Japan Institute of Metals uminum-4 mass%copper/umina Composites Produced from uminum Copper and Rice Husk Ash Silica Powders by Powder Forging Tachai Luangvaranunt 1, Chinawad Dhadsanadhep 1; *, Junko Umeda 2, Ekasit Nisaratanaporn 1 and Katsuyoshi Kondoh 2 1 Department of Metallurgical Engineering, Faculty of Engineering, Chulalongkorn University, Phyathai road, Patumwan, Bangkok 10330, Thailand 2 Joining and Welding Research Institute, Osaka University, Ibaraki , Japan This research demonstrated simple but effective process to produce -4 mass%cu/ 2 O 3 composites, by powder metallurgy method. The starting powders were aluminum, copper, and rice husk ash silica. Processing was by sintering at 650 C for 3.6 ks, hot forging of sintered billet at 600 C under 660 MPa pressure, followed by heat treatment. Hot forging of sintered billet induced plastic deformation of the matrix as well as fractured the porous silica, thus created ultimate contact between the two phases. The following heat treatment produced alumina, which was the reinforcement phase, by chemical reaction between fractured rice husk ash silica and aluminum matrix. The fabricated composite contained - and -alumina, and elemental silicon in matrix of aluminum solid solution. Addition of copper facilitated sintering by formation of liquid phase, as well as yielding a matrix material which can be strengthened by precipitation hardening. Maximum hardness obtained was 44 HRA, for composite material using 15 vol% rice hush ash silica. Peak hardness of the matrix was in range of HV, after aging for 28.8 to 43.2 ks. [doi: /matertrans.m ] (Received December 21, 2009; Accepted January 21, 2010; Published March 3, 2010) Keywords: powder metallurgy, forging, metal matrix composite, aluminum, copper, alumina, rice husk ash 1. Introduction uminum matrix composites gain much interest due to their superior properties compared to monolithic aluminum alloys. Strength, stiffness, wear resistance are among several properties which are improved by the hard reinforcement phases. Various processing methods have been reported to achieve high performance metal matrix composites. Examples are stir-casting, 1) infiltration casting, 2) direct-melt oxidation, 3) hot dipping, 4,5) and sintering of ball-mill activated mixture of powders. 6,7) Reinforcement phases are either added directly 1,2,6) or formed in-situ by a reaction occurring during processing. 3 5,7) Problems, both technical and material-wise, innate to existing processing methods are nonuniform reinforcement phase distribution, 1) weak bonding at interface of reinforcement/matrix, 8) undesirable chemical reaction at interface, 9) slow kinetics of reaction, 5) trapped porosity, 10) and others. Attempting to address the existing processing problems, this research explored powder forging as a simple but effective method for synthesis of aluminum matrix composite. The material system chosen was / 2 O 3 composite. The reinforcement phase, 2 O 3, was formed in-situ by an exothermic reaction between admixed rice husk ash silica and the aluminum matrix. In-situ formation of reinforcement phase was reported to achieve clean and strong interface with the matrix phase. 11) However, for an exothermic reaction to proceed autonomously, a critical ignition temperature was required. The key to achieving low activation temperature in this research was powder forging process which fractures the rice husk ash silica and forms ultimate contact between the silica and the aluminum matrix. Silica from rice hush ash was selected due to its amorphism and high specific surface area, *Graduate Student, Chulalongkorn University yielding higher chemical reactivity compared to crystalline silica. 12,13) Amorphous silica had a structure of a nearly tetrahedral network of SiO 4 4, with open structure where large pores surrounding atoms can play the role of vacancy in diffusion process. 14) Heat treatment of forged specimen induced the in-situ chemical reaction, in which alumina reinforcement phase was formed. Addition of four weight percent of copper as alloying element in the aluminum matrix was necessary to form liquid phase which effectively fill in any Kirkendall voids, and facilitated consolidation of the composite by typical liquid phase sintering mechanism. 15) Additional precipitation hardening treatment was performed to demonstrate the heat treatment ability of the composite. 2. Experimental Procedures uminum powders (99.9% purity, 106 mm mean size, Kojundo Japan), copper powders (99.9% purity, 70 mm mean size, Kojundo Japan) and silica powders (99.7% purity, 250 mm size) were mixed by shaking in a mby m plastic bag for 1200 s. Rice husk ash silica (99.7% purity, 250 mm size) was obtained by pyrolysis of rice husk as reported elsewhere. 16) The composition of the mixtures were -4 mass%cu/5 vol%silica, -4 mass%cu/ 10 vol%silica, and -4 mass%cu/15 vol%silica. The mixture was cold compacted in a m diameter die under a pressure of 214 MPa. The cold compact was pre-consolidated by sintering at 650 C for 3.6 ks under flow of Ar gas in a tube furnace, using heating rate of 8: C/s (5 C/ min). Hot forging of the sintered billet was carried out at 600 C, under 660 MPa with a strain rate of 17.5 s 1,in a 3: m diameter die. To induce in-situ chemical reaction to form alumina from silica and aluminum, the forged billet was heated at temperatures of 590, 630, and

2 uminum-4 mass%copper/umina Composites Produced by Powder Forging C for 36 ks. The composite was heat treatable to obtain maximum hardness of matrix. This was demonstrated by precipitation hardening procedure: solution treatment at 540 C for 7.2 ks, quenching in water, and aging at 165 C for varying time durations. X-ray diffractometer (XRD) was used to investigate the phases in the composites. Optical microscope (OM) and scanning electron microscope (SEM) with energy dispersive X-ray spectroscopy (EDS) were used to investigate their microstructure and chemical composition, respectively. Hardness of the fabricated composite materials was measured using Rockwell scale A. Thermal analysis was carried out under flow of nitrogen, using heating rate of 16: C/s (10 C/min). Measurement of area of reaction was by counting corresponding pixels in the digital micrographs, using average value from ten micrographs. Vickers hardness was measured using 9.8 N load and 15 s dwell time. 3. Results and Discussion Thermal behavior of the cold compact and hot forged specimens was investigated by differential thermal analysis. From the result in Fig. 1, the cold-compact of -4 mass%- Cu/15 vol%silica showed an endothermic peak starting at 650 C. The hot-forged specimen of the same composition; however, showed a slight endothermic reaction starting at 540 C, a large endothermic reaction starting at 610 C and an exothermic peak starting at 650 C. According to equilibrium binary phase diagram of aluminum and copper, eutectic point on aluminum-rich side was at 33 mass%cu, 548 C. The solidus temperature for - 4 mass%cu solid solution was at 580 C, above which solid solution coexisted with liquid solution. The DTA result showed large endothermic peak of the cold compact specimen, starting at 650 C. This should be the melting of aluminum powder. The large endothermic peak found in hot forged specimen, starting at 610 C indicated liquid formation of the -4 mass%cu composition. The difference between the observed solidus temperature in this experiment with that of the equilibrium phase diagram was due to non-equilibrium nature of powder compact. The cold compact was mixture of elemental aluminum and copper powders, not equilibrium solid solution composition; therefore, deviation from the DTA value, uv/mg Temperature, T / C Fig. 1 Differential Thermal Analysis of -4 mass%cu/15 vol%silica after being cold-compacted (dashed line) and hot-forged (solid line). equilibrium solidus temperature was not unusual. A delayed solidus temperature for the powder mixture was due to the slow rate of aluminum-copper solid state diffusion to achieve homogenous -4 mass%cu alloy. 17) To be of importance was the exothermic peak starting at 650 C in the hot forged specimen, which formed the in-situ 2 O 3 compound from rice husk ash silica and pure aluminum. The temperature of exothermic reaction coincided with the endothermic peak of the cold compact specimen. This suggested that melting of remaining pure aluminum powders was necessary for the reaction between rice hush ash silica and aluminum to occur. In addition, the fact that exothermic peak was observed only in the hot forged specimen confirmed that hot forging was necessary to activate the exothermic reaction. Forging caused severe plastic deformation in aluminum matrix and fractured rice husk ash. Consequently, it generated ultimate contact between fresh surfaces of rice husk ash silica and aluminum powder, facilitating the reaction to proceed. Comparison between the DTA curves of cold-compacted specimen and hot forged specimen confirmed the necessity of application of powder forging to initiate both the dissolution of copper into aluminum as well as the exothermic reaction which formed the in-situ alumina reinforcement phase. To investigate the in-situ chemical reaction, hot forged specimens were isothermally annealed at 590, 630, and 650 C, to obtain alumina reinforcement phase. Microstructure of all specimens annealed at temperatures above 590 C show reaction areas at locations of rice hush ash silica particles. Figure 2(a) and 2(b) were optical micrograph and SEM micrograph of typical microstructure of -4 mass%- Cu/15 vol%silica specimen after annealing at 590 C for 36 ks. From the optical micrograph, the matrix phase was light area, with homogeneously distributed network of dark areas of reinforcement phase. The pyramid shaped indentations were used as makers for SEM/EDS analysis of the same region which was shown in Fig. 3. Figure 3(a) to 3(d) were elemental mapping of aluminum, oxygen, silicon, and copper, respectively. By comparison between Fig. 2(a) and Fig. 3, the light matrix area consisted mostly of aluminum, and the dark area was dispersed with oxygen and islands of silicon. Elemental copper was found to be distributed sparsely in all areas. The areas where both aluminum and oxygen existed were the insitu formed aluminum oxide. It had irregular shape bordering the matrix phase, at locations where rice hush ash previously located. Phases existing in the composite were investigated by XRD, and the results were shown in Fig. 4. After sintering, the phases in the specimens were aluminum and 2 Cu compound. Rice husk ash silica was amorphous, and did not have characteristic XRD peaks. XRD pattern of asforged specimen was identical to that of as-sintered specimen. After heat treatment by isothermal annealing at 590 C for 36 ks, there existed the - and -alumina phases. The XRD results confirmed the in-situ formation of alumina as reinforcement phase in the composite. The matrix/alumina interface showed continuity, without any microcracks. In-situ formation of the alumina reinforcement created strong interface, in spite of reported poor wettability of pure aluminum on alumina. 18)

3 758 T. Luangvaranunt, C. Dhadsanadhep, J. Umeda, E. Nisaratanaporn and K. Kondoh (a) (b) 50 µm Fig. 2 Optical Micrograph (a) and SEM Micrograph (b) of a -4 mass%cu/15 vol%silica specimen after being heat treated at 590 C for 36 ks. (a) (b) (c) (d) Fig. 3 Elemental mapping by EDS of a -4 mass%cu/15 vol%silica specimen after being heat treated at 590 C for 36 ks. (a) aluminum, (b) oxygen, (c) silicon, and (d) copper. The mechanism of the in-situ formation of the alumina reinforcement phase in the present study can be summarized as diagram in Fig. 5. Cold compaction produced agglomerate of powders with large amount of porosity; therefore the low values of density. Specimen containing larger amount of added silica gave a lower density than specimen with less amount of silica, due to the low density value of silica and its porous nature.16) After sintering, aluminum powders were bonded; 2 Cu compound was formed by reaction between aluminum and copper powder. Hot forging at 600 C induced plastic deformation of the matrix as well as fractured the porous silica. This created ultimate contact between the fresh fracture surface of silica and the matrix. Remaining porosity was removed, as shown by the increase of density values. The chemical reaction found in this experiment is: 4(l) þ 3SiO2 (s) ¼ 22 O3 (s) þ 3Si(s). Thermodynamic calculation of standard Gibbs free energy of the reaction at 660 C gave 533 kj. The calculation was based on available thermochemical data of quartz silica instead of the unknown amorphous silica. The large negative Gibbs free energy values indicate that the reaction was thermodynamically favorable, and was thereby kinetically controlled. It was known that stable aluminum oxide film exists at surface of pure aluminum. Mechanical breakage of the film by hot forging was necessary to expose fresh pure aluminum surface readily for reaction. As mentioned in the introduction, amorphous silica in form of rice hush ash was reported to have higher reaction rate compared to crystalline one.12,13) Because of its metastable nature, as well as the high specific surface area, reaction with pure aluminum was enhanced by use of rice husk ash silica. In this research it was found that - and -alumina, transition alumina phases, were formed from the reaction between pure aluminum powder and amorphous silica,

4 uminum-4 mass%copper/umina Composites Produced by Powder Forging 759 Intensity, arb. unit (c) (b) # Cu # γ- 2 O 3 # κ- 2 O 3 # (a) Diffraction angle, 2θ Fig. 4 X-Ray diffraction patterns of -4 mass%cu/15 vol%silica (a) as-sintered, (b) as-forged, (c) as-heat treated. Cold compaction Sintering 650 C 3.6 ks Hot forging 600 C Density = 2.20x10 3 to 2.38x10 3 kg/m 3 Silica Density = 2.21x10 3 to 2.35x10 3 kg/m 3 Density = 2.51x10 3 to 2.81x10 3 kg/m 3 2 Cu Hardness = HV 2 Cu Cu Silica Silica Annealing C Solution heat treatment 540 C 7.2 ks Aging 165 C Density = 2.62x10 3 to 2.76x10 3 kg/m 3 Hardness = HV Hardness = HV 2 Cu Si Si 2 O 3 2 O 3 Fig. 5 Microstructural development of the composite during processing. instead of the more stable -alumina. This result compared well with solid-state reaction studies of this system reported elsewhere. 19) Transition alumina such as -or-alumina was reported to form at interface of amorphous Si film/aluminum film after heating at temperature as low as 377 C. 20) Thetaalumina with some trace of -alumina was found to occur by reaction between aluminum film and vitreous silica substrate in solid state. 13) Transition alumina phases have a more open

5 760 T. Luangvaranunt, C. Dhadsanadhep, J. Umeda, E. Nisaratanaporn and K. Kondoh Hardness, HV1 kg Fig Time, t / ks 5vol%RHA 10vol%RHA 15vol%RHA Hardness of matrix aging at 165 C for varying times. and distorted structure compared to -alumina. It was tempting to suggest that diffusion of aluminum atoms may be facilitated by the open structure of transition alumina; therefore, the chemical reaction between aluminum and silica can proceed. Addition of copper was only to facilitate sintering by liquid phase sintering mechanism. 15) A sessile drop wetting experiment by Shen et al., found that there was no reaction between liquid copper and amorphous silica. Copper has poor adhesion to amorphous silica, but was able to diffuse rapidly through both amorphous silica and aluminum oxide phase. 21) In this research, distribution of copper was in all phases in the composite, as shown by elemental copper mapping in Fig. 3(d). Upon cooling from processing temperature, due to its low solubility in aluminum, copper precipitated and formed 2 Cu compound in the matrix of the composite, which was confirmed by XRD result in Fig. 4. Silicon, as a product of the reaction has a larger diffusivity in aluminum than copper does, 17) but with much less solubility. Thereby, elemental silicon was found in the area immediate to the newly formed alumina phase, as shown in Fig. 3(c). Hardness of the matrix after hot forging was between HV, and increased to HV after annealing at C. This may be due to the complete dissolution of copper into aluminum matrix after long time exposure to high temperature. -4 mass%cu matrix was a heat treatable grade by precipitation hardening. Hardness of the matrix after aging at 165 C for varying times was shown in Fig. 6. The peak hardness of the matrix was obtained after aging for 28.8 to 43.2 ks. They were in range of HV, which were comparable to those of wrought alloy of similar composition. 22) Amount of added silica did not significantly affect either the obtained peak hardness or the aging time to obtain peak hardness. Therefore, addition of copper as alloying element was beneficial to both ease of consolidation by liquid phase sintering, and heat treatability to obtain high matrix hardness. Mechanical properties of the composite as a whole were assessed by hardness test, using Rockwell scale A. The results were shown in Fig. 7. Hardness increased with amount of added rice hush ash silica. The larger amount of added silica created larger proportion of reacted area, where alumina was present. Precipitation hardening further increased the hardness of the composite. Hardness of asannealed pure aluminum was 16 HRA, whereas those of as-annealed and as-aged composites produced from - Hardness, HRA mass%cu/15 vol%silica was 30 HRA, and 44 HRA, respectively. 4. Conclusions This research demonstrated simple but effective process to produce -4 mass%cu/ 2 O 3 composites. The method employed powder forging, which was hot forging of sintered materials, followed by heat treatment. Key to the in-situ formation of alumina reinforcement phase was hot forging which activated the chemical reaction between aluminum matrix and rice husk ash silica. The reaction produced - and -alumina phases and elemental silicon. Powder forging was shown to be an effective method to achieve full density with intimate contact between reinforcement phase and the matrix. Copper assisted densification by liquid phase sintering mechanism, and yield a composite which was heat treatable by precipitation hardening. Acknowledgement The research was financially supported by the Thailand Research Fund, grant number MRG REFERENCES Amount of RHA, vol% As-annealed hardness As-aged hardness Reacted Area Fig. 7 Hardness vs vol%rha, plotted against amount of reacted area in the micrographs. 1) A. N. Abd El-Azim, M. A. Kassem, Z. M. El-Baradie and M. Waly: Mater. Lett. 56 (2002) ) C. Badini, P. Fino, M. Musso and P. Dinardo: Mater. Chem. Phys. 64 (2000) ) V. S. R. Murthy and B. S. Rao: J. Mater. Sci. 30 (1995) ) M. C. Breslin, J. Ringnalda, L. Xu, M. Fuller, J. Seeger, G. S. Daehn, T. Otani and H. L. Fraser: Mater. Sci. Eng. A 195 (1995) ) W. Liu and U. Köster: Mater. Sci. Eng. A 219 (1996) ) S. Arakawa, T. Hatayama, K. Matsugi and O. Yanagisawa: Scr. Mater. 42 (2000) ) T. G. Durai, Karabi Das and Siddhartha Das: J. loy. Compd. 457 (2008) ) H. L. Rizkalla and A. Abdulwahed: J. Mater. Process. Technol. 56 (1996) ) J. W. Kaczmar, K. Pietrzak and W. Wosiski: J. Mater. Process. Technol. 106 (2000) ) K. H. Min, S. P. Kang, D. Kim and Y. D. Kim: J. loy. Compd. 400 (2005) ) P. Yu, C. Deng, N. Ma, M. Yau and D. H. L. Ng: Acta Mater. 51 (2003) ) J. Umeda, K. Kondoh, M. Kawakami and H. Imai: Powder Technol. 189 (2009) ) K. Prabriputaloong and M. R. Piggott: J. Electrochem. Soc. 121 (1974) Reacted Area, %

6 uminum-4 mass%copper/umina Composites Produced by Powder Forging ) V. V. Hoang, D. K. Belashchenko and V. Thuan: Physica B 348 (2004) ) R. M. German: Sintering Theory and Practice, (John Wiley & Sons, 1996) p ) T. Luangvaranunt, S. Jearwun and K. Kondoh: J. Met. Mater. Min. 17 (2007) ) D. Shang, J. E. Morral and H. D. Brody: Mater. Sci. Eng. A 447 (2007) ) A. Sangghaleh and M. Halali: J. Mater. Process. Technol. 197 (2008) ) D. H. Trinha, K. Backa, G. Pozinaa, H. Blomqvistb, T. Selinderb, M. Collinb, I. Reineckb, L. Hultmana and H. Högberga: Surf. Coat. Technol. 203 (2009) ) S. Roberts and P. J. Dobson: J. Phys. D: Appl. Phys. 14 (1981) ) P. Shen, H. Fujii and K. Nogi: Scr. Mater. 52 (2005) ) L. F. Mondolfo: uminum loys: Structure and Properties, (Butter Worths, London,1976) p. 703.