s J. Zou c,*, B.Y. Huang a, C.T. Liu d Cen Nan roan iona rm 31 ica g th the sion resistance and good oxidation resistance at elevated tem- technology and can be used for the filtration purposes. In fact, in the TieAl alloy system, there are several crystal has good oxidation resistivity and low density. As a conse- elemental powders. Al powders and dehydrided Ti powders, all with particle sizes varying between 200 and 400 meshes and having 99.8% purities, were used in the experiment. The nominal compositions were varied from Tie20 wt% Al to Tie65 wt% Al with an increment of 5 wt% Al in order to * Corresponding authors. E-mail addresses:
[email protected] (Y.H. He),
[email protected] (J. Zou). Available online at www.sciencedirect.com Intermetallics 16 (2008) peratures [4e6]. However, their intrinsic low tensile ductility at room temperature and poor high temperature strength have limited their potential applications, particularly in the aerospace field [7], although the TieAl alloy system has been used in the automotive industry in a limited scale [8,9]. On the other hand, porous materials have showed wide applications in filtration fields, such as seawater desalination, environmental protection, pharmaceuticals’ separation, metal- lurgy, and chemical engineering [10e14]. Our recent study [15] has demonstrated that the porous TieAl structures can be fabricated through the traditional powder metallurgy quence, it is necessary to study the effect of the Al content on pore structures for porous TieAl alloys. In this paper, we demonstrate that the porous TieAl alloys can be fabricated with a wide range of compositions (between Tie20 wt% Al and Tie65 wt% Al) and show different phases with different Al contents, and that the Al content in porous TieAl alloys affects their pore structure dramatically. 2. Experimental The fabrication of porous TieAl alloys was started with engineering performances such as excellent acid/alkali corro- structural materials [1e3]. This is due to their outstanding 42 wt% has low density and structure stability at high temper- atures, while the TiAl3 alloy with the Al content ofw63 wt% a2-Ti3Al, g-TiAl, and TiAl3) when using different compositions. The fundamental reasons behind these phenomena have been explored. � 2007 Elsevier Ltd. All rights reserved. Keywords: A. Titanium aluminides; C. Reaction synthesis 1. Introduction The TieAl alloy system has been investigated extensively for more than two decades as the potential high temperature structures having good physical and chemical properties [1e6]. For examples, the a2-Ti3Al alloy with the Al content from 14 wt% to 23 wt% has excellent corrosion resistivity; the g-TiAl alloy with the Al content from 35 wt% to Effects of the Al content on pore Y. Jiang a, Y.H. He a,*, N.P. Xu b, a State Key Laboratory for Powder Metallurgy, b Membrane Science and Technology Research Center, c School of Engineering and Centre for Microscopy and Mic d Metals and Ceramics Division, Oak Ridge Nat Received 14 October 2006; received in revised fo Available online Abstract Porous TieAl alloys with different nominal compositions were fabr been found that the pore parameters vary with the Al contents, indicatin contents. In addition, detailed structural characterizations showed that 0966-9795/$ - see front matter � 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.intermet.2007.11.002 tral South University, Changsha 410083, China jing University of Technology, Nanjing 210009, China alysis, The University of Queensland, QLD 4072, Australia l Laboratory, Oak Ridge, TN 37831-6115, USA 7 November 2007; accepted 17 November 2007 December 2007 ted through a reactive synthesis of Ti and Al elemental powders. It has at the nature of the pores can be manipulated through changing the Al fabricated porous TieAl alloys can have three crystalline phases (i.e., tructures of porous TieAl alloys 327e332 www.elsevier.com/locate/intermet gain a systematic understanding of the relationship between the fabricated pore structures and the Al contents. The reactant powders were firstly dry-mixed in a tumbler ball mill for 8 h with the rolling rate of 75 rpm, followed by cold pressing into compact discs with a dimension of w32 mm� 2.5 mm. The cold die pressing pressure was 240 MPa. The relative density of the discs has been determined by comparing the the- oretical volume and the measured volume, which wasw89%. The compact discs were then sintered in a vacuum furnace with a pressure of 10�3 Pa. The sintering temperature as a function of sintering time is illustrated in Fig. 1, in which three holding platforms were indicated. In general, two sinter- ing processes can be used to fabricate the porous TieAl alloys: the stepped heating method (as shown in Fig. 1) and the continuous heating method. Our extensive experiments showed that the compact discs sintered through the continuous heating method with a heating rate of w5 �C/min can cause in water according to the Archimedes method. Then, the open porosity (qp) can be calculated by the expression: qp ¼ ðWX �WSÞ=ðVX � rxÞ. And the overall porosity (qv) can be calculated by the expression: qv ¼ 1�WX=ðVX � rtÞ, where rt is the theoretical density of the synthesized alloy. The pore size can then be measured by the bubble point method [17]. 3. Results and discussions 3.1. Volume expansion As indicated in our previous study [15], the volume expan- sion during the sintering procedure for the TieAl alloy system is one of the most remarkable phenomena. Fig. 2 shows the volume and dimension expansion ratios (the volume/dimen- sion increased in the sintered discs divided by the volume/ dimension of the unsintered discs) as a function of the Al con- tent. For each measurement, five sintered discs were measured 40E s a 328 Y. Jiang et al. / Intermetallics 16 (2008) 327e332 cracks in the edges of the discs and deformations of the discs, while the compact discs sintered through the stepped heating method, in which the discs were kept at 600 �C for 2 h (the melt point of Al is w660 �C), can eliminate cracks and preserve the original shapes, which is critically vital to the practical applications. The stepped heating method can also avoid the overmuch liquid Al metal, which otherwise can cause a strenuous reaction with Ti [16] and harm to the shape of the discs, when the sintering temperature is beyond the melt point of Al. The sintered discs with different nominal compositions were characterized by X-ray diffraction (XRD: Dmax 2500VB) to identify the crystalline phases in these porous TieAl alloys. The disc dimensions before and after the sintering were mea- sured to determine the volume expansion. The pore structure was characterized by scanning electron microscopy (SEM). The open porosity was measured by the Archimedes method in water based on the following principles: the mass (WS) of a dried porous disc and the mass (WX) of the porous disc filled with the wax with a density of rx are weighted in air. The external volume (VX) of the waxed disc can be determined 0 100 200 300 400 500 600 0 200 400 600 800 1000 1200 1400 T e m p e r a t u r e / ° C Time / h Fig. 1. The illustration of sintering procedure. 20 30 40 50 60 70 0 20 Al Content / % Fig. 2. The volume expansion of the compacts with the nominal composition 60 x p a n 80 i o n r and the corresponding expansion ratio is the arithmetical average value n of expansion ratios of five discs ni (i¼ 1, 2,.5) for each composition. And the standard error s is esti- mated through the expression: s ¼ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiP5 i¼1ðni � nÞ2=5 q . As can be seen from Fig. 2, for the Al contents lower than 60 wt%, the volume expansions increase almost linearly with increasing the Al content. The maximum volume expansion ratio reaches w137% when the Al content is 60 wt%. After that, the volume expansion began to reduce to w108% at the Al content of 65 wt%. This reduction in the volume expansion is mainly due to the existence of excessive Al metal in the system. It should be noted that the TiAl3 phase is the highest Al contain- ing TieAl phase which has a weight ratio of Tie63 wt% Al. As a consequence, for the case of Tie65 wt% Al, Al is over- supplied in the system to form TiAl3 alloys. This excessive Al metal will melt after the sintering temperature exceeds the 100 120 140 160 t i o / % Radial Direction Axial Direction Volume Expansion of Tie20 wt% Al up to Tie65 wt% Al after finally sintering at 1300 �C for 30 min. melting point of Al, which may result in the shrinkage of the sintered grains due to the surface tension of liquid phase [18] and also may block the pores formed between the sintered grains. We anticipate that these synergetic effects lead to the reduction of the volume expansion in the case of Tie65 wt% Al. The axial and radial expansion behaviors are similar to the behavior of the volume expansion, and their maximum ex- pansion ratios (at the Al content of 60 wt%) are measured to be w39% and w31%, respectively. 3.2. Phase identification Fig. 3 is XRD patterns of sintered discs formed by different Al contents. As can be seen from Fig. 3, three different phases were found in four Al contents, namely, a2-Ti3Al phase found in the Tie20 wt% Al alloy, a2-Ti3Al and geTiAl phases in the Tie30 wt% Al alloy, g-TiAl phase in the Tie40 wt% Al alloy and near single TiAl3 phase in the Tie60 wt% Al alloy. It should be noted that (1) the phase changes with varying the Al content and (2), for the cases of Tie20 wt% Al, Tie40 wt% Al and Tie60 wt% Al alloy, single TieAl phase can be achieved in each case. These phenomena can be well under- stood through the TieAl binary phase diagram [19] as these compositions match with the compositions of the a2-Ti3Al, g-TiAl and TiAl3 phases. for this behavior is the agglomeration of Al particles in com- large α2-Ti3Al phase { {201} {401 { a Y. Jiang et al. / Intermetalli 10 20 30 40 50 60 70 80 90 {002} {101} {110} {004} {200} {114} {204} {220} {301} {312} {224} 2 Fig. 3. X-ray diffractograms of porous TieAl alloy filters showing the effect of raw material composition on the final phase composition of these porous TiAl3 phase 103} d C P S {111} {001} {110} {002} {200} {201} {202} {220} {113} {311} {222} Υ -TiAl phase Ti3 Al Ti3 Al Ti3 Al Ti3 Al α2+Υ phase TiAl TiAl TiAl TiAl TiAl TiAl TiAl TiAl TiAl TiAl TiAl {100} {101} {110} {200} 002} {202} {220} {203} {222} } b c TieAl alloy filters: (a) Tie20 wt% Al, (b) Tie30 wt% Al, (c) Tie40 wt% Al, (d) Tie60 wt% Al. pacts when the number of the Al powders is significantly 3.3. Pore structure As indicated earlier, the Al content can affect the volume expansion of sintered porous TieAl alloys dramatically, it is anticipated that the Al content can also have a great influence on the pore structure of these porous alloys. Since the pore structures are critically important for them to be practically useful, it is important to understand the variation of the pore structures as a function of the Al content. There are three key parameters that can comprehensively describe the pore structural properties: open porosity, the maximum pore size and the gas permeability, so that we need to address these parameters. Fig. 4(a) shows the overall porosity, open porosity and the open porosity proportion of the porous TieAl alloys as a func- tion of the Al content. There always exists a certain closed porosity since the interparticle pores in the green compacts cannot be avoided. Similar to the relationship between the vol- ume expansion and the Al content, the open porosity increases almost linearly with increasing the Al content for the Al con- tent� 60 wt%, indicating that the pores were initiated from the consumption of the Al metal during the reactive sintering. This is in accordance with the pore formation mechanism caused by the Kirkendall effect [20]. Since the diffusion rate of Al is significantly larger than that of Ti in the TiAl3 phase, which is the only phase formed during the solid diffusion reaction of Ti and Al elements below the melt point of Al [3], the net movement and consumption of Al element must be balanced by the opposite net vacancy flux, which will result in a large number of vacancies near the original positions of Al atoms. Excessive vacancies can condense into pores to reduce the Gibbs free energy for the system [15], i.e., the consump- tion of Al metal results in the formation of pores directly. The maximum open porosity of the porous TieAl alloys is w59% at the composition of Tie60 wt% Al. After that, the open porosity began to reduce gradually since the residual Al metal, when beyond its melting point, becomes liquid and coats over TiAl3 grains. Fig. 4(b) shows the maximum pore size of the porous TieAl alloys as a function of the Al content. Although it shows a tendency of the higher the Al content the larger the maximum pore size, the linear relationship, as shown in the case of the open porosity, disappears. In fact, when the Al con- tent varies from 30 wt% to 40 wt%, the maximum pore size almost remains as a constant, which indicates the uncorrelated relation between the two parameters. However, when Al con- tent is below 30 wt%, the maximum pore size increases tardily with increasing the Al content. This is mainly because the proportion of open pores increases with the Al content, as shown in Fig. 4(a), which enlarges the pore size when pores interconnected together. When the Al content is over 40 wt%, the open porosity almost remains as a constant, while the pore size increases apparently during this stage. The reason 329cs 16 (2008) 327e332 than that of Ti powders, which would enlarge the practical size of the Al particles. Since pores generated from the exhaustion tall 95 O p a 70 330 Y. Jiang et al. / Interme of Al, the larger Al particle size would lead to the formation of larger pores. Unlike the cases of the volume expansion and the open porosity, the pore size reaches the maximum (w46 mm) at the Al content of Tie65 wt% Al in this study. Compared tion- pacts 20 30 40 50 60 70 80 85 90 P o r o s i t y / % Al Content / wt % Open Porosity Overall Porosity Open Pore Proportion e n P o r e P r o p o r t i o n / % 20 30 40 50 60 70 20 30 40 50 Al Content / % b 20 30 40 50 60 70 5 10 15 20 25 P e r m e a b i l i t y / 1 0 - 5 m 3 P a - 1 m - 2 s - 1 Al Content / wt % c 30 40 50 60 M a x i m u m A p e r t u r e / m Fig. 4. Pore structure curves showing the effects of raw material composition on the pore structure properties of porous TieAl alloy filters: (a) open poros- ity, (b) maximum pore size, (c) gas permeability. r0V0ð1� q0Þ ¼ rV � 1� qp� qc � : ð1Þ where r0 and r are the theoretical densities of the compact before and after the sintering; V0 and V are the corresponding volumes; qc is the closed porosity in sintered compact. Considering V¼ V0(1þ a) with a being the volume expan- sion ratio, an expression can be finally derived as qp ¼�r0ð1� q0Þ r � 1 1þ aþ ð1� qcÞ: ð2Þ Since the pores would be blocked by the excess Al when the Al content is beyond the composition in TiAl3 phase, we only consider the composition below Tie60 wt% Al. A com- parison between Eq. (2) and the experimental data is shown in Fig. 6. The analysis implies that, based on the experimentally observed data in accordance with the theoretically calculated value, the pore structure of TieAl porous alloy can be a mathematic expression can be used to describe the rela ship between the initial overall porosity q0 in green com and the open porosity qp after the sintering: with the open porosity shown in Fig. 3(a), the pore size remains increasing even if some residual Al metal exist and leads to the reduction of the porosity. This is possibly due to the excessive Al metal tending to fill the small pores where the capillary force is relatively large. Fig. 4(c) shows the gas permeability of the porous TieAl alloys as a function of the Al content. Although the gas perme- ability monotonically increases with increasing the Al content, the increase of the gas permeability is slow for the Al content less than 45 wt%. After that, the gas permeability increases sharply with the increased ratio over an order of magnitude (from 1.44� 10�5 m3 m�2 Pa�1 s�1 for the Tie20 wt% Al to 2.56� 10�4 m3 m�2 Pa�1 s�1 for the Tie65 wt% Al). These permeabilities are sufficient for the solid-to-gas separation applications. To evaluate the pore structures, SEM was employed. Fig. 5 is SEM images of three porous TieAl alloys with the Al content, respectively, being Tie20 wt% Al, Tie40 wt% Al and Tie60 wt% Al. As can be clearly seen from these SEM images, the general pore size increases with increasing the Al content, which is consistent with results shown in Fig. 4. 3.4. Relationship between the volume expansion and the open porosity Since the generation of the volume expansion and pores is due to the exhaustion of Al through the Kirkendall effect, there must be a quantification relation between them. In fact, ac- cording to the mass conservation before and after the sintering, ics 16 (2008) 327e332 controlled through controlling the volume expansion during the sintering procedure. and rage Fig. 5. Pore microstructure of these porous TieAl alloy filters: (a) Tie20 wt% Al, (b) Tie35 wt% Al, (c) Tie65 wt% Al. Y. Jiang et al. / Intermetalli a constant related to the properties of permeation fluid macroscopic dimension of porous material. Given there is a determined relation between the ave 3.5. Relationship between the open porosity, the maximum pore size and the permeability The relationship between the permeability and the pore structure parameters can be described by the HagenePoiseuille formula [21], which gives J ¼ K� d2� q; ð3Þ where J is the permeability; d and q are the average pore size and the open porosity of porous material, respectively; K is 0.4 0.5 0.6 0.7 0.8 0.9 1.0 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 O p e n P o r o s i t y Experimental Data Equation (2) (1+ )-1 Fig. 6. Relation between the open porosity and volume expansion as theoret- ically calculated and experimentally observed. 331cs 16 (2008) 327e332 pore size d and the maximum pore size dm with dm¼ L� d, where L is the proportion factor, and givenG¼K� L, the rela- tionship between the permeability and maximum pore size can be expressed as J ¼ G� d2m � qp: ð4Þ We may now verify Eq. (4) in TieAl porous alloys by check- ing the relationship between the experimental data J and d2m � qp, as shown in Fig. 7. Since the experimental data fit well with the plot suggested by Eq. (4), indicating that the behavior of open porosity, the maximum pore size and the per- meability obeys the HagenePoiseuille formula. The constant G can be then be determined according to the linear fit shown in Fig. 7, which gives G¼ 0.0224. By now, the relationship between the permeability and the pore structure parameters in porous TieAl alloys can be described quantitatively by J ¼ 0:0224d2mq: ð5Þ a simple and cost effective way with controllable pore size in a large scale. Acknowledgements This project is financially supported by the Natural Science Foundation of China (20476106 and 50721003), the National Basic Research Program (2003CB615707), Hunan Provincial Natural Science Foundation of China (05JJ30097) and the 111 Project of Chinese Ministry of Education. References [1] Liu CT, Schneibel JH, Maziasz PJ, Wright JL, Easton DS. Intermetallics 1996;4:429. [2] He YH, Liu YX, Huang BY, Qu ZH, Lei CM. J Mater Sci Technol 1994;10:205. [3] Van Loo FJJ, Rieck GD. Acta Metall 1973;21:61. 0 200 400 600 800 1000 1200 5 10 15 20 25 Linear fit according to Equation (3)Experimental data K / 1 0 - 5 m 3 P a - 1 m - 2 s - 1 d 2 m / m 2 Fig. 7. Experimental verification of HagenePoiseuille equation in TieAl 332 Y. Jiang et al. / Intermetallics 16 (2008) 327e332 4. Conclusions Porous TieAl alloys with the nominal compositions rang- ing from Tie20 wt% Al to Tie65 wt% Al were successfully fabricated by the reactive synthesis of Ti and Al elemental powders. With increasing the Al content, the sintered porous TieAl alloys have three typical phases: a2-Ti3Al, g-TiAl, and TiAl3. The effects of the Al content on the pore structure parameters (including open porosity, maximum pore size and gas permeability) were comprehensively determined. An optimal composition of Tie60 wt% Al was acquired since the alloy has maximum open porosity of w59% with a large gas permeability of 1.7� 10�4 m3 m�2 Pa�1 s�1. The behav- iors of pore structure properties in porous TieAl alloys have been found to obey the HagenePoiseuille formula. One of porous alloy. the advantages of these porous TieAl alloys over conventional porous materials lies in the fact that they can be fabricated by [4] Lapin J. Intermetallics 2006;14:115. [5] Chan KS, Wang P, Bhate N, Kumar KS. Acta Mater 2004;52:4601. [6] Wu X. Intermetallics 2006;14:1114. [7] Henaff Gilbert, Gloanec Anne-Lise. Intermetallics 2005;13:543. [8] Gebauer K. Intermetallics 2006;14:355. [9] Sung SY, Kim YJ. Intermetallics 2006;14:1163. [10] Van der Bruggen B, Vandecasteele C. Desalination 2002;3:207. [11] Aririatu LE, Ewelike NC. Environ Protect Eng 2003;2:17. [12] Le´tant SE, Hart BR, Van Buuren AW, Terminello LJ. Nat Mater 2003;2:391. [13] Herna´ndez N, Sa´nchez-Herencia AJ, Moreno R. Acta Mater 2005;4:919. [14] Yamaguchi A, Uejo F, Yoda T, Uchida T, Tanamura Y, Yamashita T, et al. Nat Mater 2004;3:337. [15] He YH, Jiang Y, Xu NP, Zou J, Huang BY, Liu CT, et al. Adv Mater 2007;19:2102. [16] Mackowiak J, Shreik LL. J Less-Common Met 1968;15:341. [17] Herna´ndez A, Calvo JI, Prada´nos P, Tejerina F. J Membr Sci 1996;112:1. [18] Munir ZA. Ceram Bull 1998;67:342. [19] Austin CM, Kelley TJ, Mcallister KG, Chesnult JC. Aircraft engine ap- plications for gamma titanium aluminide. In: Nathal MV, Darolia R, Liu CT, Martin PL, editors. Structural intermetallics; 1997. p. 413. [20] Lee TK, Mosunov EI, Hwang SK. Mater Sci Eng A 1997;239:540. [21] Xu GB, Wang FM. J Ind Text 2005;34:243. Effects of the Al content on pore structures of porous Ti-Al alloys Introduction Experimental Results and discussions Volume expansion Phase identification Pore structure Relationship between the volume expansion and the open porosity Relationship between the open porosity, the maximum pore size and the permeability Conclusions Acknowledgements References