- Email: [email protected]
An effective method for reducing porosity in the titanium aluminide alloy Ti52Al48 prepared by elemental powder metallurgy
Scripta METALLURGICA et MATERIALIA
Vol.
26, pp. 1469-1474, 1992 Printed in the U.S.A.
Pergamon Press Ltd. All rights reserved
AN EFFECTIVEMETHODFOR REDUCINGPOROSITYIN THE TITANIUM ALUMINIDEALLOYTis2AI48 PREPAREDBY ELEMENTALPOWDERMETALLURGY
G.-X. Wangand M. Dahms (GKSS Research Centre, Geesthacht, Germany)
(Received February 24, 1992) (Revised March i0, 1992)
I. Introduction Titanium aluminide alloys based on the intermetallic compound TiAI are becoming more and more interesting due to their potential applications as high temperature materials. These alloys have been successfully prepared by elemental powder metallurgy, as reported in our earlier work (1,2). In our processing, the cold-extrusion technique is used to consolidate the elemental Ti- and AI-powder mixture with or without additives. The as-extruded material condition exhibits no intermetallics and can therefore be easily machined to complex shapes, which are then reactively sintered to obtain the desired intermetallics. In this way, the poor workability of titanium aluminides can be overcome. During reaction sintering, more AI-atoms move into Ti-particles, resulting in the formation el pores if no pressure is applied. The final pore size is proportional to the size of the AI-regions. Normally, hot isostatic pressing (HIP) is applied to eliminate pores. However, an HIP-treatment requires encapsulation of extruded pieces due to their open porosity and is therefore relatively expensive. It is thus of practical importance to reduce the size of the AI-regions and the open porosity of extruded pieces. This can be achieved by increasing the extrusion ratio, which is defined as the area ratio of specimen cross sections before and after extrusion. In the following, we report on the influence of the extrusion ratio on pore formation in TiszAI4a during pressureless reaction sintering II. ExD0rimenta,I Elemental titanium and aluminum powders of sizes smaller than 100 lain, with purities of 99.8% and 99.9%, respectively, were mixed in air to the desired composition Ti-48at.%AI. The powder mixture was pressed at room temperature to a green compact. This compact was cold extruded using two different extrusion ratios of 17 and 350. The extruded specimens were reactively sintered in a vacuum (= 10 ~ N/m 2) furnace for 6 h at 1350oc. A heating rate of 20°C/min was used. The cooling rate was approximately 130°C/min for temperatures higher than 800°C and 33°C/min for temperatures between 800 and 300°C. As-extruded specimens as well as specimens sintered at 1350oc/6 h were examined using light microscopy. A quantative determination of the porosity in the specimens was carried out with the help o1 a computer-aided image analysis system.
1469 0 0 3 6 - 9 7 4 8 / 9 2 $ 5 . 0 0 + .00 Copyright ( c ) 1992 P e r g a m o n P r e s s
Ltd.
1470
POWDER M E T A L L U R G Y
Vol.
26, No.
III. Results and Discussion 1. Microstructures of as-extruded conditions Figure 1 shows the microstructures of both as-extruded conditions, with AI having a white and Ti a dark appearance. During extrusion, particles, especially the AI-particles, are elongated in the extrusion direction to fibres. With increasing extrusion ratio, both AI and Ti-fibres become thinner and longer. Some small pores, mostly situated between Ti-particles, can be detected in the specimen extruded with the low extrusion ratio of 17 (Fig. l a and Fig. lb). The porosity is smaller than 0.5%. The specimen with the high extrusion ratio of 350 exhibits almost no pores (Fig. l c and Fig. ld). 2. Microstructures after pressureless sintering After pressureless sintering at 1350oc/6 h, both specimens are porous, with a porosity of approximately 16% for the low extrusion ratio (Fig. 2a), and less than 2% for the high extrusion ratio (Fig. 2b). The maximum pore size was measured to be about 100 pm diameter for the low extrusion ratio (Fig. 2a) and about 30 p.m diameter for the high extrusion ratio (Fig. 2b). Under polarized light, a duplex structure with TiAI- and Ti3Al-lamellae and some TiAIregions is visible (Fig. 3). The lamellar volume fraction was measured to be about 80%. The size of the lamellar grains is about 70 ~m and 50 p.m diameter for the low and high extrusion ratios, respectively. It is worth mentioning here that the lamellar grains in both specimens are much smaller than those in alloys prepared by ingot metallurgy (see for example (3)). 3. Mechanisms of pore formation and effect of increasing extrusion ratio Depending on the temperature applied, sintering of extruded elemental Ti- and AI-powder mixture can be of the nature of pure solid state interdiffusion (at temperatures below the AI-melting point), or it can be regarded as a temporary liquid phase sintering (at temperatures above the AI-melting point). In the first case, Kirkendall-diffusion takes place. The AI atoms move into Ti-particles, whereas the Ti atoms stay almost immobile (4). As a result of such one-sided diffusion, the originally dense specimens of the as-extruded conditions (Fig. 4a) become porous, with pores being situated mostly at the interface region (Fig. 4b). In the second case, AI-melt flows away along particle or grain boundaries, leaving pores behind (Fig. 4c). The AI-melt then bedews the Ti-particles so that the contact surface between Ti and AI is enlarged and the reaction between Ti and AI is accelerated. The products of the reaction are the intermetallics such as TiAI3, TiAI2 (as two intermediate products), TiAI and Ti3AI (1), which are indicated by the hatched areas in Fig. 4. After all elemental AI has been consumed, a further solid state interdiffusion between Ti and the intermetallics formed can take place. The final microstructure consists of the intermetallics TiAI and Ti3AI, and pores (Fig. 4d). The situations shown in Fig 4 can also appear successively during a sintering process. This is the case in our pressureless sintering at 13500C/6 h. The Kirkendall-diffusion (Fig. 4b) took place during heating prior to the AImelting point. Because of the relatively large heating rate of 20°C/min, the part of the porosity due to the Kirkendall-diffusion must be considered to be not essential in this work. We can thus assume that the pore sizes immediately after the flow-away of AI-melt are approximately equal to the sizes of Al-fibres of the as-extruded conditions, i. e. the specimen with the high extrusion ratio of 350 has much smaller pores at this stage than those in the specimen with the low extrusion ratio of 17. The pore sizes will be reduced or eliminated at later stages of the sintering due to the well-known capillarity. Under the action of the capillary forces, vacancy emigration (Nabarro-Herring-creep) and dislocation climb (Peach-KOhler-creep) can take place. The capillary force increases with decreasing pore size (5). Due to larger capillary forces, the pore size as well as the porosity in the specimen with the high extrusion ratio of 350 have been more reduced than those in the specimen with the low extrusion ratio of 17.
9
Vol.
26, No.
9
POWDER METALLURGY
1471
The reduction of pore size and porosity by increasing extrusion ratio can also be understood from an energetic viewpoint. Considering extrusion ratio as the only variable, then more volume energy is stored in as-extruded specimens with a higher extrusion ratio (cold working), so that more surface energy has to be removed during sintering (6). This results in a smaller pore size and porosity. Shibue (7) showed that the porosity can be reduced by replacing AI-powder with AI-7 wt.%Mn alloy powder. This fact can be also explained as result of reducing sizes of AI-fibres. Indeed, for the same extrusion ratio of 15, the as-extruded Ti- and AI-7 wt.%Mn powder mixture exhibits a finer microstructure than the as-extruded Ti- and AIpowder mixture. IV. summarv Generally, pore formation takes place during pressureless sintering of extruded elemental Ti- and AI-powder mixture because of the Kirkendall-diffusion and / or the flow-away of the AI-melt. On the other hand, the pores formed can be reduced or eliminated by Nabarro-Herring-creep and Peach-K6hler-creep due to the capillarity. With a higher extrusion ratio, a finer microstructure can be achieved for as-extruded specimens. This has proved to be an effective method for reducing not only the pore size but the porosity after sintering as well. Assuming the pores formed after reaching the AI-melting point are of the sizes of the AI-fibres in the as-extruded conditions, a larger average capillary torce and a larger reduction of pore size and porosity can be expected for specimens extruded with a higher extrusion ratio. This has been confirmed experimentally in this study. In addition, a refinement of grains, particularly of lamellar grains, has been achieved by increasing extrusion ratio Acknowledgments: The authors would like to thank Ms. W.-V. Schmitz, Mr. G. T6bke, Ms. B. Wildhagen and Mr. K. Wulf for technical assistance. V. References (1) G.-X. Wang and M. Dahms: "An Overview: TiAI-Based Alloys Prepared by Elemental Powder Metallurgy", to be published in Powder Metallurgy International (2) G.-X. Wang and M. Dahms: "Influence of Heat Treatment on Microstructure of Ti-35 wt.%AI Prepared by Elemental Powder Metallurgy", Scripta Met. et Mat., Vol. 26, No. 5, March 1992,pp. ? (3) Y.-W. Kim and D. M. Dimiduk: "Progress in the Understanding of Gamma Titanium Aluminides", JOM, Aug 1991, pp. 40-47 (4) F. J. J. van Loo and G. D. Rieck: "Diffusion in the Titanium-Aluminium System - I. Interdiffusion between Solid AI and Ti or Ti-AI Alloys", Acta Met. 21, 1973, pp. 61-71 (5) W. Schatt: "Pulvermetallurgie Sinter- und Verbundwerkstoffe", VEB Deutscher Verlag fur Grundstoffindustrie Leipzig, 1984 (6) G. Leitner, M. Dahms, W. PoeBnecker and B. Wildhagen: "Reactiive Sintering of Titanium Aluminides", to be published in proceedings of "PM Aero "91" (7) K. Shibue: "Suppression of Pores for TiAI Intermetallic Compound Prepared by Reactive Sintering", Sumitomo Light Metal Technical Reports, Vol. 32 (No. 2), April 1991, pp. 95-101
1472
POWDER M E T A L L U R G Y
(a)
(c)
,~) ~m
Vol.
26, No. 9
(b)
(d)
Fig. 1. Bright field LM-photographs of the both as-extruded conditions ((a, b) with extrusion ratio = 17, and (c, d) with extrusion ratio = 350) in directions perpendicular (a, c) and parallel to the extrusion direction (b, d)
Vol.
26, No.
9
POWDER M E T A L L U R G Y
(a)
501J'm
1473
(b)
Fig. 2. Microstructures in bright field after the pressureless sintering at 1350oc/6 h: (a) extrusion ratio = 17 and (b) extrusion ratio = 350; taken from specimen sections perpendicular to the extrusion direction
Fig. 3. Microstructures under polarized light after the pressureless sintedng at 1350°(3/6 h: (a) extrusion ratio = 17 and (b) extrusion ratio ,, 350; taken from specimen sections perpendicular to the extrusion direction t
1474
POWDER M E T A L L U R G Y
(a)
Vol.
I-- Ti-parfic[es
(b)
1 3, "?m1,' ~.
26, No.
--At-partictes
~:~ .,, intermetallics C~ ~) --pores
(c)
(d)
Fig. 4. Schematic description of microstructural changes during reaction sintering of extruded "ri- and AI-powder mixtures: (a) as-extruded condition; (b) after certain Kirkendall-diffusion; (c) after flow-away of the AI-melt; (d) after complete reaction sintering
9
Vol.
26, pp. 1469-1474, 1992 Printed in the U.S.A.
Pergamon Press Ltd. All rights reserved
AN EFFECTIVEMETHODFOR REDUCINGPOROSITYIN THE TITANIUM ALUMINIDEALLOYTis2AI48 PREPAREDBY ELEMENTALPOWDERMETALLURGY
G.-X. Wangand M. Dahms (GKSS Research Centre, Geesthacht, Germany)
(Received February 24, 1992) (Revised March i0, 1992)
I. Introduction Titanium aluminide alloys based on the intermetallic compound TiAI are becoming more and more interesting due to their potential applications as high temperature materials. These alloys have been successfully prepared by elemental powder metallurgy, as reported in our earlier work (1,2). In our processing, the cold-extrusion technique is used to consolidate the elemental Ti- and AI-powder mixture with or without additives. The as-extruded material condition exhibits no intermetallics and can therefore be easily machined to complex shapes, which are then reactively sintered to obtain the desired intermetallics. In this way, the poor workability of titanium aluminides can be overcome. During reaction sintering, more AI-atoms move into Ti-particles, resulting in the formation el pores if no pressure is applied. The final pore size is proportional to the size of the AI-regions. Normally, hot isostatic pressing (HIP) is applied to eliminate pores. However, an HIP-treatment requires encapsulation of extruded pieces due to their open porosity and is therefore relatively expensive. It is thus of practical importance to reduce the size of the AI-regions and the open porosity of extruded pieces. This can be achieved by increasing the extrusion ratio, which is defined as the area ratio of specimen cross sections before and after extrusion. In the following, we report on the influence of the extrusion ratio on pore formation in TiszAI4a during pressureless reaction sintering II. ExD0rimenta,I Elemental titanium and aluminum powders of sizes smaller than 100 lain, with purities of 99.8% and 99.9%, respectively, were mixed in air to the desired composition Ti-48at.%AI. The powder mixture was pressed at room temperature to a green compact. This compact was cold extruded using two different extrusion ratios of 17 and 350. The extruded specimens were reactively sintered in a vacuum (= 10 ~ N/m 2) furnace for 6 h at 1350oc. A heating rate of 20°C/min was used. The cooling rate was approximately 130°C/min for temperatures higher than 800°C and 33°C/min for temperatures between 800 and 300°C. As-extruded specimens as well as specimens sintered at 1350oc/6 h were examined using light microscopy. A quantative determination of the porosity in the specimens was carried out with the help o1 a computer-aided image analysis system.
1469 0 0 3 6 - 9 7 4 8 / 9 2 $ 5 . 0 0 + .00 Copyright ( c ) 1992 P e r g a m o n P r e s s
Ltd.
1470
POWDER M E T A L L U R G Y
Vol.
26, No.
III. Results and Discussion 1. Microstructures of as-extruded conditions Figure 1 shows the microstructures of both as-extruded conditions, with AI having a white and Ti a dark appearance. During extrusion, particles, especially the AI-particles, are elongated in the extrusion direction to fibres. With increasing extrusion ratio, both AI and Ti-fibres become thinner and longer. Some small pores, mostly situated between Ti-particles, can be detected in the specimen extruded with the low extrusion ratio of 17 (Fig. l a and Fig. lb). The porosity is smaller than 0.5%. The specimen with the high extrusion ratio of 350 exhibits almost no pores (Fig. l c and Fig. ld). 2. Microstructures after pressureless sintering After pressureless sintering at 1350oc/6 h, both specimens are porous, with a porosity of approximately 16% for the low extrusion ratio (Fig. 2a), and less than 2% for the high extrusion ratio (Fig. 2b). The maximum pore size was measured to be about 100 pm diameter for the low extrusion ratio (Fig. 2a) and about 30 p.m diameter for the high extrusion ratio (Fig. 2b). Under polarized light, a duplex structure with TiAI- and Ti3Al-lamellae and some TiAIregions is visible (Fig. 3). The lamellar volume fraction was measured to be about 80%. The size of the lamellar grains is about 70 ~m and 50 p.m diameter for the low and high extrusion ratios, respectively. It is worth mentioning here that the lamellar grains in both specimens are much smaller than those in alloys prepared by ingot metallurgy (see for example (3)). 3. Mechanisms of pore formation and effect of increasing extrusion ratio Depending on the temperature applied, sintering of extruded elemental Ti- and AI-powder mixture can be of the nature of pure solid state interdiffusion (at temperatures below the AI-melting point), or it can be regarded as a temporary liquid phase sintering (at temperatures above the AI-melting point). In the first case, Kirkendall-diffusion takes place. The AI atoms move into Ti-particles, whereas the Ti atoms stay almost immobile (4). As a result of such one-sided diffusion, the originally dense specimens of the as-extruded conditions (Fig. 4a) become porous, with pores being situated mostly at the interface region (Fig. 4b). In the second case, AI-melt flows away along particle or grain boundaries, leaving pores behind (Fig. 4c). The AI-melt then bedews the Ti-particles so that the contact surface between Ti and AI is enlarged and the reaction between Ti and AI is accelerated. The products of the reaction are the intermetallics such as TiAI3, TiAI2 (as two intermediate products), TiAI and Ti3AI (1), which are indicated by the hatched areas in Fig. 4. After all elemental AI has been consumed, a further solid state interdiffusion between Ti and the intermetallics formed can take place. The final microstructure consists of the intermetallics TiAI and Ti3AI, and pores (Fig. 4d). The situations shown in Fig 4 can also appear successively during a sintering process. This is the case in our pressureless sintering at 13500C/6 h. The Kirkendall-diffusion (Fig. 4b) took place during heating prior to the AImelting point. Because of the relatively large heating rate of 20°C/min, the part of the porosity due to the Kirkendall-diffusion must be considered to be not essential in this work. We can thus assume that the pore sizes immediately after the flow-away of AI-melt are approximately equal to the sizes of Al-fibres of the as-extruded conditions, i. e. the specimen with the high extrusion ratio of 350 has much smaller pores at this stage than those in the specimen with the low extrusion ratio of 17. The pore sizes will be reduced or eliminated at later stages of the sintering due to the well-known capillarity. Under the action of the capillary forces, vacancy emigration (Nabarro-Herring-creep) and dislocation climb (Peach-KOhler-creep) can take place. The capillary force increases with decreasing pore size (5). Due to larger capillary forces, the pore size as well as the porosity in the specimen with the high extrusion ratio of 350 have been more reduced than those in the specimen with the low extrusion ratio of 17.
9
Vol.
26, No.
9
POWDER METALLURGY
1471
The reduction of pore size and porosity by increasing extrusion ratio can also be understood from an energetic viewpoint. Considering extrusion ratio as the only variable, then more volume energy is stored in as-extruded specimens with a higher extrusion ratio (cold working), so that more surface energy has to be removed during sintering (6). This results in a smaller pore size and porosity. Shibue (7) showed that the porosity can be reduced by replacing AI-powder with AI-7 wt.%Mn alloy powder. This fact can be also explained as result of reducing sizes of AI-fibres. Indeed, for the same extrusion ratio of 15, the as-extruded Ti- and AI-7 wt.%Mn powder mixture exhibits a finer microstructure than the as-extruded Ti- and AIpowder mixture. IV. summarv Generally, pore formation takes place during pressureless sintering of extruded elemental Ti- and AI-powder mixture because of the Kirkendall-diffusion and / or the flow-away of the AI-melt. On the other hand, the pores formed can be reduced or eliminated by Nabarro-Herring-creep and Peach-K6hler-creep due to the capillarity. With a higher extrusion ratio, a finer microstructure can be achieved for as-extruded specimens. This has proved to be an effective method for reducing not only the pore size but the porosity after sintering as well. Assuming the pores formed after reaching the AI-melting point are of the sizes of the AI-fibres in the as-extruded conditions, a larger average capillary torce and a larger reduction of pore size and porosity can be expected for specimens extruded with a higher extrusion ratio. This has been confirmed experimentally in this study. In addition, a refinement of grains, particularly of lamellar grains, has been achieved by increasing extrusion ratio Acknowledgments: The authors would like to thank Ms. W.-V. Schmitz, Mr. G. T6bke, Ms. B. Wildhagen and Mr. K. Wulf for technical assistance. V. References (1) G.-X. Wang and M. Dahms: "An Overview: TiAI-Based Alloys Prepared by Elemental Powder Metallurgy", to be published in Powder Metallurgy International (2) G.-X. Wang and M. Dahms: "Influence of Heat Treatment on Microstructure of Ti-35 wt.%AI Prepared by Elemental Powder Metallurgy", Scripta Met. et Mat., Vol. 26, No. 5, March 1992,pp. ? (3) Y.-W. Kim and D. M. Dimiduk: "Progress in the Understanding of Gamma Titanium Aluminides", JOM, Aug 1991, pp. 40-47 (4) F. J. J. van Loo and G. D. Rieck: "Diffusion in the Titanium-Aluminium System - I. Interdiffusion between Solid AI and Ti or Ti-AI Alloys", Acta Met. 21, 1973, pp. 61-71 (5) W. Schatt: "Pulvermetallurgie Sinter- und Verbundwerkstoffe", VEB Deutscher Verlag fur Grundstoffindustrie Leipzig, 1984 (6) G. Leitner, M. Dahms, W. PoeBnecker and B. Wildhagen: "Reactiive Sintering of Titanium Aluminides", to be published in proceedings of "PM Aero "91" (7) K. Shibue: "Suppression of Pores for TiAI Intermetallic Compound Prepared by Reactive Sintering", Sumitomo Light Metal Technical Reports, Vol. 32 (No. 2), April 1991, pp. 95-101
1472
POWDER M E T A L L U R G Y
(a)
(c)
,~) ~m
Vol.
26, No. 9
(b)
(d)
Fig. 1. Bright field LM-photographs of the both as-extruded conditions ((a, b) with extrusion ratio = 17, and (c, d) with extrusion ratio = 350) in directions perpendicular (a, c) and parallel to the extrusion direction (b, d)
Vol.
26, No.
9
POWDER M E T A L L U R G Y
(a)
501J'm
1473
(b)
Fig. 2. Microstructures in bright field after the pressureless sintering at 1350oc/6 h: (a) extrusion ratio = 17 and (b) extrusion ratio = 350; taken from specimen sections perpendicular to the extrusion direction
Fig. 3. Microstructures under polarized light after the pressureless sintedng at 1350°(3/6 h: (a) extrusion ratio = 17 and (b) extrusion ratio ,, 350; taken from specimen sections perpendicular to the extrusion direction t
1474
POWDER M E T A L L U R G Y
(a)
Vol.
I-- Ti-parfic[es
(b)
1 3, "?m1,' ~.
26, No.
--At-partictes
~:~ .,, intermetallics C~ ~) --pores
(c)
(d)
Fig. 4. Schematic description of microstructural changes during reaction sintering of extruded "ri- and AI-powder mixtures: (a) as-extruded condition; (b) after certain Kirkendall-diffusion; (c) after flow-away of the AI-melt; (d) after complete reaction sintering
9