MIATERIALS GENE ELSEVIE Materials Science and Engineering A 410-411(2005)140-147 www.elsevier.com/locate/msea Structure-dependent grain boundary deformation and fracture at high temperatures Tadao Watanabe a, * Sadahiro Tsurekawaa, Shigeaki Kobayashi, Shin-ichi Yamaura a Laboratory of Materials Design and Interface Engineering, Department of nanomechanics, b Department of Mechanical Engineering, Ashikaga Institute e Institute for Materials Research(IMR), Tohoku University, Sendai, Japa Received in revised form 14 April 2005 This paper gives an overview of structure-dependent intergranular deformation and fracture of metallic bicrystals and polycrystals at high temperatures. Structure-dependent grain boundary sliding and migration during high temperature deformation are discussed in connection with grain boundary structural change by the interaction with lattice dislocations and by intrinsic and extrinsic grain boundary structural transformation due to temperature and segregation, respectively. Grain boundary engineering for improvement in creep strength, development of superplasticity and control of oxidation-assisted intergranular brittleness are introduced which were successfully achieved by engineering the grain boundary microstructures characterized by the grain boundary character distribution( GBCD), the grain boundary connectivity and grain size C2005 Published by Elsevier B.V. Keywords: Grain boundary sliding and migration; Structural transformation; Grain boundary microstructure; Grain boundary character distribution; Grain boundary connectivity; Grain boundary engineering 1. Introduction very early [1]. However, it took long to be generally recog nized until the advent of the transmission electron microscopy Grain boundaries are two-dimensional lattice defects and an (TEM) and more recently high-resolution transmission electron important origin of microstructural heterogeneity in polycrys- microscopy(HREM), which enabled us to directly observe the tals. They play essential roles in deformation, fracture and other interaction of lattice dislocations with grain boundaries and grain metallurgical phenomena occurring in polycrystalline materials boundary structures on atomic scale. In 1972, Hirth discussed at high temperatures. Grain boundary deformation and fracture the influence of grain boundaries on mechanical properties of control bulk mechanical properties of polycrystalline materials. polycrystals in connection with interaction of lattice disloc- Particularly grain boundary sliding and migration play important tions with grain boundaries [2]. However, it is only recently les in plastic deformation and fracture at high temperatures. that effects of grain boundary character and structure on grain Grain boundaries have a large variety of atomistic structures boundary-related bulk properties of polycrystalline materials depending on the orientation relationship between adjoining started to be seriously taken into consideration. More recently grains. Because of their extra energies, grain boundaries can it has been revealed that the effects of grain boundaries on be preferential sites for metallurgical phenomena controlling bulk properties of a polycrystal strongly depend on the char- mechanical, physical, chemical and electromagnetic properties acter distribution and geometrical configuration of grain bound of polycrystals. During plastic deformation in a polycrystal lat- aries interconnecting to each other, so called"grain boundary tice dislocations moving in the grain interior must inevitably microstructure", in a polycrystal system [3]. In the last three interact with grain boundaries. Effect of grain boundary struc- decades, our knowledge of grain boundary structure and prop- ture on grain boundary deformation was pointed out by Chalme erties has been so extensively accumulated as for us to be able to predict and control bulk properties of various kinds of poly nding author.lel:+81223863944;fax:+81223863944 crystalline materials, not only metallic but intermetallic, semi- E-mail address: tywata@fk9so-net ne. jp(T. Watanabe) conductor and even ceramic materials [3-51 0921-5093/S-see front matter c 2005 Published by Elsevier B.V. doi:10.1016msea2005.08.083
Materials Science and Engineering A 410–411 (2005) 140–147 Structure-dependent grain boundary deformation and fracture at high temperatures Tadao Watanabe a,∗, Sadahiro Tsurekawa a, Shigeaki Kobayashi b, Shin-ichi Yamaura c a Laboratory of Materials Design and Interface Engineering, Department of Nanomechanics, Graduate School of Engineering, Tohoku University, Sendai, Japan b Department of Mechanical Engineering, Ashikaga Institute of Technology, Ashikaga, Tochigi, Japan c Institute for Materials Research (IMR), Tohoku University, Sendai, Japan Received in revised form 14 April 2005 Abstract This paper gives an overview of structure-dependent intergranular deformation and fracture of metallic bicrystals and polycrystals at high temperatures. Structure-dependent grain boundary sliding and migration during high temperature deformation are discussed in connection with grain boundary structural change by the interaction with lattice dislocations and by intrinsic and extrinsic grain boundary structural transformation due to temperature and segregation, respectively. “Grain boundary engineering” for improvement in creep strength, development of superplasticity and control of oxidation-assisted intergranular brittleness are introduced which were successfully achieved by engineering the grain boundary microstructures characterized by the grain boundary character distribution (GBCD), the grain boundary connectivity and grain size. © 2005 Published by Elsevier B.V. Keywords: Grain boundary sliding and migration; Structural transformation; Grain boundary microstructure; Grain boundary character distribution; Grain boundary connectivity; Grain boundary engineering 1. Introduction Grain boundaries are two-dimensional lattice defects and an important origin of microstructural heterogeneity in polycrystals. They play essential roles in deformation, fracture and other metallurgical phenomena occurring in polycrystalline materials at high temperatures. Grain boundary deformation and fracture control bulk mechanical properties of polycrystalline materials. Particularly grain boundary sliding and migration play important roles in plastic deformation and fracture at high temperatures. Grain boundaries have a large variety of atomistic structures depending on the orientation relationship between adjoining grains. Because of their extra energies, grain boundaries can be preferential sites for metallurgical phenomena controlling mechanical, physical, chemical and electromagnetic properties of polycrystals. During plastic deformation in a polycrystal lattice dislocations moving in the grain interior must inevitably interact with grain boundaries. Effect of grain boundary structure on grain boundary deformation was pointed out by Chalmers ∗ Corresponding author. Tel.: +81 22 386 3944; fax: +81 22 386 3944. E-mail address: tywata@fk9.so-net.ne.jp (T. Watanabe). very early [1]. However, it took long to be generally recognized until the advent of the transmission electron microscopy (TEM) and more recently high-resolution transmission electron microscopy (HREM), which enabled us to directly observe the interaction of lattice dislocations with grain boundaries and grain boundary structures on atomic scale. In 1972, Hirth discussed the influence of grain boundaries on mechanical properties of polycrystals in connection with interaction of lattice dislocations with grain boundaries [2]. However, it is only recently that effects of grain boundary character and structure on grain boundary-related bulk properties of polycrystalline materials started to be seriously taken into consideration. More recently, it has been revealed that the effects of grain boundaries on bulk properties of a polycrystal strongly depend on the character distribution and geometrical configuration of grain boundaries interconnecting to each other, so called “grain boundary microstructure”, in a polycrystal system [3]. In the last three decades, our knowledge of grain boundary structure and properties has been so extensively accumulated as for us to be able to predict and control bulk properties of various kinds of polycrystalline materials, not only metallic but intermetallic, semiconductor and even ceramic materials [3–5]. 0921-5093/$ – see front matter © 2005 Published by Elsevier B.V. doi:10.1016/j.msea.2005.08.083
T Watanabe et al. Materials Science and Engineering A 410-411(2005)140-147 Since the time of 1970s up to now, basic studies of the inter- Ti action of lattice dislocations with grain boundaries were exten- sively performed [6-8]. Now we can more reasonably under stand mechanisms of grain boundary deformation and fracture than before. it is known that the interaction of lattice dislocations with grain boundaries strongly depend on the grain boundary haracter and structure; the roles of grain boundaries as dislo- cation source or sink, and barrier or path to dislocation motion can be affected by grain boundary character and structure [8] The absorption of"lattice dislocations" from the grain interior (not"grain boundary structural dislocations") by grain bound- ary is diffusion-controlled process because the Burgers vector of lattice dislocations normally is not always in the boundary plane and changes across the grain boundary. Generally speak 573K ing, lattice dislocations cannot easily cross the grain boundary so that they have to pile-up against the boundary. It is widely accepted that the generation of dislocation pile-up is the major source of grain boundary strengthening [9]. However, the effect of the grain size(i.e. the density of grain boundaries)on grain boundary strengthening has been explained by assuming onl that the magnitude of grain boundary strengthening due to pile Tilt Angle, e/deg up of lattice dislocations must depend on the number of piled-up Fig. 1. Misorientation dependence of the total amount of sliding during 300 min dislocations or the grain size. However, in the authors'opinion, at different temperatures in(10 10) tilt zinc bicrystals[12] the pile-up strength or the effectiveness as dislocation barrier of grain boundary is predicted to be infuenced by the grain bound ary character and structure, but unfortunately this was not taken grain boundary structural effect was explained by the difference The in the adsorption potential of lattice dislocations between high present authors have recently found that the Hall-Petch slope energy random boundary and low-energy coincidence boundary, defined by the grain boundary character distribution(GBCD) of coincidence boundaries with different 2 values on the dis. in polycrystalline molybdenum [10] sociation of lattice dislocation [18]. In general, the dissociation of lattice dislocation at coincidence boundaries becomes easier 2. Structure-dependent grain boundary deformation with increasing the 2 value and the easiest at high-energy ran dom boundaries(conventionally with E value larger than 29) The effect of grain boundary character and structure on grain In other words, random boundaries can more easily absorb lat- boundary strengthening was studied previously by tensile defor- tice dislocations and produce sliding than low-E coincidence mation of metal bicrystals at low temperature [11]and high tem- undaries and low angle boundaries peratures [ 12]. Particularly at high temperatures, grain boundary sliding takes place as one of important mechanisms of plastic 3. Grain boundary structural change during high deformation in bicrystals and polycrystals [13, 14]. It has been temperature deformation well established that grain boundary sliding behaviour strongly depends on the grain boundary character and the misorientation Lattice dislocations moving in the grain interior must ngle. High angle random boundary can slide more easily than inevitably interact with grain boundaries during plastic defor- low-E coincidence boundaries, as observed for 23 and 213 mation of a polycrystal. Simply speaking, there are two possible coincidence boundaries, which showed significant slide harden- ways of the interaction of lattice dislocations with grain bound ing in aluminium [15] ary: one is that a new lattice dislocation can be emitted from the Fig. I shows the misorientation dependence of grain bound- grain boundary into the other grain, transferring the strain field ry sliding on(1010)tilt zinc bicrystals crept at different tem- associated with lattice dislocation from one side to the other of peratures [16]. It is evident that the amount of sliding increases grain boundary. The other case is the movement of lattice dis- with increasing the misorientation angle up to 40, with a location along the grain boundary to produce grain boundary deep cusp around 55 corresponding to the 29/56.6 near- sliding[ 19]. It is reasonable to consider that the absorption of coincidence boundary demonstrating the difficulty of sliding lattice dislocations can change the initial grain boundary atom at all the test temperatures. a close relationship between grain istic structure [6-8. In fact, it was confirmed by Kokawa et al boundary sliding and crystal slip was found and discussed in [15] that the grain boundary structure can change by the absorp- details [15, 17]. Thus, we know that grain boundary sliding is tion of lattice dislocations during plastic deformation at high more difficult at coincidence boundaries and shows more signifi- temperatures, as shown by Fig. 2. Some coincidence bound- cant slide hardening during creep deformation [3. The observed aries can change to random boundaries by the absorption of
T. Watanabe et al. / Materials Science and Engineering A 410–411 (2005) 140–147 141 Since the time of 1970s up to now, basic studies of the interaction of lattice dislocations with grain boundaries were extensively performed [6–8]. Now we can more reasonably understand mechanisms of grain boundary deformation and fracture than before. It is known that the interaction of lattice dislocations with grain boundaries strongly depend on the grain boundary character and structure; the roles of grain boundaries as dislocation source or sink, and barrier or path to dislocation motion can be affected by grain boundary character and structure [8]. The absorption of “lattice dislocations” from the grain interior (not “grain boundary structural dislocations”) by grain boundary is diffusion-controlled process because the Burgers vector of lattice dislocations normally is not always in the boundary plane and changes across the grain boundary. Generally speaking, lattice dislocations cannot easily cross the grain boundary so that they have to pile-up against the boundary. It is widely accepted that the generation of dislocation pile-up is the major source of grain boundary strengthening [9]. However, the effect of the grain size (i.e. the density of grain boundaries) on grain boundary strengthening has been explained by assuming only that the magnitude of grain boundary strengthening due to pileup of lattice dislocations must depend on the number of piled-up dislocations or the grain size. However, in the authors’ opinion, the pile-up strength or the effectiveness as dislocation barrier of grain boundary is predicted to be influenced by the grain boundary character and structure, but unfortunately this was not taken serious consideration into grain boundary strengthening. The present authors have recently found that the Hall–Petch slope can change depending on the grain boundary microstructure defined by the grain boundary character distribution (GBCD) in polycrystalline molybdenum [10]. 2. Structure-dependent grain boundary deformation The effect of grain boundary character and structure on grain boundary strengthening was studied previously by tensile deformation of metal bicrystals at low temperature [11] and high temperatures[12]. Particularly at high temperatures, grain boundary sliding takes place as one of important mechanisms of plastic deformation in bicrystals and polycrystals [13,14]. It has been well established that grain boundary sliding behaviour strongly depends on the grain boundary character and the misorientation angle. High angle random boundary can slide more easily than low-Σ coincidence boundaries, as observed for Σ3 and Σ13 coincidence boundaries, which showed significant slide hardening in aluminium [15]. Fig. 1 shows the misorientation dependence of grain boundary sliding on 1010 tilt zinc bicrystals crept at different temperatures [16]. It is evident that the amount of sliding increases with increasing the misorientation angle up to 40◦, with a deep cusp around 55◦ corresponding to the Σ9◦/56.6◦ nearcoincidence boundary demonstrating the difficulty of sliding at all the test temperatures. A close relationship between grain boundary sliding and crystal slip was found and discussed in details [15,17]. Thus, we know that grain boundary sliding is more difficult at coincidence boundaries and shows more signifi- cant slide hardening during creep deformation [3]. The observed Fig. 1. Misorientation dependence of the total amount of sliding during 300 min at different temperatures in 1010 tilt zinc bicrystals [12]. grain boundary structural effect was explained by the difference in the adsorption potential of lattice dislocations between highenergy random boundary and low-energy coincidence boundary, as expected from the effect of the order of atomistic structure of coincidence boundaries with different Σ values on the dissociation of lattice dislocation [18]. In general, the dissociation of lattice dislocation at coincidence boundaries becomes easier with increasing the Σ value and the easiest at high-energy random boundaries (conventionally with Σ value larger than 29). In other words, random boundaries can more easily absorb lattice dislocations and produce sliding than low-Σ coincidence boundaries and low angle boundaries. 3. Grain boundary structural change during high temperature deformation Lattice dislocations moving in the grain interior must inevitably interact with grain boundaries during plastic deformation of a polycrystal. Simply speaking, there are two possible ways of the interaction of lattice dislocations with grain boundary: one is that a new lattice dislocation can be emitted from the grain boundary into the other grain, transferring the strain field associated with lattice dislocation from one side to the other of grain boundary. The other case is the movement of lattice dislocation along the grain boundary to produce grain boundary sliding [19]. It is reasonable to consider that the absorption of lattice dislocations can change the initial grain boundary atomistic structure [6–8]. In fact, it was confirmed by Kokawa et al. [15] that the grain boundary structure can change by the absorption of lattice dislocations during plastic deformation at high temperatures, as shown by Fig. 2. Some coincidence boundaries can change to random boundaries by the absorption of
T. Watanabe et al Materials Science and Engineering A 410-411(2005)140-14 4.1. Intrinsic effect of grain boundary structural O=1MPa transformation EGBDs observed EGBDs no Let us discuss the effect of grain boundary structural transfor- ∑23 mation originating from the entropy change of grain boundary CSL atomic configuration which may occur at a certain critical tem by Hart [20, 21]. In considerable experimental evidence which shows that the grain boundary structural transformation induced by temperature can affect significant influence on the temperature dependence of grain boundary properties such as the grain boundary energy ∑2513 .21, migration 23, 24, sliding [ 16, 25, 26] and segregation [I Therefore, it is of our particular interest to reveal whether and how the grain boundary structural transformation can affect the interaction of lattice dislocations with grain boundary. In the authors opinion, this is one of important issues associated Fig. 2. Increasing propensity to grain boundary sliding by modification of grain with intergranular deformation and fracture at high tempera- boundary structure due to the absorption of lattice dislocations in aluminium tures, remaining unsolved yet. Thus, the effect of grain boundary structural transformation on the absorption (or emission) of lattice dislocations by grain boundary and the passage of lat- lattice dislocations, or vise versa, depending on the Burgers tice dislocations across grain boundary is the most fundamental vector of lattice dislocations incorporated by the grain bound- issue to reveal mechanisms of structure-dependent intergranular ary. Asa result of such structural change due to the incorporation deformation and fracture in polycrystals so that this should be of lattice dislocations, sliding and migration characteristics of seriously considered deforming grain boundaries can be gradually modified during Shvindlerman and Straumal systematically studied the effect high temperature deformation. This is one of the origins of slide of grain boundary structural transformation on the absorption(or hardening and also slide softening. However, it is not easy to pre- disappearance)of lattice dislocation at grain boundaries,by the dict how and to what extent individual grain boundaries change transmission electron microscopy (TEM)[28]. They found that their character and structure of the initial grain boundaries can the absorption of lattice dislocations at grain boundaries could change unless operative slip systems are already known. It can occur at a certain critical temperature Te depending on the grain for dislocation adsorption or generation depends on the bound- critical temperature Te decreases almost linearly with increas- ary character and structure, the degree of the heterogeneity or ing the value of E for coincidence boundary. This suggests homogeneity of plastic de formation is very likely controlled by that the thermal stablity of grain boundary structure decreases the grain boundary character distribution(GBCD)and the grain with increasing the value of y for coincidence boundary,or boundary connectivity in a polycrystal. It is suggested that the with decreasing the degree of atomistic order of grain bound ligher degree of the macroscopic homogeneity of plastic defor- other words, high energy random boundary tends to transform mation. The importance of structural change of grain boundaries Irom a low-temperature to high-temperature structure at lower during deformation will be briefly discussed later in connection critical temperature. The effect of the initial grain boundary with grain boundary engineering for superplasticity structure on the structure transformation was also revealed from a systematic experimental work on grain boundary sliding in (1010) tilt zinc bicrystals by Watanabe et al. [16]. As shown 4. Effects of structural transformation on intergranular in Fig. 3(a), the temperature dependence of the average slid- deformation ing rate abruptl y changes at a critical temperature Tc, which depends on the boundary tilt angle. Of particular importance Since diffusion-controlled grain boundary phenomena is that the critical temperature Te is very high(0.9Tm)for depend on the grain boundary character and structure more the 29 coincidence boundary with the tilt angle of 54. and rongly at high temperature than low temperature, the character- the low angle boundary with tilt angle of 16.5 which showed istics features of high temperature deformation and fracture are actually no change of the temperature dependence of the slid very likely influenced by any grain boundary structural change ing rate up to the melting temperature. Now it is concluded caused by intrinsic or extrinsic origin. There are several possi- that these low-energy boundaries have higher thermal stabil bilities that grain boundary atomistic structure transforms from ity than high energy random boundaries, as clearly seen from one state to another, due to intrinsic or extrinsic origins. So let us the misorientation dependence of the critical temperature look at possible effects of structural change and transformation Fig. 3(b). Therefore, it is likely that the thermal stability of on grain boundary deformation and fracture at high tempera- the grain boundary microstructure of a polycrystal may be con- ures trolled by the fraction of low-energy boundaries with higher
142 T. Watanabe et al. / Materials Science and Engineering A 410–411 (2005) 140–147 Fig. 2. Increasing propensity to grain boundary sliding by modification of grain boundary structure due to the absorption of lattice dislocations in aluminium [17]. lattice dislocations, or vise versa, depending on the Burgers’ vector of lattice dislocations incorporated by the grain boundary. As a result of such structural change due to the incorporation of lattice dislocations, sliding and migration characteristics of deforming grain boundaries can be gradually modified during high temperature deformation. This is one of the origins of slide hardening and also slide softening. However, it is not easy to predict how and to what extent individual grain boundaries change their character and structure of the initial grain boundaries can change unless operative slip systems are already known. It can be said that since the effectiveness of grain boundary as site for dislocation adsorption or generation depends on the boundary character and structure, the degree of the heterogeneity or homogeneity of plastic deformation is very likely controlled by the grain boundary character distribution (GBCD) and the grain boundary connectivity in a polycrystal. It is suggested that the presence of a high fraction of random boundaries may produce higher degree of the macroscopic homogeneity of plastic deformation. The importance of structural change of grain boundaries during deformation will be briefly discussed later in connection with grain boundary engineering for superplasticity. 4. Effects of structural transformation on intergranular deformation Since diffusion-controlled grain boundary phenomena depend on the grain boundary character and structure more strongly at high temperature than low temperature, the characteristics features of high temperature deformation and fracture are very likely influenced by any grain boundary structural change caused by intrinsic or extrinsic origin. There are several possibilities that grain boundary atomistic structure transforms from one state to another, due to intrinsic or extrinsic origins. So let us look at possible effects of structural change and transformation on grain boundary deformation and fracture at high temperatures. 4.1. Intrinsic effect of grain boundary structural transformation Let us discuss the effect of grain boundary structural transformation originating from the entropy change of grain boundary atomic configuration which may occur at a certain critical temperature Tc as pointed out by Hart [20,21]. In fact, there is considerable experimental evidence which shows that the grain boundary structural transformation induced by temperature can affect significant influence on the temperature dependence of grain boundary properties such as the grain boundary energy [22], migration [23,24], sliding [16,25,26] and segregation [27]. Therefore, it is of our particular interest to reveal whether and how the grain boundary structural transformation can affect the interaction of lattice dislocations with grain boundary. In the authors’ opinion, this is one of important issues associated with intergranular deformation and fracture at high temperatures, remaining unsolved yet. Thus, the effect of grain boundary structural transformation on the absorption (or emission) of lattice dislocations by grain boundary and the passage of lattice dislocations across grain boundary is the most fundamental issue to reveal mechanisms of structure-dependent intergranular deformation and fracture in polycrystals so that this should be seriously considered. Shvindlerman and Straumal systematically studied the effect of grain boundary structural transformation on the absorption (or disappearance) of lattice dislocation at grain boundaries, by the transmission electron microscopy (TEM) [28]. They found that the absorption of lattice dislocations at grain boundaries could occur at a certain critical temperature Tc depending on the grain boundary structure or Σ value for coincidence boundary. The critical temperature Tc decreases almost linearly with increasing the value of Σ for coincidence boundary. This suggests that the thermal stability of grain boundary structure decreases with increasing the value of Σ for coincidence boundary, or with decreasing the degree of atomistic order of grain boundary structure closely related to the grain boundary energy. In other words, high energy random boundary tends to transform from a low-temperature to high-temperature structure at lower critical temperature. The effect of the initial grain boundary structure on the structure transformation was also revealed from a systematic experimental work on grain boundary sliding in 1010 tilt zinc bicrystals by Watanabe et al. [16]. As shown in Fig. 3(a), the temperature dependence of the average sliding rate abruptly changes at a critical temperature Tc, which depends on the boundary tilt angle. Of particular importance is that the critical temperature Tc is very high (∼0.9Tm) for the Σ9 coincidence boundary with the tilt angle of 54◦ and the low angle boundary with tilt angle of 16.5◦ which showed actually no change of the temperature dependence of the sliding rate up to the melting temperature. Now it is concluded that these low-energy boundaries have higher thermal stability than high energy random boundaries, as clearly seen from the misorientation dependence of the critical temperature in Fig. 3(b). Therefore, it is likely that the thermal stability of the grain boundary microstructure of a polycrystal may be controlled by the fraction of low-energy boundaries with higher
T Watanabe et al. Materials Science and Engineering A 410-411(2005)140-147 Temperature, T/K 50600 Tlt -40kJ mor t 1010>Tlt 1万 Cgb=196 kPa ∑9(56.6°) 0.9 .8 旨 06 Tilt Angle, A/deg Fig. 3. (a)Temperature dependence of the average sliding rate and the critical temperature Te for the change of the temperature dependence in(1010) tilt zinc bicrystals [12]. (b) Misorientation dependence of the critical temperature Te observed for grain boundary sliding in(10 10)tilt zinc bicrystals (12 thermal stability, i.e. the grain boundary character distribution fraction of low-energy boundaries with higher thermal stabil- (GBCD) ity, a sharp change of temperature dependence of sliding and other grain boundary-related phenomena may occur at a higher 4.2. Extrinsic effects of grain boundary structural critical temperature, as observed on bicrystal specimen contain- transformation ing a single low-energy boundary [ 16, 25, 26]. Probably this may occur in a polycrystal, which has a sharp texture and a higher In the case of grain boundary sliding of (10 1 0)tilt zinc fraction of low-energy boundaries (low-2 coincidence bound bicrystals mentioned in the preceding section, the critical tem- aries and low-angle boundaries)than a random polycrystal,as perature Tc associated with the grain boundary structural trans- theoretically predicted [30-321 formation was found to shift by 40 K to higher temperature side Since the interaction of lattice dislocations with grain bound by the presence of impurities for 40(10 1 0)tilt boundary. Ther- ary strongly depends on the grain boundary structure and can mal stability of grain boundary structure must be extrinsically directly affect plastic deformation of a polycrystal, we expect affected by grain boundary segregation of impurities. In fact, that any grain boundary structural change originating either high resolution TEMobservations on grain boundary structure in intrinsic or extrinsic grain boundary structural transformation Fe and Fe-0. 18 at.% Au alloy by Sickafus and Sass revealed that can amplify structure-dependent grain boundary deformation grain boundary segregation of Au could induce the grain bound- and fracture at high temperatures above 0.5Tm(m: the melt- ary structural transformation[29]. Moreover, it was found that ing temperature)where atomic diffusion becomes dominantly the grain boundary segregation can also affect the intrinsic grain This will bring about a new apparent complexity of roles of grain boundary structural transformation [27]. Unfortunately, until boundary sliding and migration during deformation, resulting in ecently serious consideration into possible effects of intrin- cavitation at grain boundary irregularities 33] and triple jund sic and extrinsic grain boundary structural transformation has tions [34,35]. Dynamic migration often produces grain boundary been only little taken in discussion of intergranular deformation irregularities, which can be preferential sites for intergranular and fracture at high temperatures [3, 24]. It is very likely that cavitation and cracking by sliding [3, 28] he temperature dependence of bulk mechanical properties of Another interesting grain boundary phenomenon during high olycrystalline material, which is a material system consisted of temperature deformation is dynamic grain boundary migration the network of a huge numbers of grain boundaries connected occurring during cyclic deformation, as shown for aluminium to each other in 3D space, is affected by the grain boundary polycrystal in Fig 4(a)[24]. As clearly seen, random bound haracter distribution(GBCD). When a polycrystal has a high aries (denoted by r) can migrate cyclically and extensively
T. Watanabe et al. / Materials Science and Engineering A 410–411 (2005) 140–147 143 Fig. 3. (a) Temperature dependence of the average sliding rate and the critical temperature Tc for the change of the temperature dependence in 1010 tilt zinc bicrystals [12]. (b) Misorientation dependence of the critical temperature Tc observed for grain boundary sliding in 1010 tilt zinc bicrystals [12]. thermal stability, i.e. the grain boundary character distribution (GBCD). 4.2. Extrinsic effects of grain boundary structural transformation In the case of grain boundary sliding of 1 0 1 0¯ tilt zinc bicrystals mentioned in the preceding section, the critical temperature Tc associated with the grain boundary structural transformation was found to shift by 40 K to higher temperature side by the presence of impurities for 40◦ 1010tilt boundary. Thermal stability of grain boundary structure must be extrinsically affected by grain boundary segregation of impurities. In fact, high resolution TEM observations on grain boundary structure in Fe and Fe–0.18 at.% Au alloy by Sickafus and Sass revealed that grain boundary segregation of Au could induce the grain boundary structural transformation [29]. Moreover, it was found that the grain boundary segregation can also affect the intrinsic grain boundary structural transformation [27]. Unfortunately, until recently serious consideration into possible effects of intrinsic and extrinsic grain boundary structural transformation has been only little taken in discussion of intergranular deformation and fracture at high temperatures [3,24]. It is very likely that the temperature dependence of bulk mechanical properties of polycrystalline material, which is a material system consisted of the network of a huge numbers of grain boundaries connected to each other in 3D space, is affected by the grain boundary character distribution (GBCD). When a polycrystal has a high fraction of low-energy boundaries with higher thermal stability, a sharp change of temperature dependence of sliding and other grain boundary-related phenomena may occur at a higher critical temperature, as observed on bicrystal specimen containing a single low-energy boundary [16,25,26]. Probably this may occur in a polycrystal, which has a sharp texture and a higher fraction of low-energy boundaries (low-Σ coincidence boundaries and low-angle boundaries) than a random polycrystal, as theoretically predicted [30–32]. Since the interaction of lattice dislocations with grain boundary strongly depends on the grain boundary structure and can directly affect plastic deformation of a polycrystal, we expect that any grain boundary structural change originating either intrinsic or extrinsic grain boundary structural transformation can amplify structure-dependent grain boundary deformation and fracture at high temperatures above 0.5Tm (Tm: the melting temperature) where atomic diffusion becomes dominantly. This will bring about a new apparent complexity of roles of grain boundary sliding and migration during deformation, resulting in cavitation at grain boundary irregularities [33] and triple junctions[34,35]. Dynamic migration often produces grain boundary irregularities, which can be preferential sites for intergranular cavitation and cracking by sliding [3,28]. Another interesting grain boundary phenomenon during high temperature deformation is dynamic grain boundary migration occurring during cyclic deformation, as shown for aluminium polycrystal in Fig. 4(a) [24]. As clearly seen, random boundaries (denoted by R) can migrate cyclically and extensively,
T. Watanabe et al Materials Science and Engineering A 410-411(2005)140-14 △E=+0.25% N=8 cycle 0 650 K Fig. 4.(a)Structure-dependent dynamic migration during cyclic deformation in aluminium polycrystal tested over five cycles at 623K with a frequency of 7.7x 10-3Hz Random and coincidence boundaries are denoted by r and E plus numeral, respectively. (b) The fraction of immobile grain boundaries as a function of test temperature for aluminium specimen tested to eight cycles with a frequency of 1.3 x 10-Hz and a strain amplitude of 0. 25 [24] while the Ell coincidence boundary cannot. Moreover, as seen have never been seriously taken in the design and control of bulk from Fig. 4(b), the fraction of immobile grain boundaries can mechanical properties of polycrystalline materials used at high rapidly decrease beyond T=550K(0.6Tm)resulting from temperatures a rapid increase in the fraction of random boundaries trans formed probably from low-energy boundaries by the absorption 5. Structure-dependent intergranular fracture of lattice dislocations assisted by the intrinsic grain boundary structural transformation. Thus, structure-dependent sliding and The effect of grain boundary structure on intergranular frac- migration can become more pronounced because of the intrinsic ture can be reasonably expected when intergranular fracture is grain boundary structural transformation and additional extrin- controlled by sliding, because grain boundary sliding and migra- sic effects of grain boundary structural change tion strongly depend on grain boundary structure, as already It is interesting to reveal a mechanism of dynamic grain shown. Indeed it was already confirmed by experimental works boundary migration under stress or during plastic deformation, that intergranular creep fracture is strongly affected by the grain aying particular attention to the effects of the grain boundary boundary character and structure [3, 33]. One of examples taken tructural change and structural transformation. Quite recently from early experimental observations of structure-dependent the dynamic migration of planer or curved grain boundary under intergranular creep fracture in iron-tin alloy 331, is shown shear stress was studied by Winning et al. [36] in(1 12)and in Fig. 5. It is evident that random boundaries(r)can eas- (1 11)symmetric aluminium bicrystals. It was found that a dis- ly slide and preferentially fracture, while the 27, 213, 221 tinct change of migration behaviour and mechanism from the coincidence boundaries and low angle boundaries(denoted by low-angle regime to the high angle one can occur around the mis- letter L) will not slide and fracture. Furthermore, intergran- orientation angle of 12. It is explained that the motion of grain ular crack will not propagate into and detoured around the boundary structural dislocations play important role in dynamic region in the deforming specimen where low-energy boundaries grain boundary migration. Again the dissociation of lattice dis- more densely occurred and locally formed a clusterObserved locations into grain boundary structural dislocations is likely one structure-dependent intergranular fracture has been explained by ofelementary processes of the absorption of lattice dislocations sliding-assisted mechanism or diffusional mechanism based on by the grain boundary [15, 18]. Low-energy boundaries, which the structure-dependent grain boundary sliding and diffusion and have higher thermal stability of grain boundary atomistic struc- the effectiveness of grain boundary as vacancy source and sink ture can keep their characteristics even at higher temperature. It 3]. It is well known that grain boundary diffusion also strongly is predicted that the presence of a higher fraction of low-energy depends on the grain boundary structure as well as other grain boundaries with higher thermal stability is one of important boundary phenomena 37; random boundaries are fast path for requirements for high temperature materials, which can keep atomic diffusion but low-energy boundaries are not. This leads their high performance during long service life. Unfortunately, to structure-dependent diffusion-controlled intergranular cavita- such consideration into possible effects of intrinsic and extrinsic tion and fracture(either or not involving grain boundary sliding) grain boundary structural transformations as mentioned above, during plastic deformation at high temperatures [3, 33, 38]
144 T. Watanabe et al. / Materials Science and Engineering A 410–411 (2005) 140–147 Fig. 4. (a) Structure-dependent dynamic migration during cyclic deformation in aluminium polycrystal tested over five cycles at 623 K with a frequency of 7.7 × 10−3 Hz. Random and coincidence boundaries are denoted by R and Σ plus numeral, respectively. (b) The fraction of immobile grain boundaries as a function of test temperature for aluminium specimen tested to eight cycles with a frequency of 1.3 × 10−2 Hz and a strain amplitude of 0.25 [24]. while the Σ11 coincidence boundary cannot. Moreover, as seen from Fig. 4(b), the fraction of immobile grain boundaries can rapidly decrease beyond T = 550 K (∼0.6Tm) resulting from a rapid increase in the fraction of random boundaries transformed probably from low-energy boundaries by the absorption of lattice dislocations assisted by the intrinsic grain boundary structural transformation. Thus, structure-dependent sliding and migration can become more pronounced because of the intrinsic grain boundary structural transformation and additional extrinsic effects of grain boundary structural change. It is interesting to reveal a mechanism of dynamic grain boundary migration under stress or during plastic deformation, paying particular attention to the effects of the grain boundary structural change and structural transformation. Quite recently the dynamic migration of planer or curved grain boundary under shear stress was studied by Winning et al. [36] in 112 and 111 symmetric aluminium bicrystals. It was found that a distinct change of migration behaviour and mechanism from the low-angle regime to the high angle one can occur around the misorientation angle of 12◦. It is explained that the motion of grain boundary structural dislocations play important role in dynamic grain boundary migration. Again the dissociation of lattice dislocations into grain boundary structural dislocations is likely one of elementary processes of the absorption of lattice dislocations by the grain boundary [15,18]. Low-energy boundaries, which have higher thermal stability of grain boundary atomistic structure can keep their characteristics even at higher temperature. It is predicted that the presence of a higher fraction of low-energy boundaries with higher thermal stability is one of important requirements for high temperature materials, which can keep their high performance during long service life. Unfortunately, such consideration into possible effects of intrinsic and extrinsic grain boundary structural transformations as mentioned above, have never been seriously taken in the design and control of bulk mechanical properties of polycrystalline materials used at high temperatures. 5. Structure-dependent intergranular fracture The effect of grain boundary structure on intergranular fracture can be reasonably expected when intergranular fracture is controlled by sliding, because grain boundary sliding and migration strongly depend on grain boundary structure, as already shown. Indeed it was already confirmed by experimental works that intergranular creep fracture is strongly affected by the grain boundary character and structure [3,33]. One of examples taken from early experimental observations of structure-dependent intergranular creep fracture in iron–tin alloy [33], is shown in Fig. 5. It is evident that random boundaries (R) can easily slide and preferentially fracture, while the Σ7, Σ13, Σ21 coincidence boundaries and low angle boundaries (denoted by letter L) will not slide and fracture. Furthermore, intergranular crack will not propagate into and detoured around the region in the deforming specimen where low-energy boundaries more densely occurred and locally formed a cluster. Observed structure-dependent intergranular fracture has been explained by sliding-assisted mechanism or diffusional mechanism based on the structure-dependent grain boundary sliding and diffusion and the effectiveness of grain boundary as vacancy source and sink [3]. It is well known that grain boundary diffusion also strongly depends on the grain boundary structure as well as other grain boundary phenomena [37]; random boundaries are fast path for atomic diffusion but low-energy boundaries are not. This leads to structure-dependent diffusion-controlled intergranular cavitation and fracture (either or not involving grain boundary sliding) during plastic deformation at high temperatures [3,33,38].
T Watanabe et al. Materials Science and Engineering A 410-411(2005)140-147 RS1-1 e Grain Size Fig. 5. Structure-dependent intergranular creep fracture in Fe-0.8 29. 4 MPa. Grain boundaries are denoted by R. Creep Oxidation Time, t/h angle boundary, 2 plus numeral for coincide boundaries are preferential sites for( and migration (b)T=1200K,G=10MPa RS1-2 I RSA1-2 6. Grain boundary engineering for high temperature ow freq of material So far we have discussed structure-dependent intergranular deformation and fracture on the basis of experimental works RSA3-2 performed by using bicrystalline or polycrystalline metals and Grain Size lloys in which individual grain boundaries were character ized by the modern analyzing techniques like the orientation High free of maging microscopy(OIM). Next, let us discuss the possibil- ity of grain boundary engineering of high temperature materials + Time of heating up to 1200K with desirable bulk mechanical properties and performance. We take three examples of grain boundary engineering which were recently successfully achieved: grain boundary engineering for Creep Oxidation Time, t/h (i) improvement in creep strength, (ii)development of super- Fig. 6. The creep curves obtained by creep test: (a) at 1073K for 24 h in air plasticity and (iii) control of oxidation embrittlement or rapidly solidified and annealed Ni40 at Fe alloy specimens with different frequencies of low-E coincidence boundaries and grain sizes 44] 6.1. Improvement in creep strength. From a basic study of high temperature creep deformation The specimens which had a high frequency of low-E coinci- in zinc bicrystals [12], it was found that the presence of a sin- dence boundaries showed higher creep strength, i.e. its cor gle low-E coincidence boundary can effectively increase creep responding creep curve came to lower creep strain level irre- trength of zinc bicrystal, probably because the effectiveness of spective of creep stress, as reported by other investigators for the grain boundary as barrier to dislocation motion is very high NI-Cr base alloy [40]. Until presently, the improvement in produced in comparison with the presence of a random bound- 134, 39] and Ni-base alloys [40, 43, 44]. Accordingly, this is a ary. The bicrystals with the twist angles 510 and 540 which are new approach to development ofhigh temperature material with within the range of misorientation relationship corresponding high creep strength and high thermal stability of grain bound to the 56.6(1010)/29 near-coincidence orientation showed ary microstructure, in comparison with the material produced higher creep strength than others. Moreover, grain boundary by conventional processing without any consideration into the iding is much more difficult at low-E coincidence boundary grain boundary engineering than random boundary. These findings suggest the possibility of improvement in creep strength of polycrystalline materials 6.2. Development of superplasticity by the introduction of a high fraction of low-energy boundaries such as low -> coincidence boundaries It is well known that A-Li alloys have a serious material Fig. 6(a and b) show the most recent study of grain boundary problem of poor workability due to intergranular brittleness [41 engineering for improvement of creep strength in Fe-40 at. %Fe although they have excellent physical and mechanical properties lloy by controlling the grain boundary microstructure [43, 44]. such as high specific strength suitable as aircraft structural mate-
T. Watanabe et al. / Materials Science and Engineering A 410–411 (2005) 140–147 145 Fig. 5. Structure-dependent intergranular creep fracture in Fe–0.8 at.% Sn alloy crept at 973 K and 29.4 MPa. Grain boundaries are denoted by R. for random boundary, L for low angle boundary, Σ plus numeral for coincidence boundaries. Note that random boundaries are preferential sites for creep fracture caused by sliding and migration. 6. Grain boundary engineering for high temperature materials So far we have discussed structure-dependent intergranular deformation and fracture on the basis of experimental works performed by using bicrystalline or polycrystalline metals and alloys in which individual grain boundaries were characterized by the modern analyzing techniques like the orientation imaging microscopy (OIM). Next, let us discuss the possibility of grain boundary engineering of high temperature materials with desirable bulk mechanical properties and performance. We take three examples of grain boundary engineering which were recently successfully achieved: grain boundary engineering for (i) improvement in creep strength, (ii) development of superplasticity and (iii) control of oxidation embrittlement. 6.1. Improvement in creep strength. From a basic study of high temperature creep deformation in zinc bicrystals [12], it was found that the presence of a single low-Σ coincidence boundary can effectively increase creep strength of zinc bicrystal, probably because the effectiveness of the grain boundary as barrier to dislocation motion is very high so that grain boundary strengthening can be more effectively produced in comparison with the presence of a random boundary. The bicrystals with the twist angles 51◦ and 54◦ which are within the range of misorientation relationship corresponding to the 56.6◦ 1010/Σ9 near-coincidence orientation showed higher creep strength than others. Moreover, grain boundary sliding is much more difficult at low-Σ coincidence boundary than random boundary. These findings suggest the possibility of improvement in creep strength of polycrystalline materials by the introduction of a high fraction of low-energy boundaries such as low-Σ coincidence boundaries. Fig. 6(a and b) show the most recent study of grain boundary engineering for improvement of creep strength in Fe–40 at.% Fe alloy by controlling the grain boundary microstructure [43,44]. Fig. 6. The creep curves obtained by creep test: (a) at 1073 K for 24 h in air under stress of 14 MPa, and (b) at 1200 K for 18 h in air under stress of 10 MPa or rapidly solidified and annealed Ni–40 at.% Fe alloy specimens with different frequencies of low-Σ coincidence boundaries and grain sizes [44]. The specimens which had a high frequency of low-Σ coincidence boundaries showed higher creep strength, i.e. its corresponding creep curve came to lower creep strain level irrespective of creep stress, as reported by other investigators for Ni–Cr base alloy [40]. Until presently, the improvement in creep strength has been successfully achieved for Al–Li alloy [34,39] and Ni-base alloys [40,43,44]. Accordingly, this is a new approach to development of high temperature material with high creep strength and high thermal stability of grain boundary microstructure, in comparison with the material produced by conventional processing without any consideration into the grain boundary engineering. 6.2. Development of superplasticity It is well known that Al–Li alloys have a serious material problem of poor workability due to intergranular brittleness[41] although they have excellent physical and mechanical properties such as high specific strength suitable as aircraft structural mate-
T. Watanabe et al. Materials Science and Engineering A 410-411(2005)140-147 rial. In order to solve this problem Kobayashi et al. have recently random boundary can still cause severe brittleness due to deep performed a systematic experimental work on development of oxidation penetration. Oxidation-assisted intergranular brittle superplasticity by grain boundary engineering for Al-Li alloys ness can occur as long as the grain size is very large. The grain [34]. For this purpose, the change of grain boundary microstruc- boundary connectivity was found to play an important role in ture during high temperature deformation was carefully studied intergranular oxidation-induced fracture, which is controlled by by interruption of tensile test at different stages in the course the percolation process of crack propagation of superplastic deformation. The optimal deformation condition and grain boundary microstructure for excellent superplastic ity were deduced from this systematic investigation of the grain 7. Sun boundary microstructures produced by high temperature defor mation at different deformation conditions. The changes of the Structure-dependent grain boundary sliding, migration and grain orientation distribution( texture)and of the grain boundary fracture have been discussed in connection with the interaction character distribution(GBCD)during superplastic deformation of lattice dislocations with grain boundaries and with intrinsic and extrinsic effects of grain boundary structural transforma were studied by OIM technique. It was found that the initially tion. It is suggested that the heterogeneity and the complexity sharp (1 10) texture remained up to the peak stress on the of intergranular deformation and fracture in polycrystals can be distribution with increasing strain beyond 0.8 where the steady- tural transformation. Recent achievements of grain boundary state deformation occurred at constant stress. Along with such change of the grain orientation distribution, the frequency of engineering for high temperature materials have been introduced random boundaries remained almost unchanged(37-41%)up to for improvement in creep strength, development of superplas- the peak stress and then gradually increased to higher than 70% ticity and control of intergranular oxidation-induced brittleness being almost twice higher than the initial value. This means that by controlling the grain boundary microstructure, i.e. the grain high temperature deformation at optimal test condition introduce connectivity and the a higher frequency of random boundaries in order to produce a arger plastic strain by sliding. However, random boundaries can also be preferential sites for cavitation and fracture so that the Acknowledgements optimal frequency of random boundaries is required to simulta- neously generate the optimal grain boundary connectivity, which This work was supported by the Grant-in Aid from the Min- does not allow the propagation of intergranular cracks. Further istry of Education, Culture, Sport, Art and Science( Grant No development of grain boundary engineering for superplasticity (B)(2)10555226: Grain Boundary and Interfacial Architecture with extended elongation(300%)was successfully achieved for High Temperature Advanced Materials") by the introduction of new low angle boundaries by strain rate increment during superplastic deformation [44] References 6.3. Control of intergranular oxidation brittleness [1]B. Chalmers, Metal Interfaces, ASM, 1952, pp. 299-311 [2]J. P. Hirth, Metall. Trans. 3(1972)3047-3067. Recently, we found that grain boundary oxidation prefer- [3]T. Watanabe, Mater. Sci. Eng. A166(1993)11-28 entially occurs at high energy random boundaries while low- [4]T. Watanabe, S. Tsurekawa, Acta Mater. 47(1999)4171-4185 E coincidence boundaries are very resistant to oxidation in 5T. Watanabe, Ann. Chim. Scl. Mater. 27(SuppL. D)(2002)5327-5344 Ni-40 at% Fe alloy which is widely used as base alloy of high [7]WAT. Clark,.A.Smith,JMater. Sci.14(1979)776-788 temperature materials [42,43]. Intergranular oxidation is well [8)D J Dingley, R.C. Pond, Acta Metall.27(1979)667-682 known to cause oxidation-induced brittleness in high temper- 9]TN. Baker (Ed), Yield, Flow and Fracture of Polycrystals, Appl. Sci. ature materials. There has been a strong demand for solving this problem hindering from development of a highly oxidation- [to)S. Tsurekawa, T. Watanabe, Mater Res. Soc. Symp. Proc. 586(2000) resistant temperature material. Yamaura et al. have attempted and succeeded in the grain boundary engineering for control of [12]T. Watanabe, M. Yamada, S. Karashima, Philos. Mag. 63(1991) oxidation-assisted brittleness by controlling the fraction of ran- 1013-1022 dom boundaries to reduce the extent of intergranular oxidation in [13] H. Conrad, in: J.E. Dom(Ed ), Mechanical Behavior of Materials at the Ni-Fe alloy specimens produced by rapid solidification and Elevated Temperatures, McGraw Hill, 1961, pp. 218-267. subsequent annealing [44]. As mentioned before, the specimen, [14] T.G. Langdon, Can. Met. Q. 13(1974)223-228 which had a high frequency of low-E coincidence boundaries [15]H. Kokawa, T. Watanabe, S. Karashima, Philos. Mag. A44(1981) 1239-1254. showed higher creep strength and also higher oxidation resis- [16] T. Watanabe, S. Kimura, S. Karashima, Philos Mag. A49(1984)845 nce. Random boundaries were preferentially oxidized and led [17]T. Watanabe, M. Yamada, S Shima,S Karashima, Philos Mag. A40 to fracture. Of particular importance is that the depth of oxI- [18]H. Kokawa, T. Watanabe, S. Karashima,J. Mater. Sci. 18(1983) (1979)667-68 dation penetration from the specimen surface depends on the 1183-119 length of grain boundary, roughly speaking the grain size so that [19]D. McLean, Philos Mag. 23(1971)467-472 that even if the fraction of random boundaries is very small, [20] E W. Hart, Ultrafine Grain Metals, Syracuse University, 1970, p. Il
146 T. Watanabe et al. / Materials Science and Engineering A 410–411 (2005) 140–147 rial. In order to solve this problem Kobayashi et al. have recently performed a systematic experimental work on development of superplasticity by grain boundary engineering for Al–Li alloys [34]. For this purpose, the change of grain boundary microstructure during high temperature deformation was carefully studied by interruption of tensile test at different stages in the course of superplastic deformation. The optimal deformation condition and grain boundary microstructure for excellent superplasticity were deduced from this systematic investigation of the grain boundary microstructures produced by high temperature deformation at different deformation conditions. The changes of the grain orientation distribution (texture) and of the grain boundary character distribution (GBCD) during superplastic deformation were studied by OIM technique. It was found that the initially sharp {110} texture remained up to the peak stress on the stress–strain curve, then changed to more random orientation distribution with increasing strain beyond 0.8 where the steadystate deformation occurred at constant stress. Along with such change of the grain orientation distribution, the frequency of random boundaries remained almost unchanged (37–41%) up to the peak stress and then gradually increased to higher than 70%, being almost twice higher than the initial value. This means that high temperature deformation at optimal test condition introduce a higher frequency of random boundaries in order to produce a larger plastic strain by sliding. However, random boundaries can also be preferential sites for cavitation and fracture so that the optimal frequency of random boundaries is required to simultaneously generate the optimal grain boundary connectivity, which does not allow the propagation of intergranular cracks. Further development of grain boundary engineering for superplasticity with extended elongation (∼300%) was successfully achieved by the introduction of new low angle boundaries by strain rate increment during superplastic deformation [44]. 6.3. Control of intergranular oxidation brittleness Recently, we found that grain boundary oxidation preferentially occurs at high energy random boundaries while low- Σ coincidence boundaries are very resistant to oxidation in Ni–40 at.% Fe alloy which is widely used as base alloy of high temperature materials [42,43]. Intergranular oxidation is well known to cause oxidation-induced brittleness in high temperature materials. There has been a strong demand for solving this problem hindering from development of a highly oxidationresistant temperature material. Yamaura et al. have attempted and succeeded in the grain boundary engineering for control of oxidation-assisted brittleness by controlling the fraction of random boundaries to reduce the extent of intergranular oxidation in the Ni–Fe alloy specimens produced by rapid solidification and subsequent annealing [44]. As mentioned before, the specimen, which had a high frequency of low-Σ coincidence boundaries showed higher creep strength and also higher oxidation resistance. Random boundaries were preferentially oxidized and led to fracture. Of particular importance is that the depth of oxidation penetration from the specimen surface depends on the length of grain boundary, roughly speaking the grain size so that that even if the fraction of random boundaries is very small, random boundary can still cause severe brittleness due to deep oxidation penetration. Oxidation-assisted intergranular brittleness can occur as long as the grain size is very large. The grain boundary connectivity was found to play an important role in intergranular oxidation-induced fracture, which is controlled by the percolation process of crack propagation. 7. Summary Structure-dependent grain boundary sliding, migration and fracture have been discussed in connection with the interaction of lattice dislocations with grain boundaries and with intrinsic and extrinsic effects of grain boundary structural transformation. It is suggested that the heterogeneity and the complexity of intergranular deformation and fracture in polycrystals can be enhanced by such structural change and grain boundary structural transformation. Recent achievements of grain boundary engineering for high temperature materials have been introduced for improvement in creep strength, development of superplasticity and control of intergranular oxidation-induced brittleness, by controlling the grain boundary microstructure, i.e. the grain boundary character distribution (GBCD), the grain boundary connectivity and the grain size. Acknowledgements This work was supported by the Grant-in Aid from the Ministry of Education, Culture, Sport, Art and Science (Grant No.: (B)(2) 10555226: “Grain Boundary and Interfacial Architecture for High Temperature Advanced Materials”). References [1] B. Chalmers, Metal Interfaces, ASM, 1952, pp. 299–311. [2] J.P. Hirth, Metall. Trans. 3 (1972) 3047–3067. [3] T. Watanabe, Mater. Sci. Eng. A166 (1993) 11–28. [4] T. Watanabe, S. Tsurekawa, Acta Mater. 47 (1999) 4171–4185. [5] T. Watanabe, Ann. Chim. Sci. Mater. 27 (Suppl. 1) (2002) S327–S344. [6] P.H. Pumphrey, J. Phys. C4 36 (1975) C4-23–C4-33. [7] W.A.T. Clark, D.A. Smith, J. Mater. Sci. 14 (1979) 776–788. [8] D.J. Dingley, R.C. Pond, Acta Metall. 27 (1979) 667–682. [9] T.N. Baker (Ed), Yield, Flow and Fracture of Polycrystals, Appl. Sci. Pub., 1983. [10] S. Tsurekawa, T. Watanabe, Mater. Res. Soc. Symp. Proc. 586 (2000) 237–242. [11] S. Miura, K.T. Aust, Acta Metall. 26 (1978) 93–101. [12] T. Watanabe, M. Yamada, S. Karashima, Philos. Mag. 63 (1991) 1013–1022. [13] H. Conrad, in: J.E. Dorn (Ed.), Mechanical Behavior of Materials at Elevated Temperatures, McGraw Hill, 1961, pp. 218–267. [14] T.G. Langdon, Can. Met. Q. 13 (1974) 223–228. [15] H. Kokawa, T. Watanabe, S. Karashima, Philos. Mag. A44 (1981) 1239–1254. [16] T. Watanabe, S. Kimura, S. Karashima, Philos. Mag. A49 (1984) 845. [17] T. Watanabe, M. Yamada, S. Shima, S. Karashima, Philos. Mag. A40 (1979) 667–683. [18] H. Kokawa, T. Watanabe, S. Karashima, J. Mater. Sci. 18 (1983) 1183–1194. [19] D. McLean, Philos. Mag. 23 (1971) 467–472. [20] E.W. Hart, Ultrafine Grain Metals, Syracuse University, 1970, p. 115
T. Watanabe er al. Materials Science and Engineering A 410-411(2005)140-142 (21EW. Hart, The Nature and Behavior of Grain Boundaries, Plenum Press, [36]M. Winning, G. Gottstein, L S. Shvindlerman, Acta Mater. 50(2002) 1972,p.1l5 (22H. Gleiter, Z. Metallkde 61(1970)279. 37]I. Kaur, w. Gust, Fundamentals of Grain and Interphase Boundary Dif- [23]D W. Demianczuk, K.T. Aust, Acta Metall. 23(1975)1149 fusion, Ziegler Press, Stuttgart, 1988, pp. 275-300 (24V. Raman, T. Watanabe, T.G. Langdon, Acta Metall. 37(1989)705-714. [38L. [25]M. Biscondi, C. Goux, Membr. Sci. Rev. Metall, 65(1968)167. 39G. Avramovic-Cingara, K.T. Aust, DD. Perovic, G. Palumbo, Proceed- [26] P Lagarde, M. Biscondi, Can Met. Q. 13(1974)245-251 ings of the International Symposium on Recent Development in Light (27M. Menyhard, B. Blum, C.J. McMahon Jr, Acta Metall. 37(1989) Metals, Toronto Ontario, 1994, pp. 281-29 49-557. 140V. Thaveeprungsriporn, G.S. Was, Metall, Mater. Trans. 28A(1997) [28]LS. Shvindlerman, B B. Straumal, Acta Metall, 33( 1985)1735 101-2112 (29K.. Sickafus, S.L. Sass, Acta Metall. 35(1987)69-79 [41]D. Dew-Hughes, E. Creed, W.s. Miller, Mater. Sci. Tech. 4(1988) 30 T. Watanabe, H Fujii, H. Oikawa, K.I. Arai, Acta Metall. 941-952 106-112. 31A. Garbacz, M. w. Grabski, Acta Metall. Mater. 41(1993)469-473 142S. Yamaura, Y. Igarashi, S. Tsurekawa, T. Watanabe, Acta Mater. 47 32 L. Zuo, T. Watanabe, C. Esling, Z. Metallkde 85(1994)554-558 (1999)1163-1174 33T. Watanabe, Met. Trans. 14A(1983)531-545 143S. Yamura, Y. Igarashi, S. Tsurekawa, T, Watanabe, in: A. Meike, et [34]S. Kobayashi, T. Yoshimura, S. Tsurekawa, T. Watanabe, Mater. Trans al.(Eds ) Properties of Complex Inorganic Solids, vol. 2, Kluwer Aca- 44(2003)14691479 demic/Plenum Pub., 2000, pp. 27-37 35s. Kobayashi, S. Tsurekawa, T. Watanabe, Acta Mater. 53(2005) [44]S. Yamaura, S. Tsurekawa, T. Watanabe, Mater. Trans. 44(2003) 1051-1057 1494150
T. Watanabe et al. / Materials Science and Engineering A 410–411 (2005) 140–147 147 [21] E.W. Hart, The Nature and Behavior of Grain Boundaries, Plenum Press, 1972, p. 115. [22] H. Gleiter, Z. Metallkde 61 (1970) 279. [23] D.W. Demianczuk, K.T. Aust, Acta Metall. 23 (1975) 1149. [24] V. Raman, T. Watanabe, T.G. Langdon, Acta Metall. 37 (1989) 705–714. [25] M. Biscondi, C. Goux, Membr. Sci. Rev. Metall. 65 (1968) 167. [26] P. Lagarde, M. Biscondi, Can. Met. Q. 13 (1974) 245–251. [27] M. Menyhard, B. Blum, C.J. McMahon Jr., Acta Metall. 37 (1989) 549–557. [28] L.S. Shvindlerman, B.B. Straumal, Acta Metall. 33 (1985) 1735. [29] K.E. Sickafus, S.L. Sass, Acta Metall. 35 (1987) 69–79. [30] T. Watanabe, H. Fujii, H. Oikawa, K.I. Arai, Acta Metall. 941–952. [31] A. Garbacz, M.W. Grabski, Acta Metall. Mater. 41 (1993) 469–473. [32] L. Zuo, T. Watanabe, C. Esling, Z. Metallkde 85 (1994) 554–558. [33] T. Watanabe, Met. Trans. 14A (1983) 531–545. [34] S. Kobayashi, T. Yoshimura, S. Tsurekawa, T. Watanabe, Mater. Trans. 44 (2003) 1469–1479. [35] S. Kobayashi, S. Tsurekawa, T. Watanabe, Acta Mater. 53 (2005) 1051–1057. [36] M. Winning, G. Gottstein, L.S. Shvindlerman, Acta Mater. 50 (2002) 353–363. [37] I. Kaur, W. Gust, Fundamentals of Grain and Interphase Boundary Diffusion, Ziegler Press, Stuttgart, 1988, pp. 275–300. [38] L.C. Lim, R. Raj, Acta Metall. 33 (1985) 1577–1583. [39] G. Avramovic-Cingara, K.T. Aust, D.D. Perovic, G. Palumbo, Proceedings of the International Symposium on Recent Development in Light Metals, Toronto Ontario, 1994, pp. 281–291. [40] V. Thaveeprungsriporn, G.S. Was, Metall. Mater. Trans. 28A (1997) 2101–2112. [41] D. Dew-Hughes, E. Creed, W.S. Miller, Mater. Sci. Tech. 4 (1988) 106–112. [42] S. Yamaura, Y. Igarashi, S. Tsurekawa, T. Watanabe, Acta Mater. 47 (1999) 1163–1174. [43] S. Yamura, Y. Igarashi, S. Tsurekawa, T. Watanabe, in: A. Meike, et al. (Eds.), Properties of Complex Inorganic Solids, vol. 2, Kluwer Academic/Plenum Pub., 2000, pp. 27–37. [44] S. Yamaura, S. Tsurekawa, T. Watanabe, Mater. Trans. 44 (2003) 1494–1502