Processing, properties and applications of high strain rate superplastic materials
Mamoru MabuchiNational Industrial Research Institute of Nagoya,
Hirate-cho, Kita-ku, Nagoya, 462, Japan
Kenji Higashi
Department of Metallurgy and Materials Science,
College of Engineering, Osaka Prefecture University,
Gakuen-cho, Sakai, Osaka 593, Japan
Abstract
Recent research and development on high strain rate superplasticity has led to a new concept of "accommodation helper mechanism" and a new opportunity of "superplastic forging". In the present article, recent works about processing for ultra fine-grained materials, deformation mechanisms and applications of high strain rate superplasticity are reviewed.Introduction
Superplastic materials exhibit very large elongations >500 % (in a special case,>5000 %)(Ahmed & Langdon, Higashi). However, large elongations are usually attained only in a low strain rate range from 10-5 to 10-3 /s. The superplastic strain rate range is rather low for commercial forming of structural materials and the commercial viability of conventional superplastic forming is limited. However, recent advances on superplasticity R&D have led to a new field of "high strain rate superplasticity". High strain rate superplasticity is very attractive for commercial applications because high strain rate superplastic forming can lead to high productivity in fabrication of products with complicated shapes.High strain rate superplasticity was originally found in a SiCw/2124 composite in 1984 (Nieh et al.) and in a mechanically-alloyed IN9021 in 1985 (Nieh et al.). Since 1990, high strain rate superplasticity has been studied extensively in aluminum based materials. High strain rate superplastic materials are listed in Table 1. It should be noted that superplastic behavior is attained at relatively high strain rates >10-2/s. These strain rates are considerably higher than a typical forming rate used for conventional superplastic materials, but rather close to the commercial hot working rates of 10-1 to 102/s. High strain rate superplasticity is defined as superplasticity which is found at strain rates >10-2/s in JIS H 7007 by Japanese Standards Association. It can be seen from Table 1 that most of high strain rate superplastic materials are aluminum based materials. Recently, it has been reported that some magnesium based materials exhibit high strain rate superplasticity. It is of interest to note that high strain rate superplasticity is attained in not only aluminum and magnesium alloys, but also their composites reinforced with ceramic whiskers or particles.
Table 1. Superplastic Properties of
High-Strain-Rate Superplastic Materials
| Material | Temp (K) | Strain Rate (/s) | Flow Stress (MPa) | m | Elongation (%) | Grain Size (µm) | References |
| Dynamic Recrystallization | |||||||
| Supral (Al-Cu-Zr) | 753 | 3x10-3 | 9 | 0.4-0.5 | 1270 | 5 | Grimeset al., Wattset al. |
| 2124-Zr | 748 | 3x10-1 | 35 | 0.5 | 490 | 1 | Niehet al. |
| 7475-Zr | 793 | 10-1 | 11 | 0.6 | 600 | 2.5 | Matsuki |
| 7475-Zr | 793 | 3x10-1 | 900 | 1.3 | Furushiroet al. | ||
| Al-Mg-Zr | 773 | 10-1 | 21 | 0.3 | 570 | Higashiet al. | |
| Al-Cu-Zr | 743 | 10-1 | 25 | 0.3 | 480 | ||
| Al-Mg-Mn | 823 | 3x10-2 | 5.7 | 0.45 | 500 | ||
| Thermomechanical Treatment | |||||||
| SiCp/1100 | 903 | 10-1 | 7.5 | 0.4 | 200 | 2 | Imaiet al. |
| SiCw/2124 | 798 | 3x10-1 | 10 | 0.33 | 300 | Niehet al. | |
| SiCw/2124 | 823 | 1 | 4 | 0.5 | 150 | Kim & Hong | |
| SiCw/2009 | 798 | 3x10-2 | 15 | 0.5 | 180 | Mishraet al. | |
| SiCw/2009 | 773 | 6.7x10-1 | 0.37 | 190 | Han& Chan | ||
| SiCp/2124 | 748 | 3x10-1 | 48 | 0.17 | 139 | Han et al. | |
| SiCw/6061 | 823 | 2x10-1 | 6.5 | 0.32 | 300 | ||
| SiCw/6061 | 873 | 2x10-1 | 1.7 | 0.34 | 440 | 2 | Han & Chan |
| SiCp/6061 | 853 | 3x10-1 | 5.5 | 0.33 | 207 | ||
| SiCp/6061 | 853 | 10-1 | 4.5 | 0.5 | 350 | Nieh et al. | |
| SiCp/6061 | 853 | 3x10-1 | 4.2 | 0.36 | 342 | 2 | Hikosaki & Imai |
| SiCp/7075 | 793 | 5x10-1 | 9 | 0.42 | 200 | 0.6 | Matsuki et al. |
| SiCp/7075 | 793 | 1 | 12 | 0.37 | 200 | 0.5 | Matsuki et al. |
| SiCp/7075 | 793 | 10 | 31 | 0.36 | 200 | 0.5 | Matsuki et al. |
| SiCp/8090 | 848 | 2x10-1 | 4.7 | 0.42 | 300 | 6 | Chan, Han & Yue; Han & Chan |
| Si3N4w/Al | 903 | 10-1 | 10 | 0.47 | 200 | 2 | Imaiet al. |
| Si3N4w/2124 | 798 | 2x10-1 | 8 | 0.5 | 250 | Imaiet al. | |
| Si3N4w/2024 | 818 | 4x10-2 | 2.4 | 0.35 | 280 | 4 | Mabuchi,Higashi & Langdon |
| Si3N4w/2009 | 773 | 3x10-1 | 7.5 | 0.3 | 280 | 1 | |
| Si3N4p/2124 | 788 | 4x10-2 | 2.2 | 0.4 | 840 | 2 | |
| Si3N4w/6061 | 818 | 1 | 6 | 0.3-0.5 | 700 | 1 | Mabuchi& Higashi |
| Si3N4w/6061 | 798 | 2x10-1 | 22 | 0.5 | 300 | Mabuch & Imai | |
| Si3N4w/6061 | 818 | 2x10-1 | 7 | 0.5 | 600 | 3 | Mabuchiet al. |
| Si3N4w/6061 | 818 | 2x10-1 | 11 | 0.5 | 260 | 3.1 | Mabuchiet al. |
| Si3N4w/6061 | 818 | 3x10-1 | 4.7 | 0.3 | 150 | 3 | |
| Si3N4w/6061 | 833 | 10-1 | 11 | 0.3-0.5 | 480 | 3.3 | Mabuchi& Higashi |
| Si3N4w/6061 | 818 | 2x10-1 | 6.2 | 0.46 | 600 | Tochigi, Imai & Ai | |
| Si3N4w/6061 | 818 | 2x10-2 | 15 | 0.31 | 173 | 3 | Tochigi, Imai & Ai |
| Si3N4p/6061 | 833 | 2 | 5.2 | 0.3-0.5 | 620 | 1.3 | Mabuchi& Higashi |
| Si3N4p/6061 | 833 | 1 | 5.5 | 0.3-0.5 | 350 | 1.9 | |
| Si3N4p/6061 | 818 | 10-1 | 5.3 | 0.3-0.5 | 450 | 3 | |
| Si3N4w/7064 | 798 | 8x10-1 | 18 | 0.4 | 160 | 3.5 | Mabuchi et al. |
| Si3N4w/7064 | 818 | 2x10-1 | 10 | 0.4 | 230 | 3.6 | Mabuchi et al. |
| Si3N4w/7064 | 818 | 5x10-1 | 14 | 0.5 | 240 | 3.5 | Mabuchi et al. |
| Si3N4w/7064 | 833 | 10-1 | 4.2 | 0.3-0.5 | 380 | 3.5 | Mabuchi & Higahsi |
| Si3N4w/7075 | 773 | 2x10-1 | 24 | 0.31 | 260 | Lim & Nishida | |
| Si3N4p/7064 | 818 | 1 | 10 | 0.45 | 330 | Mabuchi & Higashi | |
| Consolidation of Amorphous or NanocrystallinePowder | |||||||
| Al-Ni-Mm | 885 | 1 | 15 | 0.5 | 650 | 1 | Higashi; Higashi et al. |
| Al-Ni-Mm-Zr | 873 | 1 | 15 | 0.5 | 650 | 0.8 | Higashi |
| Mg-Al-Ga | 573 | 10-2 | 8 | 0.5 | 1080 | 2 | Uoya et al. |
| Mechanical Alloying | |||||||
| IN9021 | 723 | 7x10-1 | 5 | 0.3 | 300 | Nieh et al. | |
| IN90211 | 748 | 2.5 | 40 | 0.3 | 505 | 0.5 | Bieler et al. |
| IN9052 | 863 | 10 | 15 | 0.6 | 330 | 0.5 | Higashi, Nieh & Wadsworth |
| IN905XL | 848 | 20 | 12 | 0.6 | 190 | 0.4 | Higashi, Nieh & Wadsworth |
| IN9021 | 823 | 50 | 18 | 0.5 | 1250 | 0.5 | Higashi, Nieh & Wadsworth |
| SiCp/IN9021 | 823 | 5 | 5 | 0.5 | 600 | 0.5 | Higashi, Nieh & Wadsworth |
| Physical Vapour Deposition | |||||||
| Al-Cr-Fe | |||||||
| Intense Plastic Straining | |||||||
| Al-Mg-Li-Zr | 623 | 10-2 | 85 | 1180 | 1.2 | Valiev | |
| Al-Cu-Zr | 573 | 10-2 | 970 | 0.5 | Valiev | ||
| Zn-Al | 473 | 3x10-2 | 27 | 1970 | 0.6 | Furukawa et al. | |
It is now accepted that high strain rate superplasticity is associated with an ultra fine grain size (67,68). The variation in superplastic strain rate as a function of inverse of grain size is shown in Fig. 1. The grain size is typically about 10 µm for conventional metallic materials showing superplastic behavior at 10-5 Æ 10-3 /s. Clearly, high strain rate superplasticity (> 10-2 /s) is attributed to a very small grain size (< about 5 µm). Advances of processing technology for ultra fine-grained materials have led to a new field of high strain rate superplasticity. Also, much effort has been made to understand the deformation mechanisms of high strain rate superplasticity in recent years.
Processing and Microstructure
Advances of processing technology for ultra fine-grained materials have led to high strain rate superplasticity. Processing for high strain rate superplastic materials is classified from grain refinement routes as follows; (1) dynamic recrystallization, (2) thermo-mechanical treatment, (3) consolidation of amorphous or nanocrystalline powder, (4) mechanical alloying, (5) physical vapor deposition, and (6) intense plastic straining.Dynamic Recrystallization : for some aluminum alloys with a fine subgrain structure, dynamic recrystallization occurs and fine-grained microstructure (<3 - 5 µm) is attained at an early stage of hot deformation. The dynamically recrystallized materials exhibit superplastic behavior at high strain rates >10-2 /s (7-10). Microstructural change during deformation and the variation in stress as a function of strain is shown in Fig. 2 for an Al-Li alloy showing dynamic recrystallization during hot deformation. It can be seen from Fig. 2 that fine subgrain structure is found before a tensile test, however, the misorientation between neighboring grain boundaries increases with straining by continuous recrystallization, as a result, a fine-grained microstructure is attained at e = 1. The flow stress increases with strain at an initial strain range before the change from the subgrain structure to the fine-grained structure, however, it decreases with strain because of the microstructural change. It should be noted that the strain rate sensitivity increases because of the microstructural change (69), indicating that grain boundary sliding is significantly caused by the microstructural change. A very fine subgrain structure is required to attain a fine-grained structure due to dynamic recrystallization. Fine and stable particles such as Al3Zr play an important role in grain refinement by dynamic recrystallization. Very small particles less than 30 nm are needed to attain grain refinement by dynamic recrystallization (70).
Thermo-Mechanical Treatment : grain refinement has been attained by thermo-mechanical treatment in many materials, as shown in Table 1. In particular, a very small grain size has been attained by a combination of thermo-mechanical treatment and powder metallurgy method. It is of interest to note that many very fine-grained metal matrix composites have been produced by thermo-mechanical treatment, as shown in Table 1. Microstructure of a high strain rate superplastic Si3N4w/6061 (Al-Mg-Si) matrix composite exhibiting superplasticity at 2 x 10-1 /s (28) is shown in Fig. 3. The high strain rate superplastic composite was processed by a combination of thermo-mechanical treatment and powder metallurgy. A grain size of the composite is about 3 µm. A very small grain size of ~ 3 µm is attained by grain refinement through hot extrusion for the aluminum matrix composites with ceramic whiskers or particles (71). It is recognized that large particles (>1 µm) stimulate recrystallization. Reinforcements stimulate recrystallization during hot working (extrusion) for the composites. Furthermore, grain growth is limited by fine particles which are dynamically precipitated during hot working(72). Therefore, grain refinement by hot extrusion for the composites is attributed to both recrystallization and precipitation (72).
Consolidation of Amorphous or Nanocrystalline Powder : ultra fine-grained Al-14wt%Ni-14wt%Mm and Al-14wt%Ni-7wt%Mm-1wt%Zr have been processed by extruding amorphous or nanocrystalline powders. The grain sizes of these materials are 50 ~ 100 nm. However, the near-nano scale grained structures are unstable above about 773 K, and grain growth occurs significantly at elevated temperatures. The grain growth rate depends on the annealing temperature, the holding time and the heating rate. For the Al-Ni-Mm-Zr alloy, the very fine-grained structure remains even after annealing for 30 s at 773 K, however, when the specimen is heated to 873 K and then annealed for a holding time of 30 s, grain growth occurs significantly and the grain size becomes about 1 µm. However, it should be noted that the grain size is still very small (= 1 µm) even after annealing at a high temperature of 873 K.
Mechanical Alloying : some high strain rate superplastic aluminum based materials, i.e. IN9052 alloy (Al-4.0wt%Mg-1.1wt%C-0.8wt%O), IN905XL alloy (Al-4.0wt%Mg-1.5wt%Li-1.2wt%C-0.4wt%O), IN9021 alloy (Al-4.0wt%Cu-1.5wt%Mg-1.1wt%C-0.8wt%O) and 15 vol% SiCp/IN9021 composite, have been produced by a mechanical alloying route. These mechanically alloyed aluminum based materials consist of ultra fine grains of 350 ~ 500 nm, and exhibit superplasticity at very high strain rates (>10 ~ 100/s). The superplastically deformed IN 9021 specimens are shown in Fig. 4. It should be noted that very large elongations > 1000 % are attained at very high strain rates of more than 100/s. Fine carbide (Al4C3) and oxide (Al2O3, MgO, or LiO2) particles are dispersed during mechanical alloying. The size of the dispersed particles is less than 30 nm, and the volume fraction of the particles is approximately about 5 vol% for these materials. The structure of the mechanically alloyed materials is very stable at high temperatures because the dispersed particles limit grain growth. Therefore, the mechanically alloyed materials have high potential for high strain rate superplasticity.
Intense Plastic Straining : a submicrometer grain size has been attained by intense plastic straining through equal channel angular extrusion (ECAE) (73-85). ECAE is a special method to attain significantly large shear strain with relatively low pressure for massive bulk products. Through this technique, bulk samples with a submicrometer grain size are produced, and high strength and superplasticity are attained (64,65,73-76). In addition, it is reported that the materials processed by ECAE exhibit unique grain growth behavior (77,79,80) and unusual Hall-Petch relationship (81,82) because of non-equilibrium grain boundaries. Thus, ECAE can give new opportunities for applications of materials processing.
Deformation Mechanisms
It is accepted that grain boundary sliding is the dominant deformation process of high strain rate superplasticity as well as conventional superplasticity. Recently, some deformation mechanisms of high strain rate superplasticity have been proposed from the viewpoint of accommodation process of grain boundary sliding (21,87-96). Mishra et al. (90-93) noted that the rate controlling mechanism depends on the volume fraction of second phase particles, and it changes from grain boundary diffusion controlled grain boundary sliding to interface diffusion controlled interfacial sliding, as a result, the activation energy for superplasticity in metal matrix composites is much higher than that for grain boundary diffusion of the matrix. On the other hand, Nieh et al. (3) pointed out that a liquid phase plays an important role in high strain rate superplasticity. Recently, it was suggested that a liquid phase affects the superplastic properties and the high activation energy value in the composites is attributed to the presence of a liquid phase (97). The presence of a liquid phase was evidenced by in-situ transmission electron microscopy and differential scanning calorimetry (26,30,98). The evidence of partial melting at the interface in high strain rate superplastic Si3N4p/Al-Mg composite is shown in Fig. 5. Koike et al. (99) investigated solute segregation at interfaces and grain boundaries by microchemical analysis and they noted that melting occurs preferentially at the segregated grain boundaries and interfaces.Grain boundary sliding is a dominate deformation mechanism during superplastic flow. Under standard superplastic conditions, the grain compatibility during grain boundary sliding is maintained by concurrent accommodation processes which involve grain boundary migration, grain rotation, diffusion or dislocation motion. However, the high strain rate superplastic materials contain a large volume fraction (5 ~30 vol %) of second phase particles or ceramic reinforcements. These particles, with their thermal stability and non-deformability, hamper grain boundary sliding and/or act as stress concentrators. In these materials, therefore, a high concentration in local stress by grain boundary sliding, if it is not fully accommodated by diffusional or plastic flow, causes excessive cavity formation at the particle interfaces and at triple junctions of grain boundaries, resulting in premature fracture. Therefore, in order to obtain large superplastic elongations, special mechanisms are required for the accommodation process to relax the stress concentrations. Especially, for metal matrix composites reinforced with a high volume fraction of hard reinforcements, superplasticity is not necessarily attained only by a small grain size because excessive cavitation is caused due to the high stress concentrations around the reinforcements located on grain boundaries. In particular, it is difficult to accommodate grain boundary sliding at high strain rates by diffusion processes including diffusion-controlled dislocation movement because the times are too short.
Table 2. Partial Melting Temperature &
Optimum Superplastic Temperature for High Strain Rate Superplastic
Aluminium Matrix Composites
| Material | Temperature (K) | Maximum Elongation (%) | Strain Rate (/s) | ||
| Superplastic | Partial Melting | Solidus | |||
| Si3N4/Al Composites | |||||
| Si3N4p (0.2µm)/Al-Cu-Mg (2124) | 773 | - | 775 | 280 | 3x10-1 |
| Si3N4p (1µm)/Al-Cu-Mg (2124) | 788 | 784 | 775 | 840 | 4x10-2 |
| Si3N4w /Al-Cu-Mg (2124) | 818 | - | 775 | 280 | 4x10-2 |
| Si3N4p (0.2µm)/Al-Mg (5052) | 818 | 821 | 866 | 700 | 1 |
| Si3N4p (0.2µm)/Al-Mg-Si (6061) | 833 | 830 | 855 | 620 | 2 |
| Si3N4p (0.5µm)/Al-Mg-Si (6061) | 833 | 829 | 855 | 350 | 1 |
| Si3N4p (1µm)/Al-Mg-Si (6061) | 818 | 822 | 855 | 450 | 1x10-1 |
| Si3N4w /Al-Mg-Si (6061) | 833 | 843 | 855 | 480 | 1x10-1 |
| Si3N4p (1µm)/Al-Zn-Mg | 818 | 814 | - | 330 | 1 |
| Si3N4w /Al-Zn-Mg | 833 | 830 | - | 380 | 1x10-1 |
| SiC/Al Composites | |||||
| SiCw/Al-Cu-Mg (2124) | 798 | 776 | 775 | 300 | 3x10-1 |
| SiCp/Al-Cu-Mg (2014) | 753 | - | 780 | 349 | 2x10-4 |
| SiCw/Al-Mg-Si (6061) | 823 | - | 855 | 300 | 2x10-1 |
| SiCp/Al-Zn-Mg-Cu (PM64) | 773 | 793 | - | 450 | 2x10-4 |
| SiCp/Al-Zn-Mg-Cu (7475) | 788 | - | - | 310 | 2x10-4 |
| SiCp/Al-Zn-Mg-Cu (7075) | 793 | - | 749 | 300 | 5 |
| SiCp/Al-Cu-Mg-C-O (IN9021) | 823 | - | - | 600 | 5 |
It should be noted that for many high strain rate superplastic materials, a maximum elongation is attained at the temperature close to or slightly above the partial melting temperature where meting occurs locally at the interfaces and/or grain boundaries (26,30,100). The partial melting temperature and the optimum superplastic temperature for high strain rate superplastic aluminum matrix composites are listed in Table 2, where the optimum superplastic temperature is the temperature where a maximum elongation is attained. It can be seen that the optimum superplastic temperature is very close to the partial meting temperature. This suggests that the presence of a liquid phase is related to the origin of high strain rate superplasticity. Recently, a new model (89) was proposed from the viewpoint of the accommodation mechanism by a liquid phase. This new model suggests that a liquid phase serves both to relax the stress concentrations and to limit the build up of internal cavitation and subsequent failure. This concept has been supported by many investigations of cavitation (101-108).
However, the authors would like to emphasize that a liquid phase is not always necessary to attain high strain rate superplasticity. When the stress concentrations at the interfaces and at triple junctions of grain boundaries are relaxed by diffusion or dislocation movement, a liquid phase is not needed for high strain rate superplasticity. It has been reported that high strain rate superplasticity is attained without a liquid phase for the magnesium based composite (52) and the aluminum alloys (64, 92, 93). It should be noted that a liquid phase serves to help the accommodation process for grain boundary sliding when grain boundary sliding is not sufficiently accommodated by diffusion and dislocation movement, namely, the role of a liquid phase is the "accommodation helper" (109).
In general, a tensile stress tends to cause decohesion at a liquid phase. Hence it is required to limit decohesion at a liquid phase in a tensile stress field in order to attain superplasticity by tensile straining. Miller and Chadwick (110) considered a thin layer of liquid between two solids and they developed the critical tensile stress required to overcome the attraction due to surface tension. According to their result, the critical tensile stress required to overcome the attraction due to surface tension is given by
sc = 2gL/h
where sc is the critical tensile stress required to overcome the attraction due to surface tension, gL is the surface energy of a liquid phase and h is the thickness of a liquid phase. The variation in the critical stress as a function of the thickness of a liquid phase for aluminum based materials is shown in Fig. 6, where the surface energy of melt aluminum is taken to be 0.923 N/m. The thickness of a liquid phase at the optimum superplastic temperature is less than 30 nm (111,112). The experimental flow stresses at the optimum superplastic temperature for the high strain rate superplastic materials are ~ 15 MPa. The experimental flow stresses are much lower than the critical stress at a liquid thickness of 30 nm. Therefore, it is suggested that a liquid phase is not the sites for decohesion at the optimum superplastic temperature (112).
Commercial Applications
High strain rate superplasticity is very attractive for commercial applications because high strain rate superplasticity enables superplastic forging. Superplastic forging is expected to lead to high productivity in fabrication of products with complicated shapes. To date, there are a few of commercial applications of superplastic forging (113,114).An example of superplastic forging in an Al-Ni-Mm-Zr alloy is shown in Fig. 7 (114). The complicated component with a con-rod shape was forged superplastically from an as-extruded workpiece at a commercial production speed (less than 1 second). The forged product showed good post-deformation mechanical properties (114).
It is important to control cavitation occurring during superplastic flow for commercial applications because cavitation has a strong deleterious influence on post-forming service properties. Recently, Iwasaki et al. (115) showed that for a high strain rate superplastic Si3N4p/6061 composite, cavities are reduced by static annealing after superplastic deformation and post-deformation tensile properties are sufficiently recovered. Cavitation control by static annealing is attractive for commercial applications because of low costs.
Technologically, high strain rate superplasticity is of great interest because it is expected to result in economically-viable, near-net-shape forming techniques for the automobile, aerospace, and even semi-conductor industries. In near future, commercial applications of superplastic forging will increase in a wide range of industries.