45. Saturation Of Fragmentation During Severe Plastic Deformation.pdf

MR40CH13-Pippan

ARI

24 May 2010

ANNUAL REVIEWS

Further

15:1

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

Click here for quick links to Annual Reviews content online, including: • Other articles in this volume • Top cited articles • Top downloaded articles • Our comprehensive search

Saturation of Fragmentation During Severe Plastic Deformation R. Pippan,1 S. Scheriau,2 A. Taylor,1 M. Hafok,1 A. Hohenwarter,2 and A. Bachmaier1 1 Erich Schmid Institute of Materials Science, Austrian Academy of Sciences, A-8700 Leoben, Austria; email: [email protected] 2 Christian Doppler Laboratory for Local Analysis of Deformation and Fracture, A-8700 Leoben, Austria

Annu. Rev. Mater. Res. 2010. 40:319–43

Key Words

First published online as a Review in Advance on April 9, 2010

high-pressure torsion, grain boundary migration, nanocrystalline, ultrafine-grained material, nanocomposites

The Annual Review of Materials Research is online at matsci.annualreviews.org This article’s doi: 10.1146/annurev-matsci-070909-104445 c 2010 by Annual Reviews. Copyright  All rights reserved 1531-7331/10/0804-0319$20.00

Abstract In this review, we focus on the saturation microstructure that evolves during severe plastic deformation (SPD). These nanocrystalline or ultrafinegrained microstructures consist predominantly of high-angle boundaries, although low-angle boundaries are also present. Deformation temperature, alloying, and strain path are the dominant factors controlling the saturation grain size in single-phase materials. The saturation grain size decreases significantly with decreasing deformation temperature, although the dependency is stronger at medium homologous temperatures and less in the lowtemperature regime. The saturation microstructure is sensitive to strain rate at medium temperatures and less so at low temperatures. The addition of alloying elements to pure metals also reduces the saturation grain size. The results indicate that grain boundary migration is the dominant process responsible for the limitation in refinement by SPD. Therefore, second-phase particles of the nanometer scale can stabilize even finer microstructures. This mechanism of stabilization of the microstructure is an effective tool for overcoming the limit in refinement of single-phase materials by SPD. The improved thermal stability of the obtained nanostructures is another benefit of the introduction of second-phase particles.

319

MR40CH13-Pippan

ARI

24 May 2010

15:1

INTRODUCTION

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

Heavy plastic deformation at relatively low homologous temperatures, usually called severe plastic deformation (SPD), is an efficient method in producing an ultrafine-grained or a nanocrystalline structure in materials. Traditional grain refinement techniques, such as extrusion, forging, or rolling, do not permit the application of such large strains without either failure or limitations in the load capacity necessary to obtain such levels of refinement. In the past 25 years, various SPD techniques have been developed and improved for the synthesis of large quantities and very different types of bulk nanostructured materials (1–3). Besides achieving uniform grain refinement, the various SPD methods can accomplish a reduction in the porosity and impurities in compact samples (1–9). The enhanced properties of such materials and the development of the SPD techniques have caused the material science community to take a growing interest in this area. Such interest is clearly reflected in the large number of special conferences devoted to SPD (see, e.g., References 4–7) and in the special symposia for SPD in all conferences related to mechanical properties or microstructural phenomena in metallic or intermetallic materials. Although SPD is a new technique, many papers, as well as several reviews, have focused on the processing, microstructure, and properties of SPD materials (1–3, 9–11). Most of these papers are related to processing, the microstructural evolution at equivalent strains between 1 and 10, and property changes in this strain regime. At larger strains, often strains between 10 and 30, saturation in the refinement process occurs. The phenomena causing this saturation and the possible means of generating even finer microstructures are not often considered. The present review is devoted to this subject. We consider here results obtained primarily by the technique of high-pressure torsion (HPT), a SPD processing technique that permits the application of the largest strains in a relatively simple way.

HIGH-PRESSURE TORSION The first HPT devices (12, 13) were modified Bridgman anvil–type devices (14). Further modification (15–16) led to the system as it is now usually used; this system is schematically depicted in Figure 1a. Both anvils are provided with cylindrical, somewhat conical cavities. The diameter of the cylindrical cavities is identical to the sample diameter. The combined depth of the cavities is somewhat smaller than the initial height of the HPT sample. Thus, during loading, a small amount of the material flows between the two anvils. The friction in this thin ring confines the free flow of material out of the HPT tool. This back pressure then induces a relatively well-defined hydrostatic pressure within the processing zone of the tool. The material in this sealing region prevents the touching and, as a consequence, the failure of the anvils during the subsequent torsional deformation. Both cavities of the tool are sandblasted to clean the surfaces and to provide enough friction through the generated microroughness, which is necessary for a continuous torsion deformation. To prevent sliding, a minimum pressure is required. Generally, the minimum pressure is three times the flow stress of the undeformed material. The shear strain, γ , can be calculated according to Equation 1: γ =

2π n · r, t

1.

where r, n, and t are the distance from the center of the sample, the number of turns, and the sample thickness, respectively. The equivalent plastic strain, based on the assumption that two samples deformed via different strain paths should be comparable when the same plastic work has been done on them, is then (16, 17) γ 2π n ε = √ = √ · r. 3 3t 320

Pippan et al.

2.

MR40CH13-Pippan

ARI

24 May 2010

15:1

a

Strain gauge

b

c

Sample

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

Torsion axis

Axial

Upper anvil

r

t Radial

γ

Tangential Lower anvil

d

Figure 1 Schematic representation of (a) the high-pressure torsion (HPT) deformation of a sample, (b) the torsion deformation of the sample, and (c) the principal observation directions of the microstructure.

The strain in the center should be zero and should increase linearly with the radius for ideal torsion deformation. As long as the thickness-to-diameter ratio is smaller than 1/10, Equations 1 and 2 describe the strain in real HPT samples relatively well (15, 18, 19), except in the immediate vicinity of the edge of the sample. For more details, see References 15 and 19. Compared with other SPD processes, the HPT technique offers several advantages, as follows: 



 

Most importantly, extremely high shear strains can be achieved via very simple means. One revolution of a 0.8-mm-thick HPT sample corresponds to an equivalent strain of approximately 18 at a radius of 4 mm. One hundred revolutions, which are usually not difficult to apply, correspond then to an equivalent strain of 1800. Such large strains are impossible to obtain with any other technique. HPT permits the SPD of relatively brittle or high-strength materials at low temperatures, at which severe deformation is often impossible (see, for example, References 11 and 19–32). The strain and the strain rate at a specified radius of the sample can be precisely controlled. Heating and cooling of the anvils permit SPD at well-defined temperatures. Our device can be cooled by liquid nitrogen and heated up by an induction coil to approximately 700◦ C. www.annualreviews.org • Saturation of Fragmentation During SPD

321

MR40CH13-Pippan

ARI

24 May 2010





15:1

The total torque versus the angle of rotation can be measured. This allows for estimation of the flow stress evolution (33, 34) and for control of the process such that, for example, slip between the anvil and the sample does not occur. A change in the rotation direction permits, furthermore, a cyclic SPD (35, 36).

The largest disadvantage of standard HPT is the “limited” sample size. Our largest devices can produce samples with a diameter of approximately 40 mm and a thickness of a few millimeters. However, there are some ways to overcome this limitation (see, for example, References 9, 11, 37, and 38).

THE FRAGMENTATION OF MICROSTRUCTURE

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

Figure 2 shows the crystal orientation maps (often called inverse pole figure maps) of HPTdeformed polycrystalline pure Ni at different strains. All orientation maps are obtained in axial directions. The scale of the maps differs significantly. The maps of the undeformed Ni and those for strains of 0.5 and 1 show a 400 × 400 μm area, and at strains of 8, 16, and 32, only a 7 μm × 7 μm area is depicted. The map of the undeformed Ni indicates a polycrystalline microstructure with a grain size of approximately 70 μm. After a deformation to an equivalent strain of 0.5, the original grain structure is clearly visible. Small fluctuations of the orientation within the grains and somewhat larger changes of the orientation near the grain boundary are observed. At equivalent strains of 1 and 2, the fluctuation of the crystal orientation within the original grains increases, varying significantly from grain to grain. From these maps, it is evident that the elevated density of dislocations is not randomly stored in the microstructure but that rather the dislocations concentrate mainly into boundaries. These boundaries subdivide the initial grains into a hierarchical structure of cell blocks with large misorientations and ordinary dislocation cells with smaller misorientations. This initial process is well documented in the literature (39–43). With increasing strain, the size of these two types of structural elements decreases, the boundaries become more obvious, the misorientation of neighboring elements increases, and it becomes impossible to locate the original grain boundaries. However, at a certain strain—in the present case of pure Ni, this is a strain of between 8 and 16—the refinement process saturates. The cell block and dislocation cell structure, which are characteristic features at small and medium SPD, are transformed to a uniform granular structure. A quantitative analysis of such a map shows that in the saturation regime the fraction of high-angle boundaries, i.e., a misorientation larger than 15◦ , is approximately 80%, whereas 20% of boundaries exhibit a misorientation between 2◦ and 15◦ (44, 45). The saturation microstructure shows a typical shear texture (44, 46, 51). However, the misorientation of neighboring structural elements is approximately random (16). A common explanation for the occurrence of grain fragmentation is the grain-grain interaction, which allows continuity to be maintained during deformation. Neighboring crystals exert stresses on each other such that the local stress varies within a grain. As a consequence, different sets of slip systems are activated in different regions within one crystal, leading to different changes in orientation. There is plenty of evidence at small strains for this phenomenon (see, for example, References 47–49). To see if this grain-grain interaction is the dominant effect causing the fragmentation, we performed HPT deformation of single crystals (44, 51). Cu and Ni single crystals with crystallographic orientations 111 and 001 parallel to the torsion axis were deformed to different strains. Figure 3 shows orientation maps in the vicinity of the center of a 111 Cu single crystal and at a radius of 3 mm, where the equivalent strain is 2. The torsion of the 111 crystal could be realized by a slip in a single plane, the shear plane of the torsion sample. The slip direction, however, had to be different at different angles. In the ideal case, the crystal orientation 322

Pippan et al.

MR40CH13-Pippan

ARI

24 May 2010

15:1

ε=0

ε = 0.5

ε=1

100 μm

100 μm

100 μm

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

111 223 233

112 113

123

122 133

ε=2

ε=4

25 μm

25 μm

ε=8

ε = 16

ε = 32

2 μm

2 μm

2 μm

001

013 012 023

011

Figure 2 Illustration of the microstructural changes due to severe plastic deformation (SPD) of a typical coarsegrained material to a submicrometer-grained microstructure. Orientation micrographs were taken in the axial direction of a HPT polycrystalline Ni deformed at room temperature to strains of 0, 0.5, 1, 2, 4, 8, 16, and 32. The maps show the crystal orientation parallel to the surface normal; the color coding is also indicated. Note the large differences in the scales of the micrographs.

should change only in the radial and tangential directions from the top to the bottom of the sample. Near the center of the sample in Figure 3b, this change in the crystal orientation from top to bottom is clearly visible. The continuous change in the crystal orientation parallel to the radial direction indicates clearly the homogeneous torsion deformation of the single crystal. In the same sample, at a radius of 3 mm, which corresponds to ε = 2, such a change in the crystal orientation is visible (see Figure 3c). However, in this case a pronounced fragmentation of the microstructure has developed. The single-crystal experiments indicate clearly that grain-grain interaction is not a necessary condition for the development of such fragmentation. In Figure 4, a comparison of the torque measured during HPT deformation of 111 and 001 single crystals and a polycrystalline Ni sample is shown. For strains of less than three, at the edge of www.annualreviews.org • Saturation of Fragmentation During SPD

323

MR40CH13-Pippan

ARI

24 May 2010

111

a

001

15:1

b

c

101

RAD

TAN

100 μm

TOR 200 μm

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

r

Figure 3 Orientation micrographs (inverse pole figure maps) taken in the radial direction of a HPT-deformed Cu single crystal with a 111 orientation parallel to the torsion axis. Researchers applied 0.15 revolutions to the 0.7-mm-thick sample. (a) The orientation color code and the analyzed area. The radius where the analysis is performed and the coordinate system are indicated. (b) The continuous change in the color, i.e., the continuous change in the crystal orientation with respect to the radial direction, from top to bottom indicates that near the center (r = 0) of the sample, a homogeneous torsion occurred. (c) At radius r = 3 mm, which corresponds to an equivalent strain of 2, a block-like substructure with large misorientations to the neighboring blocks occurs in addition to a top-to-bottom orientation change.

the sample, a significant difference between the three samples is observed. The single crystal with the 111 plane parallel to the shear plane exhibits the least hardening, whereas the 001 crystal shows the greatest hardening. The polycrystal hardening is in between. The hardening strongly affects the initial orientation (43). However, the initial fragmentation is also significantly affected by the initial orientation; this is partly visible in the orientation maps in Figure 2 at ε = 0.5 and 1 in the different grains. This effect is clearer in the orientation maps of the 111 and 001 single crystals deformed to a strain of 0.3 in Figure 5. Whereas the 001 single crystal demonstrates a well-pronounced substructure at this relatively small strain, the 111 single crystal does not change significantly from the initial orientation, and only on a somewhat larger scale can the start of the fragmentation be detected (44). At strains larger than 3, the difference in the torque necessary to deform the different single crystals and the polycrystal disappears. Only the polycrystal exhibits a somewhat larger torque in the saturation region. This difference is reflected in the somewhat finer microstructure, which is caused by the greater impurity concentration in the polycrystalline sample. Apart from the aforementioned small difference in the saturation grain size, the final structure seems independent of the initial microstructure. Does this statement also hold for smaller initial grain sizes? To investigate this question, electrodeposited Ni with an initial grain size of approximately 20 nm was HPT deformed (50). Figure 6 displays the scanning electron micrographs of the initial 324

Pippan et al.

MR40CH13-Pippan

ARI

24 May 2010

15:1

Onset of the saturation region of the single crystals 600

a 500 500 400

M (Nm)

M (Nm)

400 300

300 200 100

200

0

0

1

2

0

5

10

15

εeq

20

25

Nickel <0 0 1>||TOR single crystal Nickel <1 1 1>||TOR single crystal Nickel polycrystal

30

1.2

6

Boundary length / area (μm–1)

b 1.0

d (μm)

4

0.8

0.6

0.4

0.2 0

5

10

15

20

25

eeq Nickel <001>||TOR single crystal Nickel <111>||TOR single crystal Nickel single crystals

30

35

100

c 5

80

4 60 3 40 2 20

1 0

0

5

10

15

20

25

30

High-angle boundary fraction (%)

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

0

3

ev

100

0

eeq Cells 2°–15° Cell blocks >15° High-angle boundary fraction

Figure 4 (a) Measured torque versus equivalent strain for Ni single crystals and a Ni polycrystal. The equivalent strain corresponds to a radius of 6 mm, and the sample diameter is 14 mm (44). The effect of the initial orientation at small strains is clear, and at larger strains, this difference disappears. Panel b shows the grain size obtained from inverse pole figure maps in the axial direction of the 111 and 001 single crystal, taking only large-angle boundaries into account as a function of strain. Panel c displays, for the 111 Ni single crystal, the mean boundary length of boundaries between 2◦ and 15◦ and of boundaries larger than 15◦ , i.e., high-angle boundaries as well as the fraction of high-angle boundaries to low-angle boundaries as a function of strain. This figure indicates that in the saturation region of the torque, all microstructural parameters also reach saturation.

microstructure and the microstructure deformed to a strain of 68. These micrographs make evident that the grain structure is significantly coarsened by the HPT deformation. The grain size obtained at large strains is approximately equal to the grain size of the previously described Ni samples. In summary, during HPT deformation of a single-phase material in the saturation regime, one always ends up with the same ultrafine-grained or nanocrystalline microstructure independently of the initial microstructure, i.e., independently of whether the initial structure was coarser or finer. www.annualreviews.org • Saturation of Fragmentation During SPD

325

ARI

24 May 2010

15:1

a

b

c

3 μm

10 μm

10 μm

Figure 5 Orientation maps of HPT-deformed Ni single crystals in the axial direction. Panels a and b show the maps of 111 and 001 single crystals deformed to a strain of 0.3. The difference in fragmentation on the micrometer scale is visible. The initial orientation, however, does not affect the saturation microstructure of either crystal. Panel c displays only the 111 crystal’s saturation microstructure, which is identical to that of the 001 crystal. The saturation microstructure is somewhat coarser than in the polycrystal in Figure 2 because of the higher impurity level of the polycrystal.

THE EFFECT OF TEMPERATURE AND ALLOYING Figure 7 shows transmission electron microscope (TEM) micrographs of pure Ni and Ni25Fe, both deformed by HPT at room temperature up to ε = 32. The micrographs are taken in the radial direction. The alloying significantly reduces the grain structure size from approximately 200 nm in Ni to approximately 100 nm in the Ni25Fe alloy. This size reduction is a general phenomenon observed in many metals and single-phase alloys (10, 11, 52, 53). The deformation temperature has a similarly pronounced effect on the saturation grain size. Figure 8 shows the measured torque during HPT deformation at the temperature of liquid nitrogen, 72◦ C, and 417◦ C. At all three temperatures, the torque increases with rotation angle

a

b

c Size of structural elements

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

MR40CH13-Pippan

200 nm

MC nickel

~200 nm

NC nickel

200 nm 0

5

10

15

20

25

30

Equivalent strain, εV Figure 6 Scanning electron micrograph of (a) electrodeposited nanocrystalline Ni and (b) HPT-deformed, electrodeposited Ni in the axial direction. The coarsening of the nanocrystalline Ni due to HPT is obvious. Thus, the saturation grain size in SPD is independent of the initial microstructure, as schematically depicted in panel c. 326

Pippan et al.

ARI

24 May 2010

15:1

a

b

100 nm

100 nm

Figure 7 Transmission electron microscope (TEM) micrographs of (a) HPT Ni and (b) Ni with 25% Fe taken in the radial direction. The significantly finer microstructure of the Ni Fe alloy is clearly visible, which shows the strong effect of alloying on the saturation grain size.

or strain and reaches saturation. The difference in torque reflects the difference in the resulting grain size. The finer grain size at low temperature induces the higher flow stress and the higher torque necessary to deform the sample. Not only does Figure 8 indicate that the temperature has a significant effect on saturation grain size, but it also shows that the strain to obtain the saturation microstructure depends strongly on the deformation temperature. For Ni deformed at −196◦ C, 72◦ C, and 417◦ C, saturation was observed at strains of approximately 30, 10, and 3, respectively.

200

Ni: –196°C = 0.045 Tm

Torque (arb. units)

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

MR40CH13-Pippan

150

Ni: 72°C = 0.2 Tm

100

50

Ni: 417°C = 0.4 Tm

0 0

10

20

30

40

Strain, εeq Figure 8 Measured torque versus equivalent strain for polycrystalline Ni deformed by HPT at 417◦ C, 72◦ C, and −196◦ C. The equivalent strain corresponds to a radius of 3 mm, and the sample diameter is 8 mm. The deformation temperature affects both the saturation torque (i.e., the saturation flow stress) and the onset of saturation. Tm denotes the material melting point. www.annualreviews.org • Saturation of Fragmentation During SPD

327

MR40CH13-Pippan

ARI

24 May 2010

15:1

dε/dt = 0.0625 s–1 dε/dt = 0.0025 s–1

a

Al3Mg alloy Al1Mg alloy

b

2.5

0.5

Microhardness, Hv (GPa)

Mean microstructural size (μm)

1.8

0.4 0.3 0.2 0.1 0.0 200

300

400

500

600

700

Deformation temperature, T (°C)

800

2.0

1.6 1.4

1.5

1.2 1.0 1.0 0.5

0.8 0.6

Grain size, d (μm)

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

0.6

–200

–100

0

100

200

300

400

500

0.0

Deformation temperature, T (°C)

Figure 9 Illustration of the effect of HPT deformation temperature on the saturation grain size determined from backscatter electron (BSE) images in the radial direction of (a) pure Fe (33) and (b) an Al3Mg alloy (53). Both plots indicate the strong temperature effect, which seems less pronounced at low temperatures than at medium temperatures. For pure Fe, the effect of strain rate is also displayed, which indicates a stronger effect of the strain rate at high temperatures. For Al3Mg (open symbols), the variation of hardness is also plotted, and at room temperature, the saturation grain size of Al1Mg (closed symbols) is plotted for comparison, which indicates again the strong effect of alloying.

The general effect of temperature and alloying on the saturation grain size is clearly visible in Figure 9. The saturation grain size as a function of deformation temperature is shown for pure Fe (33) and Al3Mg (53). To demonstrate the effect of alloying, the saturation grain size of Al1Mg is also indicated for room temperature deformation. The effect of alloying in the AlMg alloy is even more pronounced than in the case of the NiFe alloy. An addition of 1% and 3% of Mg to Al reduces the saturation grain size obtained from backscatter electron (BSE) micrographs from approximately 1 μm in pure Al to approximately 300 nm and 200 nm, respectively. The variation of the HPT deformation temperature in the Al3Mg alloy from −196◦ C to 450◦ C results in a saturation grain size changing from approximately 100 nm to a few micrometers. In both materials, Fe and the AlMg alloy, the deformation temperature dependency of the saturation microstructure is greater at elevated temperatures. In this elevated temperature regime, a significant effect of the strain rate is also observed (33, 53). At room temperature and below, the effect of deformation temperature is smaller, and the strain rate does not significantly change the saturation microstructure. Additionally, in the low-temperature regime, the grain structure in the radial direction exhibits elongated grains with a preferred orientation. One might suppose that this preferred direction corresponds to the shear angle and should therefore, at large strains, be parallel to the shear direction. However, this is not the case. In the saturation regime, one observes a well-defined angle between the preferred direction of the elongated grains and the shear direction (see Figures 11 –14, below). This angle depends on temperature and alloying (33, 53). At higher temperatures, the angle increases, and the grain structure becomes more and more equiaxed.

LOW-TEMPERATURE STEADY-STATE DEFORMATION In the examples given above, the measured torque and hence the shear flow stress saturate without any reductions in the entire range of deformation conditions investigated. When the deformation 328

Pippan et al.

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

MR40CH13-Pippan

ARI

24 May 2010

15:1

torque becomes constant, the grain size also becomes constant. Only the misorientation distribution of neighboring grains or grain-like structural elements is somewhat changed; they become more and more random with increasing strain and also reach saturation at larger strains (16, 35, 46, 54, 55). Despite the nearly random distribution of neighboring grains, a typical shear or torsion texture remains. One can legitimately term this condition a steady state, consistent with the terminology used for warm or hot working. We now consider the processes that lead to such a steady state during HPT at these low homologous temperatures. In the saturation regime, the generation of defects, i.e., the generation of dislocations, vacancies, and new boundaries, has to be in equilibrium with the annihilation of these defects. The simplest explanation for the occurrence of the steady state at such low temperatures, at which diffusion processes are extremely restricted, is a change in the deformation process, during which the generation of new defects is also restricted. A change from pure dislocation glide to sliding along microstructural boundaries is one possibility. Such a change was proposed for pure Cu (55) for crystallites smaller than 100 nm and for Al and its alloys (56, 57) and has been frequently discussed for submicrometer and nanocrystalline Fe (58, 59). To study whether grain boundary sliding is the dominant deformation mechanism at low temperatures, investigators performed a two-step HPT experiment (60, 61). A Ni sample was deformed to a strain of 42 at a radius of 2 mm, which is beyond saturation. The sample was cut at a radius of 2 mm. On the electrolytically polished, cut surface, a 20 × 20 μm fine grid was produced by a focused ion beam. The sample was further deformed by HPT up to a shear angle of γ = 1 at a radius of 2 mm. The as-prepared and the deformed grid is shown in Figure 10a,b. These micrographs demonstrate that during HPT deformation in the saturation regime, a rather homogeneous shearing of the microstructure, even below 100 nm, occurs, i.e., the shearing is relatively homogeneous even on the grain level. As a consequence, grain boundary sliding cannot be the dominant deformation mechanism in the saturation regime. The mechanism dominating the deformation during the fragmentation process also appears to remain dominant in the steady state. One may argue that in such small grains the condition of dislocation generation rate being equal to the annihilation rate is easily fulfilled because dislocations can move from grain boundary to grain boundary. However, such shearing increases the length of boundaries. This is demonstrated in Figure 11b. The saturation micrograph in Figure 11a is artificially sheared by a strain of approximately 2, which is very small for HPT. In real HPT, one can apply a strain of 1000, and the structure remains the same (10). Even the small artificial additional shearing in Figure 10b indicates the significant elongation of the grains and the reduction in the thickness of the grains. Boundary migration is necessary to obtain a steady-state microstructure and to avoid a continuous decrease in the thickness of the elongated grains. Furthermore, one also needs fragmentation of the elongated grains to obtain a steady-state grain length. Grain boundary migration is a key feature of the different types of dynamic recrystallization during hot working (33, 63–67). To avoid problems with the classical terminology for dynamic recrystallization used by the hot-working community, here we do not term this process dynamic recrystallization during SPD despite several similarities. This is because in classical hot working the grains are very large in relation to the dislocation cell substructure, whereas during SPD at low temperatures, the distance between high-angle grain boundaries is on the order of the size of the dislocation substructure and the mean free path of the dislocations. This difference is even more evident when one compares the size of the nuclei for discontinuous recrystallization and the typical SPD grain size. In most cases, the SPD grain size is on the same order of magnitude of or smaller than the typical recrystallization nuclei. Furthermore, the mean distance s, which a boundary moves during a certain strain increment ε, is significantly different in the steady state for hot www.annualreviews.org • Saturation of Fragmentation During SPD

329

MR40CH13-Pippan

ARI

24 May 2010

15:1

a

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

4 μm

b

4 μm

Figure 10 Scanning electron micrograph of a cut plane of a HPT sample deformed to saturation. (a) On the polished surface, a regular pattern was fabricated by focused ion beam. (b) The sample after an additional HPT deformation by a strain increment of γ = 1. The experiments indicate that, even on the submicrometer level, the microstructure is homogeneously sheared in the saturation regime (61).

working and SPD. The ratio s/ε in the typical hot-working condition is a few micrometers per strain increment of 0.1, whereas in the typical low-temperature SPD, the required s to explain the steady-state microstructure is only a few tens of nanometers per strain increment of 1. As shown in Figure 9, the saturation grain size decreases continuously with the deformation temperature. We believe that the steady-state microstructure is governed by dynamic restoration processes, which continuously change from the high-temperature regime to the low-temperature regime. In the medium-temperature regime (by which we mean in a high-temperature regime for SPD but below the typical hot-forming temperatures), the grain boundary movement is governed by processes similar to dynamic recrystallization. The boundary movement reduces the dislocation density and is controlled mainly by dislocation density and diffusion (68). In this regime, diffusion, which is necessary for dislocation reactions, boundary dragging, and dislocation annihilation, is an important process. In the low-temperature regime, medium-distance diffusion is suppressed. In our opinion, boundary movement is governed by high stress and strain. We term this process strain-induced boundary migration. In principle, one may also call this process stress-induced 330

Pippan et al.

MR40CH13-Pippan

ARI

24 May 2010

15:1

a 5 μm 111

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

001

101

b

c

Figure 11 Illustration of a homogeneous deformation of a typical saturation microstructure without boundary migration. (a) The saturation microstructure (inverse pole figure map) of HPT-deformed Ni at room temperature. (b) The micrograph is homogeneously shared to an equivalent strain of 1.7. (c) The microstructure is artificially rolled (stretched and compressed) to an equivalent strain of 1.4.

boundary migration, corresponding to the case of stage III deformation, in which recovery is predominantly a stress-driven process (43). However, external stresses as well as internal stresses cause movement of the grain boundaries. These internal stresses are caused by dislocations in pileups at the grain boundary or by the entering of dislocations into the boundary. We assume that these entering dislocations are the starting points for the stuffing process of atoms from one grain to the other. Such processes have been seen in molecular dynamic deformation simulation of nanocrystalline materials (69, 70), and experimental observations in various types of loading support the occurrence of these phenomena (71–76). In addition, the reduction of grain boundary energy may be an important driving force. Several experimentally observed features indicate changes in the dynamic restoration mechanism in the medium-temperature and the low-temperature SPD regimes. The most important points are listed below. A closer look at the dependency of the saturation grain size versus deformation temperature indicates that temperature dependency is more pronounced in the medium-temperature regime. The effect becomes more clearly visible when one plots the saturation grain size as a function Q , where ε˙ is the strain rate, R is the gas of the Zener-Hollomon parameter Z = ε˙ · exp RT www.annualreviews.org • Saturation of Fragmentation During SPD

331

ARI

24 May 2010

15:1

constant, Q is the apparent activation energy, and T is the absolute temperature in Kelvin. In the medium-temperature regime, the effect of the strain rate and temperature on the saturation grain size, expressed in terms of Z, is similar to that in the hot deformation regime, whereas for low temperatures, a significant difference is observed (33). In the low-temperature regime in monotonic SPD, i.e., when the strain path is not changed, as in ECAP (equal channel angular pressing) route A, in accumulative roll bonding (ARB), and in monotonic HPT, the grains are not equiaxed. Only in certain observation directions, as in the cases of Figures 2 and 4, do the grains appear equiaxed. In HPT, one may assume that the grains should be aligned parallel to the shear angle (16). In the saturation regime, where the shear strain γ is usually larger than 20, the grains should be aligned in the shear direction. However, this is not the case. The long axis of the grains is always somewhat inclined toward the shear direction, and this angle of inclination increases with increasing deformation temperature. Such inclination and its variation with temperature are impossible to explain with grain boundary sliding or other localized deformation processes. The aspect ratio of the grains depends significantly on the SPD temperature (33). In the highertemperature regime, even under monotonic SPD, the resultant grains are equiaxed. However, with decreasing temperature, the grains become more and more elongated, with a certain inclination to the shear direction. In the saturation regime, the aspect ratio and the inclination angle do not change. This indicates a change in the mechanism. In the high-temperature regime, the movement of boundaries seems to be isotropic, whereas in the low-temperature regime, a preferred movement is necessary, leading to the elongated grains observed. If we assume that the grains at low temperature are relatively homogeneously sheared, then they become elongated. To obtain a steady-state thickness of the elongated grains, movement of boundaries parallel to the short axis of the grains is required. The increasing length of grains requires a subdivision of the grains, which may be caused by the formation of a dislocation boundary or by a process similar to geometrical dynamic recrystallization (40). The new boundaries formed should be perpendicular to the long axis of the grains. These boundaries should be low-angle boundaries, whereas the others should be predominantly high-angle boundaries. An analysis of boundary character from orientation micrographs (Figure 12) and TEM micrographs supports this idea (77). Figure 13 shows the normalized torque measured during HPT deformation at the same homologous temperature for Ni, Cu, and Ag. The three pure metals represent high, medium, and low stacking-fault energies. γ  SFE are 125 mJ m−2 , 40 mJ m−2 , and 16 mJ m−2 for Ni, Cu, and Ag, respectively (78). The three metals exhibit significant differences in the increase in torque as a function of strain. This is not surprising because γ  SFE is an important parameter for the dislocation cell formation and, hence, for the hardening behavior. Despite the enormous difference in the hardening behavior, the saturation grain sizes in these three pure metals are very similar, as shown in Figure 13. In summary, this indicates that the stacking-fault energy strongly influences the structure formation but does not affect the final saturation grain size. The grain boundary migration–controlling effects are more important. In our view, the frequently observed alloying effect is not caused by the change in the stacking-fault energy; rather, it is related to the change in the grain boundary mobility. In the proposed mechanism by which a steady-state grain size is obtained, the mobility of the grain boundary is the key parameter. The influence of impurities mentioned for the polycrystalline Ni supports this concept. A comparison of the saturation grain size at medium SPD temperature with the low-temperature regime indicates that in the low-temperature regime, the effect of alloying is not as pronounced as in the medium-temperature regime. This indicates that at low temperatures, the strain (stress)-driven boundary migration becomes more and more dominant.

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

MR40CH13-Pippan

332

Pippan et al.

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

MR40CH13-Pippan

ARI

24 May 2010

15:1

2 μm Figure 12 Illustration of the distribution of high- and low-angle boundaries in a HPT-deformed Ni sample. The orientation map is taken in the radial direction at ε = 32. High- and low-angle boundaries are marked in black and red, respectively. This figure demonstrates that predominantly the elongated grains with large-angle boundaries are subdivided by low-angle boundaries.

Although the influence of alloying on the saturation grain size in the very low-temperature regime is limited, alloying and impurities are extremely important compared with the case at intermediate temperatures.

REMARKS ON THE EFFECT OF STRAIN PATH The effect of strain path has been extensively investigated (see, for example, References 1 and 79). In HPT, ECAP, and cyclic groove pressing (CGP), simple shear is the dominant deformation mechanism. In contrast, in ARB, cyclic extrusion compression (CEC) (80), and the various types of multiforging (81), compression stretching is the dominant deformation. Besides the difference in the dominant strain, the means of applying the strain, in a cyclic or monotonic fashion, is the characteristic feature by which to distinguish the types of SPD techniques. Dinda et al. (82) processed Ni by ARB and obtained a grain size of approximately 10 nm, which is more than one order of magnitude finer than that in HPT-deformed Ni (see Figures 3 and 5). One may argue that the different strain path is the reason for the difference in grain size or that, in regard to the proposed mechanism for the occurrence of the saturation, grain boundary movement may depend on the type of straining. We deformed a large Ni sample (of diameter 30 mm and thickness 7 mm) by HPT to the saturation regime. Then, the sample was rolled to a thickness of approximately 0.7 mm. The HPT microstructure is shown in Figure 14a. A simple stretching and reduction in grain thickness should result in very thin and long grains, as shown in Figure 11. The actual result is shown in the micrograph in Figure 14b. The grains are now aligned in the rolling direction. They are also elongated, somewhat more than in the HPT sample, but much less than expected from a simple elongation due to the rolling. The thickness of the elongated grains in the rolled sample is somewhat smaller than in the HPT-deformed sample but is not correlated to the applied reduction of the rolled sheet. After a certain deformation by rolling, the shape of the grains reaches a saturation structure, as in HPT. This indicates that the restoration processes in HPT and during rolling are the same in the saturation regime. www.annualreviews.org • Saturation of Fragmentation During SPD

333

MR40CH13-Pippan

ARI

24 May 2010

15:1

1.2

a Cu

1.0 Ag

M / Msat

0.8

Ni

0.6

0.4

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

0.2

0.0

b

c

Ag, 20°C

0

Cu, 50°C

1

2

3

εeq

d

4

5

Ni, 144°C

111

001 5 μm

5 μm

101

5 μm

Figure 13 Illustration of the effect of stacking-fault energy on the saturation microstructure in pure metals. Ag, Cu, and Ni are deformed at the same homologous temperature. The normalized torque-versus-strain curve (a) indicates the strong effect of the stacking-fault energy on the hardening behavior. However, the saturation grain size is not significantly affected, as demonstrated by the orientation maps in panels b–d.

The application of the strain in a cyclic or a monotonic manner can significantly affect the saturation microstructure. In ECAP, simple monotonic or simple cyclic shearing as well as very complex shear strain paths can be realized. A change in the rotation direction in HPT permits the cyclic application of the strain, here termed cyclic HPT (CHPT). We consider here only monotonic and simple cyclic shearing. In the saturation regime at low temperatures during monotonic HPT, grain size is so small that dislocations move predominantly from grain boundary to grain boundary. The reduction of thickness due to shearing of the grain is compensated by the movement of the grain boundary, and grain length is determined by the fragmentation process. In CHPT-deformed pure Ni and pure Fe, below a cyclic strain increment of approximately four, a significantly coarser saturation microstructure is obtained. Can this finding be explained with the proposed mechanism? In CHPT, both the geometrically required reduction of thickness and the increase in grain length are limited, as schematically depicted in Figure 15a,b. As such, grain coarsening by strain-induced grain boundary movement and the required reduction of defect density continue as long as new boundaries can form. At large ε, the monotonic processes control the resulting saturation grain size, whereas below a certain critical strain, the cyclic processes 334

Pippan et al.

MR40CH13-Pippan

ARI

a

24 May 2010

15:1

1 μm

b

1 μm

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

r

Figure 14 Representation of the effect of strain path on the saturation microstructure. The BSE micrographs show (a) the saturation microstructure of the HPT-deformed Ni and (b) the saturation microstructure of a HPT and an additionally heavily rolled (ε = 2.4) Ni sample. A change in the shape and alignment of the grains is evident. The grain volume is not significantly changed, and the reduction in thickness of the elongated grains does not equate to the reduction in thickness of the rolled disk.

become dominant. This difference is evident if one considers the microstructural evolution and the evolution of the flow stress of a HPT-deformed and subsequently CHPT-deformed sample in Figure 15c–e. In the first few cycles of CHPT, the flow stress continuously decreases until it again reaches saturation, which is associated with grain growth.

BULK MECHANICAL ALLOYING: THE BLOCKING OF BOUNDARY MOTION Grain boundary migration limits refinement during SPD. A way to overcome this limitation is to introduce phase boundaries. During SPD of fully pearlitic steels, no saturation was observed (27). More precisely, only monotonic strains of approximately 20 could be applied in standard HPT tools—at this point, the strength of the processed material becomes higher than the strength of the tool material. Coarse-grained CuW and CuCr composites (with approximately 50-50 vol%) were subjected to HPT (20, 83). These composites also show saturation, but at much higher strains. HPT transformed the coarse-grained initial structure into a nanocomposite. In both composites, a grain size of approximately 10 nm was obtained. For comparison, the saturation grain size of Cu and Cr deformed at room temperature is approximately 300 nm, and in W, it is approximately 100 nm. In the composite in the saturation regime, the grain size is more than one order of magnitude smaller. What determines the limit in this case and what are the deformation mechanisms in these nanocomposites are not clear. Perhaps, at a grain size of approximately 10 nm, the W and Cr grains www.annualreviews.org • Saturation of Fragmentation During SPD

335

MR40CH13-Pippan

ARI

24 May 2010

15:1

a

b ε = 1

2

4 Δε = 1

d

200

c

e

150

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

Torque (Nm)

100

εeq = 32

50

εeq total = 32

εeq = 32

0 –50

–100 –150 –200

0

500

1000

1500

2000

2500

3000

Time (s)

d

5 μm

e

5 μm

Figure 15 Illustration of the difference between monotonic and cyclic severe shear deformation. Panels a and b clarify the difference between monotonic and cyclic shear for a cube volume element or cube-like grain. Panel c displays the change in the torque in a cyclic HPT experiment with ε = 1 after a monotonic HPT deformation to a strain of 32 for Ni. One can observe softening and a corresponding increase in grain size during cyclic HPT. Panels d and e are orientation maps of Ni from the points indicated in panel c; the difference in grain size is clear.

are not further deformed or fractured. The W and Cr particles seem to “swim” in the Cu matrix, like the much larger SiC and Al2 O3 particles during SPD deformation in particle-reinforced Al alloys (84, 85). This significant further refinement in these metal-metal composites is an additional indicator that grain boundary sliding or similar localized deformation cannot be the dominant deformation process in the steady-state regime in single-phase materials. Such processes should 336

Pippan et al.

ARI

24 May 2010

15:1

also limit refinement in the above-mentioned metal-metal composites, and the saturation grain size should be on the same order of magnitude as for the corresponding pure metals, which is not the case. As discussed in the previous section, the different strain path in the ARB-processed pure Ni cannot be responsible for the significantly finer microstructure compared with the other SPD deformation routes. Therefore, we investigated whether the oxides introduced by the process could be a sufficient barrier for boundary migration during SPD. Ni powders with particle sizes of approximately 80 μm and 5 μm, both with a natural oxide layer, were precompacted. In addition, the 5 μm powder was annealed before compaction to increase the thickness of the oxide layer. The compacted samples were HPT deformed until the steady state was reached in torque, which also resulted in a steady state of strength or hardness. The HPT deformation led to dense bulk Ni with a small quantity of NiO nanoparticles (86). The difference in the obtained grain size was clearly reflected in the hardness differences. The hardness of the recrystallized coarse-grained Ni increased from 70 HV to approximately 300 HV, and the consolidated coarse Ni powder was somewhat harder (approximately 500 HV). The bulk Ni generated from the HPT-deformed 5-μm powder resulted in significantly larger hardness, which could, extraordinarily, be further enhanced to approximately 690 HV in the artificially oxidized Ni powder. TEM analyses revealed for the latter samples a distribution of grain sizes between 10 nm and 30 nm (see Figure 16).

Particle numbers (%)

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

MR40CH13-Pippan

30 25 20 15 10 5 0 0

20

40

60

Grain size (nm)

80

100

100 nm

Figure 16 A TEM micrograph of a HPT-consolidated Ni powder (86) indicates that the fragmented oxide films result in a significantly finer grained structure compared with the standard HPT-deformed Ni in Figures 2, 4, 7, and 15. www.annualreviews.org • Saturation of Fragmentation During SPD

337

MR40CH13-Pippan

ARI

24 May 2010

15:1

Homologous temperature, T/Tm 0.15 9

0.20

0.25

0.30

0.35

0.40

0.45

0.50

8 HPT oxidized Ni powder

Hardness (GPa)

7 6 5

HPT Ni powder 4 3

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

2 Bulk HPT Ni 1 0

0

100

200

300

400

500

600

Annealing temperature (°C) Figure 17 Illustration of the thermal stability of different HPT-processed Ni samples. The consolidated, oxidized Ni powder exhibits the highest hardness, i.e., the finest microstructure and the best thermal stability. Thus, the same mechanism restricting grain growth during thermal treatment also provides finer microstructures during SPD.

The different HPT-deformed Ni samples were then heat treated for 1 h at different temperatures to analyze the thermal stability of the obtained microstructure. Figure 17 shows the effect of annealing on hardness. The importance of introduced nanoparticles on both the obtainable hardness, or refinement, and the thermal stability is apparent. The finest microstructure has the highest thermal stability, i.e., the same mechanism, which reduces grain boundary migration during SPD and improves the microstructure’s thermal stability. Furthermore, the combination of different powders and the use of a composite as a starting material for SPD offer a completely new type of material synthesis. This type is similar to classical mechanical alloying; however, one ends up with a dense bulk material rather than with a powder. An additional advantage of the SPD process described here in generating nanocomposites, compared with powder metallurgy, is the use of standard composites or micrometer-sized powders instead of nanometer-sized powders (the former are easier to handle). The oxidation of a metal powder is the easiest way to stabilize finer microstructures and to produce a metal oxide–metal matrix composite. By the use of a carburization or coating process, almost arbitrary combinations of thin coatings and substrate metal or alloy can be achieved. Furthermore, the problems of particle cluster formation during processing, a drawback in conventional powder consolidation, can be partially avoided. Coating with fullerenes before powder compaction and SPD processing has effects similar to those of the above-mentioned oxide film (87).

CONCLUSION SPD is a technique developed in the past 20 years to refine a coarse-grained material down to the submicrometer or nanocrystalline region. At large strains, typically between equivalent strains of 5 to 30, a saturation of the refinement process is observed. We focus here on saturation during 338

Pippan et al.

MR40CH13-Pippan

ARI

24 May 2010

15:1

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

SPD, parameters controlling limits in refinement, and the responsible phenomena. The steadystate microstructure evolving in the saturation region requires equilibrium in the generation and the annihilation of defects: vacancies, dislocations, and high- and low-angle boundaries. To avoid a further size reduction of the very fine grains resulting in an increase of boundary length, grain boundary migration is the key process limiting the refinement achievable by SPD. At medium deformation temperatures, the structural restoration process performed by grain boundary migration seems to be similar to dynamic recrystallization during hot forming. At low temperatures, stressor strain-driven boundary migration becomes more and more important. The effect of alloying, deformation temperature and deformation rate, and various microstructural features support the proposed mechanism.

SUMMARY POINTS 1. Severe plastic deformation (SPD) leads to a refinement of a coarse-grained microstructure into a submicrometer and nanocrystalline microstructure. 2. At large strains, saturation in refinement is observed. Grain size, fraction of low- and high-angle grain boundaries, and texture do not change with further deformation. 3. Temperature, impurities, and alloying are the most important parameters controlling the limitation in refinement. 4. Strain rate dependency of the saturation grain size is stronger in the medium-temperature regime than at low temperatures. 5. Monotonic straining results in a finer saturation microstructure than does cyclic deformation. 6. The initial microstructure of a single-phase material does not affect the saturation grain size. A coarse-grained, single-phase material refines, and a finer nanocrystalline material coarsens. 7. Phenomena that reduce grain boundary mobility are also effective in reducing the saturation grain size obtained by SPD.

DISCLOSURE STATEMENT The authors are not aware of any affiliations, memberships, funding, or financial holdings that might be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS The financial support by the Austrian “Fonds zur Forderung der wissenschaftlichen Forschung” ¨ Projekt 10402 N16 and the Christian Doppler Forschungsgesellschaft is gratefully acknowledged.

LITERATURE CITED 1. Valiev RZ, Islamgaliev RK, Alexandrov IV. 2000. Bulk nanostructured materials from severe plastic deformation. Prog. Mater. Sci. 45:103–89 2. Valiev RZ, Islamgaliev RK, Kuzmina NF, Yong Li, Langdon TG. 1999. Strengthening and grain refinement in an Al-6061 metal matrix composite through intense plastic straining. Scr. Mater. 40:117–22 www.annualreviews.org • Saturation of Fragmentation During SPD

339

ARI

24 May 2010

15:1

3. Viswanathan V, Laha T, Balani K, Agarwal A, Seal S. 2006. Challenges and advances in nanocomposite processing techniques. Mater. Sci. Eng. R54:121–285 4. Zehetbauer MJ, Valiev RZ, eds. 2004. Nanomaterials by Severe Plastic Deformation. Weinheim: Wiley-VCH 5. Zhu YT, Langdon TG, Valiev RZ, Semiatin SL, Shin DH, Lowe TC, eds. 2004. Ultrafine Grained Materials III. Warrendale, PA: TMS 6. Zhu YT, Langdon TG, Mishra RS, Semiatin SL, Saran MJ, Lowe TC, eds. 2002. Ultrafine Grained Materials II. Warrendale, PA: TMS 7. Lowe TC, Valiev RZ, eds. Investigations and applications of SPD. Nato Sci. Ser. 3. High Technology 80 8. Zehetbauer MJ, Zhu YT, eds. 2009. Bulk Nanostructured Materials. Weinheim, Ger.: Wiley-VCH 9. Horita Z, ed. 2006. Nanomaterials by severe plastic deformation. Mater. Sci. Forum 503–504 10. Pippan R, Wetscher F, Hafok M, Vorhauer A, Sabirov I. 2006. The limits of refinement by severe plastic deformation. Adv. Eng. Mater. 8:1046–56 11. Zhilyaev AP, Langdon TG. 2008. Using high pressure torsion for metal processing: Fundamentals and applications. Progr. Mater. Sci. 53:893–979 12. Zharin VA, Shashkin DP, Yenikoponyan NS. 1984. Proc. USSR Acad. Sci., DAN SSR 278:144 13. Kuznetsov RI, Bykov VI, Chernyshev VP, Pilyugin VP, Yefremov NA, Pashegev AV. 1985. Plastic deformation of solid bodies under pressure. 1. Equipment and methods. Inst. Phys. Math., Urals Sci. Cent., Sverdlovsk, USSR, Preprint 4/85 (in Russian) 14. Bridgman PW. 1952. Studies in Large Plastic Flow and Fracture. New York: McGraw Hill 15. Vorhauer A, Pippan R. 2004. On the homogeneity of deformation by high pressure torsion. Scr. Mater. 51:921–25 16. Hebesberger T, Stuwe HP, Vorhauer, Wetscher F, Pippan R. 2005. Structure of Cu deformed by high ¨ pressure torsion. Acta Mater. 53:393–402 17. Stuwe HP. 2003. Equivalent strains in severe plastic deformation. Adv. Eng. Mater. 5:291–95 ¨ 18. Wetscher F, Vorhauer A, Pippan R. 2005. Strain hardening during high pressure torsion deformation. Mater. Sci. Eng. A 401–411:313–16 19. Wadsack R, Pippan R, Schedler B. 2002. The effect of predeformation on the ductility of chromium. J. Nucl. Mater. 307–311:701–4 20. Sabirov I, Pippan R. 2005. Formation of a W–25%Cu nanocomposite during high pressure torsion. Scr. Mater. 52:1293–98 21. Faleschini M, Kreuzer H, Kiener D, Pippan R. 2007. Fracture toughness investigations of tungsten alloys and SPD tungsten alloys. J. Nucl. Mater. 367–370:800–5 22. Ivanisenko Y, Lojkowski W, Valiev RZ, Fecht HJ. 2003. The mechanism of formation of nanostructure and dissolution of cementite in a pearlitic steel during high pressure torsion. Acta Mater. 51(10):947–52 23. Ivanisenko Y, MacLaren I, Sauvage X, Valiev RZ, Fecht HJ. 2006. Phase transformations in pearlitic steels induced by severe plastic deformation. High Press. Technol. Nanomater. 114:133–44 24. Rentenberger C, Waitz T, Karnthaler HP. 2004. HRTEM analysis of nanostructured alloys processed by severe plastic deformation. Scr. Mater. 51(8):789–94 25. Rentenberger C, Waitz T, Karnthaler HP. 2007. Formation and structures of bulk nanocrystalline intermetallic alloys studied by transmission electron microscopy. Mater. Sci. Eng. A 462(1–2):283–88 26. Concustell A, Sort J, Surinach S, Gebert A, Eckert J, et al. 2007. Severe plastic deformation of a Ti-based nanocomposite alloy studied by nanoindentation. Intermetallics 15(8):1038–45 27. Wetscher F, Pippan R, Sturm S, Kauffmann F, Scheu C, Dehm G. 2006. TEM investigations of the structural evolution in a pearlitic steel deformed by high-pressure torsion. Metall. Mater. Trans. A 60:1963– 68 28. Wetscher F, Vorhauer A, Stock R, Pippan R. 2004. Structural refinement of low alloyed steels during severe plastic deformation. Mater. Sci Eng. A 387–389:809–16 29. Harai Y, Kai M, Kaneko K, Horita Z, Langdon TG. 2008. Microstructural and mechanical characteristics of AZ61 magnesium alloy processed by high-pressure torsion. Mater. Trans. 49(1):76–83 30. Bonarski BJ, Schafler E, Mingler B, Skrotzki W, Mikulowski B, Zehetbauer MJ. 2008. Texture evolution of Mg during high-pressure torsion. J. Mater. Sci. 43(23–24):7513–18 31. Revesz A, Schafler E, Kovacs Z. 2008. Structural anisotropy in a Zr57 Ti5 Cu20 Al10 Ni8 bulk metallic glass deformed by high pressure torsion at room temperature. Appl. Phys. Lett. 92:011910

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

MR40CH13-Pippan

340

Pippan et al.

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

MR40CH13-Pippan

ARI

24 May 2010

15:1

32. Meyer LW, Hockauf M, Hohenwarter A, Schneider S. 2008. Ultimate strength of tungsten heavy alloy after severe plastic deformation at quasi static and dynamic loading. Mater. Sci. Forum 584–586:405–10 33. Vorhauer A, Pippan R. 2008. On the onset of a steady state in body-centered cubic iron during severe plastic deformation at low homologous temperatures. Metall. Mater. Trans. A 39:417–29 34. Wetscher F, Vorhauer A, Pippan R. 2005. Strain hardening during high pressure torsion deformation. Mater. Sci. Eng. A 410–411:213–16 35. Wetscher F, Pippan R. 2006. Cyclic high-pressure torsion of nickel and Armco iron. Philos. Mag. 86(36):5867–83 36. Wetscher F, Pippan R. 2009. Hardening and softening behavior by cyclic high-pressure torsion. Metall. Mater. Trans. A 40:3258–63 37. Mizunuma S. 2006. Large straining behavior and microstructure refinement of several metals by torsion extrusion process. Nanomater. Severe Plast. Deform. 503–504:185–90 38. Pippan R. 2009. High pressure torsion: features and applications. In Bulk Nanostructured Materials, ed. M Zehetbauer, YT Zhu, pp. 217–33. Weinheim: Wiley-VCH 39. Hansen N, Mehl FR. 2001. New discoveries in deformed metals. Metall. Mater. Trans. A 32:2917–35 40. Humphreys FJ, Hatherly M, eds. 1995. Recrystallization and Related Annealing Phenomena. Oxford, UK: Pergamon 41. Hansen N, Kuhlmann-Wilsdorf D. 1986. Low energy dislocation structures due to unidirectional deformation at low temperatures. Mater. Sci. Eng. 81:141–61 42. Winther G, Huang X, Hansen N. 2000. Crystallographic and macroscopic orientation of planar dislocation boundaries: correlation with grain orientation. Acta Mater 48(9):2187–98 43. Kocks UF, Mecking H. 2003. Physics and phenomenology of strain hardening: the FCC case. Prog. Mater. Sci. 48:171–273 44. Hafok M, Pippan R. 2008. High-pressure torsion applied to nickel single crystals. Philos. Mag. 88(12):1857– 77 45. Zhang HW, Huang X, Hansen N. 2008. Evolution of microstructural parameters and flow stresses toward limits in nickel deformed to ultrahigh strains. Acta Mater. 56(19):5451–65 46. Schafler E, Pippan R. 2004. Effect of thermal treatment on microstructure in high pressure torsion (HPT) deformed nickel. Mater. Sci. Eng. A 387–389:799–804 47. Tatschl A, Kolednik O. 2003. On the experimental characterization of crystal plasticity in polycrystals. Mater. Sci. Eng. A 342:152–68 48. Sachtleber M, Zhao Z, Raabe D. 2002. Experimental investigation of plastic grain interaction. Mater. Sci. Eng. A 336:81–87 49. Musienko A, Tatschl A, Schmidegg K, Kolednik O, Pippan R, Cailletaud G. 2007. Three-dimensional finite element simulation of a polycrystalline copper specimen. Acta Mater. 55:4121–36 50. Yang B, Vehoff H, Hohenwarter A, Hafok M, Pippan R. 2008. Strain effects on the coarsening and softening of electrodeposited nanocrystalline Ni subjected to high pressure torsion. Scr. Mater. 58:790–93 51. Hafok M, Vorhauer A, Keckes J, Pippan R. 2006. HPT-deformation of copper and nickel single crystals. Mater. Sci. Forum 503–504:621–26 52. Oh-ishi K, Horita Z, Smitz DJ, Valiev RZ, Nemoto M, Langdon TG. 1999. Fabrication and thermal stability of a nanocrystalline Ni-Al-Cr alloy: comparison with pure Cu and Ni. J. Mater. Sci. 14(11):4200– 7 53. Bachmaier A, Hafok M, Pippan R. 2010. Rate independent and rate dependent structural evolution during severe plastic deformation. JIM Mater. Trans. 51(1)8–13 54. Vorhauer A, Hebesberger T, Pippan R. 2003. Disorientations as a function of distance: a new procedure to analyze local orientation data. Acta Mater. 51:677–86 55. Kozlov EV, Zhdanov AN, Popova NA, Pekarskaya EE, Korenva NA. 2004. Subgrain structure and internal stress fields in UFG materials: problem of Hall-Petch relation. Mater. Sci. Eng. A 387–389:789–94 56. Xu C, Furukawa M, Horita Z, Langdon TG. 2003. Using ECAP to achieve grain refinement, precipitate fragmentation and high strain rate superplasticity in a spray-cast aluminum alloy. Acta Mater. 51:6139–49 57. Huang Y, Langdon TG. 2003. Using atomic force microscopy to evaluate the development of mesoscopic shear planes in materials processed by severe plastic deformation. Mater. Sci. Eng. A 358:114–21 www.annualreviews.org • Saturation of Fragmentation During SPD

341

ARI

24 May 2010

15:1

58. Valiev RZ, Ivanishenko YV, Rauch EF, Baudelet B. 1996. Structure and deformation behavior of Armco iron subjected to severe plastic deformation. Acta Mater. 44:4705–12 59. Takaki S, Kawasaki K, Kimura Y. 2001. Mechanical properties of ultra fine grained steels. Mater. Process. Technol. 117:359–63 60. Hafok M, Pippan R. 2007. Shear deformation of submicron structured materials. J. Mater. Res. 98:283–89 61. Hafok M, Pippan R. 2007. Post-shear deformation of high pressure torsion-deformed nickel under hydrostatic pressure. Scr. Mater. 56:757–60 62. Deleted in proof 63. Gourdet S, Montheillet F. 2000. An experimental study of the recrystallization mechanism during hot deformation of aluminum. Mater. Sci. Eng. A 283(1–2):274–88 64. McQueen HJ, Evangelista E, Kassner ME. 1991. Classification and determination of restoration mechanisms in the hot working of Al alloys. Z. Metallkd. 82:336–45 65. Belyakov A, Sakai T, Miura H, Kaibyshev R, Tsuzaki K. 2002. Continuous recrystallization in austenitic stainless steel after large strain deformation. Acta Mater. 50:1547–57 66. Kassner ME, McQueen HJ, Evangelista E. 1993. Geometric dynamic recrystallization in aluminum and aluminum alloys above 0.6Tm. Mater. Sci. Forum 113–115:151–56 67. Drury MR, Humphreys FJ. 1986. The development of microstructure in Al-5% Mg during high temperature deformation. Acta Metall. 34:2259–71 68. Montheillet F. 2004. Moving grain boundaries during hot deformation of metals: dynamic recrystallization. In Moving Interfaces in Crystalline Solids (CISM Courses and Lectures No. 453), ed. FD Fischer, pp. 203–56. New York: Springer 69. Van Swygenhoven H, Derlet PM, Budrovic Z, Hasnaoui A. 2003. Unconventional deformation mechanism in nanocrystalline metals? Z. Metall. 94(10):1106–10 70. Wolf D, Yamakov V. 2009. MD simulation of deformation mechanisms in nanocrystalline materials. In Bulk Nanostructured Materials, ed. MJ Zehetbauer, YT Zhu, pp. 165–202. Weinheim, Ger.: Wiley-VCH 71. Brandstetter S, Zhang K, Escuadro A, Weertman JR, Van Swygenhoven H. 2008. Grain coarsening during compression of bulk nanocrystalline nickel and copper. Scr. Mater. 58(1):61–64 72. Mompiou F, Caillard D, Legros M. 2009. Grain boundary shear-migration coupling. I. In situ TEM straining experiments in Al polycrystals. Acta Mater. 57:2198–209 73. Caillard D, Mompiou F, Legros M. 2009. Grain boundary shear-migration coupling. II. Geometrical model for general boundaries. Acta Mater. 57:2390–402 74. Mughrabi H, Hoppel H, Kautz M, Valiev RZ. 2003. Annealing treatments to enhance thermal and mechanical stability of ultrafine-grained metals produced by severe plastic deformation. Z. Metall. 94:1079–83 75. Mughrabi H, Hoppel H, Kautz M. 2004. Fatigue and microstructure of ultrafine-grained metals produced by severe plastic deformation. Scr. Mater. 51:807–12 76. Winning M, Gottstein G, Shvindlerman LS. 2002. On the mechanisms of grain boundary migration. Acta Mater. 50:353–63 77. Zhang HW, Huang X, Hansen N. 2008. Evolution of microstructural parameters and flow stresses toward limits in nickel deformed to ultrahigh strains. Acta Mater. 56:5451–65 78. Hafok M. 2007. Aufkl¨arung der Verformungsvorg¨ange bei der Hochverformung. Dissertation, Montanuniv., Leoben 79. Valiev RZ, Langdon TG. 2006. Principles of equal-channel angular pressing as a processing tool for grain refinement. Prog. Mater. Sci. 51:881–981 80. Richert M, Stuwe HP, Zehetbauer MJ, Richert J, Pippan R, et al. 2003. Work hardening and microstruc¨ ture of AlMg5 after severe plastic deformation by cyclic extrusion and compression. Mater. Sci. Eng. A 355:180–85 81. Mironov SY, Salishchev GA, Myshlyaev MM, Pippan R. 2006. Evolution of misorientation distribution during warm “abc” forging of commercial-purity titanium. Mater. Sci. Eng. A 418:257–67 82. Dinda GP, Rosner H, Wilde G. 2005. Synthesis of bulk nanostructured Ni, Ti and Zr by repeated cold-rolling. Scr. Mater. 52(7):577–82 83. Sauvage X, Jessner P, Vurpillot F, Pippan R. 2008. Nanostructure and properties of a Cu-Cr composite processed by severe plastic deformation. Scr. Mater. 58:1125–28

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

MR40CH13-Pippan

342

Pippan et al.

MR40CH13-Pippan

ARI

24 May 2010

15:1

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

84. Sabirov I, Kolednik O, Pippan R. 2005. Homogenization of metal matrix composites by high-pressure torsion. Metall. Mater. Trans. A 36:2861–70 85. Sabirov I, Kolednik O, Valiev RZ, Pippan R. 2005. Equal channel angular pressing of metal matrix composites: effect on particle distribution and fracture toughness. Acta Mater. 53:4919–30 86. Bachmaier A, Hohenwarter A, Pippan R. 2009. New procedure to generate stable nanocrystallites by severe plastic deformation. Scr. Mater. 61:1016–19 87. Tokunaga T, Kaneko K, Sato K, Horita Z, et al. 2008. Microstructure and mechanical properties of aluminum-fullerene composite fabricated by high pressure torsion. Scr. Mater. 58(9):735–38

www.annualreviews.org • Saturation of Fragmentation During SPD

343

AR416-FM

ARI

3 June 2010

22:44

Annual Review of Materials Research

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

Contents

Volume 40, 2010

New Developments in Composite Materials Biological Composites John W.C. Dunlop and Peter Fratzl p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1 On the Mechanistic Origins of Toughness in Bone Maximilien E. Launey, Markus J. Buehler, and Robert O. Ritchie p p p p p p p p p p p p p p p p p p p p p p p p25 Teeth: Among Nature’s Most Durable Biocomposites Brian R. Lawn, James J.-W. Lee, and Herzl Chai p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p55 Mechanical Principles of Biological Nanocomposites Baohua Ji and Huajian Gao p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p77 Optimal Design of Heterogeneous Materials S. Torquato p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 101 Physical Properties of Composites Near Percolation C.-W. Nan, Y. Shen, and Jing Ma p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 131 Magnetoelectric Composites G. Srinivasan p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 153 Self-Healing Polymers and Composites B.J. Blaiszik, S.L.B. Kramer, S.C. Olugebefola, J.S. Moore, N.R. Sottos, and S.R. White p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 179 Steel-Based Composites: Driving Forces and Classifications David Embury and Olivier Bouaziz p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 213 Metal Matrix Composites Andreas Mortensen and Javier Llorca p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 243 Current Interest The Indentation Size Effect: A Critical Examination of Experimental Observations and Mechanistic Interpretations George M. Pharr, Erik G. Herbert, and Yanfei Gao p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 271

vii

AR416-FM

ARI

3 June 2010

22:44

Plasticity in Confined Dimensions Oliver Kraft, Patric A. Gruber, Reiner M¨onig, and Daniel Weygand p p p p p p p p p p p p p p p p p p p 293 Saturation of Fragmentation During Severe Plastic Deformation R. Pippan, S. Scheriau, A. Taylor, M. Hafok, A. Hohenwarter, and A. Bachmaier p p p p p 319 Ultrasonic Fabrication of Metallic Nanomaterials and Nanoalloys Dmitry G. Shchukin, Darya Radziuk, and Helmuth M¨ohwald p p p p p p p p p p p p p p p p p p p p p p p p p p 345

Annu. Rev. Mater. Res. 2010.40:319-343. Downloaded from www.annualreviews.org by University of Wisconsin - Madison on 09/01/12. For personal use only.

Oxide Thermoelectric Materials: A Nanostructuring Approach Kunihito Koumoto, Yifeng Wang, Ruizhi Zhang, Atsuko Kosuga, and Ryoji Funahashi p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 363 Inkjet Printing of Functional and Structural Materials: Fluid Property Requirements, Feature Stability, and Resolution Brian Derby p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 395 Microfluidic Synthesis of Polymer and Inorganic Particulate Materials Jai Il Park, Amir Saffari, Sandeep Kumar, Axel Gunther, ¨ and Eugenia Kumacheva p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 415 Current-Activated, Pressure-Assisted Densification of Materials J.E. Garay p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 445 Heterogeneous Integration of Compound Semiconductors Oussama Moutanabbir and Ulrich G¨osele p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 469 Electrochemically Driven Phase Transitions in Insertion Electrodes for Lithium-Ion Batteries: Examples in Lithium Metal Phosphate Olivines Ming Tang, W. Craig Carter, and Yet-Ming Chiang p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 501 Electromigration and Thermomigration in Pb-Free Flip-Chip Solder Joints Chih Chen, H.M. Tong, and K.N. Tu p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 531 The Structure of Grain Boundaries in Strontium Titanate: Theory, Simulation, and Electron Microscopy Sebastian von Alfthan, Nicole A. Benedek, Lin Chen, Alvin Chua, David Cockayne, Karleen J. Dudeck, Christian Els¨asser, Michael W. Finnis, Christoph T. Koch, Behnaz Rahmati, Manfred Ruhle, ¨ Shao-Ju Shih, and Adrian P. Sutton p p p p p p p p p p p p p p 557 Index Cumulative Index of Contributing Authors, Volumes 36–40 p p p p p p p p p p p p p p p p p p p p p p p p p p p 601 Errata An online log of corrections to Annual Review of Materials Research articles may be found at http://matsci.annualreviews.org/errata.shtml

viii

Contents