Metallurgical and Materials Transactions A

, Volume 38, Issue 11, pp 2815–2824

Effects of Degree of Deformation and Deformation Temperature on Primary Recrystallization Textures in Polycrystalline Nickel

Authors

  • H. Chang
    • Thayer School of EngineeringDartmouth College
    • Materials Science for Continental Teves
    • Thayer School of EngineeringDartmouth College
Article

DOI: 10.1007/s11661-007-9324-1

Cite this article as:
Chang, H. & Baker, I. Metall and Mat Trans A (2007) 38: 2815. doi:10.1007/s11661-007-9324-1

Abstract

The effects of both deformation temperature and degree of deformation on the deformation texture, recrystallization behavior, and recrystallization texture were studied for cold-rolled, high-purity, polycrystalline nickel. Differential scanning calorimetry was used to determine both the stored energy of deformation and the recrystallization temperature, and electron backscatter patterns were employed to reveal both the deformation and recrystallization textures of nickel rolled to either 90 pct thickness reduction at −196 °C, 25 °C, and 200 °C or rolled to 90, 95, and 98 pct thickness reductions at 25 °C. The results show that decreasing the rolling temperature below room temperature increased the stored energy considerably and decreased the recrystallization temperature, whereas increasing the rolling temperature had no effect on either the stored energy or the recrystallization temperature. These different rolling temperatures had little effect on the cube texture produced by primary recrystallization. In contrast, increasing the rolling reduction increased the stored energy (at least from 90 to 95 pct), decreased the primary recrystallization temperature, and also sharpened the primary-recrystallized cube texture.

1 Introduction

The cube texture, i.e., {001}<100>, is a major primary recrystallization component in many fcc metals. The origin of this texture has long been debated, with discussions focused on two alternative models, oriented growth and oriented nucleation.[13] According to the oriented growth model, the growth of cube grains is dominant because they have the maximum boundary mobility with respect to the major rolling texture components. In the oriented nucleation model, regions in the rolled matrix with cube orientation are favorable nuclei of new grains.

There has been considerable recent interest in the production of nickel strips or tapes with sharp cube textures as substrates for growing high-temperature YBCO superconductors.[610] Such texture control of the nickel substrate and, hence, of the sputter-deposited superconducting oxides has allowed the production of polycrystalline tapes with {001}<100> textures that contain mainly low-angle boundaries. The critical current densities of such textured tapes are significantly higher than those in randomly oriented polycrystalline material.

A previous study[4,5] on the directional recrystallization of 90 pct cold-rolled, 99.5 wt pct purity, polycrystalline nickel suggested that a strong primary recrystallization texture—for nickel, the cube texture—is necessary for growing single crystals or columnar grains upon directional annealing at secondary recrystallization temperatures. The growth mechanism of the columnar grains or single crystals was also discussed.[5] There are a variety of factors including deformation temperature, degree of deformation, and annealing temperature that affect the deformation and recrystallization behavior of nickel.

In this article, we present the results of an investigation into the effects of both the degree of deformation and the deformation temperature on the deformation texture, stored energy of deformation, recrystallization kinetics, and recrystallization textures of high-purity nickel.

2 Experimental methods

Polycrystalline nickel of 99.995 wt pct purity was purchased from Alfa Aesar (Ward Hill, MA) in the form of 12-mm-diameter rods. The major impurities were 24 ppm C, 34 ppm Fe, and 2 ppm S. The microstructure of the as-received nickel is shown in Figure 1. Both the cross section (Figure 1(a)) and the longitudinal section (Figure 1(b)) of the material showed a nonuniform microstructure composed of a mixture of small (∼20 μm) and large (≤100 μm) grains. The {100} pole figure from the cross section (X-Y plane) of the nickel in its as-received state is shown in Figure 1(c). Note that all the pole figures have a linear scale for the contour lines with grayscale shadings indicated, and the maximum intensity on each pole figure is shown. The result indicates a fiberlike texture with major components of {111} and {100} parallel to the normal (Z) direction.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-007-9324-1/MediaObjects/11661_2007_9324_Fig1_HTML.gif
Fig. 1

Microstructures of the (a) cross section (X-Y plane) and (b) longitudinal section (Y-Z plane), and (c) the {100} pole figure from the cross section (X-Y plane) of the as-received high-purity nickel specimen. (Contour lines are shown with gray scales to indicate the relative densities. Contour lines = 1, 2, 3, 4, 5, and max = 7.45.) A schematic of the examined specimen surfaces is also shown

The nickel rods were machined into bars with a cross section of ∼9 × 9 mm. In order to study the effects of the deformation temperature, the bars were rolled to 90 pct thickness reduction at −196 °C, 25 °C, and 200 °C. For rolling at −196 °C, the specimen was cooled thoroughly in liquid nitrogen before and after each rolling pass. For rolling at 200 °C, the specimen was heated in an air furnace to 200 °C before and after each rolling pass. After obtaining the desired temperature in liquid nitrogen or the air furnace, the specimen was immediately rolled at room temperature. The time from taking the specimen out of the liquid nitrogen or from the furnace through rolling the specimen was less than 5 seconds. The rate of change of the specimen temperature during this short period can be estimated from[11]
$$ \frac{{dT}} {{dt}} = - \frac{{hA}} {{c\rho V}}(T_{i} - T_{r} ) $$
where Ti is the initial temperature of the specimen, Tr is the room temperature, A is the surface area of the specimen (typically 3 to 30 × 10−4 m2), V is the volume of the specimen (typically 2 to 3 × 10−6 m3), h is the convection heat-transfer coefficient (∼10 W m−2 K−1 for natural convection in air), c is the specific heat capacity of nickel (444 J kg−1 K−1), and ρ is the mass density of nickel (8908 kg m−3). The temperature change in a specimen with thickness ∼0.9 mm (∼90 pct thickness reduction) was calculated based on the preceding equation. The results show that the specimen that was initially equilibrium at 200 °C or −196 °C has a temperature change rate of less than 1.5 °C/s, which is negligible. Of course, some heat is also lost/gained to/from the oil covered rolls during the rolling, but this is probably also small because rolling took less than a second.

To study the effects of the degree of deformation, the bars were rolled to 90, 95, and 98 pct thickness reductions at 25 °C.

Disk specimens for differential scanning calorimetry (DSC) measurements were cut parallel to the rolling plane from the rolled strips using an electrodischarge machine and then heavily etched in a solution of 40 pct nitric acid and 40 pct acetic acid in water to remove any surface damage. The recrystallization temperatures and stored energies of deformation of the rolled specimens were measured using a Perkin-Elmer (Waltham, MA) DSC7 operating in continuous heating mode at a heating rate of 20 °C/min. For each specimen condition, at least two DSC measurements were performed.

Rolled specimens were isothermally annealed for 1 hour in air at their primary recrystallization temperatures, as determined using the DSC. Specimens for microstructural examination by optical microscopy were mechanically polished using 0.05 μm alumina powder and then etched in the nitric acid/acetic acid solution noted previously. Specimens for texture determination by electron backscatter patterns (EBSPs) were electropolished in 10 pct sulfuric acid in methanol, at a voltage of ∼14 V, and a current density of ∼65 mA/cm2 at ∼0 °C. The textures of the as-received, rolled, and primary recrystallized specimens were determined using EBSPs with HKL technology (HKL Technology, Concord, MA). The orientation imaging system is attached to an FEI XL30 (FEI Company, Hillsboro, OR) field emission gun scanning electron microscope operated at 30 keV. Typically, for the rolled specimens, the area scanned was from ∼600 × 600 μm2 to ∼1 × 1 mm2, while for the primarily recrystallized nickel specimens, the area scanned was from ∼500 × 500 μm2 to 745 × 745 μm2. The typical grid spacing used in both cases was 2 to 4 μm. For a specimen with an average grain size of 20 μm, at least 800 grains were scanned. For each specimen, the scanned areas were randomly selected. From the EBSP data, the orientation distribution function (ODF) was calculated using the series expansion method[12] with a highest harmonic lmax = 22 by the HKL Channel 5 software. The maximum values shown in the ODF maps are the peak densities directly observed in these maps. The area fraction of the relevant texture component was calculated from the orientation maps constructed from the EBSP data with maximum deviation angle of 10 deg from the ideal orientation. Because not only the maximum density values, but also the scattered width of the peaks, were taken into account, the area fractions of different components with similar peak densities in the ODF maps may differ substantially from each other.

3 Results

3.1 Effects of Temperature

Figure 2 shows the {100} pole figures of nickel specimens rolled to 90 pct thickness reduction at −196 °C, 25 °C, and 200 °C. It is clear from these pole figures that the rolling textures of nickel specimens rolled at different temperatures were similar and can be described as pure-metal type with three major orientation components, {112}<111> (C-copper), {123}<634> (S), and {110}<112> (B-brass). Figure 3 shows composite texture plots made from Φ = 0, 45, 60, and 65 deg sections of the corresponding ODF maps of specimens rolled at −196 °C, 25 °C, and 200 °C. For specimens rolled at 25 °C, the peak intensity of the B component appears to be slightly higher compared to the values obtained for specimens rolled at −196 °C and 200 °C. The area fraction of each rolling component as a function of rolling temperatures is shown in Figure 4. In all specimens, the components close to the S orientation were strongest. Increasing or decreasing the rolling temperature from 25 °C appears to have decreased the strength of B components and also slightly decreased the strength of the S components. It did not significantly affect the strength of C components. However, any effect of temperature on the rolling texture was small and probably within the measurement error.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-007-9324-1/MediaObjects/11661_2007_9324_Fig2_HTML.gif
Fig. 2

{111} pole figures of high-purity nickel 90 pct rolled at (a) −196 °C, (b) 25 °C, and (c) 200 °C. (Contour lines are shown with gray scales to indicate the relative densities. Contour = 1, 2, 3, and max = 3.98; 1, 2, 3, 4, and max = 4.65; and 1, 2, 3, and max = 3.37 for (a), (b) and (c), respectively.)

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-007-9324-1/MediaObjects/11661_2007_9324_Fig3_HTML.gif
Fig. 3

Comparison of relevant sections (Φ2 = 0, 45, 60, and 65 deg) from the ODF maps for high-purity nickel 90 pct rolled at −196 °C, 25 °C, and 200 °C. The smaller font numbers indicate the values of the relative contour lines, while the large font numbers indicate the peak density read from the ODF maps

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-007-9324-1/MediaObjects/11661_2007_9324_Fig4_HTML.gif
Fig. 4

Variation of area fraction of rolling components in the high-purity nickel 90 pct rolled at −196 °C, 25 °C, and 200 °C

Table I summarizes the DSC measurements of the specimens after 90 pct rolling reduction at temperatures of −196 °C, 25 °C, and 200 °C. The stored energy for the specimen rolled at −196 °C of 123.2 ± 2.7 J/mol was much greater than that of the specimen rolled at 25 °C of 47.1 ± 0.1 J/mol, with a concomitant reduction in the primary recrystallization temperature from 393.3 ± 1.3 °C for the 25 °C specimen to 370.6 ± 1.0 °C for the −196 °C specimen. This increase in stored energy is presumably caused by the prevention of dislocation rearrangement and recovery at low temperatures. The reduction in recrystallization temperature at the low-temperature rolled specimen is a consequence of the higher stored energy.[13] In contrast, rolling at 200 °C had no marked effect on the stored energy of 48.4 ± 0.9 J/mol and recrystallization temperature of 398.7 ± 0.6 °C compared with rolling at 25 °C.
Table I

Average Stored Energies and Primary Recrystallization Temperatures of Nickel Specimens Rolled under Different Rolling Conditions

Thickness Reduction Upon Rolling (Pct)

Rolling Temperature (°C)

Average Primary Recrystallization Temperature (°C)

Average Stored Energy (J/mol)

90

–196

370.6 ± 1.0

123.2 ± 2.7

90

200

398.7 ± 0.6

48.4 ± 0.9

90

25

393.3 ± 1.3

47.1 ± 0.1

95

25

384.6 ± 0.8

58.3 ± 0.5

98

25

378.8 ± 3.0

56.5 ± 1.9

All rolled specimens were fully recrystallized after annealing at their primary recrystallization temperatures (which were similar) for 1 hour, as shown by the optical microstructure images in Figure 5. The grain size after primary recrystallization is ∼20.2 μm for the nickel specimen rolled at −196 °C, ∼21.3 μm for the nickel specimen rolled at 25 °C and ∼22.3 μm for the nickel specimen rolled at 200 °C, i.e., essentially the same within measurement error. The primary recrystallization textures of the nickel specimens that were rolled at different temperatures are shown in {100} pole figures (Figure 6). For the specimen rolled at 25 °C, after primary recrystallization, a texture composed of both cube and other orientations was formed, which can be seen in both Figure 6(b) and Figure 7. One noncube orientation was found to be {013}<100>, which has an 18 deg/<100> relationship with the cube texture (for example, note that in the section of Φ2 = 0 deg, the spread of the contour lines shows a rotation of about 18 deg from the ideal cube orientation with respect to the <100> axis). The other noncube orientations were read from the ODF map to be (30, 51, 25), (45, 71, 43), and (65, 44, 60). The Miller indices transformed from these Euler angles are (122) \( [2\ifmmode\expandafter\bar\else\expandafter\=\fi{2}1] \), (221)\( [1\ifmmode\expandafter\bar\else\expandafter\=\fi{2}2] \), and (212)\( [\ifmmode\expandafter\bar\else\expandafter\=\fi{1}\ifmmode\expandafter\bar\else\expandafter\=\fi{2}2] \), which correspond to annealing twins of the form {122}<221>.[5,6] For the specimen rolled at −196 °C, a similar primary recrystallization texture was observed, as shown in Figures 6(a) and 7. Three major components—cube, {013}<100> and {122}<221> annealing twins—were observed. However, the cube component was found to be weaker than that observed in the primary-recrystallized specimen rolled at 25 °C. For the specimen rolled at 200 °C, a fairly weak primary recrystallization texture was produced, as shown in Figures 6(c) and 7.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-007-9324-1/MediaObjects/11661_2007_9324_Fig5_HTML.jpg
Fig. 5

Microstructures of high-purity nickel 90 pct rolled at (a) −196 °C, (b) 25 °C, and (c) 200 °C that were isothermally annealed at their primary recrystallization temperatures for 1 h

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-007-9324-1/MediaObjects/11661_2007_9324_Fig6_HTML.gif
Fig. 6

{100} pole figures of high-purity nickel 90 pct rolled at (a) −196 °C, (b) 25 °C, and (c) 200 °C that were isothermally annealed at their primary recrystallization temperatures for 1 h. (Contour lines are shown with grayscale shadings to indicate the relative densities. Contour lines = 1, 2, 3, 4, 5, and max = 5.13; 1, 2, 4, 6, 9, and max = 9.47; and 1, 2, 3, and max = 3.09 for (a), (b), and (c), respectively.)

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-007-9324-1/MediaObjects/11661_2007_9324_Fig7_HTML.gif
Fig. 7

Sections of Φ2 = 0 and 65 deg from ODF maps for high-purity nickel 90 pct rolled at –196 °C, 25 °C, and 200 °C that were isothermally annealed at their primary recrystallization temperatures for 1 h. The values of the relative contour lines are indicated

3.2 Effects of Thickness Reduction

Figure 8 shows the {100} pole figures of nickel rolled to 90, 95, and 98 pct thickness reduction at 25 °C. Even though all specimens showed similar rolling textures containing three major components––{112}<111> (C-copper), {123}<634> (S), and {110}<112> (B-brass)–– some quantitative differences were again observed. Figure 9 shows the composite texture plots made from Φ = 0, 45, 60, and 65 deg sections of the corresponding ODFs of these specimens. It is clear that, as the rolling reduction increased, the peak densities of both C and S components were increased, while the peak density for B components was slightly decreased. As can be seen in Figure 10, although the relative frequency of B decreased, the sum of the relative frequencies of the S, B, and C components increased as the rolling reduction was increased, which means that increasing the thickness reduction of rolling resulted in a strengthening of the rolling texture. Similar results were reported in a study on both initially coarse-grained and initially fine-grained nickel with >99.9 wt pct purity, cold-rolled to the same thickness reductions,[14] where it was shown that the total volume fractions of the S, B, and C components increased continuously as the thickness reduction of cold rolling was increased from 95 to 99 pct.[14]
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-007-9324-1/MediaObjects/11661_2007_9324_Fig8_HTML.gif
Fig. 8

{100} pole figures of high-purity nickel cold-rolled to (a) 90 pct, (b) 95 pct, and (c) 98 pct thickness reduction at 25 °C. (Contour lines are shown with grayscale shadings to indicate the relative densities. Contours = 1, 2, 3, 4, and max = 4.65; 1, 2, 3, 4, and max = 4.76; and 1, 2, 3, 4, and max = 4.92 for (a), (b), and (c), respectively.)

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-007-9324-1/MediaObjects/11661_2007_9324_Fig9_HTML.gif
Fig. 9

Comparison of relevant sections (Φ2 = 0, 45, 60, and 65 deg) from the ODF maps for high-purity nickel 90, 95, and 98 pct cold-rolled at 25 °C. The smaller font numbers indicate the values of the relative contour lines, while the large font number indicates the peak density of each rolling component

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-007-9324-1/MediaObjects/11661_2007_9324_Fig10_HTML.gif
Fig. 10

Variation of the area fractions of rolling components in high-purity nickel 90, 95, and 98 pct cold-rolled at 25 °C

Table I summarizes the DSC measurements of the nickel specimens after 90, 95, and 98 pct rolling reduction at 25 °C. As the rolling reduction was increased from 90 to 95 pct, the stored energy increased from 47.1 ± 0.1 J/mol to 58.3 ± 0.5 J/mol, and the primary recrystallization temperature decreased from 393.3 ± 1.3 °C to 384.6 ± 0.8 °C. A further increase of the thickness reduction to 98 pct further decreased the primary recrystallization temperature to 378.8 ± 3.0 °C, while having no significant effect on the stored energy of 56.5 ± 1.9 J/mol. In an earlier article, we also measured the stored energy and recrystallization temperature of high-purity cold-rolled nickel at similar rolling reductions.[15] In that study, the average stored energy and recrystallization temperature of the 90 pct cold-rolled specimens were (77.2 ± 3.9) J/mol and (354.5 ± 2.3) °C, respectively, while the average stored energy and recrystallization temperature of the 95 pct cold-rolled specimens were (88.0 ± 5.9) J/mol and (336.5 ± 2.1) °C, and those of the 97.5 pct cold-rolled specimens were (92.5 ± 4.9) J/mol and (335.4 ± 0.7) °C. Although the two different sets of nickel had essentially the same impurity levels and similar textures,[15] the grain structures were quite different. In the earlier study, the grains were elongated along the longitudinal direction of the rod due to the swaging, and it appeared that the as-received material probably was not fully recrystallized.

Again, all rolled specimens were fully recrystallized after annealing at their primary recrystallization temperatures for 1 hour. This is demonstrated in Figure 11, where the optical microstructure images show grains that are equiaxed from recrystallization rather than elongated due to rolling. The grain size after primary recrystallization is ∼21.3 μm for the nickel specimen that was 90 pct rolled, ∼20.4 μm for the nickel specimen that was 95 pct rolled, and ∼23.4 μm for the nickel specimen that was 98 pct rolled, which is probably within the measurement error. The primary recrystallization textures of the nickel specimens rolled to different thickness reductions are shown in both {100} pole figures (Figure 12) and ODF maps (Figure 13). For the 90 pct rolled specimen, as discussed earlier, a texture composed of three components––cube, {013}<100>, and {122}<221> annealing twins––was formed, as shown in Figures 12(a) and 13. As the rolling reduction was increased, the texture components associated with annealing twins decreased, while the cube texture strengthened (Figures 12(b), (c), and 13). It is worth noting that even at the highest rolling reduction of 98 pct, the {013}<100> texture was still observed (Figure 13). In summary, it has been found that increasing the thickness reduction of cold rolling greatly contributed to the strengthening of the primary recrystallized cube texture.
https://static-content.springer.com/image/art%3A10.1007%2Fs11661-007-9324-1/MediaObjects/11661_2007_9324_Fig11_HTML.jpg
Fig. 11

Microstructures of high-purity nickel: (a) 90 pct, (b) 95 pct, and (c) 98 pct cold-rolled at 25 °C that were isothermally annealed at their primary recrystallization temperatures for 1 h

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-007-9324-1/MediaObjects/11661_2007_9324_Fig12_HTML.gif
Fig. 12

{100} pole figures of high-purity nickel: (a) 90 pct, (b) 95 pct, and (c) 98 pct cold-rolled at 25 °C that were isothermally annealed at their primary recrystallization temperatures for 1 h. (Contour lines are shown with grayscale shadings to indicate the relative densities. Contour lines = 1, 2, 4, 6, 9, and max = 9.47; 1, 2, 4, 6, 8, 12, and max = 12.23; and 1, 2, 4, 6, 8, 10, 14, 18, and max = 18.68 for (a), (b) and (c), respectively.)

https://static-content.springer.com/image/art%3A10.1007%2Fs11661-007-9324-1/MediaObjects/11661_2007_9324_Fig13_HTML.gif
Fig. 13

Sections of Φ2 = 0 and 65 deg from the ODF maps for high-purity nickel: 90, 95, and 98 pct cold-rolled at 25 °C that were isothermally annealed at their primary recrystallization temperatures for 1 h. The values of the relative contour line are indicated

4 Discussion

The temperature of deformation has long been recognized as an important factor affecting the deformation structure. It is well known that the stored energy of deformation is generally associated with the number of defects, such as dislocations, vacancies, and stacking faults, produced by deformation.[16] The dislocation densities within metallic specimens deformed at temperatures below 25 °C have been found to be much higher, and the dislocation distributions more uniform, than in specimens deformed at 25 °C.[10,17] Similar findings for various metals, including copper, aluminum, and iron, have been noted by Cotterill and Mould.[16] According to their review,[16] the stored energy of deformation at −196 °C was always found to be much higher than that for a metal with the same purity deformed to a similar strain at 25 °C. Ginden et al. studied recrystallization in 99.98 pct purity, polycrystalline copper cold-rolled at temperatures of −269 °C, −253 °C, −196 °C, and 23 °C.[1821] Their results showed that both the primary recrystallization temperatures and the activation energy of recrystallization decreased as the deformation temperatures were decreased.

A reasonable explanation of the preceding observations is that thermally activated processes such as dynamic recovery are inhibited by the low deformation temperature, making the annihilation of dislocations more difficult during the rolling process and, thus, contributing to an increase in the stored energy. The present work on 90 pct rolled high-purity nickel showed similar results. The lower primary recrystallization temperature and higher stored energy of nickel 90 pct rolled at −196 °C, compared to those rolled to the same reduction at 25 °C, is presumably due to a higher dislocation density. The fact that the stored energy of cold work and the recrystallization temperature of the specimens cold-rolled at 25 °C and 200 °C are the same indicates that the increase in temperature over this range, from 0.17 to 0.27 Tm, where Tm is the melting temperature of nickel, produces little effect on dislocation recovery processes compared to room temperature.

Based on the preceding observations, one may expect that the deformation temperatures will also affect the recrystallization textures. However, the results of the present work indicate that lowering the deformation temperature does not strengthen the primary-recrystallized cube texture. This is presumably because the deformation temperature had little effect on the texture of the rolled nickel from which the primary recrystallization texture was derived.

The formation of a cube texture in fcc materials depends strongly on the rolling texture. It is commonly accepted that a sharp cube recrystallization texture usually forms in materials with a large component of the copper-type rolling texture, in which the major components are S and C, whereas the cube texture does not form in materials with brass-type rolling texture, in which B is the major component.[22] Lücke and Hirsch[23,24] suggested that cube grains have the best chance of growing because the cube orientation forms a 40 deg/<111> high mobility boundary with the major components of the deformation texture. For example, the S component best meets the condition to form 40 deg/<111> high mobility boundaries. It also has been suggested that recovery is expected to take place more quickly in cube-oriented subgrains due to low elastic interaction between the slip systems, which aids the rapid development of cube nuclei from these areas.[2527] For example, Sokolov et al.[28] found that cube-oriented subgrains were mostly located in deformation-band regions with orientations close to S and had borders with both S- and C-oriented regions, whereas no marked recrystallization centers were found close to B orientations. From the oriented growth aspect, the increase in the fraction of the preferred rolling components should favor the formation of a sharper cube texture.

As indicated in the present work, increasing the rolling reduction increased the stored cold work energy (at least from 90 to 95 pct) and concomitantly decreased the primary recrystallization temperature. It also increased the fraction of the S and C components and somewhat decreased the B component. These changes produced both a stronger copper-type rolling texture and a stronger primarily recrystallized cube texture, which was derived from the rolling texture.

It is known that a significant fraction of cube texture may be obtained during hot extrusion or hot rolling.[2933] Daaland and Nes[32] studied as-hot-rolled Al-Mn-Mg alloys and observed not only recrystallized, dislocation-free cube grains, but also cube bands formed by subgrains that were distributed parallel to the rolling direction. They suggested that the cube orientation is metastable during hot deformation and survives heavy reductions without rotating toward the stable texture components. The cube grains are flattened to a critical thickness and subsequently serve as nucleation sites for a new generation of cube grains.[32] Doherty et al.[31] showed evidence that the frequency of cube and near-cube grains in recrystallized warm-extruded Al was much higher than that of cold-compressed Al specimens. He also suggested that increasing the deformation temperature helped strengthen the final recrystallized cube texture.[2] The present study found a contrasting result in high-purity nickel. It is worth noting that in the present study, the rolling process at 200 °C was not continuous “hot rolling.” The specimen was heated to the temperature before and after each rolling pass, which were applied with room-temperature rolls. It was expected that recovery or dynamic recovery would occur to a larger extent during rolling at 200 °C, compared with rolling at lower temperature, resulting in a decrease of stored energy and probably an increase of recrystallization temperature. However, as shown in Figure 5, specimens rolled at 25 °C and 200 °C to the same reduction did not show significant differences in their stored energies and recrystallization temperatures. With a thickness reduction per rolling pass of only 100 to 200 μm, excessive heating of the strips during room-temperature rolling was prevented. Therefore, it is likely that the actual “rolling temperature” applied in this study may not be high enough to promote either the nucleation of cube-oriented grains or the formation of cube-oriented subgrain structures.

It is worth noting that the annealing temperature also affects the formation and the strength of the cube texture. In an earlier study,[15] we measured the evolution of the recrystallization texture with increasing annealing temperatures in 99.995 pct nickel specimens that were either 90, 95, or 97.5 pct rolled at 25 °C. The results indicated that, when secondary recrystallization was prevented, increasing the annealing temperature increased the strength of the primarily recrystallized cube texture. Furthermore, increasing the cold-rolling thickness reduction increased the onset temperature of secondary recrystallization, and, thus, increased the thermal stability of the primarily recrystallized cube texture.[15]

5 Conclusions

Through studying the effects of deformation temperature and the degree of deformation on recrystallization of high-purity nickel, the following conclusions can be drawn.
  1. 1.

    Decreasing the rolling temperature increased the stored energy of cold rolling and decreased the primary recrystallization temperature, but had little effect on either the rolling texture or the primary-recrystallized cube texture. In contrast, increasing the rolling temperature had little effect on the stored energy, the recrystallization temperature, the rolling texture, or the primary-recrystallized texture.

     
  2. 2.

    Increasing the thickness reduction of cold rolling increased the stored energy and decreased the primary recrystallization temperature, and strengthened both the rolling texture and the primary-recrystallized cube texture.

     
  3. 3.

    At lower rolling reductions, annealing twins were observed in the primary recrystallization texture. As the rolling reduction was increased, the number of annealing twins decreased and a texture containing strong cube and a small fraction of {013}<100> component was produced.

     

Acknowledgments

This research was supported by National Science Foundation Grant No. 0217565. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation.

Copyright information

© THE MINERALS, METALS & MATERIALS SOCIETY and ASM INTERNATIONAL 2007