Tension and strain annealing for abnormal grain

JOURNAL OF APPLIED PHYSICS 109, 07A941 (2011)

Tension and strain annealing for abnormal grain growth in magnetostrictive Galfenol rolled is better sheet Hyunsuk Chun,1,a) Suok-Min Na,2,b) Jin-Hyeong Yoo,2,c) Manfred Wuttig,1,d) and Alison B. Flatau1,2,e)

1 Department of Materials Science and Engineering, University of Maryland, College Park, Maryland 20742, USA 2 Department of Aerospace Engineering, University of Maryland, College Park, Maryland 20742, USA

(Presented 16 November 2010; received 29 September 2010; accepted 6 January 2011; published online 12 April 2011) This paper investigates the effect of tension annealing and of strain annealing on abnormal grain growth (AGG) in polycrystalline rolled sheet of NbC-added Fe81Ga19. Electron backscattering diffraction patterns captured by orientation imaging microscopy were used for analyzing texture and grain configuration of tension annealed samples. Tension annealing, with simultaneous application of tension and high temperature annealing, with strain rate control, was conducted to study “dynamic recrystallization” for AGG. Strain annealing, application of tension to produce uniform strain and subsequent high temperature annealing, was carried out to study the effect of “strain induced boundary migration” on AGG. A small amount of texture variation resulted, but no significant AGG was observed during tension annealing. During strain annealing, no texture or grain size variation was observed in the gauge region of uniform strain in the sample; however, abnormally grown Goss-texture was observed in regions of high stress concentration outside of the C 2011 American Institute of Physics. [doi:10.1063/1.3565419] gauge region of the sample. V

I. INTRODUCTION

Highly textured polycrystalline Galfenol (Fe–Ga alloy) with a h100i preferred orientation is a candidate magnetostrictive material used in sensor and actuator applications. High temperature annealing of NbC-added Galfenol rolled sheet at 1200 C under Ar atmosphere produces secondary recrystallization with a Goss (110) h001i texture and 180 ppm magnetostriction from as-rolled sheet with 30 ppm magnetostriction.1 Results from our previous work suggest that the differences in grain boundary energy, especially high energy grain boundaries, affect Goss-textured abnormal grain growth (AGG).2 In the present work, two different methods are used to investigate the use of tension as an approach for altering grain boundary energy and thereby influencing the secondary recrystallization process by promoting AGG in 1%–2.5% NbC-added Galfenol rolled sheet. One method, “tension annealing,” is used to promote dynamic recrystallization (DRX). This is achieved by applying tension to produce a uniform strain state in the sample while simultaneously conducting a high temperature anneal. The other method, “strain annealing,” is used to promote strain-induced boundary migration (SIBM). This involves applying tension with a uniform strain rate while at moderate temperatures and then removing the tensile load and conducting a high temperature anneal to promote secondary recrystallization. Several studies show both of these phenoma)

Electronic mail: [email protected]. Electronic mail: [email protected]. Electronic mail: [email protected]. d) Electronic mail: [email protected]. e) Electronic mail: [email protected]. b) c)

0021-8979/11/109(7)/07A941/3/$30.00

ena promote AGG in different materials, such as Mo, Cu, and Al–Mg alloys for DRX (Refs. 5–10, and 11) and Zr and Fe for SIBM.7,12 In this paper, these two methods will be investigated to assess their effect on AGG in NbC-added Galfenol rolled sheet. Although the supporting mechanism in the literature for both methods is somewhat different, it is suggested that the observed recrystallization phenomena in both methods originates from the common consequence of the retention of large amount of stored energy that is a result of the applied tension of the sample.3 During DRX, one of the mechanisms that occurs is the local structure rearrangement. During annealing, the material is constantly being deformed as new grains nucleate and grow. Once plastic deformation reaches a critical level, new dislocation-free grains form in the deformed structure. These increase boundary migration mobility by facilitating release of grain boundaries from pinning sites, such as oxide and/or carbide particles that can produce Zener pinning.4,5 The driving force for SIBM is presumed to arise from a difference in dislocation density. Differences in dislocation density form an energy balance that promotes discontinuous grain boundary motion without nucleation.3,6 The three most important factors emphasized in studies of these processes are critical strain, strain rate, and annealing temperature. A minimum strain, called the critical strain, is necessary for initiation and propagation of new grains, after which the grain size decreases with increasing strain in both cases.5,7 Strain rates between 106 and 104/s are reported as being necessary to avoid a multiplication of the dislocation density and formation of multiple Luders bands.8,9 Annealing at temperatures of over 60% of melting temperature is needed for AGG in both processes.

109, 07A941-1

C 2011 American Institute of Physics V

Downloaded 19 Apr 2011 to 128.8.202.26. Redistribution subject to AIP license or copyright; see http://jap.aip.org/about/rights_and_permissions

07A941-2

Chun et al.

J. Appl. Phys. 109, 07A941 (2011)

II. EXPERIMENTAL PROCEDURES

The initial polycrystalline (Fe81Ga19) þ 1.0–2.5 mol. % NbC rolled sheets were produced by rolling as described in Ref. 1. For tension annealing, as-rolled Galfenol rolled sheet was annealed at 1073–1273 K (60%–70% of the melting temperature) with the simultaneous application of tension and true strain rates of 2 104/s under an Ar atmosphere. The system was held at the anneal temperature for roughly 30 min prior to applying tensile loads to accommodate thermal expansion of all materials and initiate testing from a steady state thermal condition. A Terfenol-D actuator provided accurate strain rate control. The maximum displacement to the sample from the actuator was about 1.7 mm (15% of sample initial length). Strain and stress variation were monitored during tension annealing. For strain annealing, the as-rolled Galfenol samples were preannealed at 973 K under an Ar atmosphere for 10 h to fully introduce primary recrystallized grains. This provided homogeneous recrystallization with a 10 lm diameter grain size suitable for secondary recrystallization. The temperature was increased up to 573 K to soften the sample and then tension was applied with an materials test system (MTS) load stand to introduce the suggested critical strain in the sample, and thereby maximize the stored strain energy available to promote AGG during the subsequent high temperature anneal. This anneal was done at the same conditions as for tension annealing, i.e., at 1073–1273 K for 1 h under Ar. Assessment of AGG was conducted using electron backscattering diffraction (EBSD) patterns that were captured and analyzed using orientation imaging microscope (OIM) software to obtain inverse pole figures (IPF) and orientation distribution function (ODF). III. RESULTS AND DISCUSSION A. Tension annealing

IPF images and ODF plots from different anneal conditions are shown in Figs. 1(a)–1(c). The as-rolled samples with c-fiber texture and an average grain size of 5.3 lm were used as an initial state. Those samples were annealed at 1273 K for 3 h with (a) no tension, (b) uniform stress, and (c) uniform strain. After annealing at 1273 K for 3 h with no tension, some of the major c-fiber texture in the asrolled sample was changed to a {100} texture with h110i and/or h100i orientation as the grain size increased to an average diameter of 60 lm [Fig. 1(a)]. After tension annealing at the same temperature with constant strain rate of 2 104/s, a small amount of elongation of grains and lattice rotation are observed in Fig. 1(b). The length of the sample was finally increased by about 15% of the initial sample length and the average grain size was 95 lm, also larger than in the sample with no tension annealing. The area fraction of Goss grains: {110} h001i was 19.7 or 10.7% lower than in Fig. 1(a) and the area fraction of Cube texture: {100} h001i and {111} grains were 48.1 and 32.2%, respectively, or 5.8 and 4.9% higher than in Fig. 1(a). (The tolerance range that specifies the minimum and maximum lattice variation is 0 –30 .) Somewhat different

FIG. 1. (Color online) IPF images along normal direction and ODF plot at different conditions: (a) annealed at 1273 K for 3 h with no tension, (b) annealed at 1273 K for 3 h with tension and strain rate control (2 104/s), and (c) annealed at 1273 K without strain rate control but with a constant tensile load.

texture variation is shown in Fig. 1(c). In this case, a constant tensile stress of 75 MPa was applied to the sample without any strain rate control at the same temperature as in Figs. 1(a) and 1(b) until the sample failed. The gauge region of the sample started to thin and the sample failed due to fracture within the first few minutes of loading. The failed sample had elongated to about 1.5 times longer than its initial state and it exhibited a c-fiber texture similar to the asrolled sample over 47% of the sample area fraction. These experiments were also conducted at a slightly lower anneal temperature, at 1173 K, but in all cases, no significant DRX phenomenon and no AGG was observed. It is not clear if a different range of annealing temperature, critical strain and/ or strain rate for DRX would produce AGG. To resolve this, additional tension annealing at other conditions should be undertaken.

FIG. 2. (Color online) Experimental true stress vs true strain curve and optical image of before and after tension at 573 K.


07A941-3

Chun et al.

J. Appl. Phys. 109, 07A941 (2011)

mation during straining the sample and the influence of stored strain energy on AGG was demonstrated indirectly in this region. To determine the critical strain needed to produce this effect, tapered samples such as those used by Refs. 7 and 9 will be strain-annealed for the further study. III. CONCLUSIONS

FIG. 3. (Color online) IPF image of a sample annealed at 1273 K for 1 h after imposing a 5 104/s strain rate controlled tension force.

B. Strain annealing

The strain annealing study used a uniformly primary recrystallized sample that was strained with a constant 5 104/s strain rate to obtain the critical strain for SIBM. Figure 2 shows the true stress variation as a function of true strain. The steady-state strain rate was attained after a true strain of 1%. At a strain of slightly less than a true strain of 8%, the sample fractured outside of the gauge region, in an area adjacent to pinholes that were introduced to address a challenge in gripping the dog bone shaped test samples. Plastic deformation of the sample is observed only around these pinholes. The upper part of the pinhole region fractured at an applied 210 MPa true stress as shown in Fig. 2. No visible elongation was observed on the gauge section that was originally the intended focus of this test. The sample was cut in half and used to investigate the effect of anneal at two temperatures, 1073 and 1273 K, for 1 h under an Ar atmosphere. No texture or grain size variation was observed from the half-dog bone sample that annealed at 1073 K. Figure 3 shows the IPF from half-dog bone sample that underwent anneal at 1273 K for 1 h. No texture or grain size variation was observed in the gauge section of this sample either. However, interestingly, a small region of abnormally grown Goss-texture was observed near the slightly elongated, plastically deformed, and pinhole region. The maximum size of abnormally grown Goss grain is 1400 lm diameter and it is tilted about 8 from the rolling direction. Stress was concentrated around the pinhole with small defor-

The effect of tension annealing and strain annealing on AGG in (Fe81Ga19) þ 1.0–2.5 mol. % NbC rolled sheet was investigated. During tension annealing at 1273 K for 3 h, a small amount of grain elongation and lattice rotation were observed under strain rate control tests where stress was sufficient to produce plastic deformation. Tension annealing produced an increase in the area fraction of cube texture and {111} grains relative to what was observed with no tension annealing. c-fiber texture was predominant in the sample that was tension annealed without strain rate control. However, no DRX with AGG was observed during this process. During strain annealing, the sample was fractured near a small region of high stress concentration prior to the onset of any visible elongation in the dog bone shaped sample’s gauge section. The sample was cut into half and each half was annealed at a different temperature, one at 1073 K and the other at 1273 K, for 1 h under Ar atmosphere. No AGG occurred in the gauge region of these samples. However, unexpectedly, abnormally grown Goss-texture was observed in the high stress concentration region near the pinhole of sample had elongated under load. This suggests that strain annealing of this alloy at stress levels that are sufficient to cause plastic deformation (210 MPa) should be the focus of future investigation. These preliminary results indicate AGG in FeGa rolled sheet can be influenced with a strain anneal process. ACKNOWLEDGMENTS

This work was supported by ONR MURI Grant No. #N000140610530. 1

S. M. Na, J. H. Yoo, and A. B. Flatau, IEEE Trans. Magn. 45, 4132 (2009). H. S. Chun, S. M. Na, C. Mudivarthi, and A. B. Flatau, J. Appl. Phys. 107, 09A960 (2010). 3 F. J. Humphreys and M. Hatherly, Recrystallization and Related Annealing Phenomena (Pergamon, UK, 1995). 4 S. A. Atroshenko, Recrystallization and Grain Growth Proceeding of the First joint International Conference, (2001), p. 931. 5 J. Ciulik and E. M. Taleff, Scr. Mater. 61, 895 (2009). 6 V. Przetakiewicz, Metall Term. Obr. Met. 2, 60 (1982). 7 D. Chaubet, J. P. Fondere, and B. Bacroix, Mater. Sci. Eng., A 300, 245 (2001). 8 S. H. Oh, M. Legros, D. Kiener, and G. Dehm, Nature Mater. 8, 95 (2009). 9 D. J. Bailey and E. G. Brewer, Metall. Trans. A 6A, 403 (1975). 10 M. G. Ardakani and F. J. Humphreys, Acta metall. 42, 763 (1994). 11 M. D. Drury and F. J. Humphreys, Acta metall. 34, 2259 (1986). 12 P. Morgand, C. R. Acad. Sci. Fr. 257, 3598 (1963). 2
