Elsevier

Acta Materialia

Volume 61, Issue 6, April 2013, Pages 2113-2120
Acta Materialia

Texture and training of magnetic shape memory foam

https://doi.org/10.1016/j.actamat.2012.12.032Get rights and content

Abstract

Magnetic shape memory alloys display magnetic-field-induced strain (MFIS) of up to 10% as single crystals. Polycrystalline materials are much easier to create but display a near-zero MFIS because twinning of neighboring grains introduces strain incompatibility, leading to high internal stresses. Pores reduce these incompatibilities between grains and thus increase the MFIS of polycrystalline Ni–Mn–Ga, which after training (thermo-magneto-mechanical cycling) exhibits MFIS as high as 8.7%. Here, we show that this training effect results from a decoupling of struts surrounding pores in polycrystalline Ni–Mn–Ga during the martensitic transformation. To show this effect in highly textured porous samples, neutron diffraction measurements were performed as a function of temperature for phase characterization and a method for structure analysis was developed. Texture measurements were conducted with a magnetic field applied at various orientations to the porous sample, demonstrating that selection of martensite variants takes place during cooling.

Introduction

Magnetic shape memory alloys (MSMAs) have gained much attention due to the large reversible strain which can be induced in single crystals by varying the direction of a magnetic field [1]. Off stoichiometric monocrystalline Ni–Mn–Ga MSMAs, in particular, are very promising as they show strains up to 10% near room temperature [2], [3], [4], [5], [6]. The strain is produced by twin boundary motion, which is induced by a change of orientation of a magnetic field [1], [3], [7], [8], [9]. In the martensite phase, Ni–Mn–Ga has a high magneto-crystalline anisotropy, which provides the driving force for twin variant growth through twin boundary motion [6], [10], [11]. Many studies have investigated the parameters controlling twin mobility. Surface defects pin twin boundaries, requiring additional force to unpin the twins such that further deformation can occur [12], [13], [14]. Non-metallic impurities (as particles or in solid solution) also pin twin boundaries and thus increase the twinning stress [15]. The higher the twinning stress, the lower the driving force that can be imposed magnetically against an external load [15]. The twin density, hierarchically twinned microstructures and conjugate twinning have a significant impact on twin mobility and possibly affect the fatigue life of magnetic actuation [8], [10], [15], [16], [17]. Training (thermo-magneto-mechanical cycling (TMC)) leads to martensite variant selection of preferentially oriented variants, which reduces the twinning stress, for further cycles, in single crystals [3], [4], [10], [12], [18].

Recently, we have shown that the MFIS of polycrystalline Ni–Mn–Ga can be significantly increased by introducing porosity [19], [20], [21], [22]. Pores replace grain boundaries and reduce internal constraints imposed by the misorientation of neighboring crystals. Polycrystalline porous Ni–Mn–Ga with a bimodal pore size distribution (hereafter “bimodal foam”) have shown very large MFIS of up to 8.7% [20], [21]. A bimodal foam consists of struts connected at nodes forming a network surrounding the larger (500–600 μm) pores; these struts and pores contain smaller (75–90 μm) pores. The MFIS of 8.7% of these hierarchically porous bimodal foams is higher than expected for a randomly textured foam with twin boundary motion as the sole source of deformation within a strut. We have suggested that the high MFIS may be due to texture (a crystallographic effect) [20] and/or strut hinging (a geometric effect) [21]. Even if the foam were a single crystal, that would already have been a breakthrough as such a “single crystal foam” would be much easier to make than a bulk single crystal. In the case of strut hinging, struts tilt with the tilt axis passing through the node [23]. With hinging produced by twin boundary motion within the node, localized bending strain at the node could lead to large displacements further away from the hinge and therefore produce large macroscopic strains.

Neutron diffraction was employed here to identify the active mechanisms of deformation and training within the bimodal foams. Neutron diffraction was chosen due to the high interaction volume of neutrons, making it possible to observe changes in microstructure within the sample. Because the samples are oligocrystalline (they contain a small number of grains), we apply a method rarely used for neutron diffraction, but well-known in X-ray powder diffraction, namely spinning the sample to provide a better powder average of its crystal structure; this method makes structure analysis for highly textured and/or oligocrystalline materials possible.

Section snippets

Experimental procedures

Samples of similar foam studied previously [19], [21] were used here, and only a summary of the foam processing is given below. High purity elements (99.9% nickel pellets from ESPI, 99.999% gallium pellets from Alfa Aesar and electrolytic-purity manganese flakes from Alfa Aesar), forming an alloy with a nominal composition of Ni52Mn24.3Ga23.7, with melting point slightly above 1100 °C [24], were melted in a vacuum induction furnace (Reitel, Induret Compact) and cast into ingots within copper

Neutron diffraction

Fig. 2 shows diffraction spectra of sample 1a taken at 203 K from the integrated 90° detector ring (Fig. 1d) for four selected sample orientations (Fig. 2a) and for all 61 sample orientations (Fig. 2b). The integrated data from 61 orientations (Fig. 2b) showed additional peaks 0 2 0 and 1 0 5 for d spacing between 2.5 and 3.0 Å, which permitted to refine the structure. Peaks at d spacing smaller than 1.5 Å were not indexed because several high-index poles match these peaks within experimental error,

Magnetomechanical properties

Upon TMC, the sample showed an almost threefold increase in MFIS from 0.4% (H/C1) to 1.15% (H/C 2 and 3), reflecting the training effect (Fig. 5) [21]. When rotating a magnetic field during heating and cooling, the magnetic field favors two particular twin variants sharing a mobile twin boundary. Twin variants which are not contributing to the MFIS do not form. Thus, twin–twin interactions that reduce twin mobility are eliminated [16], [18], [21], [35]. During H/C 1, the temperature change was

Conclusions

Neutron diffraction measurements were performed on polycrystalline Ni–Mn–Ga foams with bimodal pore size distribution as a function of temperature for phase characterization. A method for structure analysis was developed which is useful for oligocrystalline samples and samples with strong texture. Texture measurements were conducted with a magnetic field applied at various orientations to the sample, demonstrating that selection of martensite variants took place. The following main results were

Acknowledgments

This project was funded by the National Science Foundation through grant NSF-DMR 1207192 (Boise State University) and DMR-1207282 (Northwestern University). M.C. acknowledges partial financial support through the German Research Foundation (DFG) priority program SPP 1239 (grant No. Schn 1106/1). P.M. is thankful to ETH Zürich for donating magneto-mechanical testing devices. This work has benefited from the use of the Lujan Neutron Scattering Center at LANSCE, which is funded by the US

References (52)

  • V.A. Chernenko et al.

    Scripta Metall Mater

    (1995)
  • P. Müllner et al.

    Scripta Mater

    (2003)
  • P. Müllner et al.

    Acta Mater

    (2010)
  • M. Chmielus et al.

    Scripta Mater

    (2011)
  • M. Chmielus et al.

    Acta Mater

    (2010)
  • M. Chmielus et al.

    Acta Mater

    (2011)
  • S. Kustov et al.

    Acta Mater

    (2004)
  • X.X. Zhang et al.

    Acta Mater

    (2011)
  • D.L. Schlagel et al.

    J Alloys Compd

    (2000)
  • H.R. Wenk et al.

    Nucl Instr Meth Phys Res

    (2003)
  • H.R. Wenk et al.

    Nucl Instr Meth Phys Res A

    (2003)
  • J. Pons et al.

    Acta Mater

    (2000)
  • J. Pons et al.

    Mater Chem Phys

    (2003)
  • J. Pons et al.

    Mater Sci Eng A

    (2006)
  • M. Richard et al.

    Scripta Mater

    (2006)
  • L. Righi et al.

    Acta Mater

    (2007)
  • R. Chulist et al.

    Scripta Mater

    (2010)
  • U. Gaitzsch et al.

    Acta Mater

    (2009)
  • M. Pötschke et al.

    Scripta Mater

    (2010)
  • U. Gaitzsch et al.

    Scripta Mater

    (2007)
  • I. Glavatskyy et al.

    Mater Sci Eng A

    (2008)
  • A. Stebner et al.

    Acta Mater

    (2011)
  • K. Ullakko et al.

    Appl Phys Lett

    (1996)
  • P. Müllner et al.

    J Appl Phys

    (2004)
  • P. Müllner et al.

    J Appl Phys

    (2002)
  • S.J. Murray et al.

    Appl Phys Lett

    (2000)
  • Cited by (0)

    1

    Present address: Micron Technology Inc., Boise, ID, USA.

    2

    Present address: Department of Materials Science and Engineering, Cornell University, Ithaca, NY 14853, USA.

    View full text