3.Ferromagnetic Shape Memory Alloys (FSMAs)

Sponsorship: Prior; Helsinki University of Technology (subcontract on grant from TEKES, Finnish Technology Foundation), Lord Corporation, Boeing Corporation (subcontract on DARPA grant), Midé Technologies, Office of Naval Research, ACX Corp. (subcontract on NASA grant).

Personnel:
 

R. C. O'Handley Miguel Marioni

S. M. Allen Marc Richard

David Paul Jorge Feuchtwanger

David Bono Bradley Peterson

Shape-memory alloys show a martensitic transformation characterized by significant dimensional changes. They compete with piezoelectric and magnetostrictive materials in many smart-material and actuator applications. Shape memory materials can exhibit large strains but are typically activated by a temperature change. This mode of actuation is inefficient and has strict frequency limitations.

We have demonstrated that certain shape memory materials that are also ferromagnetic can show very large dimensional changes under the application of a magnetic field. These strains occur within the low-temperature (martensitic) phase. This effect was first observed in the low-termperature, tetragonal (martensitic) phase (Fg. 3.1b) of the Heusler intermetallic Ni₂MnGa (Fig. 3.1a), where strains of 0.19% were produced in fields of 0.8 Tesla at 265K (see Fig. 3.2). These large field-induced strains (comparable to those observed in Terfenol-D) are associated with the motion of twin boundaries in the martensitic phase under the driving force of the change in Zeeman energy, M.H, across the twin boundaries.
 
 

Fig. 3.1 a) Heulser structure of Ni2MnGa.           Fig. 3.1 b) Tetragonal phase of the same composition, stable           below a martensitic transformation temperature of about 0oC.

Fig. 3.2 Strain versus field for martesnitic single crystal Ni2MnGa at -8oC. The strain is measured along [001] with the field applied along [001] and [110]. Adapted from Ullakko et al., 1996.

The potential for capturing strains approaching the 6.6% c-axis contraction associated with the transformed phase was advanced with the observation shown below. We were able to isolate a lone twin boundary in a single-crystal bar of Ni-Mn-Ga (Fig. 3.3). The off-stoichiometry composition was chosen (Murray et al., 1998) so that the material is martensitic at room temperature. In addition to observing the lone twin boundary and its associated 6% shear strain, we were able to move the twin boundary in an applied field of about 4 kOe.
 
 

The observation in Fig. 3.3 allowed us to confirm empiracally the prevailing notions that the twin boundaries occur along what had been {101} and {011} planes in the pareent phase ({211} planes in martensite), hence the c axis changes direction by about 86 degrees across the twin boundary, and the preferred direction of magnetization is along the tetragonal caxis.

In order to get the crystals to strain extensionally rather than in the manner shown in Fig. 3.3, samples were cut so that the twin boundaries lie at about 45^o to the sample axis. Fig.. 3.4 is a selection of frames from a high-speed (1200 frames/sec) video of the motion of twin boundaries in such a sample measuring 6 mm by 6 mm by 20 mm. The initial state of the sample is nearly a single-variant state with vertical c axis established by a static vertical stress of about 1.6 MPa. As the horizontal field (orthogonal to the stress axis) is increased to about 5 kOe, new, dark-shaded twins are seen to nucleate and grow, causing the sample height to increase by 6%. These new variants have their c axis and preferred direction of magnetization horizontal.

Measurement of DC field-induced strain on a single crystal of martensitic Ni-Mn-Ga at room temparature showed a 6% strain if the sample was stressed orthogonal to the field direction by about 1.5 MPa (Murray et al. 2000).  These quasistatic field-induced strain curves showed a hysteresis of about 1500 Oe.

A system was made to allow AC actuation under stress (Fig. 3.5).  The bias stress is applied by a spring (horizontal in this figure); the stress is adjusted by compressing the spring toward the sample.  A specially designed field coil and laminated core provided a field of up to 7 kOe (vertical in this figure.
 

 
 
 
 
 
 
 
 
 
 
 
 
 

The actuation shown in Fig. 3.5 has been digitally collected and analyzed for a different sample than that shown in the video. Typical results are shown in Fig. 3.6. Note that the field induced strain increases with increasing bias stress, maximizing near 1.4 MPa. For larger stresses, the field-induced strain decreases, being blocked for stresses of 2 to 3 MPa. The theoretical blocking stress is a function of the anisotropy energy, saturation magnetization and applied field as described by O'Handley (1998) and Murray et al. (2000).         Fig. 3.6 Field-induced strain taken at room temperature in the apparatus of Fig. 3.5 under different average stresses at 1 Hz drive (2 Hz actuation). This particular sample only strained by about 3.1% at optimal stress bias of about 1.4 MPa.

Data like that in Fig. 3.6 have been taken up to actuation frequencies of several hundred Hz. The limitation in these cases is not the material but the impedance match between the power supply and the drive coils.

The impact of this discovery of large magnetic-field-induced strains in FSMAs can be appreciated by the fact that the energy density in the material exceeds that in Terfenold-D, being of order 10 kJ/m³ so far. Further, power output we have already achieved in these materials far exceeds that of electric motors and is comparable to that of an internal combustion engine.

Some relevant FSMA publications:

"Large Magnetic-field-induced Strains in Ni₂MnGa Single Crystals," R.C. O'Handley, K. Ullakko, J.K. Huang, and C. Kantner, Appl.

"Field-induced strain under load in Ni-Mn-Ga magnetic shape memory materials," S.J. Murray, M. Farinelli, C. Kantner, J.K. Huang,

M.Allen, MRS Conf. Proc. Vol 604 279-284 (2000).

New York., 2001).

Return to Table of Contents