Characterization of Shape Memory Alloys by DSC and DMA, Part 2: DMA Analysis
In Part 1 of this series, we studied the properties of nitinol using differential scanning calorimetry (DSC). In this second part, we investigate the behavior of nitinol using dynamic mechanical analysis (DMA).
Introduction
The deformation of an object made from a shape memory alloy such as nitinol can be completely reversed either by heating (shape memory effect) or by elimination of the stress that produced the deformation (superelasticity). These unusual properties are a consequence of reversible transformation of the crystal lattice.
Part 1 of this series [1]) dealt with the thermally induced transformation of the crystal lattice. We used DSC measurements to investigate the shape memory effect using nitinol as an example.
This second part describes the superelastic behavior of nitinol and shows how dynamic mechanical analysis (DMA) can be used to study this effect.
The DMA experiments were carried out in the tension mode using DMA/SDTA861e and DMA 1 instruments.
Results and Discussion
Transformation temperatures
The sample was the same material as that used for the DSC measurements in the Part 1.
First, the transformation temperatures for an equiatomic nitinol wire with a diameter of 0.8 mm were determined. To do this, a 9-mm long sample specimen was measured in the DMA/SDTA861e using a displacement amplitude of 1 µm and an auto-offset of 120%. In this mode, the offset force is automatically adapted to changing sample stiffness during heating.
The results of the heating and cooling measurements are displayed in Figure 1. The upper part of the diagram shows the storage modulus and the lower part the loss factor (tan delta) as a function of temperature. During the phase transformation, the storage modulus exhibits a marked change; the tan delta curve shows a peak. Furthermore, the loss factor of martensite is higher than that of austenite. The loss factor curve for a stainless steel wire is shown for comparison. This shows that austenitic nitinol exhibits elastic properties that are similar to those of stainless steel. The DMA measurements also show hysteresis, as was observed in the DSC measurements. The transformation temperatures determined from the DMA and DSC curves are summarized in Table 1. The temperatures determined using the two techniques are in good agreement.
Superelasticity
If a spectacle frame made of austenitic wire is deformed at room temperature, for example by a blow, the spectacle frame immediately recovers its original shape after the blow. The phenomenon is called superelasticity.
Wires made of normal metals, for example steel, are already irreversibly deformed after an extension of about 0.5%. Superelastic materials change their lattice structure when they are mechanically stressed.
Figure 2 shows the effect of an external stress on the lattice of a normal metal and an austenitic shape memory alloy. With normal metals, the behavior under stress is determined by Hooke's law; the potential reversible extension is small. With austenitic shape memory alloys, the austenite is partially converted to martensite under stress. If the stress is increased, the proportion of the martensite phase increases.
In this way, reversible extensions of up to 10% are possible thanks to the austenitemartensite transformation. For superelastic behavior, the starting material must therefore be present as austenite.
To investigate the phenomenon of superelasticity, experiments were carried out in the tension mode using a DMA/ SDTA861e using the same 0.1-mm thick nitinol wire previously measured by DSC microscopy [1]. The DMA measurements were performed at 35 °C
The measurement program used is displayed in the upper half of Figure 3. The nitinol sample was subjected to different static forces for 3 minutes. The dynamic force amplitude was set to zero. The stress corresponds to the force normalized to the cross-sectional area of the wire.
The lower part of Figure 3 shows the measured deformation values recorded during the experiment. The deformation at 0 N corresponds to the offset that arises through clamping the sample. The strain at a particular stress corresponds to the deformation corrected by the offset and normalized to the original sample length.
If the different stresses applied to the sample are plotted as a function of the resulting strain, the stress-strain curve shown in Figure 4 is obtained.
The first part of the curve represents the elastic behavior of austenite. From a strain value of about 0.01, austenite changes to the twin structure of martensite (superelastic plateau). At strain values above about 0.08, the material is present entirely as martensite. A further increase of the stress leads to a (Hooke's law) deformation of the (martensitic) wire.
If the stress is reduced in steps, the martensite changes to austenite as shown in Figure 5. It can be seen that the loading and unloading curves do not superimpose. In fact, the superelastic plateau occurs at lower strain values during unloading compared with on loading. During the stress-strain cycle, a hysteresis effect is observed similar to that in a heating-cooling cycle.
The slope in the linear region of the stress-strain curve corresponds to the elastic modulus of austenitic nitinol. Numerically, this yields a value of 67.4 GPa. To verify this value, an isothermal DMA experiment was performed at 35 °C using a dynamic displacement of 1 µm and an auto-offset of 120%. The result is displayed in Figure 5 in the small inset diagram. The two values for the elastic modulus agree well: from the stress-strain curve, E = 67.4 GPa and from the dynamic DMA measurement, E = 67.1 GPa.
The same experiment was also performed using a stainless steel wire with a diameter of 0.09 mm (green curve). The slope yields an elastic modulus of 207 GPa. This value agrees well with literature values of 200 to 210 GPa
Effect of temperature on the plateau height
To investigate the influence of temperature on the beginning of the martensitic transformation under stress, the stressstrain behavior of a nitinol wire with a diameter of 0.01 mm was measured at different temperatures above the austenite start temperature (As) using a DMA 1.
Summary and Conclusions
DMA can be used to characterize the thermal and mechanical properties of shape memory alloys. In the thermally induced phase transformation of austenite to martensite or the other way around, the modulus changes and a peak appears each time in the tan delta curve during the phase change. Besides this, a hysteresis effect is observed between heating and cooling just like in the DSC measurements.
Above the austenite temperature, objects made of shape memory alloys can undergo reversible expansion of up to about 10% under stress in a restricted temperature range. This property can be investigated with the aid of stress-strain experiments using DMA.
The phenomenon of the shape memory effect and superelasticity is summarized in Figure 8. For the shape memory effect and superelasticity to occur, an object must first be conditioned to a particular shape. Conditioning is performed, for example for nitinol, at temperatures above 500 °C. An object that has been conditioned in this way can be reversibly extended as austenite up to about 10%.
It can also be deformed as martensite (twin structure), but the deformation is then permanent. If an object deformed in the way described is then heated to above the austenite temperature, the deformed object reverts back to the shape originally given to it.
Shape memory effect and superelasticity can easily be investigated by DSC, DSC microscopy and DMA.
Characterization of Shape Memory Alloys by DSC and DMA, Part 2: DMA Analysis | Thermal Analysis Application No. UC 413 | Application published in METTLER TOLEDO Thermal Analysis UserCom 41