Measuring transformation temperatures in Nitinol alloys

For almost any use of a Nitinol Shape Memory Alloy, it is highly desirable that one knows the Transformation Temperatures (TTRs) of the alloy. The TTRs are those temperatures at which the alloy changes from the higher temperature Austenite to the lower temperature Martensite, or vice versa.

There are numerous ways of performing transformation temperature testing, but three are in common use with Nitinol alloys to provide helpful data to product designers -- Constant Load, DSC, and Active A_f.

Constant load

First, it is straightforward to apply a load to the alloy and monitor its deformation and shape recovery simultaneously with temperature as the material is cooled and heated through the transformation range. For example, the elongation and contraction of a shape memory wire under tensile loading is shown in Figure 1 as the temperature is lowered and subsequently raised. Generally, the specific load that the material will see in the actual application is used for the test to simulate the conditions in practice. The temperature points noted are ones frequently used to describe the behaviour of a particular alloy. M_s is where the Martensite starts to form on cooling and M_f is where Martensite finishes; As marks the start of Austenite formation on heating while A_f identifies the finish of the transformation to Austenite and completion of shape recovery. This type of test is generally used for applications which utilise the shape memory effect (not superelasticity) in Nitinol.

Be aware that the TTRs are stress dependent parameters, i.e., the TTRs will be different under different loads. In order to determine the TTRs at zero stress, a curve such as that shown in Figure 1 must be obtained at two or more stress levels. The particular transformation point of interest can then be extrapolated to zero stress. A side benefit of the stress dependent nature of the transformation is that the transformation temperature can be precisely tuned in many actuator applications by making small adjustments to the bias force acting against the shape memory element.

Differential scanning calorimeter (DSC)

A precise method of determining the TTR values at zero stress, but one requiring an expensive instrument, is to use a Differential Scanning Calorimeter (DSC). This DSC method yields a plot such as Figure 2 by measuring the amount of heat given off or absorbed by a tiny sample of the alloy as it is cooled or heated through its phase transformations. The DSC yields excellent, repeatable results on fully annealed samples (annealed at temperatures above 700 deg.C for sufficient time to achieve a full anneal, generally about 10 to 15 minutes for small samples). One important drawback to the DSC method is that tests on partially cold worked materials, such as those used to optimise, can yield poor, inconclusive results. This same drawback also may apply to samples which have undergone a heat treatment in the range of 400 to 600 deg.C following cold working. The Constant Load or Active A_f tests are recommended for material in these conditions.

Fully annealed DSC results are often used as the basis for NiTi raw material selection since they effectively characterise the baseline properties of the material prior to cold working and heat treatment.

Active A_f

The third common method of determining TTRs is known as the Active A_f (or Functional A_f) test, also known as a Water Bath or Alcohol Bath test. This test is conducted by merely bending a sample of the alloy, such as a wire, while it is below Ms and then monitoring the shape recovery while it is heated.

For instance, if a wire is gently bent into a hairpin shape (180 degree bend) by finger pressure, then warmed slowly in a stirred liquid bath while monitoring bath temperature, one can easily measure the retained bend angle at specific temperatures. Plotting this data (as in Figure 3), one finds that the curve reaches zero retained bend angle at A_f. The radius of the initial 180 degree bend should not be smaller than 10 times the wire diameter for complete bend recovery. This method, while not very sophisticated, will yield surprisingly accurate, repeatable results if performed carefully, and it requires very little experimental apparatus.

Note that testing superelastic materials by this method requires starting at about -50 deg.C. A bath of crushed dry ice and alcohol is recommended for testing in this range. A water bath may be appropriate for higher temperature alloys. This test method is frequently used for recording the A_f of superelastic materials for guidewires and other applications.

Other methods for measuring Nitinol alloys transformation tempertures

Other methods such as measuring resistivity changes have been used to measure TTRs. However, the above three techniques will serve the needs of most researchers and development engineers who need to characterise or perform Quality Assurance tests on their materials.