Shape memory alloys (SMAs) remain one of the most commercially viable active materials, thanks to a high specific work and the wide availability of high quality material. Still, significant challenges remain in predicting the degradation of SMA actuators during thermal cycling. One challenges in both the motivation and verification of degradation models is the measurement of retained martensite fraction during cycling. Direct measurement via diffraction is difficult to perform in situ, impossible for thin wires, (< 0.5mm) and prohibitively difficult for lengthy studies. As an alternative, the temperature coefficient of electrical resistivity (TCR) is used as an indicator of martensite phase fraction during thermal cycling of SMA wires. We investigate this technique with an example cycling experiment, using the TCR to successfully measure a 20% increase in retained martensite fraction over 80000 thermal cycles. As SMA wire temperature is difficult to measure directly during resistive heating, we also introduce a method to infer temperature to within 5 °C by integrating the lumped heat equation.
Over 60% of energy that is generated is lost as waste heat with close to 90% of this waste heat being classified as
low grade being at temperatures less than 200°C. Many technologies such as thermoelectrics have been proposed as
means for harvesting this lost thermal energy. Among them, that of SMA (shape memory alloy) heat engines appears
to be a strong candidate for converting this low grade thermal output to useful mechanical work. Unfortunately,
though proposed initially in the late 60's and the subject of significant development work in the 70's, significant
technical roadblocks have existed preventing this technology from moving from a scientific curiosity to a practical
reality. This paper/presentation provides an overview of the work performed on SMA heat engines under the US DOE
(Department of Energy) ARPA-E (Advanced Research Projects Agency - Energy) initiative. It begins with a review
of the previous art, covers the identified technical roadblocks to past advancement, presents the solution path taken to
remove these roadblocks, and describes significant breakthroughs during the project. The presentation concludes with
details of the functioning prototypes developed, which, being able to operate in air as well as fluids, dramatically
expand the operational envelop and make significant strides towards the ultimate goal of commercial viability.
A series of experiments is presented examining the thermo-electro-mechanical response of commercially-available, conditioned,
shape memory alloy (SMA) wires (Flexinol, from Dynalloy, Corp.) during cyclic thermomechanical loading. A
specialized experimental setup enables temperature control via a thermoelectric/heatsink in thermal contact with the wire
specimen during various modes of testing. It allows simultaneous measurement of elongation, load, strain and resistivity in
a selected gage length. It also allows full-field optical and infrared imaging to be performed during testing. A moderately
high transition temperature NiTi-based shape memory wire (90C Flexinol) is characterized first by differential scanning
calorimetry and a series of isothermal experiments over a range of temperatures. Subsequent experiments examine the
shakedown behavior over a range of dead loading temperature cycles. Results show a significant two-way shape memory
effect, suggesting that both residual stresses and locked-in oriented Martensite are considerable in this commercial alloy.
Repeatable behavior (little shakedown) is confirmed at relatively low stress levels, but significant evolution in the response
(shakedown behavior) exists at higher stress levels during the first several temperature cycles.