CHARACTERIZATION AND MODELING OF TRAINED NITINOL TORSIONAL ACTUATORS UNDER REVERSE BIAS LOADS

摘要

Shape Memory Alloy actuator components are typically designed and trained to operate against loads in a single direction. This may lead to overly complex actuator designs, where devices such as ratchets or large biasing forces are used to isolate the shape memory alloy components from exposure to external loads of an uncertain direction and magnitude while in service. An understanding of the effects of reverse bias loads on trained SMA actuator components may enable simpler designs, improved operation, and better control. In this paper the design and training of a representative NiTinol rotary actuator is described. The trained actuator is operated over a range of isobaric loads in both the trained and reverse bias directions and the actuators performance is quantified. The stability and characterization of the actuator's performance while working against loads in the trained direction are compared to its operation against reverse bias loads of increasing magnitude. Consistent and stable operation is shown for varying loads in the trained direction and for small reverse loads. At larger reverse loads, actuator performance begins to shift indicating the reverse bias is impacting the tube training. A method of characterization which identifies the magnitude of reverse loading that may be detrimental to the actuator's performance is proposed. A new model originally derived specifically to capture the effects of training on SMA tubes via the incorporation of constant transformation back-stresses is applied to the analysis of these tubes. An FEA implementation of the model is used to simulate tube actuation behavior against loads in the trained and reverse bias directions, and accuracy of the analysis is demonstrated.