Material Characterization and Mid-Span Bending Capacity With Finite Element Simulated Predictions

Abstract

Intelligent materials have been the subject of research for many years. Shape memory alloys (SMAs) are a type of intelligent material that has been targeted for many different uses; such as actuators, sensors and structural supports. SMAs are attractive as actuators due to their large energy density. Although a great deal of information is available on the axial load capacity and on the tip force for SMA tweezer-like devices, there is not enough information about the load capacity at mid-span, especially at the macro-level. Imposed displacement at mid-span experimental evaluation of an SMA beam in the austenitic and martensitic regimes has been studied. To this end, a specimen of near equi-atomic nitinol was heat-treated (shape set) into a ‘U’ shape and loaded into a custom test fixture such that the boundary conditions of the beam are approximated as roller-roller; and the sample was deformed at different temperatures while reaction forces were measured. The displacement is near maximum displacement of the U shape without causing a change in concavity, thus full-scale capacity is shown. Additionally, Unified Model (finite element) predictions of the experimental response are also presented, with good agreement. Due to the robust nature of the Unified Model, geometric parameter variations (wire diameter and radius of curvature) were then simulated to encompass the design envelop for such an actuator. The material properties needed as inputs to the Unified Model were obtained from constant temperature tensile tests of a specimen subjected to the same heat treatment (shape set straight). The resultant critical stresses were then extracted using the tangent method similar to the one described in ASTM F-2082. It is worth noting that the specimen was trained before the stress value extraction, but the transversely loaded specimen was not trained due to the difficulty involved (inherent uneven stress distribution). The contribution of this work is the presentation of experimental results for transverse (mid-span) loading of a nitinol wire and the simulation results allowing for design of a proper actuator with known constraints on force, displacement or temperature (2 of 3 needed). In other words, this work could be used as a type of 3D look-up table; e.g. for a desired force/displacement, the required temperatures are given. Future work includes developing a sensor-less control strategy for simultaneous force/displacement control.