MODELING OF THERMO-MECHANICAL BEHAVIOR OF NITINOL ACTUATOR FOR SMART NEEDLE APPLICATION

Abstract

A large and increasing number of cancer interventions, including both diagnosis and therapy, involve precise placement of needles, which is extremely difficult. This challenge is due to lack of proper actuation of the needle (i.e., actuated from the proximal end, which is far away from the needle tip). To overcome this challenge, we propose to bend the needle using a smart actuator that applies bending forces on the needle body; thereby, improving the navigation of the needle. The smart actuator is designed with shape memory alloy (SMA) wires, namely Nitinol, due to their unique properties such as super-elasticity, shape memory effect, and biocompatibility. For accurate steering of the smart needle, there is a need to understand Nitinol thermo-mechanical behaviors. Various existing SMA constitutive models were investigated and compared. Since SMA is used as an actuator in this project, only one dimensional constitutive models are considered. Two distinct models with different phase transformation kinetic approaches were chosen. The first model was proposed by Terriault and Brailovski (J. Intell. Mat. Systems Structures, 2011) using a modified one dimensional Likhachev formulation. The second model was developed by Brinson (J. Intell. Mat. Systems Structures , 1993). Since all SMA constitutive models are empirically based, several important materials' constants such as Phase Transformation Temperatures are needed. The four Transformation Temperatures are: Martensite start (Ms), Martensite finish (Mf), Austenite start (As), Austenite finish (Af). Differential Scanning Calorimetry (DSC) was used to obtain these constants. These temperatures are also influenced by stress, defined by the Clausius-Clayperon coefficients. The coefficients were obtained by measuring Nitinol temperature and displacement response under various constant stress conditions. In order to study its actuation behavior, Nitinol wires under constant strain configuration and resistance heating were tested for their force response. The thermo-mechanical responses were then compared with numerical simulations. While Terriault and Brailovski resistance heating formulation agrees strongly with temperature responses, the model cannot be used to simulate the actuator mechanical responses. Brinson model simulations of the force responses were found to agree well with experimental results. In conclusion, Terriault and Brailovski resistance heating formulation should be coupled with Brinson model to accurately simulate Nitinol actuation behavior for the smart needle.