The design of an active catheter is an example of a challenging design problem of an shape memory alloy (SMA) adaptive structure. The objective is to find a geometry that combines the electrical, thermal and mechanical properties of the structure in such a way that optimal bending performance is achieved. This paper introduces the application af an efficient gradient-based design optimization procedure to this design problem. The specific model used focuses on the R-phase transformation in Ni-Ti, and involves multi-point constraints to implement symmetry conditions. The nonlinear mechanical analysis is carried out using an incremental-iterative approach in combination with an augmented Lagrangian technique to account for the nonlinear constraints. Sensitivity analysis is performed using finite differences in combination with fast reanalysis, where a new correction term is applied to the multi-point constraints that significantly improves the accuracy. The proposed gradient-based optimization approach is compared to an alternative direct method, and a clear advantage in terms of the number of required function evaluations is achieved. The application of design optimization yields active catheter designs that clearly outperform previous versions. It is expected that the presented method will prove useful for the design of other SMA adaptive structues as well.
|Conference||48th AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics, and Materials Conference|
|Period||23/04/07 → 26/04/07|
- conference contrib. refereed
- Conf.proc. > 3 pag