Shape Memory Home      Exposition      To Register      ASM Homepage
Back to "Session 10: Device Modeling and Testing" Search
    Back to Main Search

Tuesday, May 9, 2006 - 5:10 PM
MEM10.7

Thermoelastic Shape Memory Modeling of Medical Devices with FEA

T. K. Parnell, PEC - Parnell Engineering & Consulting, Sunnyvale, CA; S. Choudhry, MSC.Software, Redwood City, CA; T. J. Lim, Stanford University, Stanford, CA

The shape memory effect is very useful for medical devices.  Its effective use opens up many new design and treatment options.  The Shape Memory Effect (SME) allows for a particular configuration to be set in a part while it is hot, the part deformed while it is cold to a new shape, and then the original shape recovered by heating.  In this way, the part has “shape memory” of the original shape and can recover it through heating only from another stable configuration.  

 

The nonlinear FEA code MSC.MARC has a thermoelastic shape memory material model that can model this sort of behavior.  The MSC.Marc model is based on a phenomenological approach that provides a description of a wide range of the observed behavior, and which is tractable from both an analytical and a computational viewpoint.  The model describes the development of transformation induced inelastic strains by austenite to martensite transformation and by additional texturing of martensite.  The implemented material model accounts for elastic strain, conventional plastic strain, thermal strain and strain due to phase transformations through the technique of additive strain decomposition.  The phase transformation strain evolves due to the formation of oriented, stress-induced martensite arising as a result of the reorientation of randomly orientated thermally induced martensite.  The phase transformation strain is further decomposed into TRIP (transformation induced plasticity) and TWIN (twinning) strains.  

 

We outline the basis of the model and also show some numerical examples.  The numerical examples illustrate the thermo-mechanical behavior and are presented to show the robustness and effectiveness of the proposed model.