Shape Memory Alloy Simulation
-
-
Hello, I am simulating nitinol SMA shape memory effect and am having some issues with the shape recovery. I'm using a simple hollow cylindrical (10mm diameter, 10mm long) peg and I've applied a fixed constraint to one end and a displacement of 0.5mm (5% strain) to the other end. The displacement is applied in step 2 then deactivated in the following steps. I then apply a thermal condition to heat the material to and above its transition temp (37C, then 40C, then back to to 37C). As I understand it, based on this material model (see attached picture for material properties) the peg should regain its initial shape after undergoing the thermal stimulus since the applied strain is less than the recoverable strain (8.6%). I've made a stress-strain curve for conditions where the thermal stimulus is applied and suppressed to try and better understand it, but I'm confused as to why the shape is not fully recovering (expecting strain to return to near-zero). Please advise how to change the model and/or simulation conditions such that the material regains its initial shape after deformation and thermal stimulus. I'm using substeps and large deflection is currently off.
Peg, meshed with linear hexahedral elements:
Analysis, where green is displacement in step 2 (then deactivated) and purple is thermal in steps 4-6:
Without thermal stimulus:
With thermal stimulus:
Strain with thermal stimulus (shows some recovery when heated from 37C to 40C)
-
-
Thanks for the link, I'm afraid it doesn't address the same issue. Where the other post describes a non-convergence issue, my simulation converges but I'm not seeing the near-zero residual strain. I've followed the spinal spacer example quite closely in terms of setting up the model, so I'd like to clarify my understanding of the material definitions. For the reference temperature, is this the austenite start temperature? If so, how is the austenite finish temperature defined (ie. temp where phase change goes to completion)? I've achieved more shape recovery than the initial post by changing the Newton Raphson to unsymmetrical (thanks to the linked post) and as seen in the attached picture, stress reduces but there is still over 3% remaining strain even though the max strain was roughly 6% (and recoverable strain was defined as 8.5% if I did that right, please see material definition above). Also worth noting is that this shape recovery happens only when temperature significantly exceeds the reference temp (ie. by 10+ degrees). Can you please clarify these material properties and advise how to improve the strain recovery? Thanks
Stress-strain curve:
Displacement of node on the bottom of the cylinder (where displacement is applied, then removed) over time:
-
-
-
-
- The topic ‘Shape Memory Alloy Simulation’ is closed to new replies.
You are navigating away from the AIS Discovery experience