Learning Forum

[Florisvanelteren]

I am trying to model a shape-memory-wire bending actuator and I am running into some issues, and I'm wondering if anyone has tips! An illustratrion of what I am trying to model is included. The shape memory alloy wires will contract from a prestrained condition when heated. At room temperature the model looks like the one in the illustration. The SMA wires are embedded into a rubbery matrix, and attached to a (stiffer) elastic substrate. when the SMA-wires are heated up, they contract and cause the whole actuator to bend due to the recovery stress that they generate, the elastic moment generated by the matrix and the substrate will balance out the recovery stress of the SMA until a certain curvature is achieved.

- I want to model the function of this actuator with large deformations and connected to another sheet structure, so a simple model is preferred (the elastic substrate must be modeled as a sheet).

- I am using ANSYS workbench mechanical.

- I would prefer not to use the SMA-material type as this does not seem to comply with what I've found out experimentally with my own SMA wires

I have two options in mind 
1)modeling the matrix material as a solid, and create a line of nodes within the matrix material on which a stress/strain relation can be asssigned (at 0% strain between the nodes -> -500Mpa between the nodes, at 1% strain,-300 Mpa, etc. in a linear relation). This seems like the most simple option, but I can't find out how to do that.
2) modeling the matrix material as a sheet, and modeling the SMA wires as a very thin sheet which has a certain stiffness and have a thermal contraction assigned to it, however I need a thermal contraction only in the direction that the SMA wires are going, and I can't find out how to do that. Also, in that case I don't know how to properly connect the sheets to each other.

Some help would be greatly appreciated, thanks in advance!

Cheers,

Floris

[peteroznewman]

Option 2 can be realized with two materials that use Orthotropic Thermal Expansion.
One material has a positive coefficient in X and zero in Y and Z.
Another material has a negative coefficient in X and zero in Y and Z.

I created two solid bodies in a single part with Topology set to Share.
I fixed one end and applied a 100 C thermal load to get this deformation.

  

Attached is the ANSYS 18.2 archive of this model. I could easily add an elastic matrix around three wires and a sheet substrate to get what you want. You can show your appreciation by clicking "Like" if you find this post helpful.

Regards,

Peter

[Florisvanelteren]

Hi peter, 

Much appreciated! my idea is slightly different (I will use an SMA wire that simply extends, with the counter force supplied by the elastic substrate and the matrix) but this should get me going in the right direction, thanks!

Kind regards,

Floris

[peteroznewman]

Hi Floris,

Perfect! If you only need thermal expansion, then you can have a beam model built from a line body and you don't need a solid body of the wire. I had to use a solid body with two solids to get the bending, but you will get that effect with your substrate.

Regards,

Peter