Coiled muscle wire

Background

I am starting to experiment with a new type of robotic actuator, called a coiled muscle wire (among a few other terms). Essentially, they are a new class of linear actuator that contracts (or expands, depending on how you make it) when heated. Coiled muscle wires are reported to be very strong for their scale, relatively easy to fabricate, and incredibly cheap to produce.

Muscle wire I produced

This particular muscle wire is made of nylon fishing line, but people have reported success with a number of different materials, including conductive sewing thread.

Goal

The goal here is to produce some small controllable unit of robotic actuator, where a given input from a control device produces a predictable output. With some exceptions, laboratory muscle wire experiments seem to concern themselves with the material properties of the muscle wires themselves. I want to take that information and produce some usable muscle actuated device–a proof of concept for usability in robotics.

Challenges

One of the challenges of coiled muscle wire is that it needs to be heated in order to actuate. A popular way of achieving this is by resistive or Joule heating, where you simply run a current through an appropriately resistive circuit, and some of the electrical energy is converted into heat.

There are two strategies I’ve seen so far in terms of resistive heating of coiled muscle wires:

  1. Have the coiled muscle wire be made of a conductive material itself, and heat directly
  2. Have a separate resistive heater that works on a non-conductive muscle wire

These two options will clearly guide material choice. Having a conductive muscle wire brings all the requirements into one tidy package, and so it is worth exploration. On the other hand, separating the two functionalities (heating, and muscle flexing) into two materials may also be wise: getting one material to accomplish two separate tasks at once may be asking too much.

Initial Tests

I’ve tested making muscle wire with a variety of materials, and the most reliable material seems to be nylon fishing line. It is the least prone to breakage during fabrication and tolerates heat well. Obviously, nylon fishing line is non-conductive.

Conductive thread is very prone to breakage, coils in an unpredictable way, and most importantly is very intolerant of excessive heat. I did some simple tests running a current through a section of thread and once too much heat is applied the material breaks down completely.

Therefore my intention is to continue with the nylon muscle wire and design an external resistive heating element. Designs for that are forthcoming but I may create a fabric-like weave (seen here) where a heating circuit is woven into the material. This blogger has been a particularly good resource (thank you WriterofMinds!), and she found that Litz wire (a special kind of wire that is a bundle of individually isolated copper wires) is really good/flexible for this purpose.

Once I have a muscle wire setup that is externally heated, I want to integrate that into some small actuator/moving part, and begin working on the control layer.

Leave a Reply

Your email address will not be published. Required fields are marked *