CS 154: HMC Robotics Lab 2
Purpose This lab began as an attempt to design an autonomous robot using muscle wire ( NiTinol). Muscle wire is a nickle titanium alloy which contracts when heated. The lab evolved into an exploration of the limits of muscle wire.

Part 1: Matchbox Leg
We bagan with a fairly simple idea, an articulated leg which would pull and lift so that it could progress without dragging its feet. We did not have a lot of materials with which to experiment, so we created a prototype leg out of matchsticks and a matchbox. This prototype worked once, after which it refused to work again (see "What we learned" below).

Part 2: Staquito Leg Modification
We then decided to cannabalize parts from a broken Stiquito robot. This consisted mostly of a leg and a piece of grey plastic. To this, we developed a better design which uses the leg as a conductor to power the muscle wire. This part taught us some basic techniques for making the pieces for the leg (for example, use an Xacto type knife to cut copper tubing as it does not cause the tubing to collapse upon itself). This leg worked much better than the matchbox monstrosity. This leg uses about 10cm of muscle wire and about as much music wire (which is quite usefull as it can be bent in practically any shape you desire). We powered it with 3V and it worked fine. However, 9V proved a little excessive for the top wire (see "What we learned" below).

Part 3: Advanced Leg
The third leg for this project was developed from a design in the Advanced Stiquito book. This leg uses one long piece of muscle wire, a longer piece of music wire and some insulation cannabalized from some 12guage wire that was lying around. This design uses the tension of the unheated muscle wire to keep the muscle wire from contacting the conducting music wire. This leg is the first which has been able to keep working. It is also the only one which we have made so far that can lift weight (we lifted some legos as a test). This leg also does not seem to mind the 9V power supply that killed the other legs.

Part 4: Butterfly
We then decided that it would be nice to have something that is pretty. We decided that a butterfly would be nice. (click on pic for more...) One of the project books had a basic design which we modified to support our heavy wings. The basic idea is that the muscle wire which runs down the center of a tube pulls a center "wing lifting" piece. The wire contracts and the wings flap (see Bryce for a demo). This project really disliked 9V but was a little slow with 3V so we thought we would try it. Oops.

Where to now?
The idea behind the leg development is to find something that will work for an autonomous light seeking robot. We invision an robot which walks by inching along.

What we learned
First and foremost, muscle wire is fairly sensitive to an improper power supply. We got lots of smoking wires when we tried to give it too much power. Oops. The matchbox robot even glows. It seems from our experimentation, that shorter wires are much more likely to burn out quickly. Muscle wire also must be under a well balanced tension. If it is not under enough tension then it will just sit there and flex but not pull. If it is under too much tension, things do not work right and you can damage the wire. Muscle wire also has an interesting reaction to fire. We played with a short piece of muscle wire and a couple of matches. The wire would move towards the flame until it was straight. Pretty spiff. We also tried to convince the wire that it wanted to stay in a spiral, this was not a success.

Conclusions
Muscle wire is neat. It is relatively inexpensive, fairly easy to make something that works. However, it also requires some careful engineering to get a setup that will do exactly what you want and getting the wire properties right can be a touch tricky at times as well.