ROBOTIC HAND!!! Upgrades

[Souper man]

I need some Ideas for upgrading my High School's robotic hand. Right now, It runs on a motor (specs unknown) that turns a long worm gear that anchors a cable that pulls 5 wires that in turn pulls each finger closed. There is a Switch that controlls the motor going forward and backwards.

More to the point: They want me to upgrade the Hand so that it can move each finger individually. I have 2 ideas:

Muscle wire
Motors/servos

Muscle wire contracts when no current is flowing through it, but expands when high current goes through it. the website of where to get it is here. It is also called flexinol wire. www.dynalloy.com

The Motor Idea is that a servo has a spindle that pulls in the cable itself. This will take up more room, but the Muscle wire idea will conduct electricity on its metal frame and it actually heats up.

Please help!

Pictures as follow!
(this is one of the coolest robotic hands You may have seen!)

1324: Fingers of the hand. Rubberbands help the hand go back into original position. Tension springs are also shown, they help reduce over stressing the fingers.

1325: Picture of worm gear and motor. Anchor for cables are also shown in left hand side. Switch also shown.

1326: Entire hand.

1327: Thumb and Inner hand shown. Black foam to help protect from rivets and bolts. Increases grip.

1328: Too large, but showed underside of arm. Unremarkable.

1329: Central cable connector. all fingers and thumb connect to this. I plan to replace this part and spiral gear with muscle wire or motor cables.

1330: End of fingers. Shows the Rubber bands end point. I plan to keep the rubber bands in place.

I am favoring the muscle wire idea, and I am funded by the school. OH YEAH, THAT MEANS BIG CASH MONEY FOR ME!!!

I plan to have insulators between the connection points, and a plexiglass divider for the inner portion. More pictures and stuff much later on, but post your ideas here please!

A 5 button finger pad will be placed and will be able to swivel into operators choice of position.

 

[Souper man]

NOTE: Price of parts do not matter, much!

 

[dknguyen]

You could try air muscles? That's what I'd use if I had access to a compressor.

But I'd stay away from muscle wire for something like that. The forces needed to overcome the elastic bands that hold the fingers opens is much larger than a reasonable amount of muscle wire can put out (well it depends what thickness you use, but muscles wires are very inefficient so larger ones need to heat up much more and cool down much more slowly). The big problem is the displacement of the muscle wire. THe metal of the arm might also act as too good a heatsink- either way, the current required to heat up the wires (combined with the resistance of the wires) might make it tricky to find a power supply that has a high enough voltage and enough current.

I'd go for motors mounted on the forearm (if not air muscles).

 

[Souper man]

I do not have access to compressed air. The muscle wire I am using heats up to 158F, takes 610Ma and 9v to expand, pulls/operating strength of 590 grams, 0.08 in thick, takes 1 second to heat up and 2.2 - 3.5 seconds to contract. This muscle wire is efficent

 

[dknguyen]

To get efficiency compare how much power you put into the wire to how much work you get it out of it- it's probably not going to beat a motor.

Is the displacement of the muscle wire enough for what you want?

 

[Souper man]

I wanted somthing that would closely simulate tendons, and muscle wire is perfect. Efficency isnt great, but It works. I am going to use PVC separator, and I may make a model of it.

 

[Hank Fletcher]

My vote is for a motor-and-marionette combination, although I have to admit it's far from ideal. Let's face it, the best humanoid robots will be more human, i.e. more unity in the parts constituting the structure, power supply, and motivation than a metal chassis, heavy battery, and cyclical motor. So for the coolness or avant-guard factor muscle wire is the way to go. The air muscles are too noisy and bulky for what they deliver, in my opinion.

My reason for leaning towards motor-and-marionette is the same as the other folks here: more power for less effort, meaning your effort. It's much more conventional this way, so you'll have a lot easier time of troubleshooting and hitting the community up for questions. But as long as someone else is footing the bill, I'd have to say go for the adventurous route. Just make sure that their expectations (and your own) of what the final product will be able to do are realistic, so everybody's happy and you don't burn any bridges for future funding.