Questions about air muscles
Brian Mullen
I have noticed two ways to make them. McKibben's way, where the muscle
lengthens with air pressure, and the Shadow robotics way where it contracts
with air. Does anyone have any experience with these? Which way is better?
Where is a good supplier of the materials needed to make one?
mar...@shadow.org.uk
Hi!
Hm. It's all rather tricky and depends on many factors. Our early
models were very poor indeed (a few ounces). Now four of them will
lift a car. I am sorry, but I cannot divulge sources and commercially
sensitive information. We have spent a lot of money and energy
on research; we are on mark 200 or so.
I can, however, post some elementary theory if that is of interest to
people.
As for your question...
Pulling is inherently stronger, just on the theory of the braid.
In our experience the muscles are very quick and easy to use.
Attaching them with cable ties or kevlar rope allows easy adjustments.
Controlling them is trickier. You have to have your whole
arrangement in a feedback loop. We have some application notes
on our web site which may be helpful, but if you want speed
control you will have to solve this to your own requirments as
we have not had to build anything with speed control as yet.
The compliance is useful. The return force goes up with the square
of the displacement. This means that an antagonistic muscle pair
will have an energy minimum even when the valves are sealed.
Handy if you want to just hold something that someone is likely
to walk into by mistake. The compliance can be controlled by
varying the pressure in the muscles. One way to do this is to
pulse a 3/2 valve. By changing the mark to space ratio of the
signal, variable position and compliance should be posible. We
are going to test this very soon (as soon as our test rig is
finished).
yours, Marcus.
--
Marcus Baker, Shadow Robot Project, London.
EMail: Mar...@Shadow.org.uk
Web: www.shadow.org.uk
Rich Walker
In message <19970814....@shadowrp.shadow.org.uk>
mar...@shadow.org.uk wrote:
> Hi!
>
> > I have noticed two ways to make them. McKibben's way, where the muscle
> > lengthens with air pressure, and the Shadow robotics way where it contracts
> > with air. Does anyone have any experience with these? Which way is better?
> > Where is a good supplier of the materials needed to make one?
Buy them!
> >
[Snip's most of Marcus's reply, except:]
> Attaching them with cable ties or kevlar rope allows easy adjustments.
Kevlar (TM), sorry...
Cheers,
Rich.
--
Rich Walker: r...@shadow.org.uk (Shadow Robot Project)
http://www.shadow.org.uk 251 Liverpool Road
+44(0)171 700 2487 London N1 1PX