I was wondering if anyone had investigated the use of air motors for use in the reprap.
A few advantages would be less weight and it is probably something that could be
built by the reprap itself. Both the pump and the motor.

Just throwing it out there as a thought. I could not find any discussions about this
doing a search of the forums.

I did a preliminary milled design for a stepper driven by compressed air and did the drive plates. I got sidetracked by a bunch of other things that wanted doing, however.

---

-------------------------------------------------------

Hell, there are no rules here - we're trying to accomplish something.

Opportunity is missed by most people because it is dressed in overalls and looks like work.

Thomas A. Edison

Well as of recently, I've been working on designing a pneumatic stepping motor called a wobble motor for Darwin.
For an explanation of how a wobble motor works see here(site appears to be down last time I checked):
[www.act.sys.okayama-u.ac.jp]
Or if you have an IEEE account see this:
[ieeexplore.ieee.org]

Wobble motors have high torque and high resolution so it certainly
It seems like we could certainly make one that meets Darwin's requirments for a stepper motor and is printable by Darwin. Some of the disadvantages of pneumatic wobble motors are that they tend to be slower(slowness isn't an issue if you're aiming for replication...) than electrical motors and driving them; usually wobble motors are driven with expensive solenoid valves. However, we don't need to worry about solenoid valves if we replace all of reprap's control electronics with more fluidics, which are within the ability of reprap's printing capability (http://en.wikipedia.org/wiki/Fluidics). Then we could either use paper tape or some sort of acoustic PC interface(IE speakers hooked up to some fluidic amplifiers for input, a pneumatic whistle hooked up to a microphone for output) to control it. This has the principle advantage of making reprap more replicatable and thus cheaper.
Here's an example of a fluidic stepper driver:
[ntrs.nasa.gov]


So far I have begun designing a NEMA 23 size pneumatic stepper motor designed to be compatible with Darwin. For maximum replicatability I intend to use rotary rolamite bearings.(http://www.rexresearch.com/wilkes/1wilkes.htm see figure 16) for the bearings. To make the flexural bearing bands and the wobble generator I propose using reprapped molds to mold silicone sealant.

The only problems I seem to run into are:
1. Making the indented motor shaft. It might be possible to eliminate this by redesigning Darwin to use stepper motors with larger diameter replicated shafts.
2. What to do about tubing connectors, is reprap accurate enough to make small diameter tubing connectors capable of withstanding ~70-100 psi? If not, the only option is to buy tubing connectors, this is not preferable as it increases the price and introduces a closure problem.

... maybe the wobble motor in combination with reprappable pneumatic (or fluidic) muscles (see regarding threads) will be a promising way?

I'm thinking about moving ferrofluid with some surrounding solenoids around the cavities instead of air ...

Viktor

Well it might be possible to use fluidic muscles to do fine positioning of the extruder if the extruder's/toolhead's absolute position could be be determined accurately.

I want to use a fluidic control system as opposed to an electric based one is that we can currently print one using just plastic.(IE it's easier to replicate) Ferrofluidic actuators offer an interesting possibility in that it might be possible to "grow" ferrofluid. There many different species of bacteria that produce tiny magnetite grains(the prinicipal component of ferrofluid) so that they orientate with the earth's magnetic field so they swim down where conditions are more favorable for them. If we could grow these bacteria we might be able to make ferrofluid.