limit rotational speed with a mechanism?

[lysis]

Hello everyone, first post.

I'm working on a project where we need to incorporate a mechanism into a gear train which will limit the gears from spinning too fast. We will use a torsion spring to accelerate quickly to, say, 2000 rpm, but once we hit this ceiling we want to keep the rpms around 2000, while still using the energy in the spring.

We cannot use any kind of electrical speed controllers, only mechanical components.

I could really use some inspiration, does anyone have any ideas or leads?

Thanks

 

[MintJulep]

http://en.wikipedia.org/wiki/Centrifugal_governor

 

[hydtools]

Air fan drag like tape cassette doors or music box mechanisms?

Ted

 

[lysis]

Thank you guys, I appreciate the help

 

[racookpe1978]

Problem you're going to find is that gyroscopic regulators like this were very common in steam turbines and steam engines (piston and turbine) before electronics were invented.

But ... they were regulating a constant-force flow of steam (that is, they were changing a control valve position based on turbine or engine speed), not a releasing a wound-up spring slowly expanding and letting its energy come from the relaxation position of the end of the spring actuator.

 

[jmw]

A fly-ball governor was my first thought - but then I wondered, does this also limit acceleration?
Seems to me there isn't a lot of hunting or overshoot with this type of control but I guess that comes from also having a damn great flywheel?

Incidentally, Citroen's 2CV was available with a centrifugal clutch. Now that was neat - on a 500cc 2 cylinder engine.

JMW

http://www.ViscoAnalyser.com

 

[jcrags]

Another option would be a rotary damper. These things offer a resistance that is proportional to rotational rate, so at some speed everything would balance out.

http://www.mcmaster.com/#rotary-dampers/=ex7piy

 

[PersonalProfile]

Lilly controller!

http://www.winder.co.za/Lilly Controllers.htm

Regards,
Lyle

 

[IRstuff]

Some sort of eddy current braking might be possible. The braking effect is proportional to speed, so that's in the right direction.

Dashpots tend to have resistance proportional to velocity as well, since the flow is limited by a restriction orifice.

TTFN

FAQ731-376
Chinese prisoner wins Nobel Peace Prize

 

[EngMark]

I think that the type of fan used in a music box (as previously suggested) is the way to go on this. The regulation might not be perfect and it will limit your acceleration a bit. So it depends on how tightly you need to control the acceleration and speed.

The flyball governor as used on steam engines is a bit different than what you are looking for in that it operates a valve which then controls the engine speed.

But taking that idea and adding it together with the eddy current brake suggestion, how about this - are you familiar with the centrifugal advance mechanism used on pre-1980 automotive distributors, with the weights and springs? A spring(s) would retain a weight(s) in toward the center of the rotating shaft until the shaft speed was high enough that the weight(s) overcame the spring force. Similarly, you could build a magnet / spring arrangement on your shaft in which the magnet(s) were stowed at a small radius until the critical speed was reached, at which time they would move radially outboard to cause eddy current braking in a conductive annular disk. Hmmm, gotta go find my blank patent application forms...

Mark

 

[MintJulep]

The flyball governor as used on steam engines is a bit different than what you are looking for in that it operates a valve which then controls the engine speed.

No.

A flyball governor produces a linear motion output proportional to speed.

When used on a steam engine that linear motion output is used to open or close a valve.

But there is no reason why it couldn't be used to apply a friction brake, change the gap of a magnetic brake, open and close a damper on a fan, etc.

 

[CRG]

EngMark,

You are describing a similar (friction instead of eddy current) speed control system used on my 1920s Victrola. It has a governor that actuates a brake on the inside rim of the platter. Think fly-ball that actuates a friction brake instead of a steam valve.

 

[FeX32]

Well then.
IMO. To properly do what you describe you need a mechanical control mechanism. A centrifugal clutch or similar won't really limit the speed of the shaft in the same way. It may be possible but I doubt the accuracy of a simple clutch system used for this purpose. Basically you want some sort of mechanism to limit the torque the spring transmits to the shaft in such a way that the set point is calculated to be the point where inertialy the shaft is spinning at 2000rpm. There are many ways to do this. But most of them are more complicated than you may think. Not easy to construct, but really cool and work well. An example I got from a well known text on control engineering by Ogata. See attached.


Fe

 

[FeX32]

You can simply use an idea like the above and modify it to change the torque output of the spring instead of changing the position of a valve. It's basically like a governor as other engineers above mentioned.


Fe

 

[potteryshard]

The fan or friction controllers are simple and do work, but waste most of the spring energy. The escapement mechanism of a mechanical watch is highly efficient, but as usually configured results in a very low speed intermittent action. The concept could be adapted by arranging that the escapement trigger is operated by a speed-based governor instead of time-based control. Spring energy would be stepped up through a gear train, and a certain finite movement of the spring would toggle the escapement back to it's initial state. Obviously a one-way clutch and some flywheel effect would be needed so that the intermittent energy bursts from the escapement could be added to the rotating shaft.

 

[EngineerTex]

You can look at the old Ford 8N tractor governor and either use that as a basis for your design or buy one rebuilt and bolt it into your application (already set for 1800 rpm). It has a gear input and a crank arm output. Link the crank output to throttle back the gear train however you want it to (band brake or flywheel/sprag like potteryshard suggests or whatever else you like).

It shouldn't have a lot of drag to retard your acceleration, unless this application is something like a pocketwatch, in which case, you've got some serious miniaturization and probably bigger challenges than the governor.

Engineering is not the science behind building. It is the science behind not building.