Clutches

[Andrew Blair]

I was wondering what kinds of shifters people have used on their gearboxes. Dog shifting is popular, but I was thinking more about mechanical and magnetic clutches. Can you even use magnetic clutches (first legal?)? They might suck too much current in the first place.

[NoodleKnight]

I'm pretty sure they are legal, I think the main reason why people avoid using a clutch in their drivetrain is because they don't want any slipping before the wheels contact the ground.
The most common type of shifting (from what I've seen) is the dog-shiftuing method, like you said. There's mesh shifting, where you basically move the gears instead of a dog. Then there's more complex ones like a continuously variable transmission, or a CCT. Last one I can think of is where the robot actually switches around it's drive wheels. 980 did this in 2003, it was really neat to watch them shift, here's a picture, it ain't that great, but you can see that they have an extra pair of wheels; larger ones for higher speed.

Edit: Just remembered, 222's ball shifting method is another one. It's hot stuff...

[sanddrag]

Here's a quicktime movie of 980 shifting wheels. I think 258 did the same thing in 2002.

As for clutches, I don't think I've ever really seen one on a FIRST robot. It is really not necessary. For how relatively easy it is to slow down the motor(s) and for the wheels to slip on carpet, a robust shifting mechanism is all that is really needed to shift. No clutch required.

Here's one good use of clutches though. If the game requires you to get back to the endzone, but you are still out in midfield with 2 seconds left, you would want to race toward the home zone and then disengage the clutch right before the buzzer so you can coast the rest of the way.

[mechanicalbrain]

actually ive thought of using magnetic clutch. it would work allot like magnetic activated brakes. it would grab onto specific gears of release them.

[Tristan Lall]

There are a few issues regarding electromagnetic clutches (COTS or otherwise), if we use the 2005 rules as our standard.

Firstly, see the Part Use Flowchart: electromagnetic clutches are electronic components (they are designed to conduct electricity), so we ask "Is the part a motor, solenoid, pump or other actuator?" I'd say yes (either to solenoid or other actuator), and refuse to allow it.

Even if it were to pass (somehow), it would be subject to all of the electronic component rules, such as part cost, etc..

As for pneumatic clutches (COTS), I don't think that it could be argued successfully that a clutch is a cylinder (actuator or tank), rotary actuator, fitting or valve. Therefore the flowchart bans it.

[NoodleKnight]

Originally Posted by Tristan Lall

There are a few issues regarding electromagnetic clutches (COTS or otherwise), if we use the 2005 rules as our standard.

Firstly, see the Part Use Flowchart: electromagnetic clutches are electronic components (they are designed to conduct electricity), so we ask "Is the part a motor, solenoid, pump or other actuator?" I'd say yes (either to solenoid or other actuator), and refuse to allow it.

Even if it were to pass (somehow), it would be subject to all of the electronic component rules, such as part cost, etc..

As for pneumatic clutches (COTS), I don't think that it could be argued successfully that a clutch is a cylinder (actuator or tank), rotary actuator, fitting or valve. Therefore the flowchart bans it.

What about a normal friction-based clutch, which crams together two friction disks?

[Tristan Lall]

Originally Posted by NoodleKnight

What about a normal friction-based clutch, which crams together two friction disks?

If actuated with some legal mechanism (e.g. HS-322 servo, kit motor, pneumatic actuator from the list, etc.), it was legal in 2005, and ought to be legal in the future.

I think 322 used one of these a few years ago (2003?).

[henryBsick]

//NOTE\\
*Verification of this information is pending...*

Team 303 used a what I guess would be considered a friction clutch on their 2005 robot. I am not sure on the specifics but it comes down to 2 gears and one engagement device on the same shaft. The gears and the engagement device (which is in between the two) do not have flat faces. The gears are sloped inward to a line that bisects the gear and the engagement device has two faces per side that slope outward from that bisecting line. Crudely demonstrated with text symbols it looks likes this:
__ . __
l / /\ \ l
l \ \/ / l


The gears' sloping faces have highly packed felt pads which stop the engagement device (laterally powered by pneumatics) from grinding on the gears' faces. It is a friendly version of dog shifting from what I can see of it.
For any more info contact someone from 303.

[Veselin Kolev]

Shifting in cars is done similarly to the dog-shifting method. However in cars, there are syncros and there is a clutch. The clutch is just there to prevent power from being transmitted through the gears while they shift, so the load on them is less. The syncros are to match the speeds of the shifter and the gear so that they shift smoother without grinding. In robotics however, shifters are made of beefy steel dogs, so you don't really need to reduce the force of shifting with a clutch. I do not see any advantage to having a clutch on a robot shifting transmission. Heck, it might even take longer to shift if you have to engage and disengage the clutch all the time.

However, there can always be uses for something like a clutch. Sadly, the only thing that closely resembles a clutch that I can think of at the moment would be disc breaks. And a random thought, wouldnt it be awesome if you had 4 wheel drive using one power source and 2 differentials and shafts, each wheel having disc breaks, so that if a wheel loses traction, you lock the disc breaks, rerouting the power to the other wheels? It would be very impressive to actually have that many working feedback encoders on your robot... just a random thought.

[sanddrag]

Originally Posted by Veselin Kolev

And a random thought, wouldnt it be awesome if you had 4 wheel drive using one power source and 2 differentials and shafts, each wheel having disc breaks, so that if a wheel loses traction, you lock the disc breaks, rerouting the power to the other wheels? It would be very impressive to actually have that many working feedback encoders on your robot... just a random thought.

They have that on lots of full size cars and from what I read it workspretty well. On a robot, I don't think it would be practical to have all your power coming from one gearbox because then yu could not have one side going forward and the other going backward for turning in place unless you had a reverse gear, which would be overly complex.

Also, since in most robots the wheels are linked with chain, the wheel will always be driven regardless of whether it is in the air or on the ground.

But to answer your question, yes, it would be cool.

As for brakes, I seem to recall one team using a pneumatically actuated disk brake to hold their arm in position.

[Veselin Kolev]

Originally Posted by sanddrag

Also, since in most robots the wheels are linked with chain, the wheel will always be driven regardless of whether it is in the air or on the ground.

Remember that the only reason you would really have drive shafts on a robot is if you are doing 2 or 4 wheel steer, just like a car. No differentials = no turning.