Catapult?

[Rick TYler]

We rejected a catapult early on because of the inherent limitations of rate of fire. Spinning wheels or belts have such an obvious advantage in fire rate that we didn't seriously consider a catapult for long, although some students did some interesting math on one.

So, this is just FIRSTnoodling, and not a serious design proposal. I'm also not going to do the math, so forgive me for that, please.

I agree that one of the problems with catapults is the waste of energy involved. A lot of energy is used accelerating the arm, which is then usually wasted at the end of the stroke. How does this calculus change when the arm weighs essentially nothing? You could build a very stiff launching arm three feet long of hand-laid carbon fiber and fiberglass on a foam form that would weigh about about six ounces. If you molded a CF/glass dish on the end for the ball it might add another two or three ounces. I wonder how the math would work out if a pneumatic cylinder pushed on a 6-inch lever arm driving a 30-inch throwing arm on the other side of its ball bearing pivot? I have to get back to work, but perhaps some enthusiastic physics student wants to work out the numbers.

[kmcclary]

Originally Posted by Rick TYler

We rejected a catapult early on because of the inherent
limitations of rate of fire. Spinning wheels or belts have such an obvious
advantage in fire rate that we didn't seriously consider a catapult for long,
although some students did some interesting math on one.

So, this is just FIRSTnoodling, and not a serious design proposal. I'm also
not going to do the math, so forgive me for that, please.

I agree that one of the problems with catapults is the waste of energy
involved. A lot of energy is used accelerating the arm, which is then usually
wasted at the end of the stroke. [...]

Hey Rick! Email me privately and tell me how your build is going
(and what you did with the data I sent you... )

Yea even if you COULD do one, it's obviously not going to win anything
(but the crowd) this year, so don't take my comments too seriously.

But as you can see, if someone asks me to "noodle", I'll take a
real crack at it...

Now if a catapult was a serious payload widget contender, you'd
still have to power the thing, so my comments about pneumatics
vs motors (et al) stand.

Another one of "Keith's Rules":

"Stall can be managed. You must first look at the 'total energy
per round' and the 'energy transfer rate' requirements of
each device, before attempting to allocate actuators to
widgets on your machine. Right off the bat, those two criteria
may point you to the best actuator for the job."

I'm hoping at least that part of this discussion may be useful to
the next student trying to the classic "motor vs spring vs
cylinder" choice.

But yea, though it wouldn't win, I still feel a catapult, a "Cesta Wheel"
or some other "Rube Goldberg" throwing widget would offer a LOT
of "ohh, ahh" crowd entertainment value! (Especially if it whips
the ball wildly off course!)

Originally Posted by Rick TYler

How does this calculus change when the arm weighs essentially nothing?
You could build a very stiff launching arm three feet long of hand-laid
carbon fiber and fiberglass on a foam form that would weigh about about
six ounces.[...].

Well, please realize that a catapult arm is not simply a stick,
whose strength you're trying to maximize, while at the same time
minimizing weight. It's a distributed spring. Just like a bow, the
arm itself acts as an energy storage device. The material local
to the pivot rotates much faster than the load itself. In essence,
the arm "<sets>" itself (there's that banned CDF "c" word again!)
by transferring potential energy from the spring into its length.
It then re-releases that energy, converting it into Kinetic Energy
in the load mass as it accelerates. IOW, the compliance of the arm
is a huge part of the total system.

- Keith

[Ianworld]

I don't want to get into the fine details of actually making one work but the power is certainly there. A 2" bore pneumatic cylinder pushes with about 180+ pounds of force. Since work is force x distance you can multiply that out. (180 lbs of force * 9 inches of travel) inch pounds in joules is according the Google calculator 183 joules. Thats a ton of energy! I'm not sure if thats exactly how a pneumatic cylinder would translate into energy but its got to be considerable none the less. As was stated above a pouf ball needs about 13.7 joules to to go at 12 m/s(instead of the 18 joules for a single flywheel design due to the spin). That means that a 2" bore pneumatic would in theory only have to travel about an inch or two to put enough energy into the ball assuming a pretty thorough transfer of energy and a very light throwing arm. Just some numbers to think about. I'm still sticking to a fly wheel.

[Alekat]

Originally Posted by Ikeyballz

The good thing about this was that our robot did not shoot at different velocities because of the change in our battery level.

Consistency is very beneficial. But the same consistency can be accomplished with the right sensors. Of course this is only in theory for my team, seeing as how this is the first time we've incorporated a sensor into regular game play.

[KenWittlief]

Originally Posted by Ianworld

I don't want to get into the fine details of actually making one work but the power is certainly there. A 2" bore pneumatic cylinder pushes with about 180+ pounds of force. Since work is force x distance you can multiply that out. (180 lbs of force * 9 inches of travel) inch pounds in joules is according the Google calculator 183 joules. Thats a ton of energy!

understatement of the day! 183 joules could fire 13 balls all at once!

what was all that talk about cycle times and being able to fire fast enough?

[Marcel]

The problems I see with a catapult is that you can only set it to shoot at 1 angle I cant think of any way to change the angle in the middle of a match meaning you'll have to be the same distance away from the goal at all time. You cant set up to shoot 2 feet away and than be like 4 feet away and hope to make a goal. Also you'll only be able to shoot once, reload, and shoot again. There are much better designs that allow you to fire more than 1 in one load. ^_^

But thats just what I think.

[greencactus3]

Originally Posted by KenWittlief

understatement of the day! 183 joules could fire 13 balls all at once!

what was all that talk about cycle times and being able to fire fast enough?

and thats assuming that the compressor keeps the tanks at a consistent pressure throughout the stroke of the cylinder... unlikely.
-edit- i guess the tank pressure doesnt quite directly matter in all cases... it just has to stay above 60psi right?

[kmcclary]

Originally Posted by KenWittlief

understatement of the day! 183 joules could fire 13 balls all at once! what was all that talk about cycle times and being able to fire fast enough?

<SIGH> Hmmm... How do I put this?

What you're talking about is the theoretical total energy contained
in that much compressed air. That's a lot different than what a
typical cylinder can provide in output work, in a short time.

And YES, you are still ignoring cycle time of the system
(IOW, how fast it can get ready for the next shot).
Think about it... Well SURE, it sounds great! Yes, if you could
release all of that theoretical energy in an instant, you'd fire
a whole bunch of balls, and that would be pretty cool!
(You've then just "shot your wad". It'll take time to get
it charged up again.)

But that's not how cylinders work. You can't just use the
raw air (that's "making your own cylinder", a no-no), and
you're STILL limited to whatever that darn thin tubing can
supply you.

Ya can't get all of the work out, either. (Well, not in any
reasonable amount of time, anyway... ) Cylinder motion
is quite a bit like a capacitor charging in electronics.
The difference between the load's back force and the
cylinder's force is very important. If you actually tried to
get that full energy out of the cylinder by backloading it
heavily, it would take a very long time for the cylinder to
move to its final position (which in reality it wouldn't...
it'd stop from static friction first.)

The point is that cylinders work best when you can vastly
overpower the load's back force, or else they slooooooow
dooooooown to a craaaaaawwwwwllllll. Not very useful
in contest.

OOC, Have you guys ever actually tried to USE a 2" cylinder
in a contest before? We have. It's SLOOOOOOW, and you
can cycle it fully only a couple of times in a round. You just
simply don't have the air, nor the slew rate to drive it any
more than that!

Now I've used the 2", full length cylinders in several contests
with a couple of my teams. In fact, 1502 used one last year with
a multipositioning shoulder joint for our tetra arm. BUT, we
had to be very careful in our design to not ask it to move
around too much, or we'd risk running out of air. That's why
we used a motor for our arm's elbow joint. If need be, we could
use the "last gasps of air" to move the shoulder into a good
position, and leave it there. That happened a couple of times
in contest. No sweat. We could still keep lifting tetras from
the HP to the top of a goal via just the elbow joint, while the
air system slowly regenerated..

But what you're suggesting is to move it back and forth
repeatedly. Sorry to bust your bubble, but you just can't
do that more than a few cycles over several minutes with
a 2" cylinder, or it stops. (I'm getting the impression some
people here have NEVER worked with the FIRST pneumatics
before...)

But don't take my word for it. Try it out for yourself when
you get the time! You don't have to build the catapult,
but you really need to know how the kit pneumatics behave,
or one of these days you may get yourselves caught by
expecting too much out of them.

Try this: Put together the basic kit pneumatics package.
Nothing fancy. Add ANY 2" dia cylinder. Point it downward
and attach a good sized weight to the clevis. (Let's say about
100 lbs.) Now try to run it continuously up and down full strokes,
and see what happens. Does it do it quickly? Does it slow down?
How many total inches were you able to lift that 100 lb weight
up in 2:10?

Do that, and you'll gain some good insights into how much you
can get out of the kit's pneumatics. Good luck!

- Keith

[KenWittlief]

[smileing] Whenever someone comes out and says that something 'cant be done' at this point in the design cycle, sure enough when you get to your 1st regional you will find someone who found a way.

If a pneumatic shooter can launch ten balls at once, how fast does it need to cycle? how many times per match do you think you will be launching 10 balls at once?

The balls dont weigh 10 lbs each, so you are not pushing or lifting a 100 lb mass.

How do you get all that air in the cylinder 'all at once' ? You dont: you block the cylinders motion halfway down it length, precharge the cylinder up to 60 psi, pull the blocking pin out of the way and POW! its like a garage door spring with 180 lbs of force being released.

How much will ten balls weigh? with 180 lbs of force behind them what does F=MA tell us?

[bbehnke]

im not to sure wat team or if this is even the exact design but i think a pneumatic cylinder pulled back a spring or surgical tubing so a ball could be loaded from the topand then it somehow relesed it and went back to shoot another one.

[Eric Scheuing]

The only good catauplt would be a tribuchet, but if you used that, good luck aiming/firing at the goal consistently.

But seriously, if someone's team could pull off a catapult with... say 75% accuracy, I would be amazed.

[Rick TYler]

Originally Posted by Eric Scheuing

The only good catauplt would be a tribuchet, but if you used that, good luck aiming/firing at the goal consistently.

Actually, a trebuchet would be an accurate and consistent, but really impractical device for a robot. Trebuchets rely on a counterweight to pull the arm down. In a robot, where every ounce is critical, anything relying on a counterweight is a Bad Idea. Likewise, I cannot think of any practical way for a robot to reset the sling of a trebuchet after each shot. A rotating wheel, or the energy of a wheel imparted to a belt, is a much more appropriate mechanism for a robot. Even a mangonel or oranger would be easier to build than a treb, even if you have to design for the end-of-stroke shock.

[Eric Scheuing]

Originally Posted by Rick TYler

Actually, a trebuchet would be an accurate and consistent, but really impractical device for a robot. Trebuchets rely on a counterweight to pull the arm down. In a robot, where every ounce is critical, anything relying on a counterweight is a Bad Idea. Likewise, I cannot think of any practical way for a robot to reset the sling of a trebuchet after each shot. A rotating wheel, or the energy of a wheel imparted to a belt, is a much more appropriate mechanism for a robot. Even a mangonel or oranger would be easier to build than a treb, even if you have to design for the end-of-stroke shock.

I suppose, but a trebuchet would be frickin awesome.

Play Age of Empires?

[Jimbo5051]

Im am not completely sure but I dont think that catapult is legal because you can not use anything that uses a sudden stop. But I do think you could use a trebuchea.

[Matt Gent]

-a catapult design does not have to come to a sudden stop
-the energy stored in the arm's rotational motion doesn't have to be lost (it can be restored into its original form)
-an effective & accurate catapult can be made that is not a trebuchet (ie doesn't use weight @ height to store energy or an arm/sling)
-trajectory could be changed in-match, but it becomes complicated. With an intelligently selected single trajectory it is possible to score at a wide range of distances from the center goal.
-a catapult can be implemented with a rate of fire of > 1 ball/sec
-a clutch or similar device is not necessary to "k"ock the mechanism when driven by a motor
-a working prototype using a whammo-type jai-alai cesta achieved much greater than 12m/s and has many variables for adjustability.