View Full Version : Team 3489's Catapult Prototype
Kevin Thorp
10-01-2012, 12:37
Inspired by Team Titanium, Category 5 decided to try a catapult, scavenging parts from our 2011 'bot:
http://youtu.be/cyMTlgvDjZM
It works pretty well propelled by both pneumatics and surgical tubing, but we need to control the distance. And to auto-load balls.
Looks pretty good, how do you plan to auto load the balls?
Looks pretty good! As long as you can a) control distance and b) aim somehow (via turret or turning the robot), you should have a pretty solid thrower, good job with thinking outside the box and using a catapult.
What kind of distance were you getting with your mock-up?
Austin2046
10-01-2012, 13:13
i like it a lot. even if u can't score consistently, if you were doing that from the far end of the field you'll be getting lots of shots off, and getting balls to your side of the field. The four basket arrangement might also help with the variability of the shots. you might not always hit the 3 pt shots but they could bounce off and land in one of the 2 pointers.
billbo911
10-01-2012, 13:13
If you can control the distance and alignment, that will be a really killer half court(+) shooter. What a great start and good luck on refining it!
I'm thinking a robot name of "MJ" for that bot.
Kevin Thorp
10-01-2012, 13:21
What kind of distance were you getting with your mock-up?
I think the trash can was about 25' away.
We haven't figured out auto-loading yet. A belt-type ball elevator is pretty simple. The tricky part is getting it from the top of the elevator to the "cup", one at a time.
Controlling distance is another challenge. We might preload the rubber tubing by winding it up around a spool with a revolution counter. And/or adjusting the air pressure. I wonder if we're allowed to use electronically-controlled pressure regulators?
This is getting complicated. A spinning wheel launcher is pretty simple.
If you're throwing that reliably from 25 feet, maybe you could consider specializing in throwing from the other side of the court, say like in your protected alley. This would allow you access to balls automatically from your alliance inbound area, and maybe you'll just always throw from x- distance away. sure it would limit your range of throwing, but hey "The Jack of all Trades is the Master of None." Being amazingly consistent at one thing is better than being somewhat consistent at a lot of things.
Ninja_Bait
10-01-2012, 15:56
Get it to throw full field. I think it would be easier to make something for the inbounder to feed into than to make a ground-to-arm ball collector. I'm not totally convinced that it can't be done, since I haven't actually tried to design it, though.
Man you catapult is awesome.
LabVIEW Rocks!
10-01-2012, 17:39
I think the trash can was about 25' away.
We haven't figured out auto-loading yet. A belt-type ball elevator is pretty simple. The tricky part is getting it from the top of the elevator to the "cup", one at a time.
Controlling distance is another challenge. We might preload the rubber tubing by winding it up around a spool with a revolution counter. And/or adjusting the air pressure. I wonder if we're allowed to use electronically-controlled pressure regulators?
This is getting complicated. A spinning wheel launcher is pretty simple.
Get simplier, have 2 rails coming down to the launcher with a ball release barrier that can let balls through on the rail.
Blackphantom91
10-01-2012, 18:49
Glad we could help:rolleyes:
Kevin Thorp
10-01-2012, 19:15
Man you catapult is awesome.
Thanks! But remember when you watch these types of videos the robot is usually perfectly still, at a fixed distance and angle to the goal, and no other robots are harassing it.
What would really be awesome is a video of a robot making goals at 3 different distances.
Kevin Thorp
10-01-2012, 19:16
Glad we could help:rolleyes:
You could help even more by sending us your CAD files!:D
Blackphantom91
10-01-2012, 20:33
That would be no fun so you wouldnt learn. also there is a "TitaniumRule" which the GDC created we think because of us in 2009 (R78) in the rules so the design is illegal.:cool:
RedLeader342
10-01-2012, 21:14
If you're throwing that reliably from 25 feet, maybe you could consider specializing in throwing from the other side of the court, say like in your protected alley. This would allow you access to balls automatically from your alliance inbound area, and maybe you'll just always throw from x- distance away. sure it would limit your range of throwing, but hey "The Jack of all Trades is the Master of None." Being amazingly consistent at one thing is better than being somewhat consistent at a lot of things.
unfortunately, this alley isnt protected. its purely decorational. its in the rules ;D
HumblePie
10-01-2012, 21:56
unfortunately, this alley isnt protected. its purely decorational. its in the rules ;D
[G28]
Robots may not touch an opponent Robot in contact with its Key, Alley, or Bridge.
Violation: Foul; Technical-Foul for purposeful, consequential contact.
You're right about one thing..... it is in the rules ;)
Neat!
Are you venting the cylinder to atmosphere or which fittings are you using to make it double action. Last time we tried to make one like that we had issues getting it to work as a double acting cylinder due to the volume of air we were trying to move through a fitting.
CrashTestPilot
10-01-2012, 23:16
If you're throwing that reliably from 25 feet, maybe you could consider specializing in throwing from the other side of the court, say like in your protected alley. This would allow you access to balls automatically from your alliance inbound area, and maybe you'll just always throw from x- distance away. sure it would limit your range of throwing, but hey "The Jack of all Trades is the Master of None." Being amazingly consistent at one thing is better than being somewhat consistent at a lot of things.
Is there a rule that does not allow for using phisical contact with field elements to allign for shot? Could you say mount this catapult in a way that when the robot is touching the field wall and the bridge in your own alley you are perfectly alligned to make a x point shot? If the feeder was on the back end you could pick up additional balls that are rolling down the alley without moving the bot. Maybe have an allience partner feed the balls to you once you plant the robot in that perfect spot.
You may be on the path to design a "sleeper" robot that will surprise some sceptics.
Kevin Thorp
11-01-2012, 09:23
Neat!
Are you venting the cylinder to atmosphere or which fittings are you using to make it double action. Last time we tried to make one like that we had issues getting it to work as a double acting cylinder due to the volume of air we were trying to move through a fitting.
We used standard 1/4" OD push-to-connect fittings.
But we sorta cheated on the valve, using one that has more flow than is allowed in the rules. Later when we installed a "legal" valve our range dropped about 15%. I think we can get that back if we go to two smaller cylinders, each controlled by it's own valve.
Larger cylinders = more power, but more time to fill & exhaust.
martin417
11-01-2012, 09:48
We used standard 1/4" OD push-to-connect fittings.
But we sorta cheated on the valve, using one that has more flow than is allowed in the rules. Later when we installed a "legal" valve our range dropped about 15%. I think we can get that back if we go to two smaller cylinders, each controlled by it's own valve.
Larger cylinders = more power, but more time to fill & exhaust.
Multiple small cylinders, each connected to its own legal valve is an excellent way to achieve higher speed. Another trick:
(Assuming you want high speed and force on cylinder extend, and retract speed / force are not as important.) Use multiple small cylinders, each connected to its own valve. On all but one cylinder, do not put a fitting in the retract port, and plug the retract port on the valve. On the final cylinder, connect as normal. That will greatly reduce the amount of air you have to force through the exhaust ports on the valves during the extend motion, and eliminate much of the back pressure on the cylinders. We used this technique in 2008.
Third trick using multiple cylinders:
To vary speeds: Connect all cylinders to the valves in the normal fashion. High force kick, extend all cylinders.
Lower force, extend some cylinders, keep some in retract. A cylinder has more force in extend than it does in retract, so as long as you have as many or more cylinders extending than retracting, it will extend. If you use four cylinders, that will give you three different extend speeds / forces. We did this in 2010 (we also used surgical tubing to help extend, so we could actually extend with all but on cylinder in retract)
Kevin Thorp
12-01-2012, 06:35
While one group was working our a catapult prototype, another built a single wheel shooter:
http://youtu.be/LXIYwYKEA8Y
We knew this first iteration would not have a lot of range, but it'll be a good test bed for experimenting with aiming and speed control.
professorX
12-01-2012, 12:17
I would like to mention to not forget about the 60in tall limit if shooting from half court for the catapult design.
Mr. Pockets
12-01-2012, 12:36
Thanks! But remember when you watch these types of videos the robot is usually perfectly still, at a fixed distance and angle to the goal, and no other robots are harassing it.
That's actually representative though. If you just fire from the top of the key everytime and score consistantly each time, then this model works fine.
If you're getting a solid 25' with each of your shots though I'm gonna echo the others and suggest that you consider working as a resupply bot. Saves the issue of manual reloading, and makes you one valuable little pony for your team :D
mr.roboto2826
12-01-2012, 12:39
If you truly are interested in the catapult design you may want to take a look at this robot from 45 in 2006. http://www.chiefdelphi.com/media/photos/23258
RedLeader342
12-01-2012, 13:38
[G28]
Robots may not touch an opponent Robot in contact with its Key, Alley, or Bridge.
Violation: Foul; Technical-Foul for purposeful, consequential contact.
You're right about one thing..... it is in the rules ;)
oops. read wrong. saw that the lines between the key and fender were decorational and thought it was the alley. on page 5 of the arena rules
however it just says touches another robot, doesnt say you cant go into that zone ;D
You can use a lead screw with 80/20 linear slides to effectively shorten the arm when you want, thus lowering the ball speed at launch. You can also just sit at the slot and launch the balls 50 feet. However, IF YOU DO THIS, make sure your arm never extends up over 60 inches.
You may also be able to put 2 solenoid valves on the cylinder such that you can have 2 different working PSI's, giving you a high-pressure normal range and low pressure ultra-short range (for when you're right next to the thing). Double check the rules though; I know after 2008 they changed the pneumatic rules up a bit to prevent specific instances of multi-valve issues from occurring. You may also be able to hook a servo up to flow-control valves to limit the speed at which the cylinder fills if the multi-valve thing doesn't work.
The hardest part of this will be side-side accuracy, IMO.
JamesCH95
12-01-2012, 14:15
You can use a lead screw with 80/20 linear slides to effectively shorten the arm when you want, thus lowering the ball speed at launch. You can also just sit at the slot and launch the balls 50 feet. However, IF YOU DO THIS, make sure your arm never extends up over 60 inches.
You may also be able to put 2 solenoid valves on the cylinder such that you can have 2 different working PSI's, giving you a high-pressure normal range and low pressure ultra-short range (for when you're right next to the thing). Double check the rules though; I know after 2008 they changed the pneumatic rules up a bit to prevent specific instances of multi-valve issues from occurring. You may also be able to hook a servo up to flow-control valves to limit the speed at which the cylinder fills if the multi-valve thing doesn't work.
The hardest part of this will be side-side accuracy, IMO.
I was just made aware of this rule by SIRI (CD user):
<R78> The outputs from multiple valves may not be plumbed together into the same input on a pneumatic cylinder.
Just FYI.
vBulletin® v3.6.4, Copyright ©2000-2017, Jelsoft Enterprises Ltd.