Since I’m still a mechanical noob and I will be one of the main mechanical people on the team next year, could the teams who have implemented a shooter or something like a shooter please post how it works and what you used to build it. Pictures or CAD of your robot or the specific end effector would be awesome. Thanks a lot in advance.
-noob:p
Ours is very different from most you’ll see around here. It operates on a similar principle to the classic “there’s no replacement for displacement.” Ours operates off of pure power - to the tune of 2 Chiaphuas and 2 FPs, for a combined total of over 1000 watts in the shoulder. As to how well it works… lets just say you should make sure your pot mounting is really rigid when using them on parts that accelerate really fast. A little flex that caused our pot’s gear to disengage from the one on the shoulder under sudden acceleration turned the PID tuning from a 15-20mn job into a takes-the-entire-regional-and-still-isn’t-completely-done job.
There are several ways to do it; team 1726 took the approach of using pneumatics to launch the ball. The cylinders were set up in such a way that they could build up pressure while the ball was on top of the catapult and then launch it over quickly after the pressure overcame the weight of the ball. (pictures here)
39 also used a pneumatic launcher, but it works much better in our opinion. Instead of letting the cylinders build pressure, they just increased the airflow to the (single) cylinder to make it fire quicker. They had three valves running into it, allowing for a very simple and effective launcher (the best one I’ve seen so far).
The other main launching method I’ve seen involves using latex tubing to make a slingshot/catapult apparatus. 118 is noted for posting the design first, and many other teams have a similar set up.
I don’t have access to the CAD files, but if you will be at NYC stop by the pits and I’ll show you our shooter. It’s basically a forklift catapult that uses 2 FPs, surgical tubing, and an AndyMark supershifter to lock, load, and shoot.
Yeah, I’ll be there. I’ll see what team adopts me for the day as well:p
You’re more than welcome to be a SciBorg for the day.
First of all… credit where it is due, our shooter was inspired by some of the early information posted by team 1726, and although we went a slightly different route in our design, it was very comforting to know that we were “on the right track” before heading down that track.
The basic geometry is shown in the attached images… although we did make some changes from this for the final version, this should get the concept across.
While this drawing shows one 1.5"x8" cylinder, we actually use four of them, two on each “arm” of the launcher. Our choice to use 1.5x8 cylinders was based on the fact that we had five of them sitting around, unused, in the shop. You will note that when the launcher arm is in it’s rest position that the cylinder is not fully retracted. This is so that we can pre-pressurize the cylinder to 60psi.
We do this because one of the limiting constraints on a pneumatic launcher is the rate of air flow. The air flow is constricted by the tubing, the valves and by the regulator. In this design we get most of the air past the worst of the bottlenecks before we launch, allowing for a faster, more powerful, extension of the launch arms. I should also mention that we have two clippard cylinders downstream of the valve, so that not only do the launcher cylinders get pre-charged, but so do two Clippard cylinders. This helps maintain pressure in the launch cylinders as they extend.
Since the cylinders are being pressurized before we launch, we needed a way to keep them from launching before we wanted to launch. You will note that in the retracted position the cylinder is “over centre” and pushes, ever so slightly, downwards rather than upwards. This locks the launcher in position. In order to launch we first energize the single solenoid valve controlling air flow to the four launch cylinders. It takes a few seconds for the cylinders to fully charge to 60psi. Energizing the launch cylinders is done by pressing – and holding – a button on the control panel. Should the button be released, or the robot disabled, the single solenoid valve automatically vents the stored pressure, making it impossible for a disabled robot to “launch”. We had no problems with safety inspections after explaining this feature to the inspectors.
Once the system is fully energized… pressurized… charged, call it what you will, we activate the “kickers”. There are two 1.5x3" cylinders (not shown in these images) mounted perpendicular to each arm (in its retracted state) located about half way along the arm. When the launch control operator presses a large green “fire” button on the control panel, these two cylinders extend, pushing the arm “over centre” and giving the ball and arm some initial upwards velocity. From this point on the energy stored in the compressed air in the four cylinders and two clippard tanks takes over and the ball goes flying.
This system has not been optimized for air consumption… and is probably pretty wasteful. It takes about 17 seconds to fully re-charge the two high-pressure clippard tanks on the pneumatic system, somewhat longer depending on how many tries it takes for the driver to grip and grab the ball as our gripper/grabber is also pneumatic. We figured that if our re-charge rate was our limiting factor in scoring, however, that we would be doing okay. In a couple matches we did have to wait to shoot… but by and large it wasn’t a big deal.
To make sure that we had enough pressure to shoot we put a pressure transducer on the high pressure side of the system and provided operator feedback via both LEDs on the OI and a custom LabView interface. By firing at a lower pressure it was possible to place the ball on the overpass for bonus points.
One of the big benefits of this system… aside from the fact that it is relatively safe… is that it is also relatively simple. We were focussing on simple design this year. The pneumatic launch system is also showing its versatility now that we have the robot at home, as we can crank a few extra PSI out from the regulator to make for more impressive demonstration shots when showing off in larger facilities.
So far the best place to see this in action is in the Portland QFs on The Blue Alliance, or on our (award winning!) website, www.trobotics.ca under the 2008 videos section. The Portland videos don’t show much of the launcher, as we were spending too much time knocking down balls and trying to grip them to do much launching, however in Seattle we not only scored a couple times each match (when we didn’t have electrical issues), but also managed to place the ball on the overpass once or twice.
So nothing too magic here… just some good ol’ P1V1=P2V2 going on.
Jason
P.S. The “best” shot we made all year was not recorded… it was while prototyping the system using a scaled down version and a chunk of 2x4 as a “dummy mass” to represent the ball and we finally got the pressures and angles just right… and launched the block way higher than we ever expected. It wasn’t the shot that was so good… it was the way the fluorescent bulbs kind of shattered in place, then hung in the air for a brief moment, before crashing to the ground. Ooops. After that we started looking upwards a bit more carefully!
Our shooter design is fairly unique in that it uses no pneumatics or stored energy, but two CIM’s to launch the ball.
If someone else is using only motors to launch, I haven’t seen their bot yet.
The two CIM’s are on the lower left, chained to the bars atop the robot which act like a catapult and fling the ball.
The lifter-claw-thingy-which-we-never-actually-gave-a-definitive-name-to holds the ball in place directs the ball fowards as it is launched.
The launcher arm also acts as our overpass removal tool. The bent piece at the end contacts the ball just enough to pop it off the overpass.
http://www.chiefdelphi.com/media/img/b58/b584daa67636f624f2eaace53c450d4b_m.jpg](http://www.chiefdelphi.com/media/photos/30380)
Ditto for 39 giving credit to 1726 for the catapult concept.
The quantum mental leap was that a catapult with the air actator attached at a radius on the moving arm other than the ball position is basically a gearbox for pneumetics and needs to be analyzed as such. Yes it is non linear (mechanically) and this combined with the gas flow issues makes the analysis even more interesting (fun). The sims and analysis were the most challenging of my FRC experience (07, 08).
This year Team 1771 built a robot unlike any I have seen. It uses a shop vac (powered by 2 FP motors) and a large funnel to grab a ball by suction. It then uses a pneumatic cylinder arrangement to shoot the ball. The vacuum holds the ball against the funnel, and the cylinders are pressurized. When the cylinders develop enough force to break the vacuum (around 300 pounds) the ball flies off with great speed. They were able to achieve more than 6 feet vertically, and over fifteen feet horizontally. Once the bugs were out, they were able to average 5 hurdles per match, and got 7 once (I think, still waiting on video).
Unfortunately they were eliminated in the semifinals at Peachtree, and their season is done:(
Cad model pics here: (the CAD model shows four 3/4" cylinders, the actual bot has four 1.5" cylinders that extend only, and one 3/4" cylinder that extends and retracts)
http://picasaweb.google.com/mwilson417/RobotPictures/photo#5184233288827560866
Early season scrimmage Video here:
I guarantee we are the only catapult of our kind.
We use 4 springs, or about 300 pounds of force, to launch the ball. The springs are attached to a solid axle at the top, and to linear slider on the other side. A winch system using a custom gearbox for the van door motor, uses a ball lock mechanism to lock into place and winch the catapult back, thus tensioning the springs.
A limit switch tells the robot when we are cranked all the way back. We load up a ball and come around the track and hit the fire button, which engages the pneumatic piston that releases the ball lock and thus puts the gearbox into “neutral” allowing all the stored energy to be released. The results can be seen here at around 1:37 :
http://www.thebluealliance.net/tbatv/match.php?matchid=8942
…andy grady gives a great commentary
Team 95’s robot Zog used pneumatics as well, with a two-stage pneumatic cylinder arrangment and six valves to get all the flow we needed. I’ll see if I can get a good pic of it. We liked it, since Zog can hurdle the ball from floor level while driving.
Wow, some of these look sick. 125’s got air for sure. Thanks a lot for the posts so far, its cool for someone learning in FIRST to see such creative designs.
Keep 'em coming:cool:
We use two 150lb garage door springs and surgical tubing. The springs and tubing are pulled down by two globe motors and released with the small kit pneumatic.
Here’s a few pictures:
http://www.adambots.com/images/6/65/F08C-GL-084.jpg
http://www.adambots.com/images/1/10/F08B-218-09.jpg
http://www.adambots.com/images/1/14/F08C-GL-012.jpg
Here’s is some video of it at Great Lakes Regional:
http://vimeo.com/850287
(Please excuse the poor Hybrid on this match 
)
Most shooters took inspiration from the team that released their design rather early in the build period. Our team planned to build a shooter all along, and endeavored to make several prototypes before the real thing. This is what we learned.
Pneumatic pistons by themselves (directly pushing a ball) are not enough to get the ball over the overpass when firing from the ground. This comes from the fact that the pistons extend too slowly (the inherent problem of filling large chambers with small air tubes). While the pistons did provide enough force to move the ball, but they did not deliver that force quickly enough.
The question we were faced with is: what can deliver the same amount of force on the ball over a shorter period of time? To us the answer was clear: springs.
Huge springs, two of them. 240lbs of force combined. We used two 2in bore pneumatic pistons to extend these monster springs, and a small .5in bore piston to operate a latch that would release all this potential energy when retracted.
That’s the concept. The actual construction was more complicated. We were faced with more problems such as: how can we extend the springs with the pistons, and then disengage the pistons from the springs when we wanted to fire. Keeping the pistons connected to the springs would lead to an unwanted decrease in firing speed.
I will post some pictures of our spring-loaded catapult later if I have time. Any questions are welcome.
Sam
Hold on a minute! Here is a quote you should heed:
I don’t know how many time I heard on this very forum all through January and February that “so-and-so will never work” or “This mechanism cannot do that”.
Two examples I heard over and over involve the use of vacuum to acquire and hold the ball, and pneumatic cylinders (thats the proper term, the piston is the part inside the cylinder that moves) used to shoot. I read numerous posts that said things like “you can’t pick up a ball with vacuum”, “the cover is too porous”, “you can’t develop enough holding force” as well as comments like the one you just posted.
I believe that, while well intentioned, these comments discourage innovation, and prevent people from trying new approaches. Just because you haven’t figured out how to make something work doesn’t mean it’s impossible. If you read my earlier post in this thread, and visit the links there, you will see that not only can pneumatics alone “get the ball over the overpass”, they can do it very well.
I think this point is so important, I should start a new thread.
Here is a true story that illustrates what people can do if you don’t tell them it’s impossible.
A young college student was working hard in an upper-level math course, for fear that he would be unable to pass. On the night before the final, he studied so long that he overslept the morning of the test. When he ran into the classroom several minutes late, he found three equations written on the blackboard. The first two went rather easily, but the third one seemed impossible. He worked frantically on it until, just ten minutes short of the deadline, he found a method that worked, and he finished the problems just as time was called. The student turned in his test paper and left. That evening he received a phone call from his professor. “Do you realize what you did on the test today?” he shouted at the student. “Oh, no,” thought the student. I must not have gotten the problems right after all. “You were only supposed to do the first two problems,” the professor explained. “That last one was an example of an equation that mathematicians since Einstein have been trying to solve without success. I discussed it with the class before starting the test. And you just solved it!”
I always love reading quotes like this, since Team 95’s robot does exactly this, with quite a bit of margin:
http://team95.org/homebase/wp-content/uploads/2008/03/picture-2.png
Yes, we use six cylinders in a two-stage arrangement, and it took more than a little tweaking, but we are “directly pushing on the ball” with nothing but pneumatic pistons. We could get even better performance if we triggered the two stages separately at staggered intervals, but we called this setup good enough.
team 1629 uses two cylinders and two springs to fire the ball.
First the cylinders are charged to extend the springs and lower the scoop (fancy curved catapult) it is then latched with 1 small cylinder. At this point the only force is the springs but as we begin the firing sequence the opposite side of the cylinders is charged (applying force in the same direction as the extended springs) and when the latch is released up the scoop, and what ever may be on it, go. This arrangement gives us gives us the ability to hurdle from ranges back to the end of the lane dividers, as well as the ability to launch diagonally from one homestretch to the other.
Its was a tricky arrangement to control until we discovered that it was much easier to let the programming do the timing required to be efficient.
if you have any other questions you can message me or stop by our pits in atlanta
We do nothing too complicated. We have a fork lift which take the ball in. and then we use surgical tubing and tie our shooter to a bar on the robot. Then we use a wench to pull the shooting platform back…and bam.
Perhaps I seemed close-minded when I commented about using pneumatics to fire the ball over the overpass. Let me be more clear about our prototyping endeavors.
We found out that using two 1.5in bore pneumatics (directly pushing on the ball) only fired it around 6.5 feet high. That not being enough, we sought out other ways to shoot the ball higher.
So I was wrong when I said that using pneumatics to directly push on the ball wasn’t enough to hurdle. I should have said that using two 1.5in bore pneumatics (directly pushing on the ball) wasn’t enough to hurdle.
So I stand corrected because my original statement was vague. You certainly may use cylinders to push the ball directly if you have enough energy to do so. Many of the shooters (including ours) used compressed air for a reason, it was not my intention to discourage its use. (I’m also sorry if I ticked anyone off, everyone is proud of their own designs - for good reason too!)
Attached is a picture of our catapult in construction (left side of picture). Two pneumatics (as shown) are used to extend two springs of that size (only one was on the robot when the picture was taken). A small pneumatic controls a latch that releases the potential energy at operator command.
The 2in bore pneumatics are connected to the rotating structure that the latch cylinder is mounted to. When the large cylinders retract, the latch will hook on to the catapult structure. When the large cylinders extend, the latch will pull on the catapult until the springs are extended about 8 inches. When the latch disengages, the catapult will fire as the springs release their energy. The entire reloading process is automated.
It consistently shoots the ball 12 feet high.
Once again, sorry for coming off as close-minded,
Sam
