Catapult vs Wheeled Shooter

[M217]

Catapults had one big advantage this year that my team discussed in the early build season: typically, the vast majority of a catapult's variance is up-down (largely because of small inaccuracies in distance reading) while they are incredibly consistent left-right. The shape of this years goal allows for a little up-down inconsistency, since that's the longer dimension. I remember a lot of the catapult's I've seen hit a little lower than dead center from the outerworks (for example, 359's balls in tech valley seemed to touch the lower lip of the goal fairly often) but they still went in fine. For this reason, I think catapults were a lot more viable this year than they have been in previous years, and I would applaud any team that built one for making a sound strategic choice.

That being said, the packaging of a wheeled shooter and the flat linear trajectory (at least when you shoot as hard as we do) seemed to offer enough advantages that we decided to go with that. We didn't think defense would be too much of an issue because we planned for a fast release, and we'd already began work on a swerve drive that would make us much harder to pin and block.

The catapult is definitely a good choice this year, more so than I think it was in rebound rumble, but I think there are still advantages to having a wheeled shooter that make the decision nontrivial.

[sspoldi]

At team 230, we were concerned about the variability of the ball as it impacts either a spinning wheel or linear punching style shooter. In addition, we were inspired by the extremely accurate catapult used by our friends at team 177 in 2014, that was not only accurate, but extremely hard to block.

We got around the lack of adaptability of a traditional catapult by going electric, with 2 mini CIMS and a 28:1 reduction (or something like that). We found we could adapt the command profile to get pretty much any trajectory we wanted. If we went spring or bungee powered it would have been far less adaptable.

We had a lot of people comment that they couldn't believe we could shoot with a fully electric catapult. We did a fair amount of simulation to optimize the gear ratio to maximize energy delivered to the ball. A lot of teams don't do this and as a result don't get the most energy out of their device.

We lean pretty heavily on simulation written in Octave (a MATLAB like language) running on a Linux OS. It's all open source, so if anyone want's some more info please stop by our pit in champs and ask for Henry (or the drive team coach if he's off somewhere).

Cheers,
Steve S.

[marccenter]

Originally Posted by sspoldi

. We found we could adapt the command profile to get pretty much any trajectory we wanted. If we went spring or bungee powered it would have been far less adaptable.

We had a lot of people comment that they couldn't believe we could shoot with a fully electric catapult. We did a fair amount of simulation to optimize the gear ratio to maximize energy delivered to the ball. A lot of teams don't do this and as a result don't get the most energy out of their device..

Can you provide a little more information:

1) About your command profile adjustments

2) What your simulation revealed about getting the most energy out of your device (speed?).

[Littlepchan]

Our team went with flywheels because of a couple of reasons. 1: We had looked at previous games and saw how effective flywheels were. 2: We thought the balls were durable enough that there wouldn't be variance. 3: Even if the balls were going to start breaking our compression would trump any damage. 4: One of our new mentors that has been with FRC for a while said that no matter what we would get to a fixed angle, fixed shot speed, and fixed shooting spot. I think those words along with previous game footage really gave us tunnel vision. 5: We wanted to figure out how to use flywheels. Those alone kicked our butts last year on Carver and we wanted to learn how to use them effectively. Lastly 6: We imagined that by Championships defense bots were going to become less of a problem. Plus as an added bonus our drive train is so strong we can push any bot that gets in front of us and (although we haven't tested it) we could probably shoot from our batter and still hit the target.

Originally Posted by backdrive

I find it interesting how certain shooter designs this year seemed to be vary by region- for example, Texas has a bunch of amazing catapult robots (118, 148, 2848, 4587), but California, in contrast had no top-tier catapult bots (correct me if I'm wrong), with nearly all the top teams (254, 1678, 971, 973, etc) opting for flywheel shooters.

^
Just to point out, in case no one saw, 2848's shooter did start out as Flywheels. It was only at Dallas did they build and implement a catapult into their robot. Again not poking at them, just stating how catapults seem to be really good at this game.