Flywheel Shooter Speed

[kitare102]

For anyone who has past experience with flywheel shooters: how fast did you gear your flywheels to move in surface velocity (or rpm and wheel diameter if you want other people to do the math)? How fast did your balls/other objects come out of the shooter (at least nominally)?

Thanks from team 3042

[TCMJ1816]

Back in 2012 we used a 6 inch shooter wheel being driven directly off a CIM (5310 RPM) with pretty good success.

[MaGiC_PiKaChU]

we prototyped something out of 6 versa wheels mounted directly on a CIM... shoots quite nicely for a day 2 prototype!

[Ekcrbe]

68's shooter in 2012 used an 8-inch diameter wheel (on an 8-inch ball) and ran at 2000 RPM. That was good for shooting from the front of the key (~12 feet away if I remember correctly) into the top hoop (8' 2" high) with a release angle of ~50 deg above the horizontal. Unfortunately I can't quote an empirical exit velocity.

Side notes on shooter design: we determined that there were two main reasons for that shooter's success (and I believe this will carry over to this year given the similar balls). One was the fact that it had a lot of compression as the ball entered (about 2.5" or 30%) but the compression decreased as the ball moved upward and ended with the ball leaving the wheel and rolling only against the hood, which was by that point straight rather than arced. The other was that the ball had a lot of wrap around the wheel—about 135 deg, or 9.5" of surface. The high compression and long distance allowed the shooter to impart a lot of energy into the ball with a relatively low wheel speed, and the tangential extension on the hood allowed the ball to decompress gently and guided the ball out of the shooter, eliminating any inconsistency due to the "pop" as the ball rapidly returned to its normal size upon exit from the wheels, which had been a problem in our earlier designs.

[DonShaw]

Speed will need to be determined by wheel size, boulder compression and the distance that you desire to shoot along with the trajectory you desire the boulder to take.

[philso]

Ekcrbe is giving you some good tips. Just as important as what speed was used is to keep the speed constant. We did not regulate ours and the shots were inconsistent. Less than 1/3 would go in the goal, even if the robot did not move relative to the goal.

[Ether]

Originally Posted by philso

Just as important as what speed was used is to keep the speed constant.

Which means:

1) don't just give the motor(s) a fixed command and,

2) use speed feedback to control the speed, and design your shooter (motors, gearing, wheel diameter, etc) to operate at the desired shooting speed with a voltage of less than, say, 8 volts... so you have plenty of headroom for the closed loop controller to operate.

Viable controller options for controlling speed of a spinning flywheel include PID, TBH, and bang-bang.

http://www.chiefdelphi.com/forums/sh....php?p=1146353

http://www.chiefdelphi.com/forums/sh...4&postcount=17

http://www.chiefdelphi.com/media/papers/2674

http://electronicdesign.com/analog/t...trol-algorithm

http://www.chiefdelphi.com/forums/sh...4&postcount=11

http://www.chiefdelphi.com/forums/sh...6&postcount=10

http://www.chiefdelphi.com/forums/sh...3&postcount=15

http://www.chiefdelphi.com/forums/sh...49&postcount=8

http://www.chiefdelphi.com/forums/sh...d.php?t=114591

http://www.chiefdelphi.com/forums/sh...95&postcount=3

http://www.chiefdelphi.com/forums/sh...2&postcount=12

http://www.chiefdelphi.com/forums/sh...86#post1214886

http://www.chiefdelphi.com/forums/sh....php?p=1160140

http://www.chiefdelphi.com/forums/sh...8&postcount=34

http://www.chiefdelphi.com/forums/sh...3&postcount=22

[Ekcrbe]

Originally Posted by philso

Ekcrbe is giving you some good tips. Just as important as what speed was used is to keep the speed constant. We did not regulate ours and the shots were inconsistent. Less than 1/3 would go in the goal, even if the robot did not move relative to the goal.

Ah, thank you for reminding me of this point. Mechanically, I gave the reasons why our shooter was capable of putting out consistent shots, but software was equally important. We had PID control on the wheel speed, but we also had what we called the "speed check", which disabled feeding into the shooter if the wheel was more than +/- 50 RPM from its target value. This allowed the operator to simply hold the fire button as soon as we were in position and let the robot react as soon as it was ready for each successive shot. This isn't as important this year because you can only shoot one ball at a time, but it still might have some merit.