Catapult releasing mechanism

[CAVBOTS]

Here is our issue, our catapult needs to be able to lower into the ball collecting position to scoop up the ball, and them we want to be able to, depending on where along the trajectory we want to relase, hold the arm at that angle, until we need to release. Some form of electric or pneumatic clutch would work, but what we need is recommendation on which ones have been tried and how well they work. Any thoughts??

[MrForbes]

And you want to get it designed, built, and working in a week? That's going to be tough!

There isn't really enough information in your post to make specific recommendations, since we can't see what your catapult looks like. Generally, making something release at a certain position is a tough problem. We avoid it by making all our mechanisms have only two positions, and use pneumatics to move them. In past years we've used a pneumatic release on a winch, it worked so-so. And we've controlled the position of arms using potentiometer feedback and a PID loop in the program, with varying success. Both of these methods took several weeks to get working.

[fox46]

Wow- you guys are definitely pressed for time then.

You should be able to make every target in this game using one firing solution ie: one type of shot. There is no real need to be able to adjust your shot to shoot from anywhere on the field- trying to do so adds a huge amount of un-needed (IMO) complexity. Instead of being able to make all kinds of shots, focus on doing one shot really well (consistently) which can clear the truss, move the ball down the field and hit the high goal. You will learn very fast where the robot needs to be to make each shot and it eliminates having to engineer a complex system on the robot to calculate trajectories etc.

In your situation, I would recommend a spring-loaded catapult. Latex tubing changes in tension and performance over time so use springs for consistency. An arm on a high torque gearmotor to push the catapult into the "cocked" position, engage a latch to hold it there while the cocking arm rotates out of the way, then release the latch when ready. A winch could be used as well that slackens off once cocked to permit firing. The latch could be something as simple as a latch for a chain-link gate (http://www.plumbersurplus.com/images...132-178772.jpg) actuated by a pneumatic cylinder. All that is required to produce a repeatable shot is open/close the latch, rotate loading arm fwd/rev and two limit switches to stop the arm when the catapult reaches its latching position and when the arm is back in its home position. The entire process of loading can be automated in the program so all the driver needs to do is pull a trigger to shoot- the only logic required is:

If latch limit switch not triggered > rotate loading arm FWD.
If latch limit switch triggered > close latch and rotate loading arm REV
If home limit switch triggered > Stop loading arm.
When driver pulls trigger > Release latch

The driver will quickly learn where the robot needs to be to hit the high goal ie: how far away from the goal- where the trajectory of the ball is flattest to give the greatest tolerance of error. You could also integrate an ultrasonic ranging sensor with a red light/green light system to tell said driver when they are within shooting range.

[Sparkyshires]

If I can make a recommendation, take a look at jvn's bot. they use a ratchet wrench to spin it into position, then just let go and it won't spin back.

[TheKeeg]

Originally Posted by CAVBOTS

Here is our issue, our catapult needs to be able to lower into the ball collecting position to scoop up the ball, and them we want to be able to, depending on where along the trajectory we want to relase, hold the arm at that angle, until we need to release. Some form of electric or pneumatic clutch would work, but what we need is recommendation on which ones have been tried and how well they work. Any thoughts??

At this point I would have to agree that you should focus on one shooting position that you can do well. It will be very difficult to design, build, and debug in 1 week.

[Ether]

Originally Posted by Sparkyshires

If I can make a recommendation, take a look at jvn's bot. they use a ratchet wrench to spin it into position, then just let go and it won't spin back.

What "jvn's bot" are you referring to?

[wireties]

Originally Posted by Sparkyshires

If I can make a recommendation, take a look at jvn's bot. they use a ratchet wrench to spin it into position, then just let go and it won't spin back.

The wrench is just to handle back-drive. A cam stores the force in surgical tubing. And its a little tricky to build.

[rwkling1]

Originally Posted by Ether

What "jvn's bot" are you referring to?

http://www.youtube.com/watch?v=Gm96S8gdhR0

[jwfoss]

A high strength magnet, or even an electromagnet may be a suitable latch for your design.

[Ether]

Originally Posted by rwkling1

http://www.youtube.com/watch?v=Gm96S8gdhR0

I am familiar with the JVN Ri3D bot.

But are you sure that's what the poster had in mind?

That cam and linkage would be quite a challenge for an inexperienced team with one week left in the build season.

[Tungrus]

Originally Posted by jwfoss

A high strength magnet, or even an electromagnet may be a suitable latch for your design.

Magnet? Assuming that its powerful enough to hold the catapult, how would you release it? Electromagnet may be an option, it it must be seriously powerful (to hold down about 50lb equivalent force need to throw the ball).

[Madison]

Originally Posted by Ether

I am familiar with the JVN Ri3D bot.

But are you sure that's what the poster had in mind?

That cam and linkage would be quite a challenge for an inexperienced team with one week left in the build season.

It's been a challenge for our experienced team. The forces present in a mechanism that can score the ball from >20' are tremendous (and a little bit terrifying).

[Ether]

Originally Posted by Tungrus

Electromagnet may be an option, it it must be seriously powerful (to hold down about 50lb equivalent force need to throw the ball).

There are 12V COTS electromagnets with force rating higher than that.

And it doesn't have to be that powerful if you can figure out a way to use leverage. Think about the trip mechanism on a mouse trap for instance.

[wireties]

Originally Posted by Madison

It's been a challenge for our experienced team. The forces present in a mechanism that can score the ball from >20' are tremendous (and a little bit terrifying).

I can't agree more. I've warned the students a 1000 times during the prototyping process how dangerous this can be. And that a robot w/o bumpers can shatter their ankles!

We were lucky enough to get picked by Aren Hill's (now an engineer at IFI and on JVN's Ri3D team) team in 2008 at the Colorado Regional. So we were familiar with the "choo-choo" cam and decided to use it as we were walking out of the kickoff presentation. Still it is a challenging effort!

[rwkling1]

Originally Posted by Ether

I am familiar with the JVN Ri3D bot.

But are you sure that's what the poster had in mind?

That cam and linkage would be quite a challenge for an inexperienced team with one week left in the build season.

True, the poster may have meant a different robot. With your post, I assumed you weren't familiar with Team JVN.