Creating pneumatic launchers

[Mr. 1033]

My team is in the process of rapid prototyping a launcher using 2-4 cylinders with 1.5in bore and a 10in stroke to launch the ball.

We have done our hw on here and youtube to see that the concept should work.

We are having a rough time getting the "burst" needed to propel the ball.

Any tips out the as far as how to legally set a system up?

[Richard.Varone]

Take a look at http://www.chiefdelphi.com/forums/sh...hreadid=124312, it may give you some ideas

[AustinShalit]

Anytime you use a large bore the speed is going to decrease. You might want to try using a smaller bore.

In order to have an equal amount of force, use multiple cylinders with smaller bore rather than a single large bore for greater speed.

[Ether]

Originally Posted by Mr. 1033

We are having a rough time getting the "burst" needed to propel the ball.

Any tips out the as far as how to legally set a system up?

With motors, you trade speed for torque by using a gearbox.

With cylinders, you trade force for speed by using levers.

[billbo911]

Originally Posted by AustinShalit

Anytime you use a large bore the speed is going to decrease. You might want to try using a smaller bore.

In order to have an equal amount of force, use multiple cylinders with smaller bore rather than a single large bore for greater speed.

Perfect example.
1 1 1/2" dia. cylinder with 60psi of air pressure applied produces 106Lb of force.
2 1 1/6" dia. cylinders with 60psi of air pressure applied produces 106Lb of force.

[Rick Wagner]

We have a successful prototype using two two-inch diameter, three inch stroke cylinders on a lever catapult. The cylinders are attached three inches from the catapult lever pivot, and the ball is placed 20 inches from the pivot. The catapult moves from horizontal through 45 degrees. We have a reservoir downstream from the 60 psi regulator and a reservoir upstream from the regulator at 120 psi. Each cylinder has its own solenoid. The tension side of the cylinder is open. The ball goes into the high goal reliably.

[ToddF]

Originally Posted by billbo911

Perfect example.
1 1 1/2" dia. cylinder with 60psi of air pressure applied produces 106Lb of force.
2 1 1/6" dia. cylinders with 60psi of air pressure applied produces 106Lb of force.

Note that with the same stroke length, it takes the same volume of gas to fill both of these configurations. With only a single solenoid valve supplying the gas, they fill in exactly the same time. The thing that allows the second configuration to fill faster in a FIRST robot is the ability to use two valves to fill the same volume.

In the real world, using a single large cylinder and filling with two valves would be more weight efficient than the second configuration mentioned. I hate seeing people tossing around generalities like "larger cylinders are slower" without stating the underlying assumptions behind the specific application that makes it so. I fear we're teaching our students design "rules" which are only true in the restricted world of FIRST robotics, but are completely reversed in the real world.

Another approach would be to use the single cylinder from configuration 1, and use a faster flowing valve, a perfectly valid option this year. People are finding legal valves with more than twice the flow capacity than previously legal valves.

[falconmaster]

we are using (2) 1 1/2 bore 10 inch stroke and we seem to have no problem https://www.youtube.com/watch?v=fljCPwHhnHw

[waialua359]

Originally Posted by Rick Wagner

We have a successful prototype using two two-inch diameter, three inch stroke cylinders on a lever catapult. The cylinders are attached three inches from the catapult lever pivot, and the ball is placed 20 inches from the pivot. The catapult moves from horizontal through 45 degrees. We have a reservoir downstream from the 60 psi regulator and a reservoir upstream from the regulator at 120 psi. Each cylinder has its own solenoid. The tension side of the cylinder is open. The ball goes into the high goal reliably.

Rick,
when I first heard what you folks were doing, I was very skeptical it would have enough force (speed).
But as Ether has pointed out, the lever arm is the key.
I saw a video clip of the prototype.....nice work.

[Oblarg]

I would recommend checking out the geometry of the shooter posted by MrForbes in an earlier one of these threads. It's a very clever design that is very forgiving re: airflow into the cylinders (though you absolutely must remove the fittings to maximize airflow *out* of the cylinders).

[billbo911]

Originally Posted by ToddF

Note that with the same stroke length, it takes the same volume of gas to fill both of these configurations. With only a single solenoid valve supplying the gas, they fill in exactly the same time. The thing that allows the second configuration to fill faster in a FIRST robot is the ability to use two valves to fill the same volume.

In the real world, using a single large cylinder and filling with two valves would be more weight efficient than the second configuration mentioned. I hate seeing people tossing around generalities like "larger cylinders are slower" without stating the underlying assumptions behind the specific application that makes it so. I fear we're teaching our students design "rules" which are only true in the restricted world of FIRST robotics, but are completely reversed in the real world.

Another approach would be to use the single cylinder from configuration 1, and use a faster flowing valve, a perfectly valid option this year. People are finding legal valves with more than twice the flow capacity than previously legal valves.

As long as you are taking my comments out of context, then you are right. That said, I was only giving an example of equal force provided with two commonly available cylinders.

In FIRST, access to a large variety of parts is a given. Although, many teams are on restricted budgets and thus can not afford to by valves just to attain a higher flow rate. These teams need to be creative with what they do have access to.

There are ways to actually exceed the flow rate that is available from even the highest flow valves that are legal within the rules this year.

[ToddF]

Originally Posted by billbo911

As long as you are taking my comments out of context, then you are right.

Sorry, Mr. Bill. I wasn't calling you out at all. Your example was right on the money. I was reacting to this post.

Originally Posted by AustinShalit

Anytime you use a large bore the speed is going to decrease. ...

In order to have an equal amount of force, use multiple cylinders with smaller bore rather than a single large bore for greater speed.

This design advise is correct in the context of FIRST robotics, assuming both cylinders are fed by valves with the max allowable Cv. But, it's not what you would normally do in industry.

At this stage of the design process, the students are coming up with a lot of different ideas for how to execute different concepts. Some of those are great ideas in the "real world", but not in a FIRST robot, because of the restrictions imposed by the rules. Some of those ideas are great for a FIRST robot, but are not what you would do in industry. I always take great pains to explain the difference because I don't want a student to internalize a "lesson learned" from the world of FIRST robotics and apply it to real world engineering when it might only be true in the context of the FIRST robot rules.

Whether it's because of imprecise language or because they have learned the wrong lessons, you see a lot of half-true design rules quoted by students (and occasionally mentors) on CD. I think as mentors we're responsible to explain to the students how these design rules are context dependent, and how the underlying assumptions affect their design decisions.

[billbo911]

Originally Posted by ToddF

... Whether it's because of imprecise language or because they have learned the wrong lessons, you see a lot of half-true design rules quoted by students (and occasionally mentors) on CD. I think as mentors we're responsible to explain to the students how these design rules are context dependent, and how the underlying assumptions affect their design decisions.

Spot on Todd!

There are MANY variables that have to be taken into account when designing a system, and this applies to almost any "system".
The language we use to communicate those ideas has to be correct, and should be thorough. Leaving out small details can lead to misunderstandings, as can been seen in this thread.

In this vein, here is what we will be doing in the next couple weeks.

We have a prototype we are satisfied with. It will be put into CAD, and then built accordingly.

Once built, we will run a couple thought experiments (Hypothesis), run several tests (Experimentation) , and then formation of our conclusions.
The goal of this is to optimize our catapult for competition based on the actual design we are using.