Optimizing Air Cannon Range

At around 8:00 tonight my team (1726) tested a pneumatic tee-shirt cannon. We were overjoyed that the tee-shirt was launched, but found distance to be lacking. I am wondering if some of the members more experienced with tee-shirt cannons might be able to give us some advice on how to optimize our performance. Advice on how to fold the tee-shirts as well as any optimizations in our pneumatics systems would be accepted gratefully; we used the following pneumatics system.

KOP Compressor-standard tubing-7gal steel air tank-1/2in brass fitting-10’ 1" braided nylon tubing-inline rain bird valve-2in fitting-3in cannon.

When the cannon fires it makes a sound resembling a large air horn.

On our test shot we fired at a 45 degree angle and simply stuffed the tee-shirt into the cannon barrel. We are unsure of the PSI because much of our stored air had been wasted in a previous shot and no one bothered to check the valve; however it is estimated to have been between 40-50.

Thanks in Advance for the help.

OK, we get our farthest shots when we shoot with the arm elevated to about 40*. 50psi does mid-range shots with footballs and short range shots with the t-shirts. In order to get our t-shirts to travel mid to long range when we shoot them, we have to have at least 80 psi in our cannon, and even up to 110psi to get them to really go.
One thing you have to remember, though, is that when you are at football games, you are shooting into a sea of people, and because of that, you generally want to have your angle be more than 45-50*, so everyone has more time to react, even shooting in large arcs, we have still managed to hit a couple of people in the head. Another good Idea is to have the announcer tell the crowd when you are going to shoot, so they are all looking your way and can’t be caught off guard, we do this when we are shooting Shirts, but a simple honk of our horn does well while we are shooting foam footballs and nobody really stands any chance at getting hurt.

We fold our tee-shirts by folding them in half height wise, and then rolling them up tightly lengthwise, we also wrap the shirts in plastic bags with a bit of tape on the outside so they don’t unroll in flight, because if we let them unroll, they don’t end up flying more than 10 feet.

It sounds like a similar setup to ours. We just make sure that all tubing, pipe diam’s are at least 3/4" or more. We operate ours at about 80-90 psi and we get 3 or 4 good shots off with it. However we haven’t launched any t-shirts with it yet, just foam soccer balls and footballs. We get 50 to 100 feet reliably. I highly suggest placing a pressure indicator on it so you can better tune your shots. We hooked up the old analog pressure sensor (TI) that was in the kit a few years ago (2004 ?) and hooked it into the analog port of the robot controller. Next we’ll modify the program so that the operator interface will indicate with the Feedback LEDs if it’s too low of a pressure or something along those lines.

Good luck

We did the same thing, except we get a pressure readout in numbers on the OI, It really helps us tune the shots. We also have it set up so that a pot controls how much pressure is put into the tank. When we set the pot to what we made 60psi, it will charge to 60, and regardless of any minor leaks will stay at 60 by sporadically charging when the pressure drops.

has anyone tried shooting stuff with a wheel? like lots of teams did for the '06 game. I think that would be a little more efficient than the pneumatic gun, plus one would make it self-loading!

Your cannon is honking when you fire it? If so, I suspect your fancy rainbird valve isn’t working as well as you think. I’m actually surprised it’s even sealing particulary well, since the one I used for my ping-pong ball cannon refused to seal with air without heavy modifications.

At any rate, you probably need to make the whole air system more efficient. 10’ of 1" tubing is going to restrict a lot of your air flow. Chances are that your t-shirt will be mostly down your barrel before any of the air from te storage tank gets to the valve. Secondly, your valve needs to open rapidly and completely. Right now, it sounds like it’s mostly opening, and the are rusing past the membrane inside is making it vibrate and disrupt the airflow. You can either switch to a butterfly valve actuated with a pneumatic piston, or modify your solenoid valve. I sealed up the hole in the membrane, and pressurized the top chamber seperately. I also put a spring inside the valve to push the membrane up and out of the way more completely.

That’s all assuming you want to improve your design. Your other option is to simply throw more PSI at everything, and hope nothing explodes and injures you. Seriously. Your tubing and valve there are only meant to work with water, which is fairly incompressible. Filling up that 1" tube with compressed air is an accident waiting to happen. Especially since it sounds like the only pressure control in your system is the 120 psi relief valve on the compressor.

At any rate, efficiency is more fun. My ping-pong cannon was about 1’ long, total. The air chamber was 4" of 1.5" PVC, and the barrel was about 3" long. It still launched balls farther than the huge monstrosities other teams built with 2 foot long barrels and 2’x3" air chambers. Unfortunately, the goal was accuracy, and that’s a bit hard with a cannon that shoots things 30’ on 20 psi.

Three things: like Cody said, rolling the t-shirt and taping it will significantly improve your range. Secondly, the valve has an ID of like 3/4", so your tank with the 1/2" fitting may be a bottleneck, but thirdly, t-shirts are not really the best projectiles for perfomance. If you really want to see capability, find a ball that more or less perfectly fit’s the ID of your barrel. Shirts, unless rolled in a very tight, almost bullet like shape with a perfect mate to your barrel, will not fly extremely well, no matter what you do.

I’m with Team 698. We shot shirts using a modified assembly from the 2006 game. I should have a video on our website soon. (www.hamiltonrobotics.org)

Thank you all very much for your advice. Tonight I tested the cannon with 80PSI and shot a well rolled and rubber banded tee-shirt lubricated with talcum powder 25 yards. This is a significant improvement over last nights performance it is still lacking.

I’m not sure why the tubing would restrict our airflow. The opening from the tank to the tubing is 1/2 opening and I don’t believe that can be modified with out significant structural changes to the tank itself. Since the tubing is a large diameter than this opening the tubing should not at all restrict the airflow no matter how long.

I will however look into a butterfly valve. I am slightly worried about time however because the robot must be finished before 10/20 which is our homecoming game.

Also does anyone know at a glance how fast could a butterfly valve open when actuated with a KOP pneumatic system?

Also should be make a chamber of say 5" steel piping directly attached to the valve (which is directly attached to the cannon) to improve airflow; that is definitely an option. I believe (however I am often wrong) that because our cannon is 3 1/2’ long (3" diameter) a chamber of 5" diameter piping 1’4" long would allow for complete expansion into the barrel. Would this have any significant effect on our efficiency?

Thanks again, sorry for having so many questions.

In addition to the questions above; Does anyone know a reliable supplier of butterfly valves?

Your valve is probably not that big of a problem right now, and our robot makes the horn sound as well, but functions just fine… What will definitely improve your performance would be finding a tank with an opening the diameter of the pipe you are using, as that is the only (major) bottleneck in your system right now.


(Just included the gif for interest’s sake :))

If you’re using a tank with a 1/2" hole, why not move most of your system down to about that?

We decided to go with a 3/4" Rain Bird Valve and 5/8" ID tubing since we knew the tank was a bottleneck. The advantage is we now have a higher psi rating on them all, and I’m not so afraid to take it up to 120 when the time comes anymore.

Of course we’re actually testing it today, we’ll see how it goes.

There are several things here:

*You want your air to move as easily as possible = use the biggest valve you can, and the biggest pipe/tubbing from the chamber to the valve to the barrel. In a perfect world, if you were using a 2" barrel, a 2" valve and 2" pipe from the chamber to the valve would be great. That may not be possible though.

So, your 10 feet of 1" tubing is costing you a lot of power.

*You want to release all of your air at once+ get the fasting acting valve you can find. A good solution is a sprinkler valve. These are electrically controlled, so you could use them with the FRC robot components.

In a perfect world, you might use a large chamber, have a pipe come out of one end the same size as your valve, have that pipe turn 180 degrees, go into the valve which then goes right into the barrel.

Really, you want as little extra pipe as possible.

These t-shirt launchers are a lot like pneumatic potato guns. I suggest you explore this site for some quality information:

This is a god example of an efficient gun that you might consider.


You may want to peruse the rest of the site for more information, especially the “links” tab.

If you want more information about optimal chamber to barrel size ratios, PM me, there is much more info out there.

I’m still skeptical of unmodified sprinkler valves. That’s a heck of a restriction in your system if you can avoid it. Energy spent making your robot honk is energy that’s not getting transferred to your shirt. I will admit, however, that you might not have a better option before 10/20. As you’d guessed, a large chamber connected directly to your valve would be much much more efficient.

Thinking along the same lines, I traded in a Rainbird valve this evening for a 1 inch ball valve with a three inch lever handle, rotates 90 degrees.

The plan is to test this manually as a comparison to the sprinkler valve and if it yields a better performance, the we will have to build in a festo solenoid activated pneumatic piston to open and close this valve remotely.

We have a bit more time to prototype as we are shooting for, pun intended, the final game of the fall when the two schools who comprise our robotics team play each other :cool:

Our current setup is in the CD media.
bench prototype It shoots a tee shirt 25 yards downrange with a maximum altitude of 30 feet at a 45 degree angle and internal pressure of 110 psi. Performance is limited by a slow opening valve. We discovered this when we added a second valve to our set up and were testing the timing of the two opening together. We decided the mechanical operation of the pressure diaphram was not going to yeild consistant and tunable results for a simultaneous opening.


I have found in my vast experience and long history of projectile launchers. the more kenetic energy you can put into a mass before it leaves your grip the faster and further it will go. i have completely given up on Compressed air and almost only use CO2. its cheap, available anywhere with paintball. and as long as there is Liquid CO2 present in the tank the gas maintains a pressure around 1100 PSI. the more pressure you have the smaller you can make your device and the more energy you can put into your projectile’s mass. i have stop referring to my CO2 potato launchers as launchers and rather as potato sniper rifles. the vegetables come out of the barrel so quickly that at a range of about 30 feet potato’s atomize upon impact of a hard surface. If you relay want to throw a t-shirt up into a crowd with compressed air alone you are going to need allot and I mean allot of it! and a very long barrel to transfer that energy to a shirt. if you want a device you can hold in your hand and shoot out into a crowd an long distances of 150 feet and beyond you simply need more pressure. But please be carefull and remember how Maude Flanders Died.

A sabot casing will greatly increase performance. There’s allot on the web for water balloons. Co2 is much better than air, however it’s high pressure and needs allot more care. Co2 has it’s limits when when large volumes are released. The drop in pressure limits the flash rate. That’s why nitrogen has taken over for serious paint ballers.

i agree a sabot case would allow you to increace the surface area that the compressed gas pushes on witch will transfer energy to it more quickley. however I don’t think there’s much sabot casing you can do with a rolled t-shirt. witch by itself can be used as a sabot casr for another object.

We performed the test of a manual ball valve today during class. We doubled our range and added to the altitude achieved. The diaphragm based sprinkler valve opens too slowly and limits air flow as suspected. To make this a remote device on the robot we’ll use a short throw pneumatic cylinder activated by a festo solenoid. I’ve got video of the test but no place to upload at the moment that the school filters will allow me access. :ahh:

Finally got our cannon mounted, finished, and tested today. ~40 feet at 80 psi. We hope to extend the range later, but it will be more than good enough for Homecoming tommorow.