T-shirt cannon using pneumatics

Good morning eveyone…
As a rookie team going into season 2. I am trying to put together a project for the team to practice building a set up that involves using pneumatics. Can anyone who has done this before help us with your setup/blueprint or point us in the right direction for building an frc robot that launches t-shirts? Besides the kids getting practice, we can use this to connect with our community at events.
Thank you very much in advance.


I don’t have a blueprint available right now but I can describe the system we currently use.

We use a 3000PSI scuba tank for bulk storage, through the scuba tank valve to a pressure regulator and into a decommissioned fire extinguisher tank at a considerably lower pressure (IIRC ~125PSI). This is then teed off into a fast acting (~20ms) solenoid valve with a 1" outlet, through a pressure rated 1" hose, and into the backside of the barrel changer.

I’m trying to find a reference to the solenoid we used as I think it was the most expensive single component but very much necessary to dump the volume of air needed to fire it in a short enough time that you’re not wasting energy after the payload has left the barrel.

The Poofs were one of the first to make one on a FRC platform and documented it pretty well here.

Edit: 254 names the specs for the valve they use right in the link I shared. 1" NPT valve from Mac Valves


As a rule, stay away from PVC and plastic. You can search a lot of history here on people explaining compressible versus incompressible liquids (air versus a fluid) and the result when you have an unexpected failure.


Ours uses these 1" NPT, 12CV brass pilot valves: https://www.amazon.com/Electric-Solenoid-Normally-Closed-Diesel/dp/B007N06HWA/ref=sr_1_4.

They work great, just beware that if you use CO2 (like we do) they tend to freeze shut after a couple shots.


That large fast valve, and using parts rated for air pressure (not just water pressure) are the two key ingredients. When the container breaks, compressed liquids leak, compressed gasses explode.

The stored energy in a compressed fluid is \int P(v)dv. While the average value of P(v) isn’t all that different, \int dv definitely is. At 60psi (0.4MPa), air compresses by about 80%. At 4km depth (40MPa), water compresses by about 1.8%.

832’s T-shirt cannon uses a copper pipe barrel rated to over 150 PSI. Working pressure is 120.

It would never hurt to cover the barrel in layers of some very sturdy fabric (think bumper material); in an explosion it may slow the pieces down enough so the injuries are only grievous and not fatal.

The key to getting good distance is tightly-rolled T-shorts and the use of a sabot, a thing that fits the barrel just tightly enough, but doesn’t get shot as far as the short (because it’s attached to the cannon with a cord).

It’s probably a really good idea just to rent these tanks from your local dive shop instead of acquiring one yourself. First of all, getting the tank up to 3000 PSI is outside the reach of most shop compressors, while dive shops have compressors that will do that. But, secondly, there is testing and maintenance that has to happen on tanks – if you rent, the dive shop takes care of that. If you buy, then you have to deal with it. You DO NOT want to be around a tank at 3000 PSI that isn’t up-to-date on those requirements.



I’ve used wadded up cheap plastic grocery store bags for this in T-shirt and golf ball launchers in the past - it’s gets torn up after a while but it’s super low cost so no tears shed. They are super non-aerodynamic so they don’t travel very far.

It also helps protect the shirt - I’ve seen the shirts get damaged if stuffed right down to the air inlet to the barrel. The bag acting as a sabot takes the damage from the initial air application.

It’s basically the same deal as using wadding in a model rocket to protect the parachute from the ejection charge.


We built a 4-cannon T-shirt cannon robot -

  1. you should use dust collection pulse valves instead of electronic solenoids
    Dust Collector, Pulse Valve, Series JSXFA | SMC Corporation of America

  2. a FRC compressor will not be enough to be able to fire the cannon consistently - so I reccomend using an external compressor (for pneu tools) and a large tank on the robot:

I looked at the specs of the pulse valve. It says 24V DC only. What did you do for your shooter to get a 24V on a 12V system?

Take a look at the PCM/PH. Should have a setting to go from 12V to 24V.

@Cary we used a pneumatically actuated dust collector pulse valve. There’s a separate line that runs into the valve. When pressurized, the valve is closed, when it is un-pressurized, the pulse valve releases the pressure quickly into the cannon. We used regular 12v solenoids to control that pressure.

I would urge folks interested in t-shirt cannon projects to read this paper about the risk of significant injury, and possible mitigations. it’s a cool project. it is not risk-free.


1 Like