All t-shirt cannons use compressed air to fire their projectiles. Most potato cannons use the easier to set up and more powerful combusion method, in which a flammable gas such as spray on deoderant is ignited with a spark. T-shirt cannons cannot do this because the heat caused by the expanding gas will burn the shirt. I have not yet had the opportunity to test this theory, but I believe this design will remove enough heat to ensure the safety of the shirt while still giving enough of a push to make it more powerful than conventional launchers.
The problem is that the usual ways to get rid of excess heat won’t work here. An exaust valve will reduce the pressure greatly. A fan will not cool fast enough, aside from most likely being destroyed on the first firing. There is the option of using some kind of flame retardant buffer, but I am against that as it means it must be loaded each time and most likely replaced after a while and thereby makes the whole system less “self-contained”. Increasing the volume of the barrel will allow the heat to dissipate somewhat but with that comes a decrease of pressure.
My idea comes from the heat sinks used in computers to protect heat sensitive components. Throughout the barrel will be equally spaced metal grates with holes in them. The metal will act as a heat sink for the sir flowing past, absorbing the heat safely into the metal while the air is only midly hindered and keeps most of its pressure.
Other, smaller changes I want to make to the basic design are as follows:
1) Instead of a screw on cap to allow fuel to be sprayed into the cannon, which leads to the longest part of any potato launch, I wish to use a ball valve, which can open and close much faster. I believe the ball valve can handle the pressure (as they are used in almost all air-pressure cannons) although I will be doing a smaller scale test to confirm this.
Note: I still want the back of the main barrel to unscrew so that adjustments can be made as nessesary. I just don’t want to open it every time.
2) Instead of using the traditional stun gun to ignite the gas, which, simply because it uses a stun gun, makes the entire system much more dangerous and harder to work with, I wish to use the igniter from a BBQ lighter, which uses no batteries and is much safer. This idea comes from Makezine.com here.
This concept was confirmed to be possible by a professor in mechanical engineering at BU. In the summer of 2012 I built such a cannon, with a few modifications. A more detailed description can be found here.