FireseveN, thank you for your ideas.
Yes, I am aware that there is a challenge in synchronizing release and ignition. I had on thought of it, while having nothing else to think of today

: If we are very lucky, a solution where the valve piston depresses a number of BBQ igniters when it hits the extreme open position, would time well. That would be elegant. If we are not lucky, a pneumatically actuated valve could still be used: Stick a magnet in the piston, and a Hall sensor w/ circuit outside. I suggested an electronic sequencer for the inital experiments, and it shouldn't be hard to build one.
As for the fuel type, I still hope for a liquid fuel solution. It could be so easy: Just fill up the tank for every few shots. I think this ether based fuel for R/C models may have some possibilities.
The idea of having a gaseous fuel compressed along with the air in the tank is a little scary: If there is some design guarantee that the two won't mix in the tank, then I'd be happy to fire it hand held. If there is not ... oh no ... another statically-fired, potentically lethal monster. That is why I like the carburettor talk, whether that is a commercially made carb, or a simple home made venturi one.
I think the first thing to be done still is to get some data on flow through jets in venturis (for example, right after a Supah-valve is preactically a venturi) in a pneumatic gun. Both for liquid and gaseous fuels. If venturis can be made to work precisely & reliably, they would be a really nice solution.
I don't have much experience with the importance of barrel length, except, of course, that a longer barrel usually increases velocity. I can see two advantages of a long barrel in the design:
- There is more time for the gas to burn at a place where is is useful for potato propulsion
- Long barrels should go well with hot gases: Hot gas is thin, and the flow resistance should mean less than with cold gas. My super-sonic compustion gun was 117 calipers long.
Søren