My oxygen fired ABS-pipe cannon

My oxygen fired ABS-pipe cannon

Not your ordinary potato cannon! I just had to take the sport to its full level of insanity!

Having seen my friend Ron Harding's potato gun, which in itself is already more sophisticated than the average potato gun, I was inspired to build my own.

Of course, building one just like his would have been pointless. At the same time, there was still some improvements to be made to the technology beyond what Ron had done. The main thing was the fuel mixture. Ron's cannon is fuelled with propylene and air as an explosive mixture, an no means of measuring how much fuel is in the cannon.

My cannon I designed to be fired with an oxygen-propane mixture, although other gasses work just as well. One of the major design criteria was to have a fuelling system where I could see how much fuel I put in the cannon, and where I did not have to put my hands near the explosive mixture or loaded cannon. The possibility of the mixture auto-igniting at an unexpected moment, or the cannon simply exploding being a serious consideration.

What I came up with was my balloon fuel metering system. Basically, the balloon on a T-connector. One side of the T goes to a hose connected to the fueling valve on the cannon, while the other end is pushed onto the end of a welding torch.

I begin by connecting everything up, with the fueling valve on the cannon in a closed position. Next I fuel the balloon to about a 6 cm diameter with propane. Then I fill it with oxygen to twice that diameter. I do all this from a distance as large as the length of the hose will allow. Fueling the propane first makes a lot of sense, as it takes much less of it, and by adding the oxygen afterwards, I can be sure I actually have the propane in the balloon as opposed to the hose.

The next step involves transferring the fuel from the balloon into the cannon. This is accomplished by opening the valve into the cannon. I have a slot carved into the end of my ramrod, which allows me to turn the valve from a safe distance. Its generally best to let the fuel into the cannon very slowly, so that the pressure of the gas does not push the projectile forward in the cannon. The cannon also has a bit of a 'baffle' near the fueling valve to make sure the fuel does not 'shoot' through the chamber, but slowly creeps through it, displacing the air (I hope).

The ignition system took a few tries to get right. My first system consisted of two screws screwed into the side of the ABS pipe, connected to a BBQ lighter. This system typically worked for the first shot. After that, there was too much condensation in the cannon, and I couldn't get a spark anymore. The next step was to install a spark plug, which is much better at maintaining insulation once you get some humidity in the canon. Still the wimpy BBQ lighter wasn't enough. So I switched to using an ignition coil. This also has the advantage that the ignition coil can be placed near the cannon, and the low voltage wires connected to the primary side can be run for quite a distance without having to worry about the wires arcing somewhere. In its present form, the ignition system only fails if you forget to reattach some wires after the recoil from the previous shot has yanked everything loose.

Along with a more advanced fueling and ignition systems, I also did some development work on the projectiles. The problem with shooting potato slugs is that, on impact, they just turn into a fine potato mist, but often do not penetrate the target object. Having recently bought a lathe, I decided to make some nice areo-dynamic projectiles out of hardwood. Initial experiments indicated that the projectiles quickly started tumbling in the air once they left the cannon, which is bad for both aerodynamics and accuracy. I then started to make projectiles that were just short cylinders, seeing that there was no point in making one end pointed.

Later on, I started putting a tail on the projectile, which makes it aerodynamically stable and continue to point forward. When loaded in the cannon, only the front part is in the barrel, while the tail points into the combustion chamber. The above photo shows some of my latest projectiles - before I shot them. The elaborate tail on the middle projectile turned out to be a complete waste of time. With the acceleration in the cannon probably at about 2000 times the force of gravity, the tail fins were still in the cannon's chamber after the shot. The tail on the bottom right projectile also did not work very well. It wasn't made of hardwood and just ripped apart when I shot it. A straight hardwood tail works the best.

Here's a shot of me loading the cannon with the ramrod. The ramrod isn't really necessary when shooting just a hardwood projectile - I always make these so there is about a quarter millimeter of space so they slide in and out freely. When putting wadding behind a projectile, or shooting potato slugs, the ramrod comes in really handy. Note that the projectile is loaded BEFORE the cannon is fuelled.

I did some measurements of the speed of the projectiles after leaving the muzzle of the cannon, by shooting through two pieces of wire half a meter apart and measuring the time interval between when the electrical circuits were interrupted. My good measurements all come out to about 175 to 180 meters per second, or just over half the speed of sound.

With the projectiles travelling that fast, there is little fun in shooting into the air and watching how far it will go. There is a tremendous bang when you trigger the cannon, but you just don't see the projectile - its just gone! Shooting at a particular target at close range is much more fun.

Small and helpless pieces of electronics are ideal targets - there's just so many pieces to become dislodged on impact. Numerous pagers, as well as an answering machine have met their bitter but entertaining end this way. The above shot is of another pager getting shot. I think in this shot, it wasn't a total direct hit, and the projectile got deflected under the piece of carpet, thus pushing it up. The carpet wasn't that effective at stopping anything either. Numerous holes ended up in the carpet from projectiles going through. On cleaning up, we found a small 1.5" circular shred of carpet - the same diameter as the projectiles - lying around, This, no doubt, was punched out by one of the shots.

I also tested the cannon against some 3/4" plywood. When I shot a hardwood projectile against it, it went straight through, and fairly cleanly. By the shape of the hole, I could tell that by the time it hit the plywood, the projectile was travelling sideways, and pointing down to the left. This was before I started making projectiles with tails.

A potato shot at the same piece of plywood mostly became absorbed within the plys of the wood, although the back of the wood bulged. Part of the potato sprayed through the cracks out the other side.

Shooting against the side of an old pickup truck produced less spectacular results. The sheet metal, because it is able to yield, is capable of absorbing a lot of energy before it tears. Its interesting though how much paint got peeled cleanly off the sheet metal from the impact. The indentation was about 2 cm deep from the original surface, and extended to the edge of the peeled paint. Of course, as the picture shows, the cannon did less damage to the metal than a bit of rust left to its own devices for 20 years.

Its interesting that the ABS pipe is able to withstand the abuse of being used as a cannon at all. Before firing the cannon for the first time, I took numerous precaustions to shield myself from the grenade like explosion of shrapnel that might result. Through the course of numerous 'tests' I became more and more confident that the ABS would in fact hold up. Nevertheless, after maybe about 30 firings, my cannon design did experience a failure. My original design hat a barrel that was actually screwed in to the main chamber using an ABS screwthread adapter, shown above. This was naturally a weak link, but I figured that the chamber, with its larger diameter and hence greater forces, would fail from pressure before this link would. I think what eventually happened is that from the recoil, the chamber always got bounced around, and eventually snapped the thread off just from sideways banging. What is more surprising is that the 1.5" barrel actually developed a longituinal crack. The barrel should experience significantly less stress on the ABS than the chamber. At any rate, the picture above shows the failure of the barrel at the breech end.

The comforting thing about this failure is that it did not produce ANY loose flying objects. Fact is, The first thing I noticed that was wrong was that the projectile, when it hit the target, did no damage whatsoever. My brother had videotaped that shot, and you could actually see the projectile coming - it was that slow! Most of the gasses had excaped though the crack, which no doubt opened wide when the cannon was fired. I had done some tests on pieces of ABS before building the cannon to see how much it would deform before breaking, and expected to have few flying pieces in a failure. It was nice to see this confirmed. I understand that PVC (the white stuff) does not have nearly as nice a failure characteristic.

I subsequently bought a new piece of ABS pipe for a barrel, and this time bonded directly into a new 3" to 1.5" adapter, thus eliminating this weak spot. I haven't had any mechanical failures in the cannon since.

Back to cannon page

Back to Matthias Wandel's home page