PowerLabs Cannon Research


(WARNING: All the information contained on this page is for educational purposes ONLY! Do NOT attempt to replicate ANY of the experiments contained here. No more information will be volunteered by PowerLabs on these subjects than what is contained here. Be aware that each and every experiment described here is exceedingly dangerous and requires training, equipment, and skills that are not all mentioned on this page!)My sister's door with a tennis ball sized hole right through it and the cannon that made that holeMy interest in cannons began when I was still living in Holland, and was only 14 years old. During my wonderings on the web searching for useless things other people had done I found out about “Potato Cannons”, or “Spud Guns”. Back than the hobby was only getting started and there were only two pages with information on the matter, as contrasted with the hundreds of pages distributed in several web rings we have today. The basic idea consisted in spraying some hair spray inside a plastic PVC pipe and igniting it through an electric sparkler to produce pressure which would propel a potato out of the pipe. As soon as I saw that, I knew I had to build one! However mine would have to outdo anything anyone had done up until than…
First of all I chose to fire a tennis ball out of the barrel, as I didn’t believe a potato would sustain the forces I had in mind. Secondly, hair spray (which is mainly a watered down alcohol solution with all sorts of sticky polymers which produce relatively little heat, burns slowly and stink like hell when they burn) just wouldn’t do it. I decided the most energetic propellant I could use would be a 60%propane, 40% butane gas mixture which is commonly used in blow torches for small welding jobs (hydrogen was also considered, but its 500PSI combustion pressure was deemed too hazardous. Acetylene would be even more spectacular, but too prone to detonations). Also, I was going to burn it in a pure oxygen atmosphere, since that would allow me to input 5 times more fuel in, for a much higher energy output, combustion temperature, and consequently higher pressure. The combustion pressure of the mixture is quoted as being in the 300PSI region (Van Nostrand’s Encyclopedia). In order to hold that, the pipe would have to be quite thick! I chose the largest diameter (20cm, 8″) pressure rated pipe the nearby hardware store had as my combustion chamber and used a coupler to bring it down to a second pipe that was the diameter of a tennis ball (8cm). The pipe and coupler were all 4mm thick walls, which I *hoped* would sustain the pressure. I used a multiple spark piezo-electric stove igniter with a spark gap on the o-ring sealed threaded cap on the bottom of the combustion chamber and placed two wires forming an “X” at the beginning of the barrel to prevent the tennis ball from falling in. The entire device cost me some $100 (not counting the fuel) and 2 days of construction to build; a time during which while my parents were in Germany on a 2 day trip, as I knew they would not approve of it. Once the glue was thoroughly dry, the time to test it came. Unfortunately, it was raining outside so I decided to test it indoors (OK, so I was not the smartest kid back than, but come on, I was only 14 years old!). I filled the combustion chamber with pure oxygen from a 900CC welding cylinder I got from the same hardware store, and injected the calculated (slightly over stoichiometric) amount of fuel into it using a syringe. I dropped a tennis ball into it, aimed it at my sister’s door (a solid wood door, 50 meters across the living room from my room where I was standing) with a slight upwards angle, expecting it to bounce back and stop (never having experimented with fuel/oxygen mixtures before I had no idea of just how much power I was dealing with), and clicked the igniter.
A massive explosion ensued and I was knocked on my back with the recoil, which also left me slightly dizzy and with ringing ears from the muzzle blast. Still stunned by the sheer power of the shot (and amazed that the cannon was not blown to pieces on my hands), I looked at my sister’s door only to see that there was light coming in through it. I remember being worried about it, but not so much as I was hoping to find the missing piece on the other side and somehow glue it back. On opening the door I was amazed to find out that the wooden door had a foot square (30x30cm) chunk blown off from it as the ball pushed material from the front out through the back at several hundred km/h. The entire room was covered with splinters and the door had cracks on it running from side to side! As you might imagine, explaining this to my parents wasn’t very easy…

The next step: Nitrocellulose (Gun Cotton)!

(From left to right) me, Gabriel, Henry at a dance club in Brazil; the Cannon trioThe “Tennis_ball_through_sister’s_door” story has always been one of my favorites to impress friends… People usually give me an amazed look when I tell it, laugh a bit, and end up telling me I’m crazy and leave it at that. However, when I told my friend Gabriel about it, he had quite a different response: he said we had to do it again. Now I was 18 years old and living in Sao Paulo, Brazil, and during the 4 years that had passed since my first cannon experiment I had experimented with some pneumatic devices (1/4th turn ball valve and burst disk diaphragm) and some smaller, gas/air cannons, as well as experimented thoroughly with several propellants, including one of the most powerful gun propellants ever created; Nitrocellulose (guncotton). I knew that if I was to outdo my oxygen-propane/butane tennis ball cannon this time I would need to go for a solid propellant. Gabriel already had the idea for a projectile; a golf ball (he is a golfer), so now all we had to do was find the appropriate pipe. I was very weary of using PVC for any nitrocellulose experiments since I knew from experience that the combustion pressure of NC can climb to very dangerous levels very quickly if anything gets in its way, and make a pipe bomb out of the cannon barrel, but after we found a pressure rated pipe which fit the ball perfectly and had 4.2mm thick walls and a pressure cap we decided to do it just the same provided that we could find a way to fire it well away from us, so that in the case of a mishap we would be safe from flying PVC shrapnel. At first we decided to test the pipe’s strength by firing a golf ball out of it using an M-80 (a large flash powder filled firecracker). Sure enough the pipe exploded sending flying PVC pieces everywhere. 1.13mB Video demonstrating nitrocellulose as a gun propellant. The PVC pipe is the same as was used in the first PVC cannons and is full of butane gas, which is first ignited. A second match is thrown in which deflagrates .8grams of nitrocellulose in the bottom of the pipe. I was mildly discouraged by this incident, but I concluded that the reason why the pipe exploded was because the bomb caused the pressure wave to impact its walls, as opposed to cause a gradual pressure increase that one would want in a combustion chamber. We decided that before the research could be carried on to the next level, we would need *lots* of propellant… I obtained 2litres of 98% analytical grade Sulfuric Acid, and 1.5 liter of 68% analytical grade Nitric Acid, and my quest to upscale the production of Nitrocellulose began. The procedure used was similar to the one outlined in Chemlabs, but it did not come without its difficulties. Up scaling a chemical reaction is a very dangerous thing; something that works in small scale can go terribly wrong as the volume of reactants increases much more rapidly than the surface area of the container that allows it to cool: Being a very exothermic reaction, the nitration of cotton into nitrocellulose hexanitrate tended to run away several times and it took me 3 attempts to get it right. The final procedure required that the acids be cooled to -10C prior to the cotton addition, the cotton be added one at a time, and a larger excess acid be used. Temperature control was also very critical. Washing the nitrated product was also a problem. I found that merely dumping the cotton in water caused enough heat to be produced during the dilution of the sulfuric acid that the cotton would start decomposing violently producing large amounts of poisonous N2O fumes.
In the end, once all these problems were solved, a very large batch of nitrocellulose was produced: 280 grams. This required two litres of mixed acids and several packages of makeup removal cotton disks. Once all the nitrocellulose was prepared, it was thoroughly washed, neutralized with sodium bicarbonate, and stabilized with urea. This was kept under water until needed, and than dried in a vacuum dessictor prior to use.

Cotton pads used for making Nitrocellulose A very large nitrocellulose batch

 The research goal became to find out what was the highest possible muzzle velocity we could achieve from the 1meter long PVC pipe cannon. Nitrocellulose was placed in it and ignited through a length of cannon fuse that was placed into the pipe through a 4mm diameter hole on the side. We fired the cannon upwards using increasingly larger amounts of propellant (0,5gram increments) and used a chronometer to time the airtime of the projectile and than inputted that into the equation v = sqrt( 2 x g x d ) to solve for its average velocity (muzzle velocity should be higher). At 2.5grams we obtained a firing velocity of 560km/h. At 7 grams the velocity was calculated to be 960km/h. Since we were firing in a residential area (Stefan’s house) the cannon was making enough noise to really bother the neighbors and earn us complaints so we decided to fire one final shot with 14 grams. The pipe exploded into pieces and we decided that a new cannon had to be built with the objective of breaking the sound barrier (347m/s, 1137ft/s, 1249km/h).

PVC Cannon:

   Me preparing to fire the cannon. Earmuffs a must!Duck hunting anyone?After it had been tested with up to 8grams of nitrocellulose I attempted firing the cannon with one hand using only 2 grams. A 50 gram golf ball going at the 500km/h imparted into it by the 2 gram shot has a kinetic energy of 482Joules. This is about the same energy of a .38 revolver bullet (the amount of propellant used is roughly the same too), so as you can imagine there was quite a bit of recoil from the lightweight pipe I was holding (1mB video). I never attempted this stunt again, in fear of dislocating my wrist. The second video (480Kb) shows it being fired over a lake with 2.5 grams of propellant. The video camera ran out of battery before we could tape the more spectacular shots, unfortunately 🙁 The most powerful shot from this cannon imparted approximately 1.8kJ of kinetic energy into the ball (using 9.5grams of NC; over 10kJ of chemical energy); more power than a .44 magnum round fired from a rifle! I believe it *is* possible to break the sound barrier with this setup, and probably very easy if a longer barrel and around 10 – 12grams of NC are used. I did not, however, find this to be a worthwhile endeavor after the research was moved into a steel pipe setup.

Steel Cannon:

Top view of cannon barrel, showing its 1/2" thick steel walls.

 We knew that if we were to safely break the sound barrier with a golf ball on a cannon we would need a very strong barrel. We decided to use a steel pipe and went out to a couple of scrapyards to find one. After driving for almost an hour and looking everywhere for something that would have just the right inner diameter to fit a golf ball we came across an incredible finding! A very large Caterpillar truck pneumatic actuator arm; the pipe was 1m60long and had chromium plated alloy steel walls 1.3cm (1/2″) thick! Since it was designed to withstand extremely large loads in its duty, we were sure that it would withstand the firing pressure required to get a golf ball going to mach1, and beyond! Best of all, it had a 2cm thick steel end cap fitted on it!
Our first test consisted of firing a couple of balls out of it using M-80s. The balls would go about the same speed we got with 2 – 3grams of nitrocellulose on the other cannon, which is not very surprising since flash  powder has a lower energy density than nitrocellulose and is not as efficient as a propellant.
Since we did not have a proper place to fire it at yet, we decided to test the pipe with a massive blank shot; 100 grams of nitrocellulose were loaded into it, along with 30 grams of black powder and 30 grams of flash powder. A small gas can was placed over all this propellant to produce a fire ball once it was all ignited.

 On ignition, this blank shot produced a column of fire that lip up the sky from horizon to horizon and extended 30 meters (90 feet) straight up into the night sky. As the sound of the explosion thundered through the firing site, the pipe rammed itself over 1 meter  into the ground, making it very difficult for us to remove it. Once we got it out, we noticed that the 2cm thick steel cap had been blown off by the enormous pressure! It was time to make a new one.
With Stefan’s father help we welded a new 2cm thick cap on the cannon, and resumed testing.

Grinding the metal so the weld stays on better

Welding the pipe.Cooling the pipe with Liquid Nitrogen


 A few more blank shots were than attempted with smaller (20 – 30 grams) amounts of nitrocellulose and these were taped. The nitrocellulose used for these shots was slightly damp and so it burned at a slower rate, making visible fireballs. Videos of the tests are available for download by clicking the still frame captures (450 and 300kB each).20 gram blank shot30 gram blank shot

Projectile shotThe culminating point of this project was loading 30 grams of a potassium bromate/sugar/sulfur pyrotechnic mixture into the cannon with a 200 grams nitrocellulose hexanitrate propellant charge, and ramming the golf ball into it with a tight fitting cotton sabot. The cannon was fired with a long fuse and video taped from a distance of 100 meters. In this video (a PowerLabs MUST see! 890kB), a whistling sound is heard as the pyrotechnic mixture combusts and creates pressure in the combustion chamber, than the thunder of the golf ball leaving at over 2 times the speed of sound can be heard echoing through the valley. Given the excessive amount of propellant used (as evidenced by the large muzzle blast ensuing the discharge) and the light weight of the projectile, it was assumed that the muzzle velocity was around the maximum possible for the propellant used; in the 700m/s (2500+Km/h) range. The golfball than must have had a kinetic energy in the 13000J range, and flown over 10kM (6mi+) horizontally; about the same power of an anti-aircraft round. It was postulated that metal objects (truck ball bearings, “D” cells, motorcycle pistons, steel cylinders) could be fired from the cannon to achieve even more spectacular energies, but we never got around to it since I moved. The cannon now lies as a rusty abandoned pipe somewhere in rural Sao Paulo.