How Did This Happen?!!
The 'Lorenzoni' system
It seems like there is nothing new in firearms design. In the latter 17th Century, A Florentine by the name of Michele Lorenzoni is reputed to be the inventor of a seven to twelve shot repeating flintlock rifle, where the powder and ball were stored in separate compartments in the stock and loaded with the turn of a crank. Did I mention that the gun was also self-priming? It was a complicated gun, expensive to make with the hand tools of the day and had the bad habit of exploding if not properly maintained. It was, however, the first true magazine 'repeater' and was noteable enough to be mentioned in the diary of Samuel Pepys. Lorenzoni-style rifles and pistols were made up the beginning of the percussion era by notable gunsmiths such as Nock of London. The Cookson Repeater, briefly produced in this country, was a variant of this design. Needless to say, surviving originals, particularly the pistols, are rare and stunningly expensive.
I would be hesitant to build a black powder pistol of this type; I like my fingers just where they are, but a pistol of this type using smokeless powder could be designed to be perfectly safe. Black powder easily ignites and will exlode with little confinement, but smokeless powder is very hard to ignite and unless well-confined will simply burn slowly. All that would be needed to prevent a magazine explosion would be to keep the powder tube from a direct flash-over from the chamber, and even if this were to occur and ignite the powder, provide a low pressure magazine 'blow-out', away from sensitive body parts.
The old run-around – a rotary chamber
The first design attempt was to essentially duplicate the Lorenzoni geometry, where the powder charge and ball are on the outside of a rotatable wheel. This led to a number of construction problems with the electrical ignition source, so next, a rotary-valve design was used, where a rotating plate containing a small powder well and the igniter was shuttled across a spring-loaded powder tube, and a depression a few degrees away held the ball.
The image to the left is a CAD design showing the left side of the pistol, with the rotating mechanism in the 'firing' position (the rotator has been made transparent so the inner realtionships can be seen). In this firing position, a ball is pressed down the ball tube by an inner spring mechanism and rests in a space in the rotator just in front of the ball pusher, loading it into the rotator for the next shot. After firing, the rotator handle is pushed up (right image), and the ball pusher is forced forward by the ramp on the rear of the receiver, and when aligned with the barrel, a spring in the ball pusher (not shown) loads the ball into the barrel breech. This also aligns the firing chamber with the powder tube, where a small charge of powder is pushed into it. When the rotator handle is pushed back down, the now-filled firing chamber is aligned behind the ball in the barrel, and the pistol is ready to fire again.
Lighting It Up
As discussed previously on my Plink-King and electric ignition web pages, smokeless powder is very hard to ignite. Fortunately, I had already solved this problem, so it was a relatively easy matter to screw a ZTA ceramic igniter into the rotator, leaving just enough of a cavity to meter the 0.7gr powder charge. The fiddliest problem was trying to find space for all the circuitry inside a small pistol. This was solved by moving the ignition booster just behind the receiver, putting the battery in a 'magazine extension' at the end of the grip and installing the rest inside the grip.
Oops - we have a gas problem...
This first working model was assembled with the rotator having just enough play between it and the receiver to allow it to rotate - probably less than a 0.001" gap. After clamping it into a vice and aiming at a large steel backstop, it was tested with a firing chamber holding about 0.7 grains of Hodgdon's Tite-Group powder. The first shot gave a satisfying 'pow', but I didn't hear the shot hit the backstop. I cycled it again, and following another 'pow', the dirt was kicked up a few feet in front of the backstop - something strange was going on here... When the third shot didn't show anything, I removed the assembly and probed the barrel. In spite of the authoritative bang, the ball had only gone halfway down the barrel! What had been happening was the gas was escaping from the tiny gap with a loud bang, but because of the small powder charge (and the relatively small amount of gas evolved from it), most of the energy was dissipated before the shot even got out of the barrel.
Next, I tried upping the powder charge to about one grain. This allowed the shot to hit the backstop, but the blow-by between the rotator and the barrel was very noticeable, and I was getting nervous about a possible powder tube ignition. So - the powder charge was reduced back to 0.7 grains and an attempt was made to further decrease the rotator gap. The pistol seemed to fire better when the tolerance was reduced to the point where the the rotator barely turned, but it was apparent at this point that another approach was needed.
Girandoni to the rescue - Harmony at last!
I cribbed another early invention, that of Bartholomäus Girandoni. He developed the famous repeating air gun that Austria used in the Napoleonic wars and was also reputedly carried by the Lewis and Clark expedition. This repeating air gun used a harmonica-like shuttle to transfer balls from a magazine into firing position, where the shuttle formed a gas seal between the air source and the barrel. By quickly moving the shuttle back and forth, a shot could be taken every few seconds.
I designed a variant of this shuttle mechanism which contained the igniter and a ball transfer device. After firing, pushing the shuttle in loaded the ball into the barrel's chamber while simultaneously dispensing a precise powder charge into the igniter's powder chamber. Returning the shuttle to firing position aligned the powder charge with the barrel and also allowed another ball to drop into the shuttle. The image to the left shows this implementation in its firing position, where the left tube holds the balls, and the right one holds loose powder. After firing, when held in a vertical position, gravity feeds the balls into the ball transfer chamber (as shown). When the shuttle is pressed to the right, the spring presses the ball pusher forward, and when it aligns with the barrel, the ball is popped into the breech. In this position, the powder can also drop down and refill the firing chamber. Returning the shuttle to the position shown completes the loading cycle.
The left image shows the 7075 aluminum receiver and attached 416 stainless steel barrel and Picatinny rail. The 8" barrel was cut from a 24" stock length and screwed into the receiver with 5/8 x 24 threading. The passage for the shuttle is closely fitted front and back, but the vertical position can be adjusted with set screws bearing on Torlon balls to allow precise up/down alignment for both the ball feed and the chamber. This also gives some space where blow-by can easily escape without causing any problems. The right image shows the shuttle partially inserted into the receiver, showing the firing chamber closest to the receiver and the ball-loader to its right. The shuttle is easily removed by simply unscrewing the left end cap and sliding it out.
In the left-hand image above can be seem a small plastic button on the rear of the receiver with a spring coming out the back. The spring makes contact with the 'hot' contact of the igniter coil in the grip, and the solid steel electrode, attached to the spring and running through the plastic stops about 0.040" back from the shuttle. This gap is enough to keep the 600 volts of the power supply from reaching the igniter (and possibly prematurely firing the pistol). When the trigger is pressed, a high-voltage spike from the igniter coil jumps this gap and allows the main electric charge to be carried along. This gap behind the shuttle substitutes for the spark gap switch in the power supply used for the Plink-King rifle.
I just can't tolerate any more!
The first shots went much better than with the rotary version, but when the tolerance was reduced enough to minimize blowback, the shuttle became too stiff to comfortably move. Since the slide and receiver were both made from the 7075 (unanodized) aluminum, galling wear started to show up on the shuttle. It again became clear that, although this was better design than before, yet another approach was needed to contain the powder gasses.
The Bernoulli seal
Stealing ideas from yet another sage, Daniel Bernoulli and his principle said that as a fluid flows rapidly, its pressure drops. So how does this help us here? The idea is if a moveable seal were placed into the well formed between the end of the barrel and the shuttle, and was minimally compressed, it would form a tight but easily moveable bearing surface with the shuttle. When the gun is fired, pressure will be applied to both the front and back of the seal, but any gas can escape only along the surface that mates with the shuttle. As gas begins to escape there, however, it lowers the pressure on that side and the full pressure on the front side will then force the seal even harder against the shuttle. This will not completely eliminate gas leakage but will greatly minimize it.
Paper or plastic?
The problem was then what to use as the seal. The left-most image shows a first attempt, using a combination of a stainless steel disk with a partial cut-out for an o-ring. The o-ring kept the powder gasses from going around the disk and also provided a slight forward pressure on the disk to initiate the seal. This seemed to work well, but making the disk and o-ring combination was very 'fiddly', so Radel, a hard engineering plastic was tried, and this material gave a very smooth action to the shuttle. After a few dozen shots, and again after a few hundred, it was removed for inspection. The right-hand image shows the effect of gas-cutting. As seen on the left-hand side of the image, once a tiny channel is opened up, either by a scratch or by intrinsic fit, the hot powder gasses escaping along that channel will erode more and more material until the leak becomes apparent, as seen on the right-hand side. This was not too bad, but a more permanent solution would be better yet.
A common solution wins out
The failure of the Radel seal was not because of simple burning but was solely due to gas channeling along the front edge. Because it is not possible to completely eliminate this, it was clear that a refractory surface was needed, so a small ring of 303 stainless steel was cut to act as the bearing surface for the shuttle. [Note: The stainless steel was later changed to hard brass, which slides better] It was backed with a simple square hard Buna-N o-ring which formed the inner surface of the firing chamber and supplied a small push to seal the ring to the shuttle. When the gun is fired, the pressure is uniformly distributed inside the chamber, so the o-ring has no where to go but outward - but it can't do that either, since it is constrained by the unmoving walls. This outward pressure 'squeezes' the o-ring and acts to elongate it, forcing it even harder against the metal seal. Although the internal temperatures are extreme, they are of brief duration and only slightly affect its surface. After all, shotgun shells are made from plastic with even a lower melting point, and they can be reloaded many times. However, eventually a small amount of debris finds its way into the cracks and allows some gas channeling; this grows over time, and after several hundred shots, the o-ring can erode as seen in the right image. This slightly increases blowback but does no harm to the mechanism, and another ten-cent o-ring and some cleaning fixes the problem.
Testing - Testing...
Initial MV, accuracy
Even with the tiny powder charge, the pistol was surprisingly powerful, spitting out the ball at about 900 fps. Unfortunately, it wasn't very accurate, even for a parlor pistol. I tried a series of other projectiles, and found that pre-swaged #3 (0.250") buckshot (right image) shot very accurately. The difference may have been because the swaging process gave the projectile a flat base with relatively smooth edges, as opposed to the self-swaged #4 buckshot which, as seen in the left-hand image, seem to have little fin-like projections on the back. This observation lead to yet another round of experiments.
'Stop kicking my balls!'
After the last results, the working theory was that the blast of gas following the projectile out of the barrel could catch the projections at the base of the ball and 'kick' them off course. Since the basic concept for this pistol was simplicity, I didn't want to abandon the ease of procuring and loading the ammunition, so another course was considered - peeling off the gas stream prior to the muzzle by venting the barrel. Two 0.125" holes, about 0.25" apart, were drilled laterally through both sides of the barrel about 3/4" from the muzzle in the hope this would be sufficient relief.
Another round of test firing showed that these worked fairly well. Not only was the accuracy much improved, but the muzzle velocity was not measurably affected. However, like much else, there were side effects. The gas and powder debris escaping from the vents spattered hot material all over the place and was a hazard to anyone to the side of the shooter. To fix this, a barrel shroud was added to catch the debris and direct if forward. Contrary to the way it looks, it is not a moderator (silencer) - just a hollow tube to contain the debris and keep it out of everyone's eyes - a little 'Porta-Potty' - and as you can see, just like the real ones, it fills up with s***...
Several powders were selected from the Hodgdon Burn Rate Chart Available ones were tested for muzzle velocity and accuracy at 25 yards. Note that the powder weights vary substantially in the fixed-size chamber. Interestingly, the AA-Lite would barely ignite - firing only about one time in three; the others had no problem.
|Winchester AA Lite||0.54 gr||605 fps||16.7 fpe|
|Hodgdon Tite-Group||0.86 gr||870 fps||30 fpe||1.3"|
|IMR 700-X||0.43 gr||480 fps||17 fpe|
Loading up the pistol is insanely easy; up to 18 paraffin pre-lubricated buckshot are dropped down the ball tube. Rotating the front of the grip the other direction exposes the powder tube which can be easily filled from a pan-charger. After filling, the fore-piece is rotated closed, and tightening the barrel shroud holds it in place and ready to shoot. So far, the best performing powder seems to be Hodgdon's Tite-Group, which is a fast, double-based pistol powder; others may be tried when they again become available.
Getting a grip
In one of my older pistol designs, the grip was machined from a block of Delrin (an engineering plastic), but this design was much more complicated, with many internal channels for the magazines, wiring and components, so it would be very difficult or expensive to make. In the past few years, however, digital '3-D printing' has become very easy, and even impossible-to-machine parts can be cheaply made. All of the non-metal parts (foregrip, magazine, grip and component holders) were printed using a process which creates them by fusing nylon particles with a laser (SLS process). This gives a part nearly as tough as if it were moulded and allows very complicated internal geometry. The parts do not come out perfectly, but a small amount of hand-fitting is all that is required.
Volts for Jolts
The left image shows the grip with the panel removed, showing the electronics inside. The
battery (and 'on' indicator) is shown to the right. When the stud just below the trigger is pressed
as the pistol is gripped, it actuates the small switch inside the grip and sends power from the 3V lithium
battery to the CFC inverter, which ups the voltage to about 600VAC. The HV diodes at the top of the inverter
transform the alternating current to direct current, which is stored in the HV capacitors imbedded in the upper
part of the grip.
When the charging switch is active, a LED indicator in the bottom of the grip is lit, allowing the shooter to see the pistol is 'live'. The circuitry is also designed so the pistol cannot fire unless both the charging stud and the trigger are activated simultaneously - in effect, making the charging stud a 'grip safety' as well.
Even though the 600 volts stored in the capacitors sounds like a lot, it is far too little to jump the gap in the electrode, because the resistance of the air (or powder) is far too high. To start the arc, a small igniter coil in the grip is briefly energized to boost the voltage to over 10,000 volts, creating a weak spark which will jump the gap. This weak spark briefly lowers the resistance along its path enough so the 600 volts can now flow and carry the stored energy in the capacitors and ignite the powder. The electrical connection from the grip to the receiver is through a spring on the rear of the receiver which presses against the center of the coil, where there the 'hot' connnection lies. The ground return is through the small spring on the grip which presses against the metal frame of the receiver.
Making this pistol has been a mixture of great fun, incredible frustration and an amazing learning experience. For those who might want to duplicate it or experiment further, CAD files can be found in this archive, which contains them as STEP files and a GeoMagic package.
Additional data will be posted as I get more experience with the pistol and as I can design more tests for it.