A Modular, Chamber-Loaded, Black Powder Pistol

Firing modular muzzleloader

A photo of the pistol being test fired with a .240 self-swaging round ball using 3.5 grains of 3F Swiss black powder. A 3-D view of the pistol can be seen here, and the construction details are in a separate web-page here. Note that to view the PDF file in all dimensions, you will need to use the latest free Adobe Reader

The Genesis of the Project

I thought it would be interesting to make a 'muzzle loading' type pistol that could easily be reconfigured to try out various combinations, such as:

To do all of these, I needed a design which is highly modular, and I chose to do it as a pistol, simply because it was simpler to do so (and I have a fondness for handguns). A chamber-loading design seemed to be the easiest to make and would allow some reconfiguration by simply changing chambers. The barrel was designed as a screw-in so it could be quickly changed as well, and for my old eyes, an integral scope rail was machined in so it could take a 7X telescopic sight for precise group testing and a reflex red-dot sight for general use. After listing all things I wanted to do with it, the first real design consideration was what type of ignition system to try first.

A Short History of Chamber-Loading Arms

Some of the earliest cannons were breech-loading via a separate chamber, and there were breech-loading pistols and long-guns in the early 17th Century which used chambers as well. Most of these suffered from the same problem - gas leaks which fouled the action and could cause injury to the shooter. Perhaps the best chamber-loading guns were the 'Kammerladers', made for the Norwegian army in 1842 and later adapted for sporting use in Liege. I used a variant of this design to make a breech-loading flintlock pistol which can be seen here. The best example of a chamber-loading arm is, of course, the revolver, whose origins also go back to the 17th Century, where revolving flintlock smoothbores and rifles were first used. My 'Modular Muzzleloader' is in many respects just a muzzleloading revolver with one chamber, but the advantage of a single chamber is, besides simplicity, the ability to hold very tight tolerance fits; this pistol has a gap of less than 0.001" between the chamber and the bore, thus much reducing gas leakage and side-flash.

Matchlock - Wheel Lock - Flintlock - Percussion??

Chamber 3-D model viewsAlthough I really like flintlocks, that ignition system was just too complicated if I wanted a truly modular system, so the first choice for the ignition system was a standard in-line percussion system, with the chamber having a capped nipple at the rear. After making some experimental chambers like the one shown here, I decided that, although they worked well, they were too 'conventional' for me, so I went all the way back in time...

Note: If you click on the images, you can get a rotatable 3-dimensional model of the chamber if active .pdf format.

Back To The Future - A Matchlock Ignition System!

As strange as it sounds, there are a lot of advantages to it - especially if it is done this way... A matchlock can be generally defined as a system where the ignition is accomplished indirectly by bringing a hot object in contact with gunpowder external to the main charge. This is a very reliable system, and except for the 'detail' of having to carry around a burning match is very practical and versatile. Based on previous experiments, I felt that instead of a slow-match, an electrical discharge could be very effectively used to ignite the external charge and thus fire the gun. The actual design involves the construction of a tiny 'fire pit' which holds about 1/2 grain of 3F black powder, where the loading chamber sits directly on top of the pit and has a 0.040" hole leading to the main charge.

Top of .240 RB Chamber Chamber end Fire pit in receiver
Loading Chamber Chamber Rear 'Fire Pit'

A key feature of the design is a means of preventing high pressures from forming at the rear of the chamber. As can be seen in the rear view above, there is a 0.325" diameter by 0.008" deep cut-out around the touch hole at the chamber's base, and upward radiating channels extending from the edge of the receiver's fire pit to a point outside the radius of the chamber. The chamber cut-out is shallow enough to prevent the escape of any 'priming' powder but because of its area, it allows gasses from the touch hole to escape radially, keeping the pressure in the 'fire pit' relatively low. At the center of the pit is a 0.050" stainless steel electrode coming through the center of a titanium ceramic rod, which provides heat-resistant electrical insulation. When the trigger is pressed, a charge from a photo-flash capacitor is directed between the center electrode and the rim of the pit. At full charge, the capacitor holds about 20 watts of electrical energy, and dumping all of this through the tiny powder charge instantaneously ignites it, forcing most of the 'flash' through the touch hole and into the main charge. The trigger acts as a crossbar switch, and it is literally impossible to feel when contact is made, so every time the gun goes off, it is a 'surprise', and there is zero trigger overtravel - both very desirable.

Fire in the hole! Here is what it looks like when the tiny charge in the fire-pit is ignited! Click on the image to get a short movie clip of it happening.

The Finished Pistol

Pistol and accessories Although it looks a little 'industrial', the finished pistol does have a certain flair. Shown here are the pistol grip and receiver with barrels and chambers for for .22 (.224), .25 (.25 Auto) and 9mm. There are two 9mm chambers shown here, one for the .375 self-swaging round ball and one for the 9mm airgun slug, which uses a reduced charge. One of the initial design considerations was that changing the caliber - or most any other aspect - should be easy, and it is. Chambers for the different projectiles are completely interchangeable within caliber, and changing barrels takes just a few seconds. Also shown are the 3-7X scope used for accuracy testing and a red-dot sight used for plinking; both quickly attach to the rail machined into the top of the receiver. Once everything was together, it was time for performance testing...



How To Measure Performance?

Measuring and presenting muzzle velocity data is pretty straightforeward, but how does one best describe accuracy, particularly when comparing dozens of combinations? Some loads are obviously better than others, but I wanted a way to quantitate more subtle differences. I used the same system used in military evaluations, the measurement of the "circular error probable", which tells you how large a circle will encompass 50% of the hits. I present a more extensive article on this subject here, with instructions on how to make the measurements.

Testing - "You Are What You Eat"

Not only us, but the guns we shoot... I was amazed to find out just how large the accuracy differences were, and how some hoped-for 'improvements' made things much worse. First, though, a word about lubrication...

Since all loads were 'bare ball', lubrication was clearly going to be critical if a string of shots was to be fired without cleaning in between. I found that the best method was to simply put a tiny amount of 'Wonder Lube' around the top of the projectile - just as one would in a cap and ball revolver. As the projectile moves down the bore, it deposits a thin film of lubricant which keeps the fouling soft, allowing subsequent shots to scour the residue from the previous one, thus preventing build-up of fouling. Even after fifty consecutive shots, the barrel - although not quite bright - appeared squeaky clean, and subsequent cleaning only removed a small amount of residue. In the absence of this lubrication, fouling built up very quickly and considerably degraded the accuracy. I also tried lubricated felt pads behind the projectiles, but these seemed to cause other problems.

The projectiles tested fell into roughly these categories:

Except where noted, all powder charges were Swiss 3F. All accuracy tests were hand-shot into a 8" X 11" target from a sandbag at 40 yards, using the 7X telescopic sight. Abbreviations and terms used:

Performance Tables

.22 Rimfire Barrel Liner: 6" (0.218/0.222 16" twist)

Projectile .22 RB .24 RB Swaged .222 Swaged .224 Swaged .224 Reversed .224 NAA Slug E-J Dome E-J Pointed Big Boy
Weight 15.4 20.5 23.5 20.5 23.5 23.5 31 28.5 32.4 26.2
Charge 3gr 3gr 4gr 4gr 4gr 4gr 4gr 4gr 4gr 4gr
MV xxx 1039 xxx xxx xxx xxx xxx 515 512 633
fpe xx xx xx xx xx xx xx 17 19 23
Accuracy NOP 4/10 1/10 3.8" 2.7" 7.3" 2.0" 1.2" 1.5" 1.6"

.22 Centerfire Barrel Liner: 9" (0.219/0.224 14" twist)

Projectile .240 RB .240 RB .240 RB .224 swaged NAA Slug E-J Dome E-J Pointed
Weight 20.5 20.5 20.5 xx xx xx xx
Charge 3.5gr 4F 3.5gr 3F 3.5gr 777 xxgr xxgr xxgr xxgr
MV 921 819 742 xx xx xx xx
fpe 39 31 25 xx xx xx xx
Accuracy xx" xx xx xx" xx" xx" xx"

.25 Caliber Barrel Liner: 6" Colt 25-Auto (0.243/0.250 14" twist)

Projectile .25 RB .25 RB AG Slug E-J Dome E-J Point Beeman Hollow Pt
Weight 23 23 51 35 42.5 26.2
Charge 4gr 4gr 4F 4gr 4gr 4gr 4gr
MV xxx xxx xxx xxx xxx xxx
fpe xx xx xx xx xx xx
Accuracy 2.2" 2.1" 2.5" 1.4" .8" 1.6"

9mm Barrel Liner: 9" Luger (0.348/0.356 10" twist)

Projectile .360 RB .360 RB .360 RB .375 RB .375 RB .375 RB .375 RB AG Slug E-J Dome
Weight 72 72 72 80.5 80.5 80.5 80.5 100 77.8
Charge 20gr 15gr 10gr 20gr 15gr 10gr 7gr 15gr 15gr
MV 940 780 550 1030 860 733 540 xxx xxx
fpe 142 98 49 191 133 97 52 xx xx
Accuracy 1.5" xx xx 2.6" 2.8" 2.0" 2.5" 2.7" 2.0"

Not all of the data has been collected yet; I will be filling in the missing measurements as range time and conditions permit, but what was collected tells me a good bit.

There were several surprises here. Before the testing, I would have sworn that the self-swaged balls would be the most accurate, since they would be squeezed down to the exact bore size. I also assumed they would have the highest muzzle velocity because the pressure in the chamber would have to build up to a high level to expel the ball - just as in a screw-barrel pistol. I was partially right; the self-swaging rounds generated more muzzle velocity, but the accuracy was generally worse. In the round balls swaged through a die, tiny differences in diameter made quite a difference. When the .25 round balls were swaged, the 'base' had a slight flange - just like a tiny Minie ball, so I shot them in the reversed position, thinking that as the balls left the muzzle and were slammed by the powder gasses, the round base (in the reversed position) would be perturbed less and therefore more accurate - was I wrong... I was also amazed how well the short .25 caliber barrel performed; although the projectiles did not have tremendous energy, they were quite accurate. In the smaller calibers, the air gun slugs were more accurate than the swaged bullets, perhaps because they were relatively slow.

Get A Grip!!

During the course of the accuracy testing, I began to wonder just how changes in the way the pistol was gripped affected the grouping. The 'Colt 1911' style grips used are not very conducive to getting a consistent grip, and it was possible that these were affecting the accuracy as much as anything else. The .45 shooters have a saying "The tighter the grip, the tighter the group". On the other hand, spring-powered airgun shooters, where the grip is all-important in keeping accuracy, use an extremely loose hold - keeping the gun as little restrained as possible. So - why do we have these two very diverse schools of thought?

There is a persistent myth that "The bullet leaves the barrel before the gun has a chance to move". This is not just wrong, it is a physical impossibility! Issac Newton showed that "Every action has an equal and opposite reaction", which for us shooters means that a bullet going out means the gun kicks back. BUT - the 'kick' begins as soon as the ball is accellerated by the burning powder charge, and the backward velocity of the gun is determined by the formula M*V = M*V as the bullet moves down the barrel. This law is called the 'conservation of momentum' and is related to, but not the same, as the muzzle energy calculation. Therefore, if we are shooting our 80 grain ball at a muzzle velocity of 1050 fps, we have a 'momentum product' of 84,000. This pistol weighs 3.16 pounds, or about 22,155 grains, so dividing that into the 84,000, we end up with a backward velocity of 3.8 fps - and this is how fast the gun is moving exactly as the bullet leaves the muzzle. Not very much (thank goodness), but we have a few more complications. If the pistol were hanging freely by a string, it would not go straight back because the center of mass is not in line with the barrel but somewhat under it. Therefore, the barrel will flip upward as the gun goes back, and the speed of this 'flip' depends upon the backward velocity and just how far from the barrel center-line the center of mass is. For the sake of argument, let us assume the barrel tip has a 1 ft/sec upward velocity due to recoil. Again, this is not very much, but a round ball covers the first 50 yards in 0.159 seconds, so this means that the upward velocity of the bullet given by the barrel's movement will bring the impact point almost two inches above what it would be in the absence of recoil. As long as this recoil velocity is consistent, the change in impact point can be compensated for and not affect the accuracy. Unfortunately, things get a bit more complicated...

When you grip the pistol, you become a part of its mass, and since a good portion of the grip is below the pistol's center of mass, the muzzle flip-up can be increased. So - it is evident that a consistent grip is very important to getting good accuracy, but just how much in this instance? To try to get a feel for this, two groups were shot using the .375 ball with 20 grains of powder and with the .224 slug using four grains of powder. The groups were split into a 'free grip', where the pistol merely rested on the support, held in place with just my finger tips, and a 'death grip' group, where it was held very tightly, with the barrel pressed hard into the sand bag rest. The results are shown below, with offset at 40 yards in inches relative to the aiming point:

Grip and Load Loose grip - .375 ball Tight grip - .375 ball Loose grip - .224 slug Tight grip - .224 slug
Horiz.Offset -1.7 -.4 2.0 -1.5
Vert.Offset -2 -.8 -8.8 -6.4
Group size 1.6 1.5 5.1 1.7

Although the average accuracy was about the same, the .375 tight group had much more vertical than horizontal variation, and the loose grip was exactly opposite. The .224 tight grip group was very tight, with one 'flyer', and the .224 loose grip group was just a mess... If you look at the offset differences between the .224 and .375 and then at the muzzle velocities, you can see that it isn't velocity that caused the offset but the greater recoil and muzzle flip from the larger caliber. Also, because I was gripping the pistol below the center of mass, the tight grip moved the center of the group up a bit compared to a loose grip.

Summary and Musings

Right side of pistol Left side of pistol
Overall, the project was quite a success; I was able to get good accuracy from all of the barrels by finding the right combination of projectile and charge, and I learned a good bit about accuracy testing in the process. I felt that being able to place shots in a one inch group from a pistol at 40 yards was quite an accomplishment, especially considering that I can barely hold a circle of that size, even with the scope and rest.

For those interested in more construction details, there is a separate web page here.

Click on the image to see a video about this pistol. Muzzle flash

Disclaimer:

The material presented here describes the construction of a device which has the potential to cause harm through its improper construction, maintenance or use. The author presents this material solely as a record of his personal experiments and neither encourages nor discourages like experiments by the public. The information presented here is believed to correct, but the author makes no representation whether an identical or similar construct would or would not cause harm to the user. Duplicating this or similar constructs should be only done in accordance to national, state and local laws and further should not be undertaken by persons inexperienced in such work.

Waiver of rights:

The material presented here may be used by any person or organization for either personal or commercial use without any payment to the author of any kind. No patents have been applied for, nor does the author intend any commercialization of this work. If some party wishes to commercialize any or all of this material, they have my blessings. Anyone wishing to contact me may do so through my email.