Hill pistol recreated from patent drawing

Road testing the Hill pistol 

With thanks to John G. 

He wrote on the airgunbbs.com:

The reproduction of the 1933 A.H. Hill pistol described in my previous post ( http://www.airgunbbs.com/showthread....tion-challenge )
has now been put through its paces, and here is a summary of my findings.

The cocking and loading sequence involved with this unusual two-stage cocking design can be explained by the following picture sequence. 
Pics (1) and (2): The spring loaded button on the grip edge (shown by the red arrow) is depressed, releasing the trigger guard/cocking lever. Pics (3) and (4): The cocking lever is then pulled back until it is vertical to the cylinder, at which point the sear engages the piston with a loud click. An automatic safety simultaneously locks the trigger. At this stage the hand can be released from the cocking lever. Pic (5): The cocking lever is then forced back to its original position by the palm of the hand and is locked back in place by the spring loaded button. Pic (6): The lever on the loading gate is rotated and the gate pulled outwards to access the breech. The pellet is loaded, the gate is slid back and locked in place by the cam lever, and the gun is ready to fire.

To my relief (and surprise) the pistol cocked and fired perfectly first time. Initial testing with .177 Webley Special pellets (7.6 grain) was disappointing and power was obviously low, with an average muzzle velocity of only 150 fps. An improvement was achieved by investigating the breech seal, which was a simple leather ring, as shown in the patent. An “icing sugar test” revealed that, with a pellet in the breech, very significant air loss was occurring from the breech on firing. (This test involves daubing icing sugar over the suspected leak areas and watching for clouds of powder on firing. Icing sugar is non-hazardous and easily wiped off the gun afterwards leaving no traces). Replacing the leather washer with two rubber O-rings showed no air leakage in the icing sugar test, and the muzzle velocity increased to a more respectable 245 fps. 

Interestingly the cylinder dimensions and swept volume of the Hill pistol turned out to be almost the same as for the straight grip Webley Mark 1 pistol from the same era as the patent. The cylinder diameters of the Hill and Webley are 21mm and 20.2mm respectively, and the piston strokes are both 6.5 cm, giving a swept volume in each case of 22.5 cc. As both guns take the same size spring this meant that I could make a direct comparison between the performance of my Hill piston and a good condition 1930’s Webley Mark from my collection. In fact, in all my tests I swapped the same spring between the two pistols. The Webley gave an average muzzle velocity of over 300 fps with the above pellets, so there were obviously still some wrinkles to iron out with the Hill pistol. 

One obvious factor was the weight of the piston. As defined in the Hill patent the piston was necessarily very heavy, and mine weighed in at 106 grams, compared to only 64 grams for the Webley Mark 1. A decision was made then to skeletonise the piston to bring its weight down, and as shown in the next picture a final weight of 85 grams was achieved, still heavier than the Webley but a significant reduction.

The pistol as described by Hill in his patent had no piston washer and relied on just metal-to-metal contact for the piston seal. Not being very confident about this I took the decision to use a simple flat leather washer to improve the seal, as shown in the previous picture. This was presumably not as good as it could be, as the Webley Mark 1 uses the more efficient cup type leather seal. So a further modification was called for and I made a seal of this type from leather, with a central PTFE retaining washer as shown next:

With the lighter piston and improved seal there was an immediate improvement in power output, and using the tight unsized Webley Special pellets I was getting a reasonably consistent average of 310 fps (over 10 consecutive shots the variation was between 322 and 302 fps). The pellet weight was 7.6 grains, equating to a muzzle energy of 1.62 ft lbs. The Webley Mark 1 with the same spring and same swept volume gave very similar results, averaging about 312 fps with a similar spread. 

Accuracy tests were carried out using a bench rest. The following picture shows 10 consecutive shots from the Hill pistol over a distance of 6 yards.

The next picture shows 5 consecutive shots over 11 yards.

So accuracy and consistency seemed to be acceptable for this level of pistol.

Having put a few hundred pellets through the gun I have now formed an opinion about the unique cocking system. 
The cocking method and automatic safety feature have so far had worked well with no problems.
When the gun is cocked by the first outward stroke, the 14cm spring is compressed by only 4.5cm, thanks to the floating muzzle plug which at that point is now protruding from the muzzle by about 2cm. You can see this with these before and after pictures.

As expected, this outward stroke is quite light compared with the Webley Mark 1, which at this stage would have completed its cocking action and its spring would have been compressed by 6.5cm. The trigger guard cocking lever is then pushed home, so forcing the roller unit back into the cylinder and compressing the spring a further 2 cm. The total spring compression is then 6.5cm, the same as for the Webley Mk 1. This return stroke is also easy to achieve, and although the spring is under very high compression at this stage, the curved roller system provides a high mechanical advantage. So in summary, the outward cocking stroke accounts for about 70% of the total spring compression, and the return stroke tops up by the final 30%.

When shooting several consecutive shots, the Hill pistol is noticeably more comfortable to cock than the Webley Mark 1. However, the two steps do make it slower to use and unless one has weak muscles and really appreciates the reduced cocking effort, the time factor would offset the comfort factor for the average adult shooter. Another downside to the Hill pistol is the breech sealing system. Although this is very neat and efficient, when the gate is fully extended, access to the breech is still somewhat restricted, and inserting the pellet can be fiddly, especially if you have large fingers. As you might imagine, loading the Mark 1 is much easier.

All in all, I preferred the Webley system, and I doubt that the Hill pistol could ever have been a serious competitor to the Webley pistols of the time. The advantages of the unique two stage cocking system over the Webley Mark 1 were not that great, and certainly would not have justified the extra manufacturing costs involved. By 1933 the Webley Senior had already been introduced, and Webley had been able to reduce the cocking effort of the Mark 1 by a much more economical means. So the Hill pistol design never came onto the market. Nevertheless it remains an ingenious and interesting part of vintage British air pistol history. It was the hardest project I have tackled, but probably the most rewarding.

The making of the Hill pistol 

With thanks to John G, 

He wrote on the airgunbbs.com:

Some of you have expressed an interest in how I made the 1933 Hill prototype pistol that was recently discussed in a couple of posts, so here is a potted illustrated account. If it sounds too glib, I must admit that I have left out details of all the cussing and sweating and the (too many) machining mistakes made along the way, and these can be taken as read. I am no engineer, just persistent. Everything was done on my old Myford lathe fitted with a vertical slide for the milling operations; tools and materials courtesy of Fleabay.

For me the starting point in all these build projects is to make lifesize copies of the patent drawings. This is easy with scanners and Photoshop, except for the fact that patent drawings never seem to give dimensions. So there is usually some guesswork involved. Fortunately vintage British air pistols seem to consistently use cylinders with an outside diameter of about 1 inch (e.g. all the Webley pistols, Warrior, Lincoln, Acvoke, Certus etc.) so it was reasonable to assume the same for the Hill pistol. The second stage is to work out how to build the principal carcase, which is often a major head-scratcher, as the pistol inventors always depict the carcase as a one piece steel unit. This is OK for someone with access to steel casting facilities, or drop forging, but is not helpful for the amateur, and so improvisation is called for. The carcase plan can be broken down into sections that can made from solid steel by a combination of hacksawing and filing, followed by brazing the pieces together. Precision steel tubing can often be used for the cylinder.

For the Hill pistol, this was how I dissected the carcase, with the intention to braze the various components together at some appropriate stage. 

The first component to tackle was the cylinder housing block (A). As shown in the following picture sequence, this began as a rectangular block of mild steel. 

The block was drilled with two parallel holes, one to accept the cylinder and the other the barrel. The drilled block was then milled to give two steps (pics 1 and 2). The slot for the trigger was then milled (pic 3). Producing the round profile of the cylinder required some thought, and the method finally adopted was to mill a series of closely spaced grooves longitudinally until an approximation to the circular profile was achieved, as in pics (4) and (5). Finally all the ridges were filed down to give a smooth profile (pic 6). 

The next component tackled was the barrel housing block (C). Precision seamless steel tubing was sourced as the cylinder (B), and the block had to be attached to this very strongly, as it also had to carry the cocking lever pivot pin and would be subject to a lot of stress. It was decided to use a combination of brazing and pinning to do this, and the sequence of events to achieve attachment is shown in the next series of pictures. After cutting out the shape of the housing (pic 1) with the appropriate curved base to fit snugly on the cylinder, the base was drilled to accept two pins (pics 2 and 3). Corresponding holes were drilled into the cylinder to accept the pins, and the block was brazed tightly into place (pics 4 and 5).

The grip frame (D) was cut from a slab of steel. The upper edge of the grip frame and the lower edge of the cylinder housing (1) were profiled to give a dado -type joint ready for brazing. 

With the basic carcase now sorted, the next stage to tackle was the particularly difficult one of the trigger guard/cocking lever. This had to be forked at one end, with the fork pivoting snugly against the cylinder and the muzzle barrel housing. It also had to have a precisely curved extension with a central groove to receive the roller wheel on the floating muzzle plug, as well as a second deeper slot to accommodate the cocking link. The approach used is summarised in the next sequence of pictures:

First, a slice had already been taken off the drilled cylinder housing (A), which had deliberately been made longer than necessary for this purpose, and was then cut to be a perfect fit around the cylinder and barrel housing (pic 1). It was drilled through to provide the cocking lever pivot holes as also shown in pic 1. It was then cut into two halves to provide the two jaws of the cocking lever (pic 3). A slot was milled in the end of a square section steel bar, which would form the cocking lever, to provide a track for the roller wheel of the floating muzzle plug. The bar was hot forged to give a curved end (pic 4). The two jaws were drilled for pins (pin 5), and the jaws then pinned and brazed to the end of the cocking lever, as shown in pics 6 and 7. Further heat bending of the of the bar and trimming was carried out to form the trigger guard (pics 8 and 9). 

The floating muzzle plug was turned down from steel round bar and consisted of three sections: a long narrow section which acted as a spring guide, a short wide section that was a sliding fit in the cylinder, and a short narrow section which carried a small roller wheel. This is shown in the following picture:

The piston was turned to size, slotted and a notch added for engaging with the sear. 

This next picture shows the various components loosely fitted together, including the roughed-out trigger and the cocking link, which was made of heat hardened steel. 

The trigger/sear/safety unit was a particular problem as the patent gives no information about it for the pistol. One can infer the general principles from the information that is provided for the rifle version, and the position of the pin holes in the drawing of the exterior of the pistol help define the relative locations of the trigger, sear and safety components. The shape and size of these components is very critical, as all three have to act in perfect unison. This had to be worked out from scratch, and involved a lot off head scratching and trial runs. Eventually a satisfactory conclusion was reached, and it actually managed to improve on the original by enabling a simple coil spring to be used instead of the special horseshoe spring used by Hill in his rifle, which would have been very difficult to insert into the limited space in the pistol. This shows the shape and positions of the three components, which were made of tool steel and were finally heat hardened:

The breech closure was next considered, and was based on the drawings shown in Hill’s second patent. This consists of a sliding gate, which is locked in place by a cam lever.

The barrel had been turned down from an old air rifle barrel and his was reversibly locked in place above the cylinder by grub screws let into the barrel housings at the breech and muzzle. 

When all the major components has been finished, various smaller tasks were undertaken, such as drilling the air transfer port, shaping the front and rear sights, drilling and threading holes in the grip frame for the grip plates, making the grip plate screws, making the spring-loaded catch in the grip, making the cocking lever pivot screw, and finally making the walnut grip plates themselves.

At this point it was necessary to think about any lettering I wanted to apply to the body. This needed to state the patent origin of the gun, and also needed to make it clear that the gun was a modern reproduction and not an original prototype. Having decided on what form this would take, the lettering was applied by electro-etching.

I then had to decide on the final finish to be applied to the gun. I could have gone for rust bluing to give a vintage look, or hot bluing for a deep black modern finish. I was also tempted to try nickel plating, for no better reason than I had never nickel plated a whole gun before and it seemed like an interesting challenge. In the end I opted for plating, as I also thought a plated gun would make a nice contrast with all the other guns in my vintage British air pistols display. I used what is called an “electroless nickel plating” technique as it gives a very strongly attached nickel layer, and uniform thickness is easily achieved on awkwardly shaped objects like this, unlike normal electroplating. With the grips varnished and everything put back together the final pistol, despite its rather ungainly chunky shape, looked quite smart.