The air cane was a must-have accessory for the British Victorian gentleman. As a single shot muzzle-loader it was not the best tool for self defense, yet these items were popular as mechanically interesting novelties that were able to fire a projectile without smoke or flame. This particular example was acquired in a relatively poor state. None of the accessories were present such as the pump to pressurize the chamber and key to cock the mechanism. The air chamber itself (the upper half of the cane) was corroded on the inside and deformed, and therefore deemed unsafe for use. In order to replace it, an adapter was machined to allow a 9oz CO2 bottle to be used instead.
Here is the cane assembled along with the adapter and bottle:
The separation between chamber and barrel is roughly at the center and is marked by the middle part of the painted "bamboo" finish which is rather well done.
Here's what's inside the barrel section, there is a lock mechanism connected to an outer tube into which a rifled brass liner is inserted. The rifling appears to be pressed rather than cut, and an odd choice to even rifle it in the first place given that these were mostly used with round balls and there wasn't really a way to aim it with any precision.
Close-up of the lock mechanism and adapter. I could not find any maker's marks anywhere, however the design is typical of the air canes made and sold by James Townsend of Birmingham est. 1845.
Here is a diagram roughly to scale of the lock, the crank is rotated clockwise to cock the mechanism until it is caught by the trigger.
When the trigger is pressed, the crank is released and is rotated quickly anti-clockwise by the main spring, striking the hammer which transfers the movement to the firing pin that opens the valve.
Detail of the internals of the valve, of note is the original animal horn seal that just needed a little polishing and sealed perfectly.
The adapter was machined from C12L14 steel.
Valve assembly put together and the other end of the adapter with a male 5/8-18 UNF thread to fit the paintball bottle
Detail of the connecting faces of barrel and chamber assembly. Note that there is some damage to the firing pin housing that would allow some air to escape while firing and probably interferes with the proper striking of the valve, however I made no attempt to repair it.
Filling adapter, cocking key and some Hornady 32 cal (0.310" diameter) lead ammunition
Initial testing video:
Above 850 psi or so the valve simply wouldn't open, below that it was extremely sensitive to how much the barrel assembly is screwed into the valve body.
If it's tightened as far as it can go (not too much or it opens the valve), then it will give a short sharp burst of air and the ball comes out with very little velocity, and won't even pierce the bottom of a soup can.
If on the other hand a 1mm gap or so is left between the two faces, the valve is blown wide open and it virtually empties the reservoir.
You can see this happen in the slow motion video where the first can is picked up by the airflow and blasted downrange.
The problem turned out to be that the valve stem is actually proud of the base of the thread, this was the reason it leaked if the barrel assembly was fully tightened. I figured that after more than a century the seal had deformed allowing it to come to rest further than it was supposed to.
I made a new valve stem with a Delrin seal to rectify the issue, here it is next to the old one:
Installed in the valve body you can see that it's slightly recessed at rest:
I did some testing and what a difference, completely different animal!
Pumped it to 1000 psi just to see what it would do and while the valve opened, it took a few shots before it even pushed the ball out of the barrel.
When I eventually got it to fire I ran some shots over the chrono and it was laughable:
76 fps - 0.6 ft lbs
130 fps - 1.7 ft lbs
156 fps - 2.4 ft lbs
Took it down to 600 psi and the results were very different:
634 fps - 39.8 ft lbs
623 fps - 39.4 ft lbs
616 fps - 37.6 ft lbs
602 fps - 35.9 ft lbs
Those are some impressive figures for a device of this vintage, but I thought I could do better so I decided to make a "balanced valve" - continued below