Steam cars are a thing, and they don't use wheels and cranks like a steam locomotive (and like what we're aiming for). Why not? What problem was encountered that we are going to run into by trying to build a road vehicle with a locomotive-style drive train?
The power band for reciprocating steam engines is limited, depending on your source, to 0-500 or 0-1000 rpm
Typical steam locomotives achieved higher top speeds using larger drivers, but 48"+ wheels on a car are a no-go unless you're doing something silly like our trike
The same size cylinder on a larger wheel produces less torque. One might think this would be an unacceptable tradeoff for the locomotive because acceleration would suffer, but 1) acceleration on trains is much lower than a road vehicle, and 2) the limiting factor on locomotive torque isn't the cylinder, it's the adhesion of the steel wheel to the rail. Thus, increasing driver diameter does not incur the same tradeoffs as it would on a road vehicle.
Because the acceleration tradeoff is unacceptable in a road vehicle, transmissions with multiple gear ratios were used to allow good acceleration but high top speeds
To get a high top speed on the trike, we will need a large wheel or a clever transmission that is acceptably locomotive-like and crank-based. To still have reasonable acceleration, the trike's steam engine will be significantly oversized (much larger cylinders than necessary given the power output). Unnecessarily large cylinders would be a deal breaker in a mass produced vehicle, but in our case, they're a plus!
The age of steam locomotives predated modern plastics and other materials used in modern bearings and seals. These traditional mechanisms required frequent maintenance and lubrication, and although many choose to model these mechanisms faithfully to the prototype, we can substitute modern equivalents to reduce the workload needed to operate the trike:
It is important to realize that a single wheel driven by a crank looks much different than the typical locomotive with four or more drivers, as the latter have connecting rods. The horizontal connecting rods moving in their circles make up a substantial amount of the moving parts that make locomotives fun to watch. There are two possible directions here: 1) Use an internal valve gear like Stephenson, leaving just the crank on the outside and maybe a link or two for the valve slide; 2) Use an external valve gear like Walschaerts, and have the eccentric crank, eccentric rod, and the remaining parts right in there with the main crank rod, possibly obscuring that basic motion in chaos.
The loco equivalent of the single driven wheel in our reverse trike is the British Stirling Single. There is just the cross-head and the crank; no connecting rods. It's a clean look, but almost too simple.
A Stephenson or other internal valve gear would be mostly hidden or even completely hidden like on the Stirling Single. Depending on the frame, it might still be able to provide visual interest separate from the main crank.
An external valve gear such as the Walschaerts or Baker gears puts a whole bunch of linkages front and center, but with the exception of the eccentric crank and rod, they have limited total movement. If these parts are placed far away from the main and eccentric crank, they might add to instead of detract from the main crank action. I have not found an example of a X-2-X locomotive with external valve gear to reference.
In the model locomotive world, cylinders can either be machined from castings or fabricated from machined parts. The size of the cylinders for the steam trike are large enough that machining a casting would be extremely difficult for even a full size bridgeport mill and floor lathe. Thus, we pursue an assembled design. There are a few ways to join the pieces of an assembled design:
Brazing / Welding
It is, in fact, possible to join cast iron pieces using brazing or even welding, but it is considered difficult. These methods are also usually last resorts for repairing parts for which there is no other hope of replacement. Brazing may be doable with a little practice, welding is probably out. The steps are 1) Get the whole part hot, 2) Do the joint, 3) Let the part cool as slowly as possible to prevent it from cracking. Getting a workpiece such as a cylinder made of thick-walled tube all the way to the proper temperature may be beyond a standard oxy-acetylene torch.
Fasteners + Gasket
Certain components are always designed to be affixed with fasteners and a sealing gasket, such as the cylinder end caps. Screws and a gasket can also be used for "permanent" parts. This often makes the part design more complicated to account for fastener access and extra fasteners needed to secure the gasket.
Overview
Fabrication Analysis
Steel boring bars have maximum overhang ratio of 4D. So an 8" deep bore would require a 2" diameter bar, which would require an initial trepanned hole of more than 2" to account for the tool. So the pictured 8" cylinder length made out of stock with a 2" initial hole is unrealistic, but other combinations of stock and basic cylinder dimensions may work.
Moving the center of the bore on successive passes would be easiest on a mill, using a vertical boring bar for the cut and X or Y travel to control the offset.
Typical mill boring bar heads are much larger than lathe boring tools yet have a smaller boring bar diameter capacity; only a very short cylinder length could be accommodated
The mill's quill travel, knee travel, or combination of both needs to support the total cylinder length
Boring on a lathe requires a way to accurately increase the offset of the bore while holding the workpiece. A four-jaw chuck won't work to offset the cylinder block; it would need to be clamped using T-blocks. We would then still need a way to ensure we move the assembly in the correct direction every time we increment the offset by hand.
The valve chest would need to be fabricated by connecting the four walls together in such a way that they can still be secured to the cylinder body. The other alternative involves machining out the center of a very thick (and expensive) plate, preferably by drilling holes around the interior perimeter.
Other than the steam chest, assembly simply bolts together with gaskets
Centrifugally cast iron "dry" sleeves or liners are used to repair engine blocks by allowing a cracked cylinder to be repaired. They are also used in aluminum engine blocks as the actual bearing surface. For ICE use, "universal" liners (not made for a specific engine) need to be bored and honed after installation, but it may be possible to get by with just honing for a steam cylinder.
Overview
Liner used for main cylinder
Liner mated with separately machined end plates, end caps, valve slide plate, and any other fabricated part to complete the assembly
Fabrication Analysis
Brazing likely best bet for assembly as the thickest liners are only 1/8" thick and this may be too thin to weld without distorting the liner
Fairly intricate braze operations
Join cylinder port channels and exhaust port channel (not shown in picture) to valve plate
Join valve plate + channels assembly to end plates while ensuring main cylinder liner remains aligned
Join main cylinder liner to end plates
Valve chest construction as in integrated valve plate versions above
Unlike slide valves, cylinder valves allow for a configuration where the valve ports are a straight shot down from the valve to the ends of the cylinder; this is why the valve cylinder is longer than the main cylinder on some engines. Cylinder valves do not need additional space between the valve and the main cylinder for the exhaust port. As an added bonus, with inside admission (the typical setup for cylinder valves), the rod seal/gland at the end of the valve cylinder is only exposed to exhaust pressures.
Overview
Liners used for main cylinder and valve cylinder
Main and valve cylinders supported by brackets that contain the flowpath for steam between the valve and cylinder and also secure the assembly to the vehicle
Unlike the slide valve with its exhaust port, the ends of the cylinder (exhaust) are separate from each other and must be connected together outside the assembly on the way to the exhaust stack
Basic design can scale from smaller to larger engines
Fabrication Analysis
Design depends on multiple socket-weld brazing or welding operations to secure liners to their respective ends, as well as perimeter brazing/welding to secure valve cylinder to mounting brackets
Cylinder valve inlet/exhaust ports and cylinder end caps made from bar stock
Mounting brackets with integral cylinder ports made from flat plate
Round bar cast iron stock readily available up to 6"
5/8" plate cast iron stock readily available for brackets
Inlet sleeve on cylinder valve allows use of two shorter liners to form valve cylinder
Valve inlet/exhaust port adapters are formed with simple lathe operations to form the socket, followed by slightly more involved mill work to create the inlet/exhaust ports themselves and any threaded fasteners
Brackets formed with squaring then boring operations, followed with threaded face holes for fasteners and operations on edges to create port and engine mount fastener threaded holes
Possible alignment issues with two liners used in valve cylinder; they need to be aligned pretty well for the valve spindle to not jam
The ICE Cylinder Liners as Cylinder and Cylinder Valve appears to be the best choice; at least pending some physical testing of liners and the brazing process. I see how to do all the machining tasks with available equipment.
The piston assembly will be a pair of PTFE cup seals facing away from each other, sandwiched between metal disks. The cup seals are forced against the cylinder walls by the pressure and don't need a backing o-ring or spring.
Piston Assembly project on OnShape
Although there are certainly geared locomotives, part of the mechanical aesthetic we're going for is directly driving the rotating wheel via a large crank. As strokes per minute or linear piston speed is the limiting factor for the reciprocating part of the engine, the final road speed for the trike is then limited by the wheel diameter. As illustrated in Calculations, locomotives designed for express passenger service could have drivers up to 6 or 7 feet in diameter, far larger than any highway vehicle wheel.
Planetary Chain Drive
If we can't get a reasonable wheel size out of the normal calculations, there is a sort of planetary or concentric chain drive that can provide a gear ratio while still allowing for directly driven crank motion, shown here.
Large sprocket is fixed to crank arm and does not rotate independently
Small sprocket is on main axle and rotates freely with respect to crank arm
As main crank rod from engine rotates crank arm around, large sprocket drives the smaller sprocket at an increased speed
This sort of chain drive is not covered in any chain design manuals I have seen, as the smaller sprocket is the driven sprocket and one of the sprockets is "fixed".
Concentric Transmission project on OnShape