Ok, after reading up on Engineering in the Ancient World, by JG
Landels, I'm convinced that Heron of Alexandria (or someone shortly
after him) could have invented a workable steam engine during the Roman
Empire.
If I may, I'll divide the engine into 4 parts, a boiler, the pistons, a
condensor, and an output device.
1) The boiler's easy, just scale up the boiler from the aeolipile or
any of his other devices.
2) The pistons aren't quite as easy, but there were plenty of examples,
including some in Heron's own Pneumatica (the fire engine and syringe,
for example). An impressively complex example, the Valverde Huelva
Pump, was found in Spain. Four pumps were found in a shipwreck dating
from the 1st century AD. Tests suggest a 95% capacity. Also
noteworthy is an 8 cylinder force pump found in Brittany.
Heron evidently had much experience with such pumps, mainly those in
use for fire engines. Vitruvius seems to describe pumps made through a
'lapping' technique, whereby the piston is first tooled in a lathe,
then rammed down the cylinder until it fits right. Vitruvius also
comments on air pumps (for organs), suggesting wrapping the piston "in
sheep's hide with the woold still on it."
Clearly, the Romans and Greeks had an extensive knowledge of pumps and
pressures and how to get them to work, even for gases. If an air pump
could be built, surely a steam pump could be as well. The biggest
challenge would be would be creating a single action piston (or, if
were were to be ambitious, a double action piston). Its a jump,
admittedly, but its not too far a jump.
3) The condensor is probably the biggest stretch of them. Simply
because, as far as I currently know, there's nothing contemporary like
it. Even then, you don't absolutely need it, if you're near a source
of fresh water. But, its handy, and pretty simple. Just run the
exhaust steam pipes through a tank of water. Pretty simple.
4) The output device will be, for this example, a paddlewheel for a
steamship, as ship propulsion was one area that was simultaneously:
labor intensive
done by free men
valuable to the military
Now, we have to convert the force of a piston to a wheel. Seems simple
to us, but would it be simple for Heron and his contemporaries?
Suprisingly, it probably would be fairly simple. If we look at Heron's
wind powered organ, the setup is basicly a windmill (the only example
of such a device in antiquity, in itself a good starting point for AH)
used to force a piston up and then release it down (radial rods on the
mill, which would catch a lever attactched to the piston, pull the
lever down, forcing the piston up, before slipping past the lever,
allowing it to rise and the piston to fall) pumping air into the organ.
Basicly the problem now is to:
reverse the design (of course replacing the wind mill with a
paddlewheel)
develop a better method of converting the work besides that "catch and
release" system.
Reversing the design is pretty simple. In fact, that is quite possibly
exactly how overshot waterwheels (known in that time) were likely
developed. Bucket wheels, where a wheel was turned in water in order
to scoop the water up and raise it to a height. In other words, apply
power to a wheel and get water to the top. It is likely that some
bright guy simply decided to put water at the top, and get power out of
it. Simply reverse the process. Bada bing, literal reverse
engineering. So, its not hard to imagine that someone looking at a
wheel powering a pump, could think of a pump powering a wheel.
Oh, and the Romans knew of paddlewheel boats. Somebody even had an
idea of an ox powered paddlewheel boat. It likely wouldn't have
worked, but the fact remains that the general idea was there.
Now, as for replacing the trip mechanism, its a small hop, but I think
its doable.
So, there you go. Put it on a galley, and you've got a Heronic
steamship.