how does a steam locomotive increase speed? - Model Railroader Magazine
A reciprocating steam locomotive is a 'pressure engine' as far as its producing power is concerned. The thing producing that pressure is heated steam, and while we discuss 'thermodynamics' in terms of heat, all the heat in the world won't propel a locomotive if it is not correctly applied to make pressure the engine can use.
If anyone has watched the Walschaerts discussion from the C&O group (in another current thread) they will appreciate how piston thrust is not evenly applied to the driver rim proportional to rotation, and it will not be far from there to understand why there are torque peaks (four times per revolution) with fixed or short cutoff at high boiler pressure and low speed. Near starting, you easily trade throttled lower peak pressure and larger mass flow achieved through longer admission for spiky torque, but (as with Tuplin) you will have increasing mass flow to go with your lower pressure to make the same effective average torque, and at some rotational speed you will start running out of steam-generation capacity to supply that mass flow. You could, at that point, open the throttle full, but all that does is remove any pressure restriction between the boiler and the valves, and if the cutoff is set 'late' you are filling a larger and larger volume of cylinder with full-pressure steam, which will still have considerable pressure as the exhaust valves start to crack open. A great deal of the potential work, and heat content, of that steam then gets more or less wasted as it starts to thunder into the exhaust tract... and chokes it, still at relatively high pressure, so there are back-pressure and compression issues out the wazoo.
For all intents and purposes, on a well-designed engine the pressure ANYWHERE won't exceed what the pops are set for. (In fact on large engines there are multiple pops set 2psi apart to provide larger relief capacity for very large boilers that are abruptly throttled while on solid-fuel higher firing rate... but that is a story for a different question.) We can arrange compression (of the residual steam in the cylinder after the exhaust valve closes on a stroke) to be higher, and in some cases dramatically higher, than nominal MEP, but it can waste momentum and cause wear, tear, tribological and sealing issues as the piston goes through the subsequent dead center -- it has a use other than 'cushioning' the rod mass through center; it assures that the cylinder pressure and saturation temperature is close to admission pressure as the valves come open, so there are no flow problems in the 'dead space' that would decrement performance at high cyclic. This can be a bit technically involved to design, but very simple to operate in practice.
In order for speed in the example to increase from 20 to 30 mph:
1) The engine will have to have enough 'steam' to accelerate itself and its train against the running resistance, including that approximated by the Davis formula. That is done by increasing the thrust per stroke, which in turn is accomplished by admitting more steam THROUGH THE VALVES.
2) To run steady at the new speed, the mass flow of steam has to produce drawbar pull that 'just balances' train resistance at the new speed. That will be less than what was required to accelerate it that last mph from 29 to 30, and accordingly you can MOVE THE REVERSE slightly so the engine settles at the new speed economically.
3) To get the necessary mass flow, your boiler has to be capable of producing it. If you have even remotely competent firing and reasonable engine maintenance, the boiler will have been reasonably near popping off the whole time -- which may or may not imply that the fireman had to fire more intensively, or 'trade water for steam', or conversely use the injector a bit more aggressively to keep the engine from popping (and wasting steam mass and its heat content) when the demand for acceleration is relaxed. But that is done entirely relative to gauge pressure and level... no one but an idiot thinks you can pressurize the boiler to "104%" of its safety-valve capacity, like the engine rating on the Space Shuttle, for a little emergency wartime power or whatever. This is part of the reason why it Does Not Matter If The Throttle Is Fully Opened ASAP. On any modern engine with long-lap, long-travel valves ALL the effective admission of steam, and hence consumption of steam, is regulated via the valve gear cutoff once you're above relatively trivial speed.
There are places of course where a limited throttle opening is highly desirable -- one such just came up in a discussion on RyPN involving 614T, a couple of days ago. If the engine is drifting, or producing relatively limited power keeping a train stretched on a downgrade, you'll find benefit in opening the throttle only a relatively slight amount (some engines in fact have a designated 'drifting throttle' or setting that maintains a limited pressure in the steam chests to exclude air and keep cylinder oil warm). We have had a number of discussions over in the Trains forums about 'snifting valves' over the years, and while those aren't relevant directly to this discussion, they will aid understanding of precisely how these locomotives make power and run smoothly.
