Trains.com

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

throttle vs reverser/cutoff?

9746 views
85 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Tuesday, February 9, 2021 12:44 PM

gregc
no, steady state

But it has to fluctuate, by definition, for your effect to be real at all, and it's a reasonable example of why you might want to indicate the chest and passage pressures as well as cylinder pressure under some conditions.

If the steam chest pressure 'recovers' during the cutoff time, either it produces the same 'loss' in admission as you'd see from the tract, or there is some restriction in steam flow sufficient to starve admission MEP.  The second HAS to be the case in your example, and moreover the 'reservoir' effect in the chest volume needs to be large enough that what enters during cutoff exceeds the mass flow from the chest down the tract during admission.  Outside that envelope there is obviously falling horsepower on the low hand, and lower pressure loss over admission on the other.

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Tuesday, February 9, 2021 12:10 PM

Lastspikemike
It occurred to me that the pressure change in the steam chest due to opening the valve to the cylinder would be proportional to the volume added by opening the cylinder to the steam chest.  It looks to me like that would be quite a small effect.

It is small in terms of a number like lb/hr overall mass flow, but it is important as you get to higher speed and shorter cutoff where the physical flow into the cylinder over the course of admission becomes more and more physics and time affected.  One thing that became noted was that maintaining a reservoir of relatively high mass adjacent to a large (and quick-opening, cf. circumferential port opening in long-travel valves, the use of Trick or Willoteaux valves, the better sort of quick-lift poppets) and ideally low-volume but well-flow-streamlined passage that directs the flow as well across the face of the piston as possible, was a functional advantage in a critical part of early admission.  This is a thing that proper Carter-style reversible compression control simplifies (by having high mass at nominal pressure equalized across the piston face at the start of admission).

The ratio of exposed cylinder volume to total steam volume back into the boiler (throttle/ regulator unchanged in opening amount) surely must be small.

It is not so much that it is small as that it can be functionally flow-restricted, a lesson that repeatedly bit the Franklin people in practice.  Very little flow resistance is necessary to hold up the small early mass flow for good cylinder filling in the rest of admission (then, plus expansion in cutoff)

More so if the cutoff restricts that effective volume.

The principal evil is not so much 'volumetric' I think as the smaller effective port opening as the cutoff is shortened -- this is another thing that the long-travel valves can help with -- and the other impediments to 'mass flow' that trade differential pressure for velocity (which is of little good if it results in lower density in the charge steam).

Note that the turbine has a clear advartage here in addition to continuous flow: any pressure-to-velocity can be captured in the initial impulse stage, then expansion through the subsequent reaction stages uses the expansion energy.

Once the cut off closes the steam entry the volume increases by movement of the piston, heat is converted to kinetic energy.

But what's of more interest here is the things that happen on the other side of the cutoff valve, since that is now mechanically divorced from the cylinder expansion.

In a good long-lap long-travel setup you'll get throttling of the admission flow as the steam edge approaches cutoff, but with a good valve this can again be quick -- something to note here is what happens to the flow from the boiler through the throttle restriction gregc posits through the tract and steam chest to the port area when the valve shrouds and then closed.  I had to argue with people discussing 'momentum' effects in the steam but there is clearly a mass in motion that has to be 'decelerated' -- it just does it with gas kinetics and not elastic/inelastic impact.  I will not be surprised to see reflected shock in various places, perhaps even flow reversal, and this may become more than usually important at high cyclic if any of that 'action' persists at the next admission.  Now, in this context... 

the kettle upstream of the valve maintains the pressure in the steam chest at the previous value.

Except that a better way to put it might be tries to maintain the pressure -- gregc has intentionally made the throttle restriction great enough to retard flow in admission, or there wouldn't be a question, but low enough that a net amount of pressure recovery, and corresponding density close to the valve, will be available at the moment of admission, but not supporting the mass flow which that pressure would produce if it were actually 'kettle pressure' at full kettle generating mass flow.

Put another way, since the volume of the steam from boiler to valve remains the same throughout...

... but it doesn't, in this constrained situation...

...to gain pressure would require a temperature rise throughout the volume.

I do not think this follows: the pressure rise past the presumed throttling restriction is a flow equilibration, and if the tract is heated and reasonably well insulated, as it would likely be in sustained running,  the superheated steam should heat very little as it is pushed into the now-closed volume of the steam chest.

The superheater does not increase system pressure so it is hard to see how closing the cylinder valve could.

Here, because it is no longer freely linked to the boiler; it is on the other side of a throttled opening.

Opening the throttle more would not change the pressure.

But by reducing the restriction, it allows 'quicker' as well as more pressure rise at the chest volume both during admission and in rise to equilibrium pressure after admission has ended.

Closing the throttle some should not either.

That's common sense, because 'even more' throttling restriction over what gregc had already will further cause starvation pressure drop over admission and (probably) limited rise in 'pressure' (here, mass ready to flow at next admission exerting that higher pressure).  This is the area where the question may give a counterintuitive answer to observed practice...

The selected throttle opening would likely be fully open  most of the time.

In part because that 'removes all doubt' that the best possible mass flow down the tract will be occurring at any point over a cycle.

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Tuesday, February 9, 2021 9:53 AM

Lastspikemike
Does the superheater act as a kind of buffer rising and falling in temperature in opposite phase to any pressure effects?

Not to my knowledge.  Remember that it is open at the 'back end' via the dry pipe to the ebullient zone (usually in the convection section) under the dome, so any tendency for pressure rise in the elements is largely balanced (it turns into lower density instead, another Tuplin plaint).

To the extent that the steam loses density as it gains temperature, it might be considered a 'buffer' but I don't think there are analogic benefits from considering it as such.  

I'm running around this morning but I'll give a little more thought to the idea when I can.

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Tuesday, February 9, 2021 8:06 AM

how close to boiler pressure is steam chest pressure or MEP cylinder pressure at max speed on level ground?

can this be quantified if we know the rolling resistance at that speed (see Armstrong's charts)?

Overmod
gregc is wondering if the transient rise in steam-chest pressure

no, steady state

1496

greg - Philadelphia & Reading / Reading

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Tuesday, February 9, 2021 6:46 AM

Since perhaps the whining English schoolboy claims to be resolved on creeping like snail reluctantly to knowledge, let's revisit those original questions:

gregc
does a reduction in cutoff without any change to the throttle result in an increase in steam chest pressure (since less steam is being drawn into the cylinders)?
As discussed, this is complex, probably has an answer that is time-variant (and cyclic-speed variant) unless one develops an analogue of MEP for steam-chest pressure, and already contains one semantic whopper in steam being 'drawn' into the cylinder -- although this is really 'critical' only in understanding how the physics of flow will work.  [The only time steam can be said to be 'drawn' into a cylinder is in the context of certain types of drafting/snifting, where 'displacement steam' is admitted just at atmospheric pressure to keep oxygen out of the steam tract and its vulnerable tribology, and even there a drifting throttle customarily provides a slightly higher pressure to ensure no air 'leakage'...]

Steam is always 'pushed' into the cylinder (more saltily 'mass flow is achieved through pressure differential) and it should be clear that EP, as an index of extra table work via piston thrust, will be lower than chest pressure if mass flow during admission is to take place.  Now EP at the piston face is partially determined by flow through the passages, so it may not necessarily be 'higher' than chest pressure, but any (necessary!) tendency to flow reversal at any such time would be a Bad Thing for subsequent expansive work.  So in all probability a situation in which chest pressure cycles higher than passage EP would be a poor thing to be in initially.

If the cutoff is shortened (we have amply seen by now the point I keep trying to make about using clearer terms for cutoff) with the throttle wide, the chest pressure will now peak a bit higher, closer to boiler pressure, since there is less mass flow into the cylinder and (assuming a radial gear) its interruption from admission to admission will be longer.  I would expect this to be minimal in an actual locomotive design, but for looking at the OP's question it will bear consideration.

If the throttle is not fully open, the steam-chest pressure will tend to fall, as tract pressures back to the throttle will fall, during admission, but then start to rise again during the subsequent period of cutoff as pressure tries to equilibrate in the tract.  The issue is whether there is adequate 'mass flow' into the chest volume during that period, and then adequate port and passage capacity to admit it, to give better net mass flow over admission to provide higher MEP (and hence higher work) over the ensuing stroke.

If there is 'inadequate steam' even at the reduced cutoff, the steam-chest pressure will cycle, but with decaying MEP in proportion to the decreasing tract mass flow, probably with a different point in the stroke where each successive MEP is achieved but now with some effective 'throttling' even during admission (as the swept volume increases but there is inadequate mass flow to hold EP up).

We would expect the effect to be roughly the same at x diameters down the tract from the throttle whether the lack of mass flow comes from inadequate boiler capacity, lower firing or firing efficiency, improper use of injection, or excessive throttle closure or wiredrawing effects at the throttle.

I still am unsure why the OP is concerned with the chest pressure, other than as a concern for understanding complex physics, as (including for the reasons I provided) it is incidental to how the locomotive would be practically driven.  I would agree that it can be helpful to see indicated pressure there during testing, but there were relatively few attempts to provide a desirable variable valve and chest capacity historically (and some of those attempts were decidedly unsuccessful) and in normal design the valve and chest proportions are fixed in service.  Even Wardale's proposal to use double piston valves to achieve different exhaust timing and duration from inlet did not change the valve proportions dynamically or alter the chest volume.

... can reducing cutoff w/o a throttle change result in an increase in power (i.e. tractive effort) due to higher steam chest pressure?

The simple answer is 'no', and it shows the benefit of using mass flow rather than indicated pressure as criterion.  What happens in the steam chest is completely incidental to power being developed in the cylinders; gregc is wondering if the transient rise in steam-chest pressure (during the period of cutoff) will produce higher MEP and hence more extractable work at the now shorter cutoff.  My suspicion is no, that the rise in pressure will not result in 'enough' additional mass flow to make the developed EP higher over the (unchanged) piston stroke expansion, but in any case the situation would be highly artificial.  Certainly historically there has been advocacy for providing a 'reservoir' of high-pressure and well-superheated steam close to the valves for better high-speed operation, but that presumes the throttle has long been wide open, and gregc's question already posits that it is flow (and hence pressure-recovery) restricted.

... and does this mean that if cutoff were reduced should there be a corresponding reduction in the throttle to limit to the increase in steam chest pressure to maintain a desired speed?

No, in almost any case.  It's highly counterintuitive that reducing mass flow into the cylinders could possibly produce higher power; it is only the extreme conditions in the example that might produce it, and of course even slight change in throttle-valve position would likely result in a proportionally greater collapse of peak and average steam-chest pressure, probably within a stroke or two if there is no condensate liquid-to-gas evolution as the pressure falls, resulting in the need either to lengthen cutoff again (for more mass flow) or increase the MEP (by opening, not closing, the throttle as would be instinctive).

The catch here, as I noted, is that most throttles don't work at all like automotive 'accelerators'.  Pre-Wagner dome throttles are particularly awful as, to get reasonable area through them, the force holding them shut even at turn-of-the-century boiler pressure is enormous, and as noted will remain 'evident' over the range that downstream tract pressure 'just the other side' remains higher than boiler pressure.  The mechanical advantage necessary in a grapevine throttle, and some of the evolved reflex involved in having to slam such a throttle open but then bring it back without overshoot to near desired position will convince you that depending on fine adjustment other than at tricky starting is not much fun even if your haptic skills are high.

With a good proportional direction servo (both the Franklin Precision and Throttlemaster designs are capable of the physical position but might have needed better linear sensing precision, for gregc e.g. with LVDTs) you would have a somewhat better chance at the position, but you then have the fun that a multiple is almost certainly not quite linear in flow relative to angular cam position, probably worse and less predictable as the cams wear.  This of course wasn't as much of an issue in the 'get it open to where you can drive on the reverser' era, but if you were experimenting with fine steam-chest cycling effects it would be an additional measuring and control concern.

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Tuesday, February 9, 2021 2:42 AM

looks to me that it's the average pressure (mean) during the cycle.   see Method of Ordinates which is what i calculated in my plot as well as using calculus

1448

 

greg - Philadelphia & Reading / Reading

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Monday, February 8, 2021 3:16 PM

Lastspikemike
A "mean" is halfway by definition.

Yes, but you picked the wrong metric even if you didn't mistake the technical term correctly.  What you meant was halfway down the cylinder, and that's just plain wrong.  "I'm a lawyer, Jim, not a statistician!" ... but for someone who values silently-sniggering semantics as much as you do so often, terms sometimes do matter.

If you could take the necessary pressure readings half would be higher than MEF and half would be lower.

Now you're mixing your units.  The issue is that "MEP" is that pressure that, applied consistently from admission to release (technically a bit different but it will work for this purpose) would produce the same OVERALL work as the varying pressure profile 'in reality'.  That says nothing about where in the stroke the 'average' pressure would be reached, although when you look at the actual events it's not that hard, and as noted a good indicator will give it to you even at high cyclic.

Meanwhile MEF measured someplace logical, for example at the key between the piston rod and the crosshead, is likewise an average over the stroke; we won't yet have converted it into wheelrim torque (which is really, I think, the 'figure of merit' that ought to matter here if the translation to 'force' is going to be made).

MEF isn't calculated that way anyway. It's a deduced number from data that are not in themselves pressure numbers.

Which is why it doesn't matter to indicated measurements on a steam engine (where it is nugatory, 'cuz the pressure measurements actually mirror cylinder performance, which 'pressure' data in a gas engine do much less well). 

As far as I know locomotive engineers have a boiler pressure gauge. There's no pressure reading taken from anywhere else in the steam system.

Yes, that's what you know.  Time now to know more.  There would be little point in having a pressure gauge 'further down' in the tract (except at indicating time, where there are several to many manometers or gauges taking readings at various points there) because any lowering of boiler pressure in service would be simply accommodated by lengthening the cutoff until the engine is observed to perform as expected -- the 'efficiency' function then being rendered by the goodness of the cam profile in a Valve Pilot device, if present.

But there ARE a couple of places where additional pressure gauging is important, and many modern engines have it there.  One of these, paradoxically, is in the steam-chest area, and it measures back pressure and compression, two things we haven't really taken up yet (because they matter after release).  It is valuable to know the latter in particular, because with improper manipulation of valve gear (and it's not difficult to do it wrong!) the compression pressure can peak wildly higher than boiler pressure, with its risetime being related to the mass retained as release ends being heated as the resultant of those last long-travel inches to dead center, and "MEF" no longer being meaningful in determining the piston (counter)thrust change in that distance.

An analogy to a garden hose was used earlier in this thread. I suggest the boiler system is more like a household water system. Municipal line pressure is limited by a pressure reducer at the main service tap.   This would be the throttle. The valves would be the taps.

That should work for this purpose, provided the 'pressure reducer' is an orifice type and not a diaphragm type that allows regulated high flow at a set pressure.  Certainly some of the extended analogies of that 'thought experiment', such as where the instantaneous pressure drops along the tract take place, would follow.

The chief concern I have with it is that the nature of hydraulic flow in a water system doesn't fully mirror that for high-pressure steam.  So you can't extend the analogy all the way to be a full predictor.  There are more interesting analogues in the convection section during certain types of boiler explosion...

If the system is designed correctly there should be in practice an "infinite" supply of steam until the water or fuel runs out.

That is essentially the principle behind the 'big boiler' theory of high-speed locomotive -- it's partially distorted by the premise that cutoff continues to be decreased, and mass flow lessened, as the engine goes faster and faster until 'mere wisps of steam are flicking in and out of the cylinders', which is romantic as hell but not at all representative of what is needed to make the necessary horsepower to accelerate to high speed and then sustain it.

Incidentally it is possible to cheat on draft, using characteristics of the locomotive slipstream to produce any degree of firing and also pull a partial vacuum on even a fairly large mass flow of exhaust steam.  It remains to be seen precisely how effective this might be in practice, but it promises to be no less than interesting... 

A Newcomen engine was an atmospheric vacuum engine.

Precisely why I said what I did.  Look at the context.

Both types didn't rely on steam pressure to do any work. Trevithick was da man.

He was da man in 1802.  Da real man was Perkins - he showed the way.  Some people still haven't followed, all this time later...

Superheaters run at boiler pressure.

And you need to think a bit more carefully than you have so far as to why.  

Keep in mind that there is, of course, volumetric increase in the steam mass flow as it transits the superheater, and this can be dramatic at periods of highly excessive superheat (some of the figures I've gotten from operators who knew how to use their superheat gauges -- yes, there are such things on good modern engines -- definitely qualified as 'crazy high').  I leave the solution to the reader if any reader doesn't understand why this does not produce a "pressure increase" at the throttle.  Astute readers will already understand; hopefully that includes you.

I had understood that there were some locomotives with superheaters that placed the throttle between the boiler and the superheater.

Any locomotive with the throttle in the dome or the dry pipe, by definition, has that throttle between the boiler and the superheater.  Much British practice intentionally puts the 'regulator' there as it simplifies ... issues with recirculation and 'crazy high' levels, superheater dampers, etc.  (I leave explanation of this to the British steam community, as they can do it much better than I can, and I don't see any point in not having the throttle ahead of the superheater under any operative circumstance.)

A number of engines are equipped with two throttles: one in the dome, and the second (usually a multiple type) as a front-end arrangement.  One overt purpose is to provide a solid cutoff of steam when the engine's valve gear provides imperfect isolation at mid -- you may be familiar with 'nightwalking', which occurs where pressure builds up in the tract and something like the combination lever in Walschaerts develops enough valve motion to permit slow but continued creeping as pressure builds up past a leaking throttle poppet.  Another is to allow -- in theory, at least -- access to parts of the steam circuit, including the superheater and front-end throttle, with the engine still at least partially in steam.  Of course with a proper direct-steam system and proper gas blanketing it would be unnecessary... but that's yet another digression.

I have been looking to see if there are online minutes from the National Board ESC meetings that discussed the issue of superheaters as separately-fired pressure vessels lo! these many years ago now.  One of the reasons for the ESC was that there were 'steam safety experts' who swore that the superheater on a Lima Super-Power pattern engine could constitute a separately-fired pressure vessel under certain conditions, just as it can on a locomotive with dome throttle -- they, too, didn't quite know where the throttle actually was on such a design as opposed to where they thought it to be.

  • Member since
    February 2005
  • From: Vancouver Island, BC
  • 23,330 posts
Posted by selector on Monday, February 8, 2021 2:35 PM

Lastspikemike

 A "mean" is halfway by definition.  ...

...

Ummm....no. A mean is a figure derived by adding the values in a data set, and then dividing that total by the number of data points, by count.

You're thinking of the 'median', as Greg has pointed out.  That is the value which has an equal number of data points on either side of it.

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Monday, February 8, 2021 2:33 PM

Lastspikemike
If you could take the necessary pressure readings half would be higher than MEF and half would be lower.

isn't that the median?

greg - Philadelphia & Reading / Reading

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Monday, February 8, 2021 12:34 PM

Lastspikemike
Roughly speaking, cutting the cut off time in half should cut the MEF roughly in half. Throttle position should have no effect.

ideally when steam expands, PV=C, illustrated below for 10, 40 and 80% cutoff.   i calculate the average pressure (MEP) across the cylinder cylce as 33, 77 and 98%.   

at 10%, for example, steam consumption is reduced by a factor of 10, but MEP is only reduced by a factor of 3.

greg - Philadelphia & Reading / Reading

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Monday, February 8, 2021 11:16 AM

Lastspikemike
MEF is the "halfway" pressure on the piston face.

In steam it's MEP (because measured by indicator rather than strain gage; it's highly instructive to make the actual piston-thrust measurement, but this argument doesn't contain that...)

And it is NOT a 'halfwit' pressure; it's the integral of the instantaneous pressure over the length of the stroke from admission to exhaust cutoff.  That's a highly different thing because of the 'double salience' imposed by the rod geometry.  If you wondered how an engine with an indicator diagram like, as Angus Sinclair so wittily noted, a small leg of mutton could show better over-the-road performance than one with a textbook-clean diagram -- here is part of your answer.

In brief, even just with reference to a good indicator diagram (which if you're lucky will actually have an indication of MEP calculation on it) ... look at the rectangular hyperbola from cutoff to exhaust release?  the MEP of concern is the line that produces equal AREAS under that curve to either side of it, and on a good diagram you can almost read that off or extrapolate it to get a working number.  The nice part is that this adds to the portion of MEP you read off the admission, which ought to be nearly linear (see below) to calculate your overall effective pressure...

Reducing the cutoff time will lower the MEF because the steam pressure is now stretched for a proportionally longer portion of the piston stroke.

So far, you are correct, but unfortunately your observation is self-evidently correct.  That you haven't looked at it well enough yet is clear when you then say

Roughly speaking, cutting the cutoff time in half should cut the MEF roughly in half.

Don't be ridiculous.  You haven't even remembered that the period of admission up to cutoff is not only finite, but rather long; MEP during this entire part of the stroke is not only 'reasonably' constant, but a high percentage of throttle pressure.

The MEP during expansion is instructive to study, and as Professor Loeb said to my father in medical terror school, I recommend it to your attention.

Throttle position should have no effect.

The only 'effect' it has is via the absolute mass flow that passes the throttle obstruction during the admission time.  As noted, if that shows evidence of wire-drawing effects reducing pressure for (unusable) velocity, the amount of "pressure" in the steam at the point of cutoff will be proportionally reduced, and hence the subsequent expansion will be over a lower range of pressure (again using pressure as the metric of convenience, with entropy its companion for that reason)

I looked a little at the superheater effect and much to my surprise the steam is superheated in expansion tubing so as the keep the pressure of the dry steam at boiler pressure.

Where you derived this, I'm not quite sure, as the thing regulating the superheater overpressure is a relief valve, intended to continue circulation in the elements (to preclude their relatively rapid burnout 'from outside' should circulation drop, as with the, er, throttle closed) and assuredly there is no proportional expansion in the elements; that they are open to the dry pipe and thence to the pool of overcritical water in the boiler regulates any particular pressure rise in the elements quite nicely otherwise.

Now, if you do as older practice did, and put the throttle (or "a" throttle) in the dome or dry pipe itself, you can have a separately-fired vessel overpressure issue in the elements, especially if there is any carryover into them which then is 'trapped, PEA-like, if the throttle is slammed closed.  The National Board ESC was quite justified in being alarmed by the possibilities should the valves or valve gear actually be arranged so that less flow is permitted 'through' at mid than is being evolved in such a case -- and it can be surprising just how much that is.  I'm sure "Blue Peter" rings a bell to some people here -- and that incident involved uncontrolled reverser excursion toward longest, not shortest, cutoff...

Adding heat can cause fluid flow without necessarily increasing pressure.  At least, not by much.

You're going to confuse the other Thermodynamicist Jr.s by saying this, because the whole point of heat expression in fluid flow shows up in expansion of some kind, which 'reads' as pressure.  In a gas (which, however imperfectly, is what produces the thrust in a steam locomotive) the kinetic expansion effects orthogonal to effective flow (or thrust generation) increase, and this can (and does) cause the tract to dilate and hence 'take more steam', but as you point out, Who Cares?  Even if the effect occurs and relaxes at the 'chuff' rate, it's relatively small, and would be incorporated into most of the indicator 'losses' anyway because there is no practical way to measure it dynamically independent of 'thrust-producing' mass flow.

Of course there must be a pressure difference for fluids to flow but a steam engine boiler system is a closed system until the exhaust valve opens.

Technically, it depends a bit on where you draw the system boundary.  One thing you're forgetting is that some of the work is extracted both in admission and expansion -- remember that the push on the piston isn't just 'expansion' of the steam, as in a Newcomen engine, it is also doing physical work on "something" outside the thermodynamics of the steam.  One thing that happens when people forget this is that they overconcentrate on wall-condensation effects and forget about nucleate condensation in the expansion.

Steam expands from the boiler through the throttle valves into the superheater...

You grew up in the wrong Britain.  The whole modern generation of steam technologists there accept the American pattern; the throttle is at the outlet of the circulating superheater, where it belongs ... or where Porta put it in the ACE3000, as multiple proportional Wagner throttles with coaxial control valves, directly in the steam lines ahead of the admission valves...

... expands into the steam chest and expands into the cylinder when the cutoff-controlled valve opens where it is trapped inside a chamber of changing volume. Only there  should there be any appreciable pressure drop.

But you have conflated two very different actions again, and the mistake rises to bite you.  The system remains in equilibrium, and mass flow, all the way past the moment of admission, and ideally, with little if any practical diminution of 'final' effective pressure at the moment of cutoff.  Here is the period where any action of the throttle, flow losses in the tracting, reserve volume in the steam chest, etc. has any influence on pressure whatsoever -- and believe me, it can be substantial depending on aspects of the design.  

On the other hand, once the steam edge closes admission, wherever the cutoff sets that to happen, you enter the happy world of actual intended expansion, where the effect of superheat and jacketing become more highly relevant (and the effect of open cylinder cocks, gland blow, and the like become more substantial influencers of subsequent EP and hence thrust, and make even numerically calculating the MEP integral a mess, but I digress) and the importance of using mass flow as an index of subsequent 'steam behavior' rather than pressure comes into its own.  Part of the important thermodynamics that made Schmidt such a hero is in keeping the expanding steam, a two-phase material at part of the expansion 'when it does any work', closer to acting like an ideal gas ... and that, rather than 'packing more expansive enthalpy into the steam mass', is the principal purpose of the superheat.  If you have a CRC 'rubber bible' or equivalent, you can look up the enthalpy added in superheat and see how dramatically small it is per degree rise compared to that in saturation under pressure -- Tuplin never failed to try to point this out, bless him.

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Sunday, February 7, 2021 6:43 PM

Lastspikemike
Disregarding leakage the system is closed until the locomotive begins to move. Theoretically, the steam only "leaks" out of the cylinder  after the steam chest is again a closed system. How much can the pressure in the steam chest actually drop during the time the valve is open?

The chest pressure only 'drops' because flow into the cylinder exceeds flow into the tract.  At starting 'cyclic' it may actually represent a flow impediment as the volume has to be filled from the boiler before higher pressure will be 'seen' at the piston head.

The concerns with steam-chest pressure are, I think, much more relevant in the range of speeds that the time the long-lap valve is closed to steam cause pressure effects in the steam tract.  I think of this as analogous to some of the effects with proportional relieving and recovery of compression.

...does the overlap between the two cylinders (among the three?) effectively mean there is constant flow through the throttle into the steam chest and up the exhaust. Just posing the question answers it, never a constant flow, always a series of discrete lumps of steam.

But you have to answer that question as fairly as you asked it, and the answer is 'no' but only incidentally related to 'chuffs'.  Remember that you don't hear the admission (this is why the South African 25 class is virtually silent with the fans turned off, and a Mallet working compound has only one audible exhaust signature) and the 'chuff', being a release to nominal back pressure, is not a comparable rate to admission mass flow.

One way to look at this is to examine a torque diagram like those in Wardale's Red Devil book to see where the little transition at admission cutoff comes, then compare this to indicated pressure (which also reflects tract and passage pressure to within a small amount).  But it is complicated because the effective branch point to the cylinders occurs downstream of the throttle but appreciably upstream from the chest, and it is that whole volume and mass flow of steam that is accelerated/decelerated, then stopped and started, as the piston valves work.

This thread seemed to me to be about what happens to pressure in the steam chest as the throttle is opened as contrasted to when the cut off is changed.

This is perceptive, and of course relatively irrelevant to low-cyclic operation.

Pressure transfers virtually instantly, speed of sound in the fluid actually, which is pretty damn fast.

This by another name is the reason that 'matching compression' to admission pressure at the moment of admission works, when the actual 'velocity' of the steam is still close to zero either side of the valve.  It also explains why unshrouding the valve enough to get good mass flow even a short time after steam-edge opening is important ... and the 'mass reservoir' represented by higher steam-chest pressure can give some reserve there.

The bangs you hear from unmuffled exhaust (or badly muffled as in a Harley V twin) are sound waves traveling down the exhaust pipe much, much faster than the gas flows.

Incidentally, so do some of the sharp 'bangs' or rifle-shot barks in poorly-designed front ends; in fact there is a whole discussion of supersonic nozzle design in high-performance locomotive front ends to get around shock losses.

For ICE piston engines using valves it is thought there is an rpm limit dictated by the speed air can move through the inlet and out the exhaust. At say 22,000 rpm, the maximum achieved by F1 V10 before the big rule changes to V6 turbo hybrids, engineers thought they might be getting close. Even then the gas flow is "particle" more than "wave" but wave effects have enormous effects.

I think we were just starting to 'get there' in a macro sense with the Q2 arrangement; there was enough actual 'mass flow' for pure inertia effects to show above the gas kinetics in steam flow in long pipes.  On the other hand, the Q2 was pushing practical limits for a single-unit road locomotive for other reasons, so it's doubtful that is a true limiting factor for high-cyclic operation.

Practical high cyclic is limited by a couple of other factors before pure steam kinetic effects become theoretically limiting -- balance effects and relative unsafety of precise valve-gear arrangements among them.  But a major reason attributed to the better 'efficiency' of Lenz-type poppet valves at high speed is that very-high-pressure admission (resulting in shorter mass transfer and thrust risetime) can be balanced by earlier progressive release, asynchronous timing, and much longer duration -- you rapidly get into the range where turbines are much more effective 'users' of the high inlet pressure and required compounding staging than piston engines are, probably for anything about 130mph or above.

In theory a steam engine does not need the discrete volumes required by a piston engine.

But on the other hand, remember part of why Parsons was so laughed at before he had so much fun with Turbinia: at practical railroad speeds and certainly practical railway operation in his day, the advantages of positive displacement and low-pressure plena in the absence of effective (as in ships) condensing and the disadvantages of high fabrication expense and high windage/tip losses at low speed have generally made the turbine a hard sell.  (I would argue we got most of the details right with the final version of Turbomotive 2, which had symmetrical turbines, either a multispeed or Bowes drive, and no separate reverse-turbine nonsense, but many of the fabrication and maintenance costs were still 'out of the water' for real-world express locomotive performance up to about 100mph).

What peak rpm is reached by a high speed steam locomotive bearing in mind only two cylinders and related valve mechanisms are involved in the majority of engines.

There are some peculiar numbers in the literature.  In practice the 'peak rpm' is limited by other factors before you get into theoretical steam transfer -- for example, valve tribology in the case of N&W J, or wheel bounce in the case of the C&NW E-4s.  One indication is that Kiefer, testing a J-3 (with Baker valve gear and unmodified tracting) spun the engine up to a recorded 161mph on greased rails during testing -- this obviously doesn't reflect train resistance or even locomotive resistances in service, but does indicate that you can spin the drivers up to a considerable speed net of all losses, provided the running gear can take the shock (and the track proves stiff enough!)

Roosen and Besler both thought high speed would require great reduction of reciprocating mass and great shortening of steam passages; they combined this with inherent low final-drive gearing to keep machinery speeds relatively low at high speed (which I think is a vital element about ~125mph).  I have no results pro or con regarding any of the French motor-locomotive testing, although I have tried on and off (and many of the results are at best confused by strange French operating conditions and attention to design detail).  Certainly PRR had firmly decided by 1944 that much of the coal future involved steam turbines, and by 1947 that it would involve Bowes drive for passenger speed.

Incidentally both the steam turbine and the gas turbine are of much more considerable antiquity than generally thought: the principal reason the steam turbine came first was as ship propulsion (where it is admirably suited, TIT is comparably low, and operation down to a low fraction of atmospheric pressure is easy) and gas turbines were impeded by temperature limitations for many decades.  (There is a fascinating paper, I think in the SAE repository, on how different the world would have been if more-practical gas turbines had been adopted for early automobiles in place of tetchy and complex spark-ignited carbureted engines... you'll get a kick out of it if you haven't yet read it...)

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Sunday, February 7, 2021 6:03 PM

gregc
when the throttle is first "cracked", with the locomotive standing still, steam flows thru the throttle raising pressure in one or both cylinders (depends on valve timing).

I think some of the discussion will be more straightforward if you recognize what the throttle does, and where it is.

An American multiple throttle is located at the superheater header, and consists of a series of relatively flow-streamlined poppet valves that open sequentially via cam as the lever is moved.  Since even this would be hard to move initially, the first poppet is purposely made very small, so there is less need to 'hog' the throttle open and then slam it most of the way shut for making smooth starts -- I trust you recognize what about a dome throttle makes that practice less critical in producing excess torque.  The older Wagner throttle (crca 1912) was an early and very effective form of fluidic amplifier, using a relatively small coaxial control valve to regulate differential steam pressure on the main throttle valve, and inspection of that patent and its claims will go a long way toward explaining to you what happens as steam traverses a partially-open throttle valve into a 'draining' load.

Keep in mind that as you open the throttle, the steam enters a fairly long, fairly cold tract, probably with high speed and associated flow and reflection shocks, and will suffer wall condensation to a higher degree until the tract comes up to reasonable temperature.  (I am uncertain the degree to which hot smokebox gas actually heats this tract; it may be considerable, but one notes that all the economizers like the Pielock that attempted to use smokebox-gas heat for steam superheating were more or less functional failures).  So you don't have immediate boiler pressure 'through' the throttling orifice, and then again you don't have immediate boiler pressure evident at the valve chest.  Further condensation then occurs in the ports and passages, and all this is the reason for cylinder-cock actuation, which of course lowers cylinder pressure while open.  A gauge of this is the 0.85 formula allowance for pressure in calculating starting TE.

Now it is true that

Flow thru the throttle decreases as pressure rises.
but this is only a balancing effect; almost a truism.  If a valve with adequate lap were centered and the throttle opened, pressure would make it as far as the admission chest, and once the thermal losses were overcome the 'back pressure' would be close to boiler pressure and hence mass flow would diminish.  Note that the rate of this pressure equalization is proportional to the throttle opening, as you would expect.  We will see this effect again in a moment.  

At some point, there is sufficient pressure to move the cylinder piston. when the piston moves, cylinder volume increases which impacts the increase in pressure.

Let's look very carefully, because there are several things actually changing comparatively quickly at this moment.  

In a good engine with minimized dead space, only a very small volume exists around the head of the piston.  Once this opens even slightly the 'throttling' effect of steam entering this volume drops greatly.  Now look back along the passage, valve, and chest toward the throttle: you will comparatively shortly fill up this new volume with additional mass flow, and only the amount of 'steam' that does this will pass the throttle.  (Remember that condensation in the newly-opened volume, plus the exhaust through the open cocks, account for some of the flow).  Very quickly we can expect the steam pressure against the whole of the head of the piston to reasonably equilibrate (this because superheated steam is a relatively good insulator and there is little nucleate condensation in it yet).

Think for a moment about what happens if the throttling orifice is not open far enough for the 'necessary' mass flow -- the pressure 'available' will not be at full boiler pressure, and may steadily fall below the fixed loss (decreasing as things warm up) until things balance at some lower nominal pressure.  Likewise if the throttle is cracked open past the loss rate, the pressure will rise until, again, balance across the throttle orifice is achieved.

In an ideal world the product of pressure and volume is constant, PV = C

For an ideal gas.  Even superheated steam is not an ideal gas... but ignore that for now; we're making very slow changes to the 'system'.

Pressure decreases as volume increases.  This is only true for a closed system (no change in steam mass).

Yes, but... we haven't gotten to expansion yet; there is still mass flow required, and quite a bit of it... which is why much of what you now state is so:
Pressure remains constant if the flow thru the throttle is equal to the rate of change in volume (constant density)
.  Pressure decreases if flow is less and increases if greater than the rate of change in volume. And any change in pressure will affect the flow thru the throttle.[/quote]Note that you're confusing yourself by looking at hypothetical rises and falls in indicated pressure as the engine begins to turn; the important thing is that boiler steam is 'attempting' to fill any increase in the cylinder's (or any other part of the system's) volume through the flow restrictions of the throttle, then tract, then passages.   Therefore
... If the flow is more than enough to accommodate the increase in volume, pressure will increase, applying more force on the piston and increasing the change in volume.
The first is true, the second is true, but the third is only incidentally true: you have forgotten that this is not a free piston, it is constrained by the rodwork and so its advance is a different 'balancing act', one that is governed by pressure work extracted from the steam.  Very likely at or near starting, the throttle will be open 'further' than the rate of increase in cylinder volume as the stroke commences, and therefore the portion of stroke from the moment of admission to the moment of cutoff will have indicated mep reasonably close to some proportion of boiler pressure ... at the low speed corresponding to the first few degrees of wheel rotation at starting where the 'starting TE' is figured and indicated.

Now you will see this cut all sorts of fascinating didoes at very high cyclic, but we aren't anywhere near there yet.

If the throttle is open too wide, pressure will increase rapidly and excessive piston force can result in slip.

But remember that the locomotive was designed for the empirical boiler pressure that 'gets as far as the cylinder' as it slowly increases in volume, so even at full nominal boiler pressure you're unlikely to have 'undesigned-for' piston thrust.  And, so far, the torque rise varies only with rod angle (remember here that you have to average the two components from the quartered pistons in the engine as a whole, or the numbers won't make any more sense than the performance trying to start a one-cylinder engine close to FDC or BDC) as any tendency to 'stall' would be overcome by cracking the throttle enough wider' to give the appropriate pressure.

So far there is no mystery.  Likewise it is evident that pressure and 'mass flow' remain proportional as the effective cylinder volume increases: only the areas of head and piston 'matter' in the pressure calculation, and those do not change with volume (the hoop stress in the cylinder certainly does, but we give few craps about that).  So if there is no cutoff, you eventually wind up with a cylinder full of "near boiler pressure" steam (depending as you noted on the percentage of throttle opening vs. losses and volumetric expansion) at the moment of release ... at which point you have a huge mass flow of still high-pressure steam blowing down into what is the effective flow restriction of the exhaust tract.  That, like discharging vs. charging a capacitor, is determined only by the flow characteristics of the tract as the steam edge exposes port area, with a few exceptions like jumper caps or variable nozzles that were uncommon and a bit 'tetchy' to maintain in practice...

In any case a colossal waste of the energy expensively put in the steam.  So instead, we cut off admission at some percentage of stroke, at which point flow through the throttle quickly equilibrates to near-zero at high pressure, and the trapped volume of steam does its business.  (Note that the rate of cutoff is influenced by increasing shrouding between the steam edge and the port, the opposite of steam inlet to 'infinite volume' at the moment of admission -- this is the counterpart of what happens at the exhaust)

as speed increases, there is more consistent steam consumption.

It might be better to look at this physically, a greater number of strokes-per-minute resulting in less full stoppage and starting of the elastic mass of steam in the tracting.  At some point the valve-chest volume begins to act a bit like the receiver in a compound, providing additional mass reserve over what the tract will 'flow' during the shorter and shorter period of admission; an additional part of this effect is inherent with larger valve diameter.

The need for greater steam flow necessitates the need for both less restriction between the boiler and steam chest (more open throttle) as well as a greater pressure difference between the boiler and steam chest.  (I believe this is analogous to Ohm's Law where pressure is like voltage and steam flow like current).

But be careful; you have a false analogy going.  Remember that, as in an IC engine, each stroke is a complete power event, but MEP is still a factor of cutoff and not energy release, so power extracted goes up considerably with rotational speed and the practical mass flow from which that power comes likewise has to increase.  (If it does not, the engine will balance at lower speed; if you do not shorten the cutoff you'll use more water mass to produce the same power at the wheelrims)

if there is a need to increase speed, the throttle may be opened as speed and consumption increases.

But very quickly the overall effect of throttling becomes outweighed by moment-of-cutoff concerns, to the point that it's recognized that best water economy involves opening the throttle as far as possible (nominally removing its flow resistance as far as possible) and then 'driving on the reverser' for all changes, both power and speed (since the second is a consequence of the first given the machine and consist resistance)

Depending on cutoff, cylinder ports will not be opened all the time and steam chest pressure will be allowed to recover (increase more so) when valves are not opened.

Now we get to the question of what physically happens in the valve chests during operation -- and this is of relatively peripheral concern in flow-streamlining the steam tract effectively; it can be instructive to read what Chapelon found in his practice over the years.  The principal difficulty is that at higher cyclic the physical flow of the steam becomes more and more quickly interrupted and resumed, with all the effective mass flow (this is not a weasel-word use of the term) occurring in a few milliseconds, but the geometry around the valves not being optimized for intermittent flow at high or 'tuned' rate, and reflected shock back up the column of steam increasingly becoming important.

At higher speeds, when less force and acceleration are needed, cutoff can be reduced to both reduce consumption of steam and use it more efficiently (expansion).  The reduction in steam consumption allows a higher steam chest pressure, higher starting cylinder pressure and possibly higher mean effective pressure.

But, in context, this is almost a truism ... and we don't care what the steam-chest pressure "rises" to if our steam distribution is controlled entirely by valve modulation, as the implication is always 'reduce the resistance in the steam tract up to the valves' and obviously WOT contributes to this if the designer was even halfway competent...

If this is the case, a reduction in both cutoff and throttle are necessary to maintain an MEP resulting in constant speed.

This would be nice, but in practice remember the MEP becomes peakier at higher speed and always stays in the range where it can be modulated entirely via the valve gear at sensible cyclic.  The only time you see MEP being influenced by throttle is in starting of engines with relatively low (or ininformative) FA, with the PRR T1 being one of the poster children: there, partial throttle can be looked at as an alternative to sliding-pressure firing in keeping the MEP broad enough to avoid excessive torque peaking and its various, usually evil, effects.

Of course, cutoff would likely be increased when going up a grade to increase MEP especially at lower speeds where steam consumption is lower and steam chest pressure is not as low as at higher speeds.

Hence the use of the 'company notch' (which I understood to be short, not long, cutoff when I was much younger, not realizing how the object was to wring maximum drag out of the engine over a given profile, not make the most of fuel and water...)  Here again nothing is gained from closing the throttle a bit: as you encounter a grade the only good way to increase MEP is to lengthen the cutoff and hope the effect of the added mass flow is sustainable.  (Incidentally, it might be interesting to indicate the steam-chest pressure at higher mass flow under this condition, but again, it's incidental to what happens with the throttle position and with what is adjusted to make more power at the cylinders.)

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Sunday, February 7, 2021 10:26 AM

Lastspikemike
This thread seemed to me to be about what happens to pressure in the steam chest as the throttle is opened as contrasted to when the cut off is changed.

??

gregc
does a reduction in cutoff without any change to the throttle result in an increase in steam chest pressure (since less steam is being drawn into the cylinders)?

greg - Philadelphia & Reading / Reading

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Sunday, February 7, 2021 6:48 AM

i know a steam engine is far more complicated than my simple explanation below which explains the basics

 

when the throttle is first "cracked", with the locomotive standing still, steam flows thru the throttle raising pressure in one or both cylinders (depends on valve timing).  Flow thru the throttle decreases as pressure rises.  At some point, there is sufficient pressure to move the cylinder piston.

when the piston moves, cylinder volume increases which impacts the increase in pressure.  In an ideal world the product of pressure and volume is constant, PV = C, Pressure decreases as volume increases.  This is only true for a closed system (no change in steam mass).

Pressure remains constant if the flow thru the throttle is equal to the rate of change in volume (constant density).  Pressure decreases if flow is less and increases if greater than the rate of change in volume.

And any change in pressure will affect the flow thru the throttle.

If the flow is more than enough to accommodate the increase in volume, pressure will increase, applying more force on the piston and increasing the change in volume.

If the throttle is open too wide, pressure will increase rapidly and excessive piston force can result in slip.

 

as speed increases, there is more consistent steam consumption.  The need for greater steam flow necessitates the need for both less restriction between the boiler and steam chest (more open throttle) as well as a greater pressure difference between the boiler and steam chest.  (I believe this is analogous to Ohm's Law where pressure is like voltage and steam flow like current).

if there is a need to increase speed, the throttle may be opened as speed and consumption increases.  Depending on cutoff, cylinder ports will not be opened all the time and steam chest pressure will be allowed to recover (increase more so) when valves are not opened.

at higher speeds, when less force and acceleration are needed, cutoff can be reduced to both reduce consumption of steam and use it more efficiently (expansion).  The reduction in steam consumption allows a higher steam chest pressure, higher starting cylinder pressure and possibly higher mean effective pressure.  If this is the case, a reduction in both cutoff and throttle are necessary to maintain an MEP resulting in constant speed.

Of course, cutoff would likely be increased when going up a grade to increase MEP especially at lower speeds where steam consumption is lower and steam chest pressure is not as low as at higher speeds.

 

greg - Philadelphia & Reading / Reading

  • Member since
    February 2005
  • From: Vancouver Island, BC
  • 23,330 posts
Posted by selector on Sunday, February 7, 2021 2:16 AM

Lastspikemike

... 

The cut off device is closed more then it is open. 

 

Not so. When lifting heavy tonnage, the reverser will be 'in the corner', meaning the valve is slower in action and leaves the inlet port open longer so that full steam can work longer against the piston face.  When a steamer is skipping along at limited speed, that is when the valve covers the inlet longer and restricts admission so that the steam has to work longer in the piston's stroke.

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Sunday, February 7, 2021 12:53 AM

Part of the equalization involves the steam circuit design.  In the years before the ASME 'restored' the locomotive-boiler code, there was some discussion whether a superheater needed its own separate safety-valve arrangement under Part 230 as a 'separately fired pressure vessel' -- the logic being that with good long-lap valves at mid at one end, and a tight throttle at the other, carryover 'trapped' in the elements might build up dangerous pressure.  Now this is true of a dome throttle, but a modern front-end multiple is ahead of the superheater elements, so any pressure increase that might occur in them relieves back along the dry pipe to the vast reservoir of overcritical water 'under the dome' so no explosion is possible.  

Carryover into the tract between the multiple throttle and the valve chests is certainly possible, but that's much more a bursting hazard in the liquid phase than it could ever be as a 'fired' vessel in the smokebox...

Something else that is fun is to see how much of the effect of very high boiler pressure is wasted with poor front-end design: it's fine to get it into the cylinder faster, but then you have to get it out again after it's volumetrically expanded...

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Saturday, February 6, 2021 8:08 PM

of course pressures will equalize in a closed system and as they do, fluid flow will decrease

but while in motion a steam engine is not a closed system.   steam is constantly exiting thru the exhaust ports of the cylinder and out the stack.

greg - Philadelphia & Reading / Reading

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Saturday, February 6, 2021 4:35 PM

Lastspikemike
Why would there be a pressure difference on each side of the throttle? 

why would a fluid move between two volumes if they are at the same pressure?

greg - Philadelphia & Reading / Reading

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Saturday, February 6, 2021 3:47 PM

The single biggest difference between IC and external combustion is how the heat makes pressure across the length of the stroke.

In the IC engine, the heat is developed entirely within the closed volume connecting with the cylinder, and the heat rise producing effective pressure occurs after the valves are effectively closed.  In fact when pressure-charging an IC engine the purpose is to get higher density of oxygen (and then stoich fuel) into the limited volume for the limited time available, and useful heat is actually (and often somewhat lavishly!) removed from the charge via intercooling.  The two issues of peak firing pressure and EGT limit the power output per cylinder far more than any MEP concern.

Meanwhile with external combustion, you 'could' have MEP  at essentially throttle pressure (for reasonably low cyclic) all the way to exhaust release and likely relatively high for a bit thereafter merely by keeping the inlet valve open -- in fact there is a weird approximation on limited-cutoff engines by fitting slot ports (Porta calls them 'Weiss' ports because he liked to recognize who he thought were people with key ideas) which are small ports in parallel that open to steam earlier but that can build up high equilibrium pressure given time... which will stay high, and exert "starting TE" type pressure on a piston, as long as the piston speed stays relatively slow.

Of course this also implies that nearly a whole swept volume's worth of high-pressure steam, still with considerable superheat, will blow to exhaust ... and yes, there are momentum effects in the steam just as there are in partial-vacuum ramcharging or tuned ports in IC; some very interesting ones in PRR Q2 testing... and very heavy loads applied by the piston to the motionwork and parts of the suspension and frame will be joined by heroic periodic induced draft if the front end is any good... leading to all sorts of effect from fire throwing to increased nozzle gas cutting, assuming the boiler can source that volume... here indistinguishable from mass-flow because at a reference pressure throughout... without pressure drop in parts of its structure.  A version of this could be seen in the pre-planetary-transmission PRR external steam-turbine (6200) if the throttle were opened too briskly while there was still considerable slip in the turbine... I forget the record number of staybolts pulled out, but it was distressing...

Instead of going for Brobdingnagian horsepower ratings, though, we find that metering admission mass precisely gives you that 'volume' of steam that just gets done expanding to the desired exhaust conditions to produce draft in your front end ... this being an average back pressure in the low 20s in the low '20s (see the fascinating experiments into automatic cutoff control servo-based using that as a control signal!) dropping to as low as 5 or so in late steam planning.  (Naturally you will need both extended entrainment area and better combustion-gas flow shaping at zone of entrainment to make that low a back-pressure work with the volumes of steam and combustion gas needed at high speed, but we can revisit that later...)

Now we can easily trade throttle pressure against water rate for quite a range within the constraints of EC displacement and stroke (a point that Tuplin never quite got tired of advocating).  Quite a bit of the nominal case for high (above about 300psi) throttle pressure evaporates when additional costs involved with boiler construction and maintenance are factored in (and modern water treatment is not).  But there have been interesting examples where sliding-pressure firing can produce interesting performance 'at lower required peak horsepower'.

IDEALLY a certain amount of pilot injection technique, both during the admission period and (for reasons like those justifying the N&W 'booster valve) part of following expansion, can be valuable, just as they are for good compounds.  There were some interesting discussions in late French practice, some probably at least partially mythical, regarding the best way to optimize steam distribution.

But we now come to the other end of the engine discussion, which involves how to get rid of the mass in the exhaust.  IC engines with trunk pistons have little concern in this regard, but large DA engines with SHM or radial valve gear trap a certain amount of steam at exhaust cutoff, which is then compressed (and compression-heated) to what may be a substantial counterpressure, many psi above peak boiler pressure (see some of the Okadee literature for an inkling of just how much).  Some of this is recognized as essential in overcoming inertial force on the running gear; as I noted, I consider some of it essential to get good effective mass flow at high cyclic.  But if the necessary mass is retained in a larger-bore cylinder with low port-to-passage ratio that attempts to minimize physical 'dead space'... you can get some ridiculously high rates of increase in both pressure and temperature.

Jay Carter (in SACA) handled this with proportional 'overpressure' valving at the cylinder head that meters the surplus into an insulated reservoir space, then uses this to keep pressure up in the early part of the physical return stroke.  There is little reason not to apply this to a modern (small-bore, longer-stroke) steam locomotive...

The idea that mass flow would increase as a locomotive starts accelerating downgrade with constant (full) throttle is an interesting thought experiment.  Naturally the acceleration in such a case will still 'balance' as in the Davis formula, but now because there is more 'contribution' from the steam since gravity has taken over some of the work.  (You could likely figure out just how much if you knew the machine losses for gravity and thrust respectively and integrated them across the profile, but... in practice the engineman would wind back the cutoff to keep the engine from going over timetable speed, and save water, rather than use the 'larger' amount of steam going down the inlet tract thanks to the higher swept volume at the higher achieved speed...)

Now the idea that MEP in the cylinders would increase if gravity is doing some of the work is interesting to contemplate, except that the portion of 'contribution' up to cutoff is proportional to stroke, not number of strokes per unit time, and the amount of expansion in the cylinder is proportional to geometry, not power extracted.  So yes, the steam might have higher temperature (by the amount of work 'not extracted' by expansion) and hence the indicated MEP a tad higher... except this is heat and pressure that get promptly blown out the exhaust tract (or in residual form contribute to compression issues) if it is not 'equilibrated' into producing further acceleration.  (Note that something artificial eiukd have to be done to make the train in the thought experiment accelerate 'only' at the proportional gravitational rate with the throttle and cutoff unaltered...

A more interesting discussion arises out of how you practically DO handle an engine that 'turns downhill' and you don't want to have accelerate.  That gets into the realm of drifting using sniffing air vs. steam, working steam downhill vs. using a special 'drifting throttle' with a pressure regulator (set not coincidentally around 15psi) vs. using a bypass arrangement like Wagner or Nicolai or Trofimov valves.

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Saturday, February 6, 2021 10:06 AM

PM Railfan

gregc - "and i'm suggesting that the mean effective pressure (MEP) can be higher if cutoff is reduced while keeping the throttle constant"

you could say that, yes.

sounds like an affirmation of my quesiton

greg - Philadelphia & Reading / Reading

  • Member since
    January 2013
  • 1,034 posts
Posted by PM Railfan on Friday, February 5, 2021 8:23 PM

"how could you be using more steam than was passing the throttle?" - Your clipping along at say 40mph, which lets say equals 1/2 throttle and your running a mild cutoff. Your pressure is steady, and under load. Everything nominal.

You hit a section of track thats downhill. Youll pick up speed without ever adjusting throttle - gravity + momentum. So now the cylinders are working a pinch faster than they just were on level track.

Pressure will drop from throttle to valve cylinder because cylinders are taking more steam because they are cycling faster. They wont bring it to a vacuum, but more steam (volumetrically) will be passing the throttle than your original amount was on level track. 

And only for a second as the first thing youd do going downhill is close the throttle to a pinch (or a bit more) and grab a little train brake. Now, thats little to no steam, but cylinders are cylcing cuz your doing 45mph now - downhill.

Thats one way.

The other way is you arent running a super powered locomotive and your boiler isnt keeping up with the tremendous amount of steam your feeding the cylinders. Which is the whole reason super power locos were designed.

 

"and i'm suggesting that the mean effective pressure (MEP) can be higher if cutoff is reduced while keeping the throttle constant" - you could say that, yes.

 "isn't 'bring you back up in pressure' another way of saying pressure would increase?" - yes. However, "bring you back up to" is meant to infer that instead of pressure just rising, your rising from a lower point than normal instead of starting at normal then raising pressure. But its technically still "rising'.

 PMR

 

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Friday, February 5, 2021 10:04 AM

Part of the thing to remember about the throttle is that it restricts mass flow - pulsating mass flow.  That means that the "pressure" effect at part opening is an effective pressure drop at times during the admission.  Look up 'wire-drawing' for an older explanation of the effect.

This is related to cutoff, since cutoff also controls effective mass flow.  It may help to think of cutoff as metering a given mass of steam (at whatever pressure profile describes its flow) which then expands for the remainder of the contained stroke -- the superheat in the contained steam keeping the water from condensing to liquid phase at effective pressure at any point in that expansion.  Incidentally only about .007" of the wall metal cycles temperature along with the steam, which is why superheat has a fighting chance against wall condensation ... and why jacketing and Wardale-style insulation work, by keeping the wall metal just the other side of that .007 at a higher temperature so there is less heat exchange as the steam (which is a pretty good insulator even at these pressures) expands and cools.

"Conventional wisdom" for many years is to reduce any sources of flow restriction or drag in the inlet tract, all the way to the valves -- so the idea is to get the throttle open as wide as possible, as early as possible, and then 'drive on the reverser'.  This is lovely from a thermodynamic perspective, but often lousy in real-world locomotives, where anything shorter than about 15% (in older engines, 25% or more) produces little advantage as the torque peaks and condensation losses become severe near the end of the stroke.  In true high-pressure engines this becomes more severe as the peak pressure stays high but mep gets lower and lower; even with a relatively high mechanical factor of adhesion, this can produce both low- and high-speed slipping.  There was a pretty good discussion of this with respect to the British Caprotti on the Duke of Gloucester, a gear capable of 3-5% cutoff with all the necessary mechanical precision -- the advantage of that precision then being in timing and duration rather than in 'overlap' of admission and cutoff resulting in the analogue of short 'shutter speed'.

In any case, the cutoff at highest speed will not be short, any more than the highest vehicle speed in fast cars is in the highest overdrive.  This seems counterintuitive until you realize how demanded horsepower rises with speed.  Interestingly, across a very wide range of historical examples, 'fastest' cutoff has been in a very short range, in the low 40% range, with some of the effects of compression control taking on what I consider to be dramatically heightened effect.

Note that there is another wire-drawing effect once you get to the physical valves.  One reason long-lap long-travel piston valves remain competitive with 'poppets' as long as they do is that the steam edge opens a comparatively large area to steam comparatively quickly, as the ports can extend around the periphery of the valve except for the bridges, and proper flow-shaping of the volume in the passage 'away' from the axis to the cylinder can assist in early effective flow to the piston head (which is of great importance at high cyclic rpm and short cutoff, where admission is measured in milliseconds).  Here no amount of 'reserve capacity' in the steam chest improves admission, but proper compression management does because it can provide reasonably 'hot' steam at admission pressure right down the port to the swept volume at the moment the valve opens to steam, effectively eliminating the effect of dead space in the ports and passages and any great gas-cutting when the valves are still effectively 'shrouded' and would otherwise require high gas speed at small net opening -- not a recipe for avoidance of wire-drawing!

gregc
does a reduction in cutoff without any change to the throttle result in an increase in steam chest pressure (since less steam is being drawn into the cylinders)?

You can always spot the internal-combustion men because they think steam flow is analogous to getting air and fuel into one of those engines.  Steam is not 'drawn' anywhere -- at least not if engineers have much to say about it, or unless wire-drawing is wanted (and, curiously enough, there are places it becomes highly desired, indeed, as in Franklin type D poppet gear!)  I say this not to split hairs but because the steam 'goes where it pushes' and in a pulsating flow that starts and stops with porting and unporting.  

So shorter cutoff results in 'less steam leaving the valve chest' per stroke, and hence the likelihood that both the average and 'peak' pressure in the chest will be closer to boiler pressure (or effective pressure with the throttle partway shut, which I'll get back to later.)  With the throttle open and a properly-designed inlet tract, the steam-chest pressure will be higher in proportion to the lower mass flow going into the cylinders, but ideally this will not matter significantly as there will be adequate boiler supply to keep the chest pressure high at any point in the stroke.  

(Incidentally you now see part of why 'underboilered' locomotives run out of 'horsepower' as they run out of the ability to convert water to steam -- they are plenty fast, and plenty economical, until the demanded mass flow exceeds what the boiler can source from its overcritical water (note that this need not concern firing rate yet!) and the pressure begins to fall.

can reducing cutoff w/o a throttle change result in an increase in power (i.e. tractive effort) due to higher steam chest pressure?

Only circumstantially, but it would not be 'higher steam chest pressure' other than incidentally.  It is more the steam-chest volume that matters, and we start to get into the discussion of 'too much valve' (which was one of the somewhat-misguided concerns of designers in the dark ages before mass flow was better understood).

What matters to "engine" power is mep, and this is only incidentally concerned with nominally higher starting pressure.  If that pressure falls 'too quickly' across admission, then developed mep will be lower, and that isn't anything you can gauge by nominal chest pressure from an instantaneously-reading gauge or indicator.  

Two peripheral observations:  This is where the glorious promise of the French ethyl-ether bottoming cycle for compound ship engines comes apart: you might think with that colossal vapor pressure at comparably low temperature it would be a nifty working fluid, but take your rubber bible and see why the pressure disappears nearly as soon as it has to do any practical expansive work...

And if cutoff is short, the excursion of mep that does the expansive work will be across a comparatively short range of piston travel, which is then related to wheelrim torque through the piston rod/crosshead and main rod.  Higher initial pressure means a variety of increased machine losses, and perhaps peak torque assisted by the mechanical advantage of the rod throw.  This provides greater possibility of slip four times per revolution on a 2-cylinder DA, followed by fairly sharp fall in torque resulting in shock when adhesion is restored ... assuming the slip doesn't propagate around to the next torque peak, which especially with precise and low-mass poppet valves can give you rapidly increasing spin with perhaps no easy way to control it.

Something people easily forget is that a conventional throttle, even a modern multiple-poppet front-end throttle, is not an easy thing to open and close, or to adjust precisely.  The situation is worse if you use different throttles on different engines of a locomotive (as some of the naive proposed for the 5550 T1 'betterment' program) unless you have expedient fine-adjustment methods built in; this is not a particularly difficult design exercise, but to my knowledge even the Franklin Precision air servo throttle control did not have the combination of rapid excursion and fine settling precision necessary for good effective throttle control of slip.  And the usual cutoff control is an 'average', set by lever or wheel position, and incapable of being worked longer and shorter 'in time' with the engine, let alone in appropriate phase with it.

and does this mean that if cutoff were reduced should there be a corresponding reduction in the throttle to limit to the increase in steam chest pressure to maintain a desired speed?

Not for thermodynamic reasons.  There might be mass-flow reasons in the tracting that would benefit from partially closing the throttle, but not for any putative 'increase in steam chest pressure' and it is difficult to imagine a situation in which closing the throttle at high speed would give better performance than slightly shortening the cutoff further even if there were.

The 'correct' conventional approach is what is mirrored in the Valve Pilot operation: the chosen cam relates 'road speed' to some ideal or empirical degree of cutoff, and the most efficient operation of the engine is then to 'match needles' on the Valve Pilot's instrument, something relatively easily done.

Now, one of the theoretical things that might be necessary with a T1 experiencing true high-speed slipping is the analogue of the Australian starting method in slippery conditions: if there is a limit to how short the cutoff can be set without torque-peak slip induction, then the throttle might be closed to reduce the admission pressure at peak flow.  Of course the 'correct' answer here is precisely what the PRR crews learned to do: sliding-pressure fire the engine so that you trade water rate for slip at starting or when high-speed slip becomes evident; there's no guarantee that attempting to induce wire-drawing without modulation will even help much (other than reduce the power of the engine to the point it no longer slips... which is not that different from just slowing down out of the critical speed range.  But that is a more sophisticated combination of effects than your question is concerned with.

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Friday, February 5, 2021 8:02 AM

PM Railfan
gregc
it's pressure fluctuates as the valve open/closes.

greg - Philadelphia & Reading / Reading

  • Member since
    January 2013
  • 1,034 posts
Posted by PM Railfan on Friday, February 5, 2021 5:50 AM

"my understanding is that the steam chest is the part of the cylinder that the cylinder valve slides in. " - The steam chest is the part of the loco that encompasses both cylinder sets, each side.

The upper cylinder is called the 'valve' cylinder and the lower is the 'piston' cylinder.

 

"it's pressure fluctuates as the valve open/closes." - If i read that right.... the 'piston' cyclinder pressure fluctuates as the 'valve opens and closes'. Im not sure im following you on this sentence.

 

"with throttle constant, doesn't the steam chest pressure depend on the amount of steam "consumed" by the cylinders?"  Well if nothing is changing, then yes you could measure it that way i guess. There is a measurable pressure at constant throttle vs constant rotation of wheels. Slow up the cycle, the pressure goes up. Quicken the cycle, the pressure drops. Thats is, no throttle change.

 

"wouldn't it increase if the cylinders drew less steam because of reduced cutoff?" Not neccessarily. If you were already using more steam than was passing the throttle, the cutoff would only bring you back up to pressure. That or ease off throttle because your going faster than youve given it throttle to go.

 

PMR

 

 

 

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Friday, February 5, 2021 4:49 AM

PM Railfan
Not in the chest, but the pipe from throttle to valve cylinder in chest will back up slightly.

what does "back up slightly" mean?

my understanding is that the steam chest is the part of the cylinder that the cylinder valve slides in.   it's pressure fluctuates as the valve open/closes.

with throttle constant, doesn't the steam chest pressure depend on the amount of steam "consumed" by the cylinders?   wouldn't it increase if the cylinders drew less steam because of reduced cutoff?

255

 

greg - Philadelphia & Reading / Reading

  • Member since
    January 2013
  • 1,034 posts
Posted by PM Railfan on Friday, February 5, 2021 4:15 AM

"A steam loco throttle is nothing more than a valve controling the size of an opening." - Essentially correct. Its a few openings actually. But youve got the idea.

"it controls the pressure only to the extent that it limits the flow and depends a lot on the consumption of steam by the cylinders" - Essentially correct. If the cyclinders are taking more than a wide open throttle/boiler can deliver, yes that pipe down to cyclinders will be lower pressure than a loco wide open throttle that has a boiler that can keep up. (Thus..... the term "Super Power" locomotive came to be. Locos with boilers that could keep up at wide open throttle.)

But know this - the pipe from throttle to cylinders will ALWAYS be lower pressure than the boiler even at wide open throttle, and no matter what the cylinders are doing. If it wasnt lower, steam would never leave the boiler in the first place. Like electricity takes the path of least resistance, steam takes the path of lower pressure.

Only two times is steam a higher pressure than the boiler: 1) when injecting water using an injector (venturi effect), 2) the compression stroke of the main piston.

"reducing cutoff more efficiently uses steam by allowing the lesser amount of steam to expand." - 1/2 correct. The whole idea behind cutoff was efficiency. Steam locos could only travel so far before needing a refuel. It wasnt beneficial for railroads to have to stop a fast express freight then start it again. If they could stretch those stops out, well that = better operation, less fuel costs, less tie ups on the line etc etc. You get the picture. Steam expands no matter what the cutoff is. Thats the nature of steam to begin with. So cutoff is efficiency based.

Think of a garden hose. At one end its attached to your house via a spiggot (a valve like a throttle). At the other end of hose is a sprinkler. And sprinklers have that little adjustment (which is essentially their 'cutoff') that attenuates the time allowed for water to shoot at that little prong that moves the sprinkler head around.

Say that it takes full pressure (wide open throttle) to make your sprinkler work. But at the sprinkler you dont want it flying around like a top because of that much pressure. So you dial down the 'cutoff'.

Its the same principle as "its takes more energy to get something moving than to keep it going". You can use full water pressure to turn the head, but limited cutoff so it only moves so far at a time.

Starting a train = lever in the corner! When you get up to speed you can ease the reverser lever back. Just how far depends on how good your ear is and how worn the seat of your pants are.

"does a reduction in cutoff without any change to the throttle result in an increase in steam chest pressure (since less steam is being drawn into the cylinders)?" - Not in the chest, but the pipe from throttle to valve cylinder in chest will back up slightly.

"can reducing cutoff w/o a throttle change result in an increase in power (i.e. tractive effort) due to higher steam chest pressure?" No, see above.

"and does this mean that if cutoff were reduced should there be a corresponding reduction in the throttle to limit to the increase in steam chest pressure to maintain a desired speed?" - This is strickly an operations based question. As in what is going on at that moment. There are many reasons to open/close a throttle, and the same for the cutoff. those reasons change from moment to moment depending on grade, slack, danger ahead, signal change, etc etc.

Normally, you get to speed under full forward or as such a time in the upward gaining of speed that you sense your loco doesnt need a full amount of steam. If shes in a place where she can hold her own without adjusting the throttle then you ease up on the cutoff till she has just what she needs to keep moving at the desired speed.

"what would the impact of higher steam chest pressure be on tractive effort." - well if the math of your loco is on, and she didnt slip, the TE would probably go up.

But you dont raise chest pressure. You raise boiler pressure and figure valve and main piston size and travel.

 

PMR

 

 

  • Member since
    July 2009
  • From: lavale, md
  • 4,678 posts
Posted by gregc on Friday, February 5, 2021 3:15 AM

of course steam chest pressure can't exceed boiler pressure

my question is does steam chest pressure increase when cutoff is reduced because consumption is less?

what would the impact of higher steam chest pressure be on tractive effort

Semmens provide more of an overview and history than provide a complete technical description of steam engines

greg - Philadelphia & Reading / Reading

  • Member since
    February 2001
  • From: Wyoming, where men are men, and sheep are nervous!
  • 3,392 posts
Posted by Pruitt on Thursday, February 4, 2021 10:02 PM

Basically the throttle controls the rate of steam flow into the cylinders, while the cutoff controls the duration of the steam flow. So when starting out the cutoff will be wide open to allow steam flow into the cylinders for the entirety of the piston stroke. Once up to speed, the cutoff is reduced to allow steam into the cylinders only for a very short period of time, to maintain speed while minimizing steam use.

  • Member since
    January 2009
  • From: Maryland
  • 12,897 posts
Posted by ATLANTIC CENTRAL on Thursday, February 4, 2021 9:34 PM

gregc

as far as i know, a steam engine throttle is nothing more than a valve controlling the size of an opening.   it doesn't regulate the pressure (i.e. limit the pressure).    it controls the pressure only to the extent that it limits the flow and depends a lot on the consumption of steam by the cylinders

reducing cutoff more efficiently uses steam by allowing the lesser amount of steam to expand.   in other words, when cutoff is reduced by half, there may only be  a ~~20% reduction in power from the steam.

(the other advantage of decreasing cutoff is reducing the consumption of steam which means conserving water, fuel and manpower).

 

my questions are:

does a reduction in cutoff without any change to the throttle result in an increase in steam chest pressure (since less steam is being drawn into the cylinders)?

can reducing cutoff w/o a throttle change result in an increase in power (i.e. tractive effort) due to higher steam chest pressure?

and does this mean that if cutoff were reduced should there be a corresponding reduction in the throttle to limit to the increase in steam chest pressure to maintain a desired speed?

 

Steam chest pressure is not going to exceed boiler pressure in any event, steam chest pressure will equalize to boiler pressure quickly at all but the lowest throttle settings.

Until you run out of boiler pressure, than both drop......

Cutoff is simply valve duration timing, similar to the variable cam shaft in your modern car.

Wide open for max power, reduced flow for crusing.

Throttle and cut off need to be within reasonable limits of each other for smooth and safe operation.

Sheldon 

    

Subscriber & Member Login

Login, or register today to interact in our online community, comment on articles, receive our newsletter, manage your account online and more!

Users Online

There are no community member online

Search the Community

ADVERTISEMENT
ADVERTISEMENT
ADVERTISEMENT
Model Railroader Newsletter See all
Sign up for our FREE e-newsletter and get model railroad news in your inbox!