Overmod find this interesting because when the Q2 was tested, one of the complaints involved excessive dead space (REALLY excessive -- up to 15%). But that locomotive had roller bearings and Franklin wedges, and would never have 'excessive play in the bearings', so the dead space was due to some other cause. I cannot tell what that 'cause' might have been.
Had To Fail?
Juniatha That PRR system of crews and engines - uhm - dispersion if you don't mind my expression seems to have been rather rough and inconsiderate . The T1 engines would desperately have wanted a specially trained brigade of crews composed of a rank of more interested or brighter minded people - .
Like the Caledonian Railway Cardean Class? I believe that each one had only one driver at a time.
Looking forward to the new thread, Overmod. "Modern Steam"?
Just a few replies here, seriatim...
Hi Juniatha! Your mentioning the T1's valve covers being removed and not replaced, probably for ease of maintanance. Good point, but the answer probably is the shop crews knew the diesel replacements were coming so there was no point in keeping up appearances. As steam was winding down most shop crews only did enough work on the steamers to keep them alive until the diesels showed up, probably by direction, these were proud craftsmen after all. More than likely the bugs could have been worked out of poppet valves but at that stage of the game what was the point?
Hi friend611! True, the N&W tested a T1, found little to nothing wrong with it, but as they had (and we KNOW they had!) the best 4-8-4 around in the Class J there wasn't much point in buying T1's. A real "thanks, but no thanks!" situation.
Hi Overmod! I see the name "Lady Firestorm" is enough to strike fear into your stout heart as well! She's the only one in history to make Marines run in terror! Another story...
It's been said before, but the fact is the Pennsy stayed with the K4 as a passenger hauler just a little too long. Had they started thinking of a replacement around 1935 or so things might have been different.
I am delighted to see we agree far more than disagree.
That picture at the end is one of the best things I've seen all week! (We could Photoshop it to include various railroad paraphernaila... And YES the more stress the better, keeping it on 'the knowledge'... ;-}
Only a VERY few comments
Juniatha ... viscous clutch - yet not a mechanical clutch , for sure. Still , it will not react on the spot since viscous media can only transmit forces under giving , not one-to-one.
... viscous clutch - yet not a mechanical clutch , for sure. Still , it will not react on the spot since viscous media can only transmit forces under giving , not one-to-one.
Well, not exactly; they are not transmitting 'force' when running synchronized, but at even slight difference in rotational speed *provided the change occurs relatively quickly* there will be torque transfer. This should be easily enough to arrest just-past-incipient slip when a relatively small amount of 'correction' can re-establish adhesion -- i.e. before the engine can wind up substantial relative motion at the wheelrim and increasing rotational momentum in the 'flywheels' that are the drivers.
We then engage the sprag clutch in the appropriate direction, relatively smoothly (as opposed to a Maybach clutch, although I suppose you could use one if you wanted) and that provides positive torque capacity well in excess of what the differential thrust would provide at the engine's most 'efficient' cyclic speed.
So the one unit that slips , still slips – if not spinning as fast
It will not spin more than a revollution before the other pair would either arrest the slip or start to slip itself, at which point the restoring force is no worse than on the equivalent 4-8-4.
I'd like to see what dimensioning you'd propose for a viscous clutch to hold a low rpm high torque power output of a locomotive or half of it by 1950s technology – *g*
It is actually not that large, becsuse (as with Withuhn's conjugating rods) the actual power through the differential shaft never comes up to anywhere close to the peak power of the engines.
I should probably add for the sake of completeness that my current design for this clutch is magnetorheological. I do not think that technology was developed, let alone sufficiently mature, in the '50s, but it does solve most of the difficulties quite well.
The other approach that can be used for conjugating a duplex AT STARTING is to provide the gear shafts as described, but connect them with a locking clutch (the Maybach claw clutch being an ideal approach) that makes the two engines work together with no play other than the lash and tooth deflection in the geartrain (which is very small). It is possible that a Bowes drive could be externally excited so that the 'higher-speed' shaft is very effectively braked relative to the non-slipping one (the differential motion essentially providing the buildup of exciting current.) This is a non-contact device so no degree of increasing slip would cause shock or friction, and it is relatively easy to modulate so that (if desired) the rate of re-conjugation can be controlled dynamically. (That is established and documented 1947 technology, by the way -- the description of the layout is preserved at the Hagley Museum).
Enough about the T1: the 'slippery' part of it (that is, such part of it that was real, and not an exaggeration or canard -- I agree with you on conspiracy theory but will mention it later) could have been fixed without requiring permanent conjugation of the engines, and the theoretical advantages of keeping the two engines separate under normal conditions could be preserved to a sufficient practical degree.
(Oh yes, I thought I'd mentioned that the 45-degree alignment is provided by detenting, with the idea that small changes that produce slipping would tend to cause the engines to align relative to each other. Does not need to be perfect. (I was recently treated on another list to someone's opinion that an irregular-beat engine develops better draft -- which is unusual, and might actually be substantiable. There is going to be a discussion of this point between two of the top front-end theoreticians in the near future and I'll keep you informed on how that plays out...)
>> Duplex (modern sense) - two separate engines in a common frame under one boiler<< Yep - I hope we are not *duplicating* that kind of discussion that was going on about Art Deco ?
Yep - I hope we are not *duplicating* that kind of discussion that was going on about Art Deco ?
Note that I have NOT gotten into that -- having already made it in some detail a few months ago, and 'let it go' after being told it was somewhat ... inadvisable ... to make the point you did to Lady Firestorm. Sometimes it is best for knowledge to keep its peace strategically.
>> an approach, like Deem's or mine, << Overmod , I remarked on it already and may I repeat it here : I do *not* know *all* of the engineers that will have had a hand in writing the history of steam at one time or other – mostly rather earlier and of arguably lesser impact – and I'm afraid other readers may not necessarily fare significantly better.
Overmod , I remarked on it already and may I repeat it here : I do *not* know *all* of the engineers that will have had a hand in writing the history of steam at one time or other – mostly rather earlier and of arguably lesser impact – and I'm afraid other readers may not necessarily fare significantly better.
But I have mentioned Riley Deem repeatedly, described that he worked for Lima, that he developed an approach to conjugating duplexes that he said would work, that the approach involved 'gears', and that nothing has come to light regarding exactly how he did it.
Beyond that, I cannot say more. If anyone happens to have researched the subject, they may come across his name as being someone, perhaps the only someone in the United States, to take up 'fixing' the Q2 in a way that would make it a stable locomotive in service. (Or perhaps "less slip-prone" might be a better way to describe it! ;-})
I *deem* it wise you may want to fill that name-dropping with some life or we may as well skip it.
Yeah. I was proceeding with the general scientific shorthand of just citing whoever originated or was known for something. I tend to forget that not everyone is utterly fascinated by steam technology and therefore just loves to read everything they can as soon as hinted... well, they are NOT. (In my own defense, I conceded that point already; this was just a special case... keep reminding me whenever something is unfamiliar and I will explain what I intended the reference to mean...)
>> 'duplexes' is the only correct normal plural in English.<< .. only that 'duplex' itself isn't an English word but Latin - *g*
.. only that 'duplex' itself isn't an English word but Latin - *g*
Well, as applied to steam locomotives, it is an English word. Meanwhile, Fowler (not name-dropping here; he is still the acknowledged best authority on modern English usage, to the point where his name is as applied to the subject as Roget's is to thesauri) specifically refers to this particular plural form of this particular word.
Just trust Uncle Bob on this -- he knows.
'Duplexes' – oooh , that sounds just awful , sorry . In Latin , an added ending -i is often used for plural.
Well, even if this applied to the current situation, and you were to take the adjective 'duplex' as if it were a masculine Latin noun, it would be 'duplexi' (as the word is not 'duplexius', is it?) and in fact, if you were to actually look at Latin, you would find that the "proper" plural form in that language is 'duplices' or something very like it. (Again, just trust Uncle Bob on this -- it will save time googling or wikipediating.)
Sorry to go into the boring detail, but YOU were the one that kept it going... ;-}
("Duplexii" is a humorous way that Trains Magazine used to refer to the things ... but it isn't a proper term to use when referring to them technically.) And now let's let it be.
<valid discussions snipped for brevity>.
>> This is not what was actually observed with T1s, however. << It was ! - see youtube vids .
It was ! - see youtube vids .
I was still referring to the high-speed slip, not the situation at low speed and starting .. which of course also did occur, and probably had maintenance consequences.
The phenomenon I was addressing is observed mainly under conditions where there is ample overcritical* water to make steam, the throttle is substantially open, and the valve gear set around the 'ideal' cutoff for maximum road speed. Under these conditions the efficiency of the valve gear in producing meaningful admission and effective exhaust even at very high cyclic rpm contributes to the degree of overspeed. That's really the only situation in which a significant short-term spin-up is likely, particularly after T1 engineers learned to start as slowly as Amtrak AEM-7 engineers do ;-}
*I use the word "overcritical" for water that is held in liquid phase only by overpressure, to distinguish it from 'supercritical' which I use to refer to 'steam in liquid phase' over the critical pressure (about 3206 psi). I find this a useful enough distinction to recommend it.
>> If you actually calculate the rotational inertia for the drive unit, and then calculate the thrust profile as exerted by the drive (even with the smaller pistons and shorter stroke of the T1 design), you will not be surprised at how rapidly the rotation accelerates in the absence of correction.<< No , I'm not because I know what the relations are and actual behavior turns out just that way as expected . You just *cannot* cannonball some 20 t of congregate moving mass by input of some 10 to 15 t of t.e. – the power is too small !
No , I'm not because I know what the relations are and actual behavior turns out just that way as expected . You just *cannot* cannonball some 20 t of congregate moving mass by input of some 10 to 15 t of t.e. – the power is too small !
But that is not what I said, or at least thought I said. The point begins when you posit that the MEP over, let's say, the expansive range with 40% cutoff does not fall off significantly with increasing rpm. That means that the effective instantaneous thrust (vectored through the main) still has very nearly the same value at high rpm that it would at lower rpm. I believe you observe a rapid spinup of rotational motion in your beloved YouTube clip -- perhaps you care to explain that? And that is at lower initial speed... probably below the speed corresponding to max ihp...
I will put some material on this together and put it in its own thread, to stop injecting distractions into this one.
... t does not exactly help when you blamed me of comparing ‘old’ to ‘modern’ taking my proposed draft for ‘modern’ - which by default it is not.
This reminds me a bit about the old Aesop's fable (I trust that is not name-dropping to this audience) about 'blowing hot and cold'. Modern is relative to context; in architecture, we went through 'postmodern' a long time ago. There are plenty of people who have 'modern' approaches to the steam locomotive; Porta was one who developed a taxonomy of levels. We are restricting tech in this thread to the period you proposed, but "modern" steam in that era dated from considerably later than the early days of Super-Power; I'd suggest 1934 or so as the earliest, and around 1938 as a 'mature', date for what we are discussing 'improving'.
I will confess that I had thought your steam-locomotive designs were often intended in a 'modern' context, meaning that they were intended to work in present-day service, not just a hypothetical alternate history. If that is not so, I apologize (and think you would be selling yourself and your work short in making such a limitation).
... you assume the 'modern' ( steam ) loco to have lightweight rods - these would enable high rotational speed to be attained.
I interpret this slightly differently; the lightweight rods allow augment to be kept within tolerable levels at higher speeds, which (to me at least) isn't the same thing as 'higher rotational speed' at all except circumstantially. (In any case, lower rotational inertia would give faster, not slower, acceleration with a fixed applied level of force...)
Even the modern (21st-Century) analysis of high-speed truck design recognizes how quickly instability and nonlinearity can enter dynamical systems. I find it likely, as you do not seem to, that there will be critical speeds and resonances capable of inducing damage long before the system were to spin up to force/speed balancing equilibrium. Again, I will run numbers on this in a separate thread.
Well , that's correct - and that's exactly why it meant taking chances to allow a drive unit to go spinning and just sit there waiting for a pickup of adhesion - it may or may not have worked and if it didn't while the spinning rather revved up than cooled down , then it *was* high time for action .
I hesitate to weigh in on this point, because I have seen many instances where controlled slipping was used during starting, and you are correct in what you say. The only point I would make is that by the time the slip is accelerating rather than self-correcting, it may be too late for prompt recovery, and the slip may propagate up to where there is damage before the 'action' takes effect. (This is not to detract from the sense of what you said.)
One further note I was actually waiting for you to bring up : although beyond a certain medium to upper speed the engine will *normally* be safe from slipping *as long as adhesion is regular* and *as long as working output is regular* such a condition may get disrupted , if by a sudden lack of adhesion at rail or / and by a sudden surge of effort , say by gulping water and , with throttle up-stream of superheater , this could effectively increase an intentional 2/3 throttle setting into a 1/1 for a moment - both could be enough to momentarily surpass adhesion limit and although conditions may quickly return to normal the engine once having slipped will continue to do so because of the now much reduced factor of adhesion under sliding condition steel on steel .
I confess that to an extent I deprecated this point, as mere assumption on my part, because in the T1 and other locomotives of the kind we are discussing the throttle is after the superheater, so any priming carryover into the elements would not produce high immediate pressure that cannot be controlled by the throttle. What happens instead is that, with the throttle open, the steam conditions in the superheater balance both forward and backward in flow, meaning that a sudden 'surge' of priming cannot produce a great increase in MEP only in the cylinders; it must do so in a comparable volume back through the dry pipe into the steam space. The practical effect of 'that much' priming would essentially be to produce what was generally observed: a loss of effective superheat, rather than a dramatic increase in mass flow.
In Blue Peter's case, you had carryover making the superheater an uncontrolled (except by reverser setting) watertube boiler, with the added draft from the runaway exhaust providing dramatic increase in combustion gas to that 'boiler'. I don't see that situation being likely on any... I am sorely tempted to say 'correctly-designed locomotive' but we should find more neutral words ... locomotive with properly applied front-end throttle arrangement?
[quote]That should indicate clear enough , a 'laissez-faire' attitude towards a slippage at speed was at any rate stark ill advised . I think we agree on that .[/quote.
We do. Except I can't help but think that if I'd been knocked silly by the handle I might have been just as slow to react. Not sure it is fair to use 'laissez-fiare attitude' toward the poor man. One thing that would have saved the locomotive was a Franklin Precision reverse... but I digress.
>> the entire engine slips up to a critical speed (undefined, but probably within some fraction of a full revolution) << No , not *that quick* , absolutely not!.
No , not *that quick* , absolutely not!.
I meant 'up to the speed where the sliding friction conditions are thoroughly established, and the coefficient has dropped to a (reasonably) steady lower level. (And more importantly, a speed at which nothing will re-establish an effectively higher coefficient of friction, ceteris paribus, until a net reduction of speed, to say nothing of reduction in acceleration, is made...)
>> the increased rotational momentum will resist deceleration to 'matching' speed<< Momentum is fundamentally different from output since it runs on ‘what’s there’ and therefore it cannot 'resist deceleration’ or else you'd have a Perpetuum Mobile there .
Momentum is fundamentally different from output since it runs on ‘what’s there’ and therefore it cannot 'resist deceleration’ or else you'd have a Perpetuum Mobile there .
Equal and opposite reaction. If you like, think of it as rotational inertia overcoming braking. If you are familiar with what happens when you step on a treadmill equipped with a flywheel, you will recognize what I meant... I think we agree on this but are using different terminology.
However there is no need to 'decelerate to matching speed' because the balancing speed will automatically be reached as *limit of acceleration* of rpm speed
I was talking about the point where coefficient of friction begins to increase substantially toward static, which will only take place when the mutual speed of the contact patch on the tread and the loaded spot on the rail underneath is substantially minimized.
<snipped>
... the point I was about was ‘octoring’ of the cranks logically must have resulted in some weird forces to work on crank axles and coupling rods ( mind, you a cylinder *not* at dead center and lively kicking the con rod while coupling rods do go through dead center on that side … I read nothing about that having been taken into account ; also , I would have liked to have seen how that crank axle had been contoured and manufactured – probably a built-up type
I confess to getting a bit of a headache where any 135-degree system is in use and I have to start looking at fluctuations in torsional stress. I believe that to be part of what you're getting at, and yes, I can see it leading to problems. The crank in the long versions of the Alco 244 was prone to failing in torsion -- I disremember the exact cylinder-firing order that predisposed it , and ISTR there was a harmonic involved. Do you have a source for frequency/severity of crank-axle failures on the LNs? (I can ask the resident Britons over on steam_tech if not.)
I'm pretty sure that the crank was built up -- I have a suspicion the throws were only indifferently balanced themselves (their counterbalance weight being incorporated in the wheelrims, as if the whole thing were near-perfectly rigid. And that might easily be a mistake, from the standpoint you mentioned.
Repeated slips at starting still did nothing for dependable acceleration nor for high maintenance free mileage .
I concur, without question.
Me for one , I consider this as a reluctantly taken burden , imposed by some PRR high official seeking to get rid of the Duplxii and get some return , too . I have a picture in my mind of some C&O manager calling the test engineer and ask : “Now did it fail already ? Not yet ? - well then , get done with it lively , will you !”
Certainly happened a lot, just as you say. Clearly happened on Bulleid's Leader, on the Duke of Gloucester, and (according to Haas) on NYC Niagaras right at the end of their service lives*. If you have read Angus Sinclair's "Development of the Locomotive Engine" -- get your copy here if you do not have one:
http://books.google.com/books/about/Development_of_the_Locomotive_Engine.html?id=rrcpAAAAYAAJ
it is difficult to read the account of the 'Falcon' without thinking 'the fix was in' somehow.
*My father used to have a joke regarding statistics: he would point out why it's a spurious metric to note that 'most severe skiing accidents occur on the last run of the day' -- well, of COURSE if you are severely injured, it would be the last run of that day for you...So if there is intentional sabotage (in Haas' example, being done by working either excessive on inadequate cutoff) it is likely to curtail the effective service life dramatically... which was the avowed intent, and it would seem to have worked...
The notion about short /long levers or stroke is probably taken from our own handling of heavy things better by levering than by sheer strength , however this does not at all apply to engines ...
In my view, the short-stroke issue applies to engines because the whole process of admission and exhaust does not 'scale' with the decreased stroke. That is one of the reasons the Lima engineers gave for increasing the stroke of their locomotives so greatly compared to historical wisdom. (Meanwhile, I am still amazed at how smart Golsdorf was to use short stroke and tall drivers on his 2-6-4 ... for reasons more or less completely unrelated to achieving high speed.)
In particular, the short stroke hampers torque application, which makes the engine somewhat more likely to stall. This in fact was one of the complaints C&O had. (Now, designing a duplex to be a mountain engine could be done, but the T1 was rather emphatically not that locomotive, or really meant to be. (What might be fun, along the lines of comparing K4 and M1 performance, would be to see whether a T1 equipped with lower driver size ... say, 70" for amusement ... and longer stroke could have been more competitive on C&O and N&W.
But as you said, and I agree, the fix was in, the locomotives "Had To Fail" (and, in fact, had to break irrecoverably, in the case of both the T1 and Q2, if PRR were to get out of the trust agreements...) I can't say if Baldwin could successfully claim this as an excuse' when attempting to 'sell' the by-then-discredited duplex principle to skeptical railroads... ;-}
Also , I read some engineers at N&W ‘believed’ long stroke would promote lower specific steam consumption – again this is not true.
Do you have the reference for that? It seems strange to me that N&W engineers would say such a thing -- surely they knew better! So I'd think this is a matter of interpretation, in the account that you read, or perhaps of something subtle that I'd like to see in context if possible.
In any case, long stroke is more 'acceptable' if you can do it with low rod angularity ... which is one reason those Berks drove on the third coupled axle, even with the excess mass (both revolving and reciprocating!) in the mains.
The idea may have arosen from a need of keeping a relatively large clearance *distance* between piston and cylinder head for safety from banging piston against head when developing play in bearings ( excessive play , in my view )
I find this interesting because when the Q2 was tested, one of the complaints involved excessive dead space (REALLY excessive -- up to 15%). But that locomotive had roller bearings and Franklin wedges, and would never have 'excessive play in the bearings', so the dead space was due to some other cause. I cannot tell what that 'cause' might have been./.. but would SURE like to know.
[quote ... then all European SE engines with but some 26 in of piston stroke would have been ‘steam guzzlers’ – the opposite it true : not even a T1 came near to ssc efficiency of a simple DR 03 Pacific having had an optimum around 75 – 80 mph of 12.9 lb/ihp [metric] or the Chapelonized 141.E.113 two cylinder simple of but 203 psi having peaked 11.8 lb/ihp [metric] .[/quote]
[Out of curiosity: why are we using lb. with 'metric' hp? If you convert the one, convert the other too. If that wasn't done by the author of the source you used for the figures -- this applies to them, not you...]
I have a repeated issue with relatively low ssc on large engines -- Pennsylvania engines, in particular. I can only say, a bit lamely, that absolute fuel and water consumption were not as much of a design priority in the United States as they were in contemporary France...
>> I *think* this is an example of the situation I mentioned above, where the engine crew was not aware that the engine was slipping << I think he could not have had too much difficulty to distinguish between a regular “woowoosh – woowoosh – woowoosh – woowoosh .. “ and a lame “woosh – woosh – woosh – woosh with a “wrrrrrrrrrrrrrooooooohh” overlaying
I think he could not have had too much difficulty to distinguish between a regular “woowoosh – woowoosh – woowoosh – woowoosh .. “ and a lame “woosh – woosh – woosh – woosh with a “wrrrrrrrrrrrrrooooooohh” overlaying
Historical accounts report that crews often claimed to have great difficulty hearing the exhaust, and feeling the vibration, before extreme conditions had been raised. Only one of the two exhausts in the T1 would show the faster beat, and it would "usually" be the one furthest from the enclosed cab. I have been in a number of situations where sound characteristics like this are not recognized above (high) ambient noise, and even if a majority of crews could hear the runaway and correct it .. it may only take one instance where it went unrecognized to cause problems. (I believe the staybolt issue on the S2 is another example of this sort of thing... it might have taken only one bad start to pop all the bolts that were reported, and even if there was cumulative damage to the bolts in the affected area from thermal cycling, it might have taken relatively few bad starts to develop the failure conditions...)
the driver [of 18 201] described an incident at 185 km/h ... when the engine suddenly started trembling scaringly ; so he opted to bring the throttle in ( not close it fully ) and put the brakes on but gently – that made her steady herself
Unsurprisingly (for a true professional) exactly the right response. And speaking of that very thing ...
BTW -- the word 'driver' predates the invention of railways, and is certainly a more accurately descriptive term for what the person does than 'engineer'. (This even before we contrast the usual American employee with a French 'mecanicien' who surely better merited the word 'engineer' in its current professional sense!) On the other hand, 'engineer' is the typically accepted term in these United States, and in this particular sort of discussion it IS possible to confuse the operator with the driving wheel...
So the short answer is 'yes, both are right' and you would be justified in using the same argument as for the SE Mallet if there are complaints about nomenclature....
Sure – it was sine-qua-non practice with most of European steam ; as of DB / DR standard engines it was regular procedure to drop down to 60 % or slightly over – few drivers would drop to full c/o – and start with ½ of b.p. , that was 8 bar at steam chest , enlarge to 2/3 or some 10 bar of 16 bar at steam chest, never use more than ¾ even on dry rails as long as c/o was above 50 %
To me, that's good common sense too. I brought it up because there is a sizable community of 'believers' that think the One True Way to drive steam is to open the throttle/regulator ASAP and then wind back the cutoff... no matter how the engine starts to slip or buck.
>> I am not saying this was being practiced in the youtube clip,<< I think it was – at least in the first and second scene of the vid liked here : http://www.youtube.com/watch?v=ihCHzS4jT_M
I think it was – at least in the first and second scene of the vid liked here :
http://www.youtube.com/watch?v=ihCHzS4jT_M
In the first clip, we could probably guess by the characteristics of that colossally bad piston-rod blow during the forward stroke! I am not good enough to estimate it, though.
The long slip -- HAD to be the engineman couldn't hear the roar. He couldn't just have let it go like that, could he?
RME
rfpjohn-
That confirms my suspicions. Being an "oddball" was often a death sentence on large rosters, especially when shop crews had to learn a whole new set of procedures for a small class.
Thanks,
NW
This stuff is fascinating to read and every once in a while, my dull brain squeaks out a thought which I feel may be worthy of sharing!
As for other Poppet power on the Pennsy, they equipped two K4s' with two different Poppet arrangements prior to the T1 fleet. The 5399 showed extremely positive results, with radical increases in performance over conventional K4's. However there were other changes in that locomotive which may have greatly enhanced the perception of the Poppet valves superiority.
Thank you Juniatha!
Had the T-1s been built with piston valve gear, they might not have slipped as much, due to slower valve opening, and better maintenance. Am I correct in that the T-1s were the only PRR locomotives built with Poppet Valves? If so, the Pennsy may not have invested in properly maintaining them, as E7s and E8s could do the job for less . I agree, with impending layoffs, shop morale would have been low.
rfpjohn and friend611, thank you for your contributions. Three people isn't enough to keep a great thread like this going.
Watching the lead set of drivers slip on a T1, it occurred to me that this was not unlike the lead truck of a diesel. The lead truck catches the bad rail conditions, not necessarily low joints, frogs or poorly surfaced track, but factors like frost, oil or oil mixed with rainwater or dew, or my favorite, leaves. These gremlins can cause the lead truck to lose its' grip with little or no warning to the Hoghead. On diesels equipped with ammeters for each traction motor (some GE's have this information available) one can watch the lead or second axle break loose on hard pulls or even at 60mph in high throttle positions.
Perhaps the T1 would have been less likely to be plagued by lead unit slips had they reduced the available power to the lead unit. Maybe decreased cylinder diameter of the lead set ?
Hi NorthWest
Piston valve gear would have helped in so far as it would have avoided obviously rapid and severe *deterioration* of actual poppet valves and valve gear . Steam circuit having been too free and causing slippage is a common yet ill-conceived picture . Engines with free steam flow could be as sure-footed as dull mules since at slow speed the difference between the two was tolerably small and at higher to top speed range there was no alternative because the throttled engine would just not go up to these speeds or not by exerting useable power . Compare effective work range of Chapelonized 141.P four cylinder compound with 141.R of tolerably same engine and adhesion mass : the 'R' would pull much the same freight train mass up a given ramp at some 40 mph , yet only the 'little P' would rise to speeds above 70 mph with an 800 t 20 coaches passenger train - being as sure-footed as the two cylindered 'R' from a standing start . In the T1 I think it probably was erratic valve events and likely steam leakages too by malfunctioning poppet valves that caused the engines to be unpredictably performing at low speed pulling and in fact all through the speed range as errors and leaks in valve gear got more and more severe with accumulating mileage . I also expect performance to have been widely differing from engine to engine since with deterioration besetting them there could be any degree and amount of malfunctions present in any engine , except - perhaps - those fresh from overhaul , although even overhauls seem to have been of uncertain quality . I know that getting poppet valves seating tight was an issue of due professional workmanship with many poppet valve engines , even in Austria where many engines had in later years been rebuilt to or equipped with Lenz poppet valve gear . It was also one major point that had led the Reichsbahn to decide not to introduce poppet valves in spite of repeated tests with different variants in different engines , the last ones I think on a poppet valve 03 Pacific and a little 64 2-6-2 tank engine . If with dieselization anticipated and coming up fast there were maintenance problems with poppet valves adjustments I can lively imagine what 'inspired' devotion to high quality work such engines could have expected when in shop for an overhaul of valve gear ( ironic ) Ad to it some crews neither understanding nor interested and you rather wonder how the T1 did at all manage to struggle on the way they did .
Walschaerts hard to fit on a four-coupled drive unit ? Naw - why ? Return crank on drive pin , expansion link console linked to back of cylinder and off you go .. no problem - be happy .
= J =
Hi Firelock
That's about what I think .
Or to a certain degree ..
Poor track condition seems to have had an influence , too , or otherwise it would be hard to see why on the Norfolk & Western which I understand had well maintained trackwork , the T1 would have behaved *that* inconspicuously concerning slippage - even with better than average handling granted .
That PRR system of crews and engines - uhm - dispersion if you don't mind my expression seems to have been rather rough and inconsiderate . The T1 engines would desperately have wanted a specially trained brigade of crews composed of a rank of more interested or brighter minded people - if not the system of fixed double crew manning that was widely popular with top engine classes in Europe until the 1950s should have been applied . This system often gave rise to inter-crew competition of incident free , steady punctual running or making up lateness inherited from - here - K4s double header or GG-1 having brought the train . It might have been difficult to establish with very long locomotive through running , yet would not have been impossible while in my view this very long through running was perhaps bought at a higher price than it was worth as concerns increased coal consumption , demanding common user system , engine neglect et all .
Although dieselization almost followed the last batch of T1 engines on their (w)heels , it still asks for explanation why diesel park building up did not bump K4s off long distance passenger work first - very prompt decline of the T1 indicates substantial problems , as does that regretful general 'open' running of the valve gear cam box after 1947 which visually bespeaks frequent maintenance done on these components . With dieselization gaining momentum a conspicious indifference in regard to their still substantial steam park seems to have beset most US RRs - consequent squandering of financial means must have been tremendous in regards of running what traffic was there to be run by a prematurely decomposing fleet of locomotives of thereby inevitably compromised efficiency , performance and reliability , demotivated shop and engine crews and a general air of carelessness that did nothing to promote an according sharpening of procedures and methods in preparation for diesels coming up , thus likely I should think to give rise to considerable trouble with the new , much more exacting type of motive power , since one thing was for sure : that sort of sloppiness many steam locomotives did tolerate was a no-go with diesel traction .
Regards
Juniatha
Several months ago "Trains" published one of their special issues "Steam Glory 3". There was a very fine article about the Pennsy's T1 where the slipping issue was addressed. Slipping wasn't so much a function of a locomotive design flaw as much as it was poor engineer preparation. You couldn't handle a T1 like you could a K4. Engineers who were trained in the handling of the T1 didn't have a problem with it. Those who were plopped in the cab with a "here's your train, take it!" DID have a problem.
The T1 was a good locomotive. It never really had a chance. Pennsy decided to dieselize their passenger trains in 1946 so the T1 was out of a job before it really got a chance to start it. C'est la vie.
Juniatha, from my watching of the video, it seems that the engines are slightly out of phase until about 1:21 in, and by 1:23 they are shooting sparks... I can't calculate it, but that is my observation.
Overmod, regarding long wheelbases- the Russian 4-14-4- I can hear the flanges now...
Regarding duplexes/duplexii (whatever) would Walschaerts valve gear help? I know one of the issues with the T-1 was that the poppet valves faster opening contributed to slip, and with the valves being more difficult to access...although Walschaerts would be hard to fit on only two drivers...
( wording revised evening of July 3rd )
Hi , Prof Overmod
>> Why are you assuming that a Ferguson clutch works like a fluid flywheel/torque converter? It does not; it is a viscous clutch and acts more or less immediately on rapid differential.<<
I didn't - that's your own interpretation . Ooh-kay - my wording / your wording .. viscous clutch - yet not a mechanical clutch , for sure . Still , it will not react on the spot since viscous media can only transmit forces under giving , not one-to-one . So the one unit that slips , still slips – if not spinning as fast – but still way above the borderline of the very small creep slip rate that is present when exerting effort parallel to surfaces and have it transmitted by mechanical friction of steel on steel . This friction factor severely decreases at change over from adhering condition to sliding condition . So , IMHO there is not too much gained by installing that element . I'd like to see what dimensioning you'd propose for a viscous clutch to hold a low rpm high torque power output of a locomotive or half of it by 1950s technology – *g*
>> Duplex (modern sense) - two separate engines in a common frame under one boiler<<
Yep - alright , there you are , they are , we all are ! And I hope we are not *duplicating* that kind of discussion that was going on about Art Deco ?
>> an approach, like Deem's or mine, <<
Overmod , I remarked on it already and may repeat it here : I do *not* know *all* of the engineers that will have had a hand in writing the history of steam at one time or other – mostly rather earlier and of arguably lesser impact – and I'm afraid other readers may not necessarily fare significantly better . So if you don't care to put down in words what you are aiming at I'm not making up for it by spending time google-wickipeding through the endless space of the internet to try and find out . So I *deem* it wise you may want to fill that name-dropping with some life or we may as well skip it . Besides , there is more to sound engineering than just 'patching up' a design by taking an A-component and a B-type of unit from the shelf and blend it with some flavor of a C-engine with a D-head and an E-exhaust to work on an F-drive connected , sorry conjugated to G-wheels . You will have to make it all work together and that implies you will not get around *designing* proper . For an illustration , what an automobile would it turn out to be if one were to take the empty body of a 1955 Corvette , install a 1968 air-cooled flat six Porsche 2 ltr engine in the trunk , blend it with a used Rover autotrans and a recent BMW rear axle , replace the GM front axle with a VW one of say 1970 which should not be too difficult and most of all will be sure to offer but very limited improvement , last not least put high roller rims and tires of a royal Rolly to it , finish it up with a hydro-dynamic suspension from a contemporary Citroen and dress the cab with a slightly smelly , orange-brownish interiour of a '70s vintage Volvo ..?
Ok , I know that's not what you mean – yet I wished you might want to care to describe your proposals more to the point .
>> 'duplexes' is the only correct normal plural in English.<<
.. only that 'duplex' itself isn't an English word but Latin – *g* . 'Duplexes' – awrgh , that sounds just awful , sorry . In Latin , an added ending -i is often used for plural and that makes plural of Duplex --> Duplexi or even Duplexii - you see ? *gee*
I think you're a bit nitpicking , really this gets tiring – or in the words Paul McCartney once put into a song at the time the Beatles were about to break up : Let it be !
>> The mere existence of the slip is not the principal issue <<
Oops ? Now I'm taken by surprise !
>>-- in some cases, controlled micro-slipping produces better effective average TE than working entirely within the adhesion limit <<
Ahh ! But , sorry – naw ! – never mix those two !! There is *always* a degree of micro-slip present when exerting a force longitudinally to surfaces steel on steel because only by this micro-slip can the micro-roughness of surfaces interlock to transmit a force . Absolutely no slip is only present with an engine at rest or – tolerably – when rolling slowly on straight track at no power output .
Micro-slip must be totally kept apart from a visible slip which always but always implies borderline of micro-slip transmitting forces has been surpassed and we have sliding surfaces transmitting force but at a much reduced rate of friction – a transition which actually *has* happened when a slips just starts to become visible .
>> This is not what was actually observed with T1s, however. <<
It was ! - see youtube vids , one link see below . Besides , Overmod , the T1 were steam locomotives and as sure as not I betcha they behaved very much like steam locomotives . As far as I know there never was a classic reciprocating steam locomotive of standard gauge size that ever came close to doing what you propose , because masses involved were *much* to heavy and forces were *much* to small ! Full stop .
>> If you actually calculate the rotational inertia for the drive unit, and then calculate the thrust profile as exerted by the drive (even with the smaller pistons and shorter stroke of the T1 design), you will not be surprised at how rapidly the rotation accelerates in the absence of correction.<<
Still , I'd appreciate if you'd post your calculation of how such a T1 drive set had actually accelerated from , say , 15 mph to 120 mph in about 0.1 seconds ( quoting you >> beyond human reaction << ) or , ok , let's say in 1 second flat . Looking forward to it , really .
>> This is scarcely characteristic of what I have observed with modern engines<<
That is a totally different story - that's true ! ( Same with your hot rod examples – such engines may destroy themselves at the twinkle of an eye because they’re loaded to the nines with energy and with everything working as anticipated they are forced to run at their very limits ; let alone anything goes wrong : puff , there goes your money , time and effort !)
Hooww-evvver ..
we are discussing in this thread nineteen-fifties technology as it may have been applied to ‘Old Slowhand’ steam - classic steam that is , no super-critical Indianapolis Red Rocket type of steam . I don’t get rid of conceiving a certain suspicion this may for some incidents have slipped your notice ? That earlier remark of yours does not exactly help when you blamed me of comparing ‘old’ with ‘modern’ taking my proposed draft for ‘modern’ - which by default it is not .
>> Yes, there are limits on the ultimate rotational speed that can be attained, and yes, the approach to those limits is usually gradual, not 'hard.' But I do think that limit on a fully modern locomotive, with lightweight rods and precise low-mass valve gear, is likely to be high enough to induce damage<<
IMHO , a contradiction in itself : in the first part of the paragraph you assume a 'modern' ( steam ) loco to have lightweight rods - these would enable high rotational speed to be attained and sustained ; in the second part you take that high rotational speed to point to damage thereby caused .
I'm with you that careless handling – here in the way of allowing high speed slip to continue – could and surely did cause immediate damage - or as my dad used to say : "*Any* engine can be ruined with *no* problem !" So , I really do not know what's your point in that paragraph ?
>> There can be some consideration of the damage that can occur -- both to the rail and wheel tread, and to parts of the running gear -- if overspeed occurs even for a comparatively short time -- especially if the performance is frequently repeated between tire turnings.<<
Well , time must be long enough for it to -
- go to extremes concerning speed
- run at abortive rpm speed long enough to develop play by elastic or rather by plastic deformation and then quickly enlarge play , develop ruptures or have rods torn or jack-knived and finally disintegrate .
>> Lest anyone bring up the Blue Peter incident: that was a special case <<
Oops , why ? that was a *schoolbook* case of what happens if high speed spinning is allowed to continue -- and there you have all your effects on wheel tire and rail head .
>> I am not certain of this.<<
Just believe MomJune -- *gee*
>> However, the lower coefficient of sliding friction, even with adhesion weight applied, means that the effective load once a slip develops is shifted from the train resistance to only the sliding friction, machine resistance, and rotational inertia load. That implies that high-speed slipping can propagate to the point where effective static-friction conditions cannot be reestablished, and some means of physically decelerating the drive then made. <<
Well , that's correct -- and that's exactly why I called it taking chances to allow a drive unit to go on spinning and just sit there waiting for a pickup of adhesion - it may or may not have worked and if it didn't while the spinning rather revved up than cooled down , then it *was* high time for action .
One further note I was actually waiting for you to bring up : although beyond a certain medium to upper speed the engine will *normally* be safe from slipping *as long as adhesion is regular* and *as long as working output is regular* such a condition may get disrupted , by a sudden lack of adhesion at rail ; this could be enough to momentarily surpass adhesion limit and although conditions may quickly return to normal the engine once having slipped will continue to do so because of the now much reduced force transmission factor under sliding conditions steel on steel . That should indicate clear enough , a 'laissez-faire' attitude towards a slippage at speed was at any rate stark ill advised . I think we agree on that .
>> the entire engine slips up to a critical speed (undefined, but probably within some fraction of a full revolution) <<
No , not *that quick* , absolutely not !.
>> the increased rotational momentum will resist deceleration to 'matching' speed<<
Momentum is fundamentally different from power output since it runs on ‘what’s there’ and therefore it cannot 'resist deceleration’ or else you'd have a Perpetuum Mobile here . Under condition of output brought down or made nil , mass inertia will define the rate of negative acceleration resultant from relation between adhesion ( sliding ) acting as a braking force and kinetic energy which but logically is higher with larger rotating mass than with smaller mass . However there is no need to 'decelerate to matching speed' because the balancing speed will automatically be reached as *limit of acceleration* of rpm speed which cannot be surpassed .
>> There are, in fact, known places that could become slip inducers -- crossover frogs, for example <<
Sure - 'the known suspects'
>> That is the point of the 45-degree phasing between the two engines.<<
Yes - however that's not 'quartering' if you come to think of it - it would be 'octoring' or you name it !
>> But it does not reflect the torque peaks of the same engines if running at very high speed and short (poppet-valve-enabled) cutoff.<<
Variations of torque tend to become pretty small at wheel rim because of the increasing influence of kinetic energy of the revolving masses - same also acts as a certain protection against sensitivity to sudden low adhesion spots because before engine unit can even start to accelerate adhesion is back to normalcy again . However , actual adhesion of wheels traveling at speed is clearly lower than at starting and that again may cause a borderline adhesion situation in a very powerful locomotive . With modern electrics this is regularly the case which is why they depend on very effective wheel slip control and very careful design to squeeze out all there can be obtained of available adhesion - a degree of effective power engineering that is light years away from vintage steam design which basically left engines to keep or loose sure-footedness .
>> There the torque peaks will be relatively high and sharp, and the 'smoothing' contribution more significant. <<
Not really so , because of the smoothening effect of the rotational masses – there would be significant torque peaks if using very short cut-off at slow speeds . That is what David Wardale must have encountered with his using full throttle all the time and adjusting power by cut-off . A certain medium cut-off is however rather smoothening torque profile over wheel turn , that is why with the 52 class I liked to use some 50 .. 40 % c/o right after getting the engine started and open throttle wider and then to 1/1 while gaining speed ( as long as the driver(s) would tolerate it because when wondering why I could open up fully without engine sounding as loud as they expected at full throttle they noticed my short c/o and brought settings back to the usual ‘a bit of everything’ running .
>> If that story is not related to the grate configuration, which is what I thought was the principal problem with the Lord Nelsons operationally, I would be very interested to hear it, at whatever time is appropriate. <<
Grate problems notwithstanding , the point I was about , was ‘octoring’ of the cranks logically must have resulted in some weird forces to work on crank axles and coupling rods ( mind, you have two cylinders *not* at dead center and lively kicking con rods while coupling rods *do* go through dead center … I read nothing about these things having been taken into account ; also , I would have liked to have seen how that crank axle had been contoured and manufactured – probably a built-up type
>> FORTUNATELY I can achieve eight peaks per revolution without going to fancy angles; my conjugated duplex is just two 90-degree quartered engines phased 45 degrees apart. <<
Yes – I see what you mean – only I’m afraid with your visco-clutch your 45 degrees won’t last too long and how would you propose to re-establish it ?
Still if you continue with that ‘octopushing' action I’m inclined to take the opportunity to post one little joke on Her Majesties best reknown secret (? Secret with all those explosions ?) agent : 007 disarmed in Octopussy’s eight arms - *g*
>> the 'real' problems with T1 slipping were not low-speed, but high-speed slips <<
The effect of high speed slips could be dramatically more severe – especially if not detected within a reasonable time of reaction , I agree . Repeated slips at starting still did nothing for dependable acceleration nor for high maintenance free mileage .
>> An argument could be made that problems with the T1s slipping in general service could be associated with consequences of the short stroke as well as the duplex arrangement. That specific factor was mentioned in the results of the C&O testing, for instance. <<
First of all , in my view dig ‘C&O testing’ . Me for one , I consider this as a reluctantly taken burden , imposed by some PRR high official seeking to get rid of the Duplexii and get some return , too . I have a picture in my mind of some C&O manager calling the test engineer to ask : “Say , did it fail already ? Not yet ? - well then , get done with it lively , will you !” There is more than one question arising with some very peculiar , if not downright *** engine behavior on these ‘replacement runs’ as I prefer to call them .
The notion about short / long levers or stroke is probably taken from our own preference of handling heavy things better by levering than by sheer strength . However this does not at all apply to engines – why ? Simply because we as human beings have but the same limited body power for anything we want to do and we all know by ourselves we have better control of muscles when we allow for a long levering to accomplish some heavy work , i.e. rather using a 'long stroke' at but medium effort than a 'short stroke' at full effort . So many students in practicum have made that experience when trying to turn a short levered wrench on a frozen bolt .´, pulling at full effort until all of a sudden – ZING--! bling-ga-ling the tool went skyrocketing about the shop while the guy fell on his back like a disabled bug . Note : rather take a longer lever ( and still watch what you’re doing and what the back of your hand might be hitting , just in case ) However , that did not apply to a steam locomotive because the cylinders were *by default* so designed as to put up just the desired piston thrust fitting for the applied stroke-to-wheel-diameter ratio . Also , with reasonable maintenance , fear of snapping con rod off pin should have been tolerably small – *gee*. So , provided all the drive is slack-free or nearly so and in sound mechanical condition there was no difference between short or long stroke in resultant effect nor in efficiency of piston thrust on rod system to turn wheels . I read some engineers at N&W ‘believed’ long stroke would promote lower specific steam consumption – again this was not true . It was expansion rate that made the difference and that rate was measured in percentage , not absolute length of stroke . The idea may have arosen from a need of keeping a relatively large clearance *distance* between piston and cylinder head for safety from banging piston against head when developing play in bearings ( excessive play , in my view ) – in that case an engine of longer stroke / smaller cylinder diameter had an advantage over one of the same cylinder volume but by shorter stroke , larger diameter . That , however , meant it really was smaller clearance volume which saved steam and that’s only logic . If shorter stroke ( by American mainline standards ) would have been detrimental to ssc , then all European SE engines with but some 26 in of piston stroke would have been ‘steam guzzlers’ – the opposite was true : not even a T1 came near to ssc efficiency of a simple DR 03 Pacific having had an optimum around 75 – 80 mph of 12.9 lb/ihp [metric] or the Chapelonized 141.E.113 two cylinder simple of but 203 psi having peaked 11.8 lb/ihp [metric] or the Britannia class Pacific or CSD 556 séries light Decapod at about 12.5 - 13.0 lb/ihp [metric] .
>> I *think* this is an example of the situation I mentioned above, where the engine crew was not aware that the engine was slipping <<
I think he could not have had too much difficulty to distinguish between a regular “woowoosh – woowoosh – woowoosh – woowoosh .. “ and a lame “woosh – woosh – woosh – woosh with a “wrrrrrrrrrrrrrooooooohh” overlaying .
>> -- at that speed there would not be any appreciable grinding or augment sensation, and the relatively quiet exhaust might not have been enough of an alert <<
Oh-yeah , it seems there *is* a noticeable sensation . I never had a chance to experience it myself , however as early as back at the Vienna 150 years celebration of Austrian Railways I remember the Saxon driver of 18 201 ( the 7ft 7in drivered DR one-of-a kind Pacific ) talk with one knowledgable person in the cab of that loco when they approached test runs on the Velim test ring of the CSD ( back then of course as unknown to me as the Sea of Tranquility – it helps the matter was later briefly touched in a German rail mag ) the driver described an incident at 185 km/h ( 115 mph ) when the engine suddenly started trembling scaringly ; so he opted to bring the throttle in ( not close it fully ) and put the brakes on but gently – that made her steady herself . Intensive inspection afterwards brought evidence the engine had turned into a high speed slippage and was supposed to have revved up to over 200 km/h ( 125 mph ) for a short moment , mass inertia and vibration having caused cracks in spokes . I was but a child of eleven back then in 1987 and the only reason I still remember it at all is just because I was listening eyes wide open to the white haired driver with this enlightened good-natured grand-pa air about his round face , sitting placidly at his bay in the clean cab which smelled slightly of fuel oil , fresh varnish and lubricants , smiling like Santa Claus himself while telling the story in his idiosyncratic Saxon diction as calmly as if speaking of a tour with a stopper train .
However – sure – whatever that may say about the reaction of a vastly more massive T1 in a comparable situation I wouldn’t want to guess .
>> cutoff relatively long and controls the engine acceleration with the throttle. This technique has been demonstrated to prevent slipping in a wide range of contexts.<<
Sure – it was sine-qua-non practice with most of European steam ; as of DB / DR standard engines it was regular procedure to drop down to 60 % or slightly over – few drivers would drop to full c/o – and start with ½ of b.p. , i.e. 8 bar of 16 , at steam chest , enlarge to 2/3 or some 10 bar at steam chest , never use more than ¾ even on dry rails as long as c/o was above 50 % .
>> I am not saying this was being practiced in the youtube clip,<<
I think it was – at least in the first and second scene of the vid linked here :
NorthWestAbout the 4-10-6, that would be pretty long for many railroads. I can see why they didn't sell a "Curve Straightener Class" to anyone.
With the 66" drivers it would not be all *that* bad -- shorter effective rigid wheelbase than a Q2, and perhaps (with the pivot for the six-wheel truck in the right geometrical location, and some care in the placement of its lateral-motion devices) relatively low effect on ability to negotiate curves. 2-10-6 would likely be better...
bigduke76Lima had proposals for a 2-8-6 and a 4-10-6 which didn't interest anyone
About the 4-10-6, that would be pretty long for many railroads. I can see why they didn't sell a "Curve Straightener Class" to anyone.
Thanks for the video! (Link Activated)
JuniathaThe first and second scene , engine is being handled very cautiously , very slow acceleration - no slip ; third scene starting at 1:12 acceleration looks indifferent , both sets are fine and close to in phase working until *first* set gains some advance and then at 1:20 starts to spin wheels and continuing - driver obviously *not* closing throttle - at some 50 mph plus x until settled by a pick up of adhesion . This sort of allowing the unit to be caught by normalized adhesion is very rough handling of the locomotive because pick up of adhesion , like a slip , will not work on all wheels the same time but effectively again on but the leading set first which causes coupling rods having to transmit that stopping action to second coupled axle against continuing torque input by piston thrusts ; since likely the pick up of adhesion is not symmetric neither , this also causes high torsion moment on axle from side picking up to other side not picking up , since these forces in contrast to steam forces by pistons do not have a dead center but continue all around turn of wheels - since they are acting on drive from wheel tire , not pin in wheel - they may or not cause an overload to coupling rod roller bearings as this side's rod goes through dead center ( it is true the other side holds axles in phase , yet this is only via torsion of axles - forces that can easily reach destructive levels .
I better understand the T-1's wheelslip problems now. Thank you.
On the man on the right side of the cab, I understand now, thanks.
Glad we finally got the turbine designation cleared up...
Arturo, I am unaware of a 2-8-6 from Lima -- can you provide source material?
The proposal I am aware of is a 4-8-6. It is documented in Eric Hirsimaki's "Lima, the History" (pp.225-229 in the 1986 edition), with the following statistics:
Driver size 70 inchescylinders 28” x 32”Weight on drivers 280,000 lbsWeight on engine truck 95,000 lbsWeight on Trailer truck: Front & intermediate axles 90,000 lbs Rear axle 60,000 lbsTotal weight of engine 525,000 lbsWeight of Tender (4-10-0 pedestal; 2/3 load) 337,000 lbsTractive effort (long compression) 74,600 lbsTractive effort (with booster) 87,000 lbsFactor of adhesion 3.75Grate area 132.9 sq. ft.Combustion chamber length 108” (Double Belpaire)
Information from an unconfirmed source indicates that the "ten-coupled" variant would have had lower drivers (the figure given was 66"), but with modern balancing techniques, and the gear-driven type B relieving any contribution of a radial valve drive, it should have been capable of reasonable freight speed before augment became critical (almost certainly no worse than a NKP Berk produced).
I have long suspected that it would be comparatively simpler to fit a 2-10-6 arrangement under the same general boiler size than to lengthen the eight-coupled design for the added driving axle. Could this be part of what you have observed?
Quote >> So , only the first engine would start to slip? Could an independent throttle for each engine help? <<
The leading one usually started the slip - but didn't have to remain the only one ..
It was believed the issue '1st to slip first' was because of mass transfer to the rear under the influence of tractive effort - the picture being that of an old Buick sitting down on its rear suspension when the *driver* (sic!) steps on the gas . That wasn't the problem . It's just the leading engine unit that would hit a low adhesion spot first and so it was first to slip .See for instance
"T1" 4-4-4-4 Duplex Steam Locomotives - 1940's Pennsylvania Railroad
The first and second scene , engine is being handled very cautiously , very slow acceleration - no slip ; third scene starting at 1:12 acceleration looks indifferent , both sets are fine and close to in phase working until *first* set gains some advance and then at 1:20 starts to spin wheels and continuing - driver obviously *not* closing throttle - at some 50 mph plus x until settled by a pick up of adhesion . This sort of allowing the unit to be caught by normalized adhesion is very rough handling of the locomotive because pick up of adhesion , like a slip , will not work on all wheels the same time but effectively again on but the leading set first which causes coupling rods having to transmit that stopping action to second coupled axle against continuing torque input by piston thrusts ; since likely the pick up of adhesion is not symmetric neither , this also causes high torsion moment on axle from side picking up to other side not picking up , since these forces in contrast to steam forces by pistons do not have a dead center but continue all around turn of wheels - since they are acting on drive from wheel tire , not pin in wheel - they may or not cause an overload to coupling rod roller bearings as this side's rod goes through dead center ( it is true the other side holds axles in phase , yet this is only via torsion of axles - forces that can easily reach destructive levels . This is why such kind of relentless handling was strictly banned in European steam traction a driver violating this rule , especially if causing damage to the engine could easily have found himself back on the 'left side' - i.e. swinging the shovel instead of the throttle .
Separate throttles and handles : It could help - that's why I included individual throttles in my proposal , with levers bundled so they can normally actuated by one hand to vary the degree of opening only in cases where adhesion is uncertain or sensitive .
>> I have a request- can you call them engineers, rather than drivers? That part confused me the first time I read it <<
Sorry , no , I won't - I know they are traditionally called engineers , yet properly speaking they're not - unless a railroad would ask their *drivers* to pass diploma ( or at least bachelor nowadays ) for mechanical engineering . I'm not disrespecting their often considerable amount of acquired practical knowledge - the like , how to get an unwilling diesel to shape up and resume work out on the road in the middle of a thunderstorm and what have you of incidents challenging your wits and stamina . And I wouldn't want to call them 'Lokführer' since that would translate as 'leader' - if 'locomotive leader' - the one 'leader' attempting to lead all Germany from 1933 to '45 certainly didn't even have a driver's license -*g* . How about the French 'mécanicien' - no , it’s not 'chauffeur' , that's the one who keeps the fire 'chaud' ( hot ) ? Well , I think I'll stay with the British expression 'driver' because that's what they really do .
.. 'hogger' - uhm , yes .. no , not really an alternative , either . ( what about ‘hugger ? *g* )
>> I got a good laugh out of the story, thanks for sharing.<<
.. although I'll delete it now since it doesn't really relate to steam we haven't seen and was just one of my spinning up and down the alley once in a while . It had been taxing sometimes to get these cab rides because I always had to ask for two - Michael, a classmate back then and me, or my dad and me . This one was with Michael who wasn't really a steam fan as such - he sort of appreciated when seeing one , yet he didn't miss anything when not . I think he was just enjoying an opportunity to see a little bit of Poland comfortably by accompanying us in dad's Merc . and I'll stop that , too .
Steam-Turbine-Electric or 'STE' : Oh , sure it is steam - if certainly not classic steam as I want to focus on .
Only , if they burn pulverized coal not in a boiler but combustion chambers directly feeding to a turbine - it's not .
Alright ?
firelock76. somewhere in my vast, neglected (unmounted) slide collection i have a slide taken in Oct. 1956 outside the Albuquerque shops of an ATSF 4-8-4 set aside midway through the application of an all welded boiler. also as of 1949 Lima had proposals for a 2-8=6 and a 4-10-6 which didn't interest anyone. -arturo
Juniatha And I think we might agree that Duplexii are only Duplexii if *not* coupled inside / outside / by rods or gears , let alone semi-coupled with a fluid resistance clutch which does by default *not* prevent slipping of one unit because it can only built up torque when there is a difference in rotational speed - that is when the slip has already happened.
And I think we might agree that Duplexii are only Duplexii if *not* coupled inside / outside / by rods or gears , let alone semi-coupled with a fluid resistance clutch which does by default *not* prevent slipping of one unit because it can only built up torque when there is a difference in rotational speed - that is when the slip has already happened.
Why are you assuming that a Ferguson clutch works like a fluid flywheel/torque converter? It does not; it is a viscous clutch and acts more or less immediately on rapid differential.
Anyway, so we are all singing happily on the same page, I propose we use this nomenclature:
Duplex (modern sense) - two separate engines in a common frame under one boiler
Conjugated duplex - an approach, like Deem's or mine, where the two engines are connected mechanically in some way. (I will explain later why rigid conjugation is not a good idea whatever method is used)
Withuhn conjugated duplex - the design with inside 'conjugating rods' best recognized from the ACE patent.
(I will mention in passing that explicitly according to Fowler, who is still the accepted authority on these things, 'duplexes' is the only correct normal plural in English.)
The mere existence of the slip is not the principal issue -- in some cases, controlled microslipping produces better effective average TE than working entirely within the adhesion limit, but that is a digression here. The critical issue is to avoid either propagation of the slip to higher speed, or the continued duration of the slip. I'll take this up again in a moment.
In any case, there are better ways to implement 'traction control' on steam locomotives if your object is just to work the locomotive right to within a few lb. of the adhesion limit -- some of these methods could actually be implemented with 1940s control technology. (I have described my approach elsewhere.)
One word on slipping : no matter how ( hopefully ) free the steam circuit may be , what prevents a slip to speed up as rapidly as to be beyond human reaction is the enormous rotational mass involved in a steam locomotive's drive unit
This is not what was actually observed with T1s, however. If you actually calculate the rotational inertia for the drive unit, and then calculate the thrust profile as exerted by the drive (even with the smaller pistons and shorter stroke of the T1 design), you will not be surprised at how rapidly the rotation accelerates in the absence of correction.
The relation mass to torque is much less favorable than in an old 1955 Chevy V8 engine - and that certainly didn't rev up as instantly as to beyond human recognition , even if you would step on the clutch at gas floored - it would go like "Woooooooooouuuuuuuuuaaaaaaaeeeeeeiiiiiiiiiiih - and then if you still don't lift your foot it would start having valve flutter and that would arguably give it another lease of life until it may or may not disintegrate for reciprocating mass inertia.
This is scarcely characteristic of what I have observed with modern engines. A particular example can be observed in 'truck pulls' with highly-modded light diesel motors (Duramax, 6BT, Power cerebrovascular accident) where the driveline breaks with the truck producing high output. The engine will very promptly rev beyond the ability of the usual sorts of limiter (which act more quickly than a human response) to control it. The result can be ... interesting.
Yes, there are limits on the ultimate rotational speed that can be attained, and yes, the approach to those limits is usually gradual, not 'hard.' But I do think that limit on a fully modern locomotive, with lightweight rods and precise low-mass valve gear, is likely to be high enough to induce damage.
<snipped> ... if you boil down to it there is no chance for it to go overrevving - well at least not if you move to close the throttle tolerably promptly .
There can be some consideration of the damage that can occur -- both to the rail and wheeltread, and to parts of the running gear -- if overspeed occurs even for a comparatively short time -- especially if the performance is frequently repeated between tire turnings. (Juniatha mentions this in a different sense a bit later in the previous post)
Lest anyone bring up the Blue Peter incident: that was a special case, and could indeed have been arrested short of likely 'damage' even though closing the throttle was no help, simply by winding the reverser to mid. Had the reverser handle not started to unwind, and clocked the driver hard, it is likely that the situation could have been addressed as indicated, within the range of reaction time.
... at some more or less high rpm , torque is being so much abased it cannot overcome even a sensibly reduced slip adhesion factor (actual rail adhesion) present with gliding of steel on steel.
I am not certain of this. It is common sense that on sound rail, the point at which an engine loses the ability to 'slip' will be reached before the same power 'balances' in acceleration. However, the lower coefficient of sliding friction, even with adhesion weight applied, means that the effective load once a slip develops is shifted from the train resistance to only the sliding friction, machine resistance, and rotational inertia load. That implies that high-speed slipping can propagate to the point where effective static-friction conditions cannot be reestablished, and some means of physically decelerating the drive then made.
... that's why at regular rail conditions any ( reasonably designed ) locomotive when accelerating will pass a speed beyond which it cannot be made to slip (in regular adhesion conditions , I repeat !!) This was around 80 km/h for an 01 class two cylinder standard Pacific and it may have been around 80 mph with a T1 in reasonably sound mechanical condition.
It is difficult to arrive at a 'neutral' value for the T1, because as things turned out it was easy to exceed 'regular adhesion conditions' comparatively easy and often. I believe you could fairly approximate the 'ideal' speed by modeling an equivalent 4-8-4 with comparable masses and piston thrust, an on average that *should* apply to a proper conjugated duplex.
It is a different story if wheels hit a slip spot . Since this is an *irregular* condition , no prefixed definition can be given as to what size or what reduction of adhesion a 'slip spot' will present.
This is true, but the problem is that if the slip spot conditions are such that the entire engine slips up to a critical speed (undefined, but probably within some fraction of a full revolution) all the wheeltreads will be in the full sliding state, with the lower effective resistance to piston thrust meaning the system is rapidly diverging from its initial state. Even after the correction to thrust will have been made, the increased rotational momentum will resist deceleration to 'matching' speed (or at least a speed at which adhesion weight will re-establish proper frictional conditions.
There are, in fact, known places that could become slip inducers -- crossover frogs, for example. That does not detract from the point being made -- as even if a crew knows where the problem is, they have to compensate for it over a much longer distance than just that where the drivers are passing over the defect.
... the note on high torque peaks being smoothed out by 'quartering the units to each other' is questionable because each the units still features the same amount of variations of torque over wheel rotation and if they are so phased that one is low when the other his high they will but tend to slip in separate instances , yet slip just the same.
This would be fully true if the engines were not conjugated. Because (in my design) they are, even though the slip propensity is the same as in the regular duplex at the start, the harmful effects of a propagated or sustained slip are mostly minimized, while maintaining the advantages of the duplex-drive arrangement when the engines are rotating in general sync.
However , 'quartering' exactly does *not* obtain right that effect : because each a two cylinder SE unit with cranks at 90 degrees will have four high points and four low ones , 'quartering' between front and rear *again* has both units run on high / low peak torque at the same time - if by different cylinders -*g*.
That is the point of the 45-degree phasing between the two engines. There is a valuable diagram in Wardale's Red Devil book that shows the torque at the wheelrim of a 2-cylinder simple in 15-degree increments, taking into account factors such as expansion at tested cutoff. The result showed less torque peakiness than I expected, and this would characterize most such engines, including most probably the two engines of a high-speed duplex. But it does not reflect the torque peaks of the same engines if running at very high speed and short (poppet-valve-enabled) cutoff. There the torque peaks will be relatively high and sharp, and the 'smoothing' contribution more significant. (I have not yet calculated the effective wheelrim torque peak for the conjugated duplex as acting through the viscous clutch, at very short running cutoff and high effective steamchest pressure, but that is the result of sheer laziness <vbg>)
That is why the venerable Southern Railway in GB made a venture into 135 / 45 degrees crank settings with their Lord Nelson four cylinder SE 4-6-0 (it was not successful because they had not taken into account ... but that's another story.)
If that story is not related to the grate configuration, which is what I thought was the principal problem with the Lord Nelsons operationally, I would be very interested to hear it, at whatever time is appropriate.
FORTUNATELY I can achieve eight peaks per revolution without going to fancy angles; my conjugated duplex is just two 90-degree quartered engines phased 45 degrees apart...
However , there are youtube vids showing T1 engines marching past the viewer , accelerating an express and hitting a low adhesion spot : you can see that the lead engine rather lazily , in one case at first 'reluctantly' ( almost regaining steady pulling again ) start to spin - waaay far from any alleged 'instantaneous beyond human reaction' .
Yes, but the 'real' problems with T1 slipping were not low-speed, but high-speed slips, where the throttle was wide open and the enginemen not expecting a slip to take place, let alone propagate. Recovery from such a situation was remarkably difficult for a number of reasons, including the lack of separate throttles for the engines or other apparatus that would act directly to slow an overspinning engine *to a rotational speed corresponding to the other engine* in a reasonable period of time.
An argument could be made that problems with the T1s slipping in general service could be associated with consequences of the short stroke as well as the duplex arrangement. That specific factor was mentioned in the results of the C&O testing, for instance.
However , what can also be seen is that the driver in at least two cases does not react at all but allows the front unit to go spinning along ( at about - my figuring within unpredictabilities of youtube vids - some 40 - 50 mph spinning speed ) until or not it will settle down again ...
I *think* this is an example of the situation I mentioned above, where the engine crew was not aware that the engine was slipping -- at that speed there would not be any appreciable grinding or augment sensation, and the relatively quiet exhaust might not have been enough of an alert. Sometimes too much refinement can turn out to be a BAD thing operationally.
This may also be a situation where the engineman, contrary to the 'usual' driving idea of two or three chuffs and then drive on the reverser, keeps the cutoff relatively long and controls the engine acceleration with the throttle. This technique has been demonstrated to prevent slipping in a wide range of contexts. I am not saying this was being practiced in the youtube clip, but it is possible that it was an approach that some enginemen might use to get maximal available acceleration out of the locomotive, particularly if they were aware of the slipping characteristics.
What a great story this is! (I can't help but think what a great story it would have been in the old DPM days of Trains Magazine -- in fact I think it would be a great story (with some suitable pictures) in the current Trains. (Be good as a story in a book, too, but that's a whole 'nother thing to wish for... ;-} )
Hello Juniatha!
Just a couple clarifications:
JuniathaHowever , there are youtube vids showing T1 engines marching past the viewer , accelerating an express and hitting a low adhesion spot : you can see that the lead engine rather lazily , in one case at first 'reluctantly' ( almost regaining steady pulling again ) start to spin - waaay far from any alleged 'instantaneous beyond human reaction' . However , what can also be seen is that the driver in at least two cases does not react at all but allows the front unit to go spinning along ( at about - my figuring within unpredictabilities of youtube vids - some 40 - 50 mph spinning speed ) until or not it will settle down again -
So, only the first engine would start to slip? Could an independent throttle for each engine help?
I can imagine how often new drivers were needed with that much slipping...
I have a request- can you call them engineers, rather than drivers? That part confused me the first time I read it.
I got a good laugh out of the story, thanks for sharing.
On clarification of the drive, I wasn't sure if a steam turbine-electric counted as "steam", but I got a good laugh out of your second post, too.
(P.S. 13 pages and 187 posts... must be a record for longest average post! Thank you all for wearing out your keyboards helping me learn more about steam.)
Uhm , if they have steam coming up out of their eclectic dive - that's probably Medusa having missed the Hellenian Sea by a long way ..
Never mind
And I think we might agree that Duplexii are only Duplexii if *not* coupled inside / outside / by rods or gears , let alone semi-coupled with a fluid resistance clutch which does by default *not* prevent slipping of one unit because it can only built up torque when there is a difference in rotational speed - that is when the slip has already happened .
One word on slipping : no matter how ( hopefully ) free the steam circuit may be , what prevents a slip to speed up as rapidly as to be beyond human reaction is the enormous rotational mass involved in a steam locomotive's drive unit . The relation mass to torque is much less favorable than in an old 1955 Chevy V8 engine - and that certainly didn't rev up as instantly as to beyond human recognition , even if you would step on the clutch at gas floored - it would go like "Woooooooooouuuuuuuuuaaaaaaaeeeeeeiiiiiiiiiiih - and then if you still don't lift your foot it would start having valve flutter and that would arguably give it another lease of life until it may or may not disintegrate for reciprocating mass inertia .
A steam locomotive will do much more slowly - and I have witnessed some tremendous slipping in the cab of 52 class engines which can produce a tractive effort *waayy* above what adhesion can handle because their (US) factor of adhesion is pretty low - i .e. cylinder maximum torque is relatively high in comparison with adhesion mass . It sounds alarming when in the middle of a steady heavy pulling the engine looses feet and goes thundering away with footplate trembling - yet if you boil down to it there is no chance for it to go overrevving - well at least not if you move to close the throttle tolerably promptly .
As concerns free steam circuit , this is relative , i.e. at some more or less high rpm , torque is being so much abased it cannot overcome even a sensibly reduced slip adhesion factor (actual rail adhesion) present with gliding of steel on steel . This does *not* mean in regular running the locomotive could not get above that slip balancing speed ( there is no definite one because of widely varying slip adhesion factor ) - only it does so on exerting a tractive effort *below* the one necessary to pull wheels through against adhesion on rails - that's why at regular rail conditions any ( reasonably designed ) locomotive when accelerating will pass a speed beyond which it cannot be made to slip ( in regular adhesion conditions , I repeat !! ) This was around 80 km/h for an 01 class two cylinder standard Pacific and it may have been around 80 mph with a T1 in reasonably sound mechanical condition . It is a different story if wheels hit a slip spot . Since this is an *irregular* condition , no prefixed definition can be given as to what size or what reduction of adhesion a 'slip spot' will present .
Overmods note at a one-wheel-size slip spot causing much more severe reduction of total adhesion in each an ( independent ) four coupled drive unit as compared to one eight coupled holds true . However the note on high torque peaks being smoothened out by 'quartering the units to each other' is questionable because each the units still features the same amount of variations of torque over wheel rotation and if they are so phased that one is low when the other his high they will but tend to slip in separate instances , yet slip just the same . However , 'quartering' exactly does *not* obtain right that effect : because each a two cylinder SE unit with cranks at 90 degrees will have four high points and four low ones , 'quartering' between front and rear *again* has both units run on high / low peak torque at the same time - if by different cylinders -*g* . That is why the venerable Southern Railway in GB made a venture into 135 / 45 degrees crank settings with their Lord Nelson four cylinder SE 4-6-0 ( it was not successful because they had not taken into account ... but that's another story )
However , there are youtube vids showing T1 engines marching past the viewer , accelerating an express and hitting a low adhesion spot : you can see that the lead engine rather lazily , in one case at first 'reluctantly' ( almost regaining steady pulling again ) start to spin - waaay far from any alleged 'instantaneous beyond human reaction' . However , what can also be seen is that the driver in at least two cases does not react at all but allows the front unit to go spinning along ( at about - my figuring within unpredictabilities of youtube vids - some 40 - 50 mph spinning speed ) until or not it will settle down again - this sort of running , if it was likewise representative of sluggish reaction to a slip at *upper* speed range , must surely have been asking for over-revving to happen or not - just on what incidentally happened first : pick up of adhesion or injuring of the ( rather delicately built ) valve gear ( at a traveling speed of some 80 mph spinning for instance to a speed in the vicinity of 120 mph ) . If that was a condition any T1 would haphazardly have had to suffer , say , a couple of times a trip at varying degrees of severeness , then it was a sure bet on how long an engine could stand it until it had to suffer premature and abnormal wear or downright wrecking of parts .
( note on 52 running - deleted because not to the topic - = J = )
JuniathaIt generates steam and it uses steam in a traction power generating engine unit .
Yes, but I wasn't sure if they had to be mechanical drive only or electrical drive. Thanks for clearing that up.
Overmod- thank you very much. I understood that, and thank you. Better emission control is one of the issue with steam I feel like would have needed to be developed in the '50s.
Juniatha, what I found more shocking about the UP turbine was not its appearance but that it was missing body panels...
Uhm - NorthWest - there is a unmistakable sign of what qualifies for steam ( steam locomotive , that is )
It generates steam and it uses steam in a traction power generating engine unit .
Since most of the description of the photo has been reproduced here , let's have the full text ,
to quote :
>> Union Pacific Railroad coal turbine 80 or 8080, depending on where you look, at Council Bluffs, Iowa on October 31, 1965, Kodachrome by Dick Rumbolz, Chuck Zeiler collection. Originally numbered 80 and 80B, the locomotive was renumbered 8080 and 8080B in April 1964 to avoid conflict with newly ordered DD35's.
The cab unit was former UP PA1 607 built in January 1947 (c/n 76311), retired in March 1961. It was rebuilt, retaining its Model 244-16 prime mover, adding 2,000 horsepower to the locomotive consist. The B-unit was former Great Northern Railway class W-1 electric locomotive built by GE in June 1947 (c/n 28448), retired by the GN in August 1956, sold to the UP as scrap in 1959. Between 1959 and 1962, the UP designed and built a coal-fired turbine, using a modified turbine power plant from a retired 50-75 series turbine locomotive, producing 5,000 horsepower. The coal tender was from Challenger (4-6-6-4) 3990. The completed 7,000 horsepower locomotive entered service in October 1962, and tested until May 1964, making its last run on May 12. It was deemed a failure, due to turbine blade erosion and soot build-up caused by the coal fuel. It ran less than 10,000 miles before being struck from the roster on March 15, 1968. The cab unit was traded to GE, the B-unit and tender were scrapped by the UP at Omaha. <<
.. and about that note >> Does anyone really care if the result looks like something out of a Mad Max movie? << just read the last sentence - that's usually the quintessence of making up a test bed in this 'cheap thrill' patch-up way .
NorthWestNext topic: Emission control. I've heard a little bit about "overfire jets", designed to reduce smoke. How do these work, and could there be more practical smoke control with '50s tech?
The usual arrangement is to provide ports in the firebox walls. High-speed jets of steam or air are introduced through nozzles at the center of these ports, and this is intended to induce airflow through the ports, somewhat after the fashion of the action in a locomotive front end.
The guns can use either compressed (brake) air or steam. There are of course advantages and disadvantages for each approach. Some of the disadvantage of using air can be reduced if some method of preheat can be applied to the secondary airflow -- but this can be difficult to accomplish in the limited space between the outer firebox sheets and the permissible loading-gage clearances.
A significant problem with either system is the noise from the jets. The large can-shaped objects are actually mufflers to cut the jet noise as much as possible -- which in practice might not be enough!
The devices were much more effective at reducing smoke with the locomotive standing or operating at low speed or intermittent operation. One principal use was to reduce smoke in urban environments, or where large numbers of locomotives standing would cause a smoke nuisance.
Here is my contribution, more or less 'off the cuff', on using this sort of system for pollution reduction. Other opinions may differ, and should be given due consideration when they are presented.
I think you are correct in thinking that proper injection of near-stoichiometric secondary air on a working locomotive using this system would be difficult with '50s technology. The situation is complicated because some method of introducing secondary air to the 'inner' portions of the combustion-gas plume needs to be made, and turbulent mixing is not always sufficient to assure this, particularly if the introduced secondary air chills part of a relatively fast-moving part of the plume below transition temperature.
To oversimlify things vastly -- excess secondary air is undesirable for a number of reasons, including that it consists of approximately 78% nitrogen which does not participate gainfully in combustion reactions but must be heated with energy liberated during the combustion process. EITHER insufficient oxygen in part of the flame plume OR dropping of the temperature below the effective transition temperature before all the 'fuel' has reacted with oxygen will result in unburned fuel, which when visible is 'hazing' or smoke. Balancing the situation between too much air and not enough oxygen, when the draft and firing conditions may change fairly dramatically in a short time during normal train handling, is not trivial
If you can accomplish primary and secondary air induction without active means, you'd be ahead, of course, which is why quite a bit of thinking and research have gone into where and how secondary air (as opposed to primary air controlled by dampers) can be introduced. One interesting idea is to introduce combustion air -- preheated, if possible -- at or near the throat, or under the arch, and using damper control if necessary rather than active injection. (This is also the logical place to introduce some portion of recirculated combustion gas (it is called FGR for internal combustion, but it's essentially the same principle as the EGR you're familiar with from automobile emission practice. I won't discuss that here, but it is a technique involved with much clean-coal technology, and should be discussed... in a proper place, probably in a new modern-steam thread). To my knowledge, no system that attempted to do this was successful in the '40s or '50s, probably because no suitable cost-effective refractory materials were available for its construction.
Is that enough to, at least, get you started on understanding what's going on? If not, or if you need a better-English translation for any of the wording, say so. I have fairly easy access to Frisco 1351, which was equipped with an overfire system that is still fairly complete, and can provide some detail pictures if requested.
Thanks for your detailed response.
Next topic: Emission control. I've heard a little bit about "overfire jets", designed to reduce smoke. How do these work, and could there be more practical smoke control with '50s tech?
I appreciate your description of the duplex, and your solution. I think I understand it.
Thank you for your answers, this has been fun for me, and a great learning experience, hope it has been fun for you.
NorthWestDuplex drives- as I understand them, they were useful for balancing reciprocating valve gear at extremely high speeds. Earlier these were touched on, but I'm curious what you guys think could have been built with duplex drive. Also, is there a way to help solve the wheelslip issues with the T-1?
We covered this very extensively in past threads.
We should put something in context first, which was the trend (notably at Lima) of using very long stroke (up to 34"!) in conjunction with high pressure, high superheat, and reduced piston diameter. The T1, of course, used very short stroke (26") primarily (imho) to reduce inertial augment forces at high speed -- less for reciprocating balance than for the vertical component that induces 'hammer blow'. But short stroke was recognized on at least one railroad (C&O) as being deleterious to the locomotive's performance at anything less than high speed. Freight duplexes ... at least the rear engine of freight duplexes, which is presumably where the three-axle engine would go ... would be given the longer stroke and lighter construction that went with it; the result would likely be more effective at speed than the Qs were as-built. But let me take up the T1 as a better example of what needs to be done to get the trick to work...
The benefits of divided drive can be put in two categories. For ultra-high-speed passenger service, the duplex represents the only way to hold vertical augment within sensible bounds. For freight service, the approach lightens augment, but also divides piston thrust across multiple main pins. In both cases, through dephasing, the concept limits the instantaneous torque rise that predisposes to slipping.
The principal drawback, of course, was that a twelve-drivered duplex (think a Challenger or Allegheny without the hinge between the engines!) would have been too long for most railroads, Eastern railroads at least (and in any case, the articulated, or SE Mallet to use Juniatha's term, would be preferable given the complications of the long rigid frame). That meant that one of the engines would be only four-coupled. In the case of the T1 (and the B&O 5600) BOTH engines were four-coupled. As I recall, there was some correspondence at the Hagley Museum indicating that the T1 was considered as a 'double Atlantic' -- remember that the PRR was said to prefer the E6 over the K4 for running above 100 mph, for more than one technical reason -- without understanding a critical limitation of a four-coupled engine.
When an eight-coupled engine is heavily loaded and accelerated near the adhesion limit, and one driver finds a low spot or frog, there is still 7/8ths the adhesion helping arrest any slip. But on a standard Duplex, this situation reduces you to only 3/4 the adhesion ... abruptly. Even with what looks like a sensible FA (the T1 as I recall was 4.21) that degree of unloading will produce a slip at high or low speed. At which point the internal streamlining of the steam passages, and the low inertia, high precision, and relatively large and quick-opening poppets in the valve gear, conspire to happily produce increasing HP with speed -- ensuring that once a slip starts, it will propagate (rapidly) up to essentially the point where steam supply (very large) is outstripped by inertial load (very low until you get to LARGE cyclic rpm). This is much faster, by the way, than a human engineer can close his throttle (even with fancy air assist) or react to adjust the reverser setting. It did not help that the Ts were well-enough balanced that it could be difficult to hear or feel that a slip had set in until it had wound up to 'interesting' levels...
As you are probably aware, the PRR Q2s were equipped with a sophisticated (for the Forties) servo system to arrest slipping, This consisted of sensors that controlled balanced butterfly valves in the steam lines; if a slip was detected (differential rotation of the engines) the 'faster' one was momentarily cut off. As you might suspect, there were control instabilities in the thing, and the 'stuffing boxes' for the butterfly actuators, being exposed to steam at full superheat, proved extremely difficult -- verging on impossible, in that era -- to keep lubricated.
The logical 'fix' for this inherent problem with the duplex drive was originally postulated by Riley Deem at Lima (I believe in the late Forties). I have not found any hard technical details about his solution, other than that it involved 'gears' (Tom Blasingame knew Deem, and said he talked with him at some length, but had never heard details of the solution). Meanwhile, in the early '70s Bill Withuhn came up with a solution (described in Trains and then included in the ACE 3000 patent) which used cylinders in the 'outside corners' (where the 5600 and the PRR Q1 had them) and a set of inside rods between the two mains that kept all the rods in strict opposite phase, giving true longitudinal balance of reciprocating force (and hence, no need for any overbalance in the wheelrims, which would give reasonable rotating balance as well).
Withuhn's solution has a couple of obvious issues, the first of which is that cylinders under or against the firebox have proven very poor engineering practice. The inside rods require that the mains (which take the most torque) be cranked -- not with one throw, as with a three-cylinder engine, but with two. This in turn complicates the arrangements to provide roller bearings and then keep them in requisite alignment. There are some interesting technical problems involved in making those inside rods with roller bearings -- a little reflective thinking will divulge some of them.
In my opinion (expressed a number of places, one quite recently) the 'correct' approach is to provide gearcases, similar to large Spicer drives, on the two axles adjacent to the rear-engine cylinder block. These are then joined by an appropriately-splined (to permit suspension movement) Cardan shaft arrangement, with a center fluid coupling (similar to a Ferguson clutch) and an underrunning sprag clutch (oriented away from the engine most likely to slip -- usually the forward one for reasons I omit here).
On the T1, this goes through a center bearing located within the 'stretcher' in the cast bed between the rear-engine cylinders. There is already a lightening hole in this location adequate for the required shaft dimension; this can easily be modified to take a double mounting flange for the required clutch construction. I further provide some synchronizing detents that preferentially 'lock' the phase of the two engines 45 degrees apart (so the net effect is to produce eight torque peaks per revolution, as on some conventional four-cylinder locomotives like the British Lord Nelsons, with only marginal compromise of balance).
In operation, the two engines work separately, as intended for a Duplex, except insofar as the detenting tends to keep the phasing around where it is wanted. If one engine begins to slip, the Ferguson clutch creates drag; if the slip persists, within a fraction of a revolution the roller sprags engage and lock the engines together (but without producing hard shock to the gearing or the universals/Rzeppa joints in the conjugation shaft). A different form of centrifugal clutch, or an active overriding clutch, can of course be provided, but very quick automatic action is essential.
This is a fully balanced system, and it is easy to incorporate the gearcase into a cannon box. I have designed ways to allow the elements of the Spicer drives to be installed and adjusted without removing either the drivers or the bearings from the axles. It is interesting to note that the effective cost of providing the gears involved has dropped dramatically over the past decade or so, as machining facilities for large wind turbines have become sophisticated, widely available, costed-down, and familiar to run.
Of course, the first cost is greater than an 'equivalent' 2-cylinder 4-8-4, which in its late incarnation was perfectly capable of running at high speeds. (There are people who believe the 115-mph test run of a N&W J on PRR was completely fictional, but it does appear that dynamometer records substantiate it; of course, the same people disbelieve that the zero-overbalance method of balancing used on the Js actually worked, so YMMV...). It may be difficult to justify the added cost and maintenance expenses of a duplex unless the better high-speed performance is required -- and I can think of a number of railroads that were anticipating services in the late '40s that would require just that sort of high-speed performance without destroying the requisite track geometry (well, at least not excessively...)
The 520s were covered in Trains many years ago, and it's fairly common knowledge that they were styled after CME Harrison saw a picture of a T1 in Railway Age and was inspired. There is no direct connection between either the design of the locomotives or the 'manufacturer' (the 520s were 'homebuilt' at Islington).
But I still love both types!
Okay-to steer back on topic, sorry for my turbine jaunt, just wasn't sure if it qualified as "steam" or not.
Duplex drives- as I understand them, they were useful for balancing reciprocating valve gear at extremely high speeds. Earlier these were touched on, but I'm curious what you guys think could have been built with duplex drive. Also, is there a way to help solve the wheelslip issues with the T-1?
(As a side note to the T-1, I found the SCOA-P wheels when looking at late Australian steam, and found that the South Australian Railway 520 class looks a lot like the T-1, but I can't find a connection to Baldwin...)
Our community is FREE to join. To participate you must either login or register for an account.