Ywolfgl I do know that Europe had over 200 of these engines. Germany, Austria, Poland, Hungary, Romania, Italy Persia, Czech, France, and Romania all used them.
I do know that Europe had over 200 of these engines. Germany, Austria, Poland, Hungary, Romania, Italy Persia, Czech, France, and Romania all used them.
Which engines are you referring to?
If GM "killed the electric car", what am I doing standing next to an EV-1, a half a block from the WSOR tracks?
Bringing the drawing up for reference:
https://www.martynbane.co.uk/modernsteam/ldp/ldp.htm
I had not even considered the 'shoulder' of a Belpaire firebox, but it certainly makes the best sense of the potential alternatives!
What I now wonder about is the faint line just above the tubes that extends nearly to the backhead, with an interruption near the rear. This can't be the 'rest' waterline, and almost certainly isn't a plate or vane like that wacky mid-boiler D&RGW 'thing' in the Twenties; it almost has to be some sort of way to break up the evolving steam bubbling to reduce carryover effects ... or yet another drawing artifact I'm misinterpreting, although it can't be the base of the 'Belpaire shoulder' as drawn.
Paul MilenkovicThe Lewty booster was supposed to transfer power hydraulically?
That was the premise in the early 1970s when it was proposed.
In part I think this was due to the relative expense of both custom and OTS electrical alternatives at that time, and their perceived operational 'fragility' in a steam-locomotive environment. (Not to insult anyone, but I think practical knowledge of hydraulic power transmission might have been handwaved a bit, too.)
So instead of steam and exhaust connections to the booster (or those auxiliaries), you had inlet and return high-pressure hydraulic hoses?
Yes, and the advantages of this over any Franklin booster arrangement are immediate and very large. Very well-armored hoses with relatively low flexibility are practical (note that they could be formed hoses as well) and integrity at the joints is easy to make, assess, maintain, and break without damage when necessary.
The other great advantage is that many of the auxiliaries on the locomotive can benefit from some form of hydraulic drive, whether or not of the same kind adopted for the actual truck-mounted booster motor(s). Again the power transmission is nominally cheap, relatively easy to inspect and maintain, and can be well armored against damage. But the more significant aspect is that it reduces parasitic losses of steam, and therefore that percentage of water treatment costs, etc.
And hydro-static booster motors?
Well, this is where the handwaving starts. Obviously (to me) a standard hydrostatic motor would work even less well than a typical Franklin booster, requiring at least a couple of mechanical-transmission speeds or some kind of CVT to achieve practical volumetric flow, and almost impossible to control quickly enough in slip or skid without an expensive system of sensors, logic, and actuators. So I think it would have to be an axial-piston motor (see below), and I'd suspect it would be worthwhile to see if any of the designs of CVT for automotive use could be scaled up or made workable as a final drive.
(As a possibly amusing side note, several people I know have looked carefully at the use of large industrial Gates belts to 'conjugate' multiple truck axles from a single motored one. Yes, it's practicable; yes, it looks funny; yes, OSHA would have a fit if you tried running one without full guards...)
I guess there is experience with [hydrostatic] transmission of power on garden tractors and some larger units, but would something like that hold up in railroad service?
To answer part of this directly: I think anywhere hydrostatics of the 'usual' garden-tractor kind have been tried in actual line service, they've been failures. The great use (and relatively great success) is in hydrokinetics, particularly the Voith approaches, but these are usually employed 'shaft to shaft' without pump and turbine/runner being separated and I cringe a bit thinking about the losses if you separate them with any practical hose or pipe system. I'd envisioned using something like a bent-axis axial-piston/swashplate motor -- and a variable-displacement axial-piston pump at the 'steam motor' end -- but there may be better alternatives. Theoretically you could reverse the almost-necessary overrunning clutch between motor and axle, to get actual torque 'in reverse' (which of course the "reversible" Franklin booster never did), and equally theoretically you'd want at least a little Ferguson-clutch-like (probably magnetorheological in a practical railroad design) compliance between motor and axle in drive.
The real 'fun' of the Lewty booster is not so much its operation as a booster but as the analogue of the approach taken (but not carefully enough designed!) "APU" system on the SPV2000 railcars, or in a different sense on the extended version of the infamous Hancock Turbo-Inspirator, where one turbine drove several pumps that, at least in principle, could be separately clutched-in or declutched.
A key to doing this practically on a steam locomotive is the ability to synthesize working-frequency AC for auxiliaries powered by, say, commercial 220V AC motors and drives. Obviously you're not going to keep a triple-expansion steam motor 'going all the time' at synchronous speed even with an expensive polyphase-wound alternator, but once the cost of inverter systems (particularly sourced in the aftermarket) comes down enough, developing an appropriate feed for what amounts to a 220V power and control bus on the locomotive becomes interestingly practical. Note that operation of this system can be done, I think effectively, in parallel with any prospective 'hydraulic' (or mechanical) demands on the engine output.
BTW, note the 'flap' extending between the part of the steam space over the crown sheet and that over the tubes. I suspect there is a difference in the behavior of the 'boiling milk' over each section that makes this of benefit for the steam separation here.
To be honest, I beieve that particular line to be the representation of the "shoulders" of the Belpaire firebox, since that is where the stayed area of the firebox transitions into the cylindrical boiler barrel. So it does not represent anything inside the boiler.
Peter
Paul MilenkovicMaybe me and everyone else are reading too much into one of Porta's drawing reproduced in Wardale's book, but I guess the oddball type of superheater I had asked about was part of steam locomotive history. Yes, it didn't produce a high level of superheat, but maybe Porta thought by truncating the boiler section it could?
If you have not read Sinclair's 'Development of the Locomotive Engine' already, you should do so at your earliest convenience (I have previously posted links to download it via Google Books or read it on line, and can probably do so again if you need them). You should look for 'steam dryers' as that was the empirical purpose of the contemporary arrangements as perceived.
I think it was well recognized by the mid-1870s that saturated steam in a locomotive suffered great condensation losses, the issue for operators at the time not being so much "can we solve this" as something along the lines of "what is the cheapest 'intersection' of technological cost vs. fuel, water, and maintenance expenses". Jacketing cost money, therefore wasn't widespread in adoption; ingenuity went into using the 'waste heat' that wasnt "used to make steam in the boiler" for this useful purpose (after all, even in the smokebox the combustion gas was ever-so-much 'hotter' than the steam at saturation) and so the question became where you tapped the heat flux from the gas to 'dry' the steam.
Powerplants happily run their superheaters right up close to the furnace exhaust, and spray in water to 'desuperheat' the result down to what the heat-balance diagram indicates. However, they do this with relatively continuous load operation, and use exotic metals in the superheater, neither of which is particularly likely on a railroad, then or now. So elements in or near the firebox would routinely soften, melt, fail (likely on the road) to say nothing of the effect on the crew when this happened. Periodically the idea crops up again; you can almost make it work with a [url=https://www.steamautomobile.com/northea/lamont.html]Lamont circulating waterwall system for the radiant section, but as with pulverized-coal firing sooner or later physics and chemistry will catch up with you. (And typical methods of desuperheating are a very, very bad idea with proper water treatment in your feedwater system...)
As noted, the next place to get your 'free heat' is off the firetubes somewhere. Here you are vying with steam generation out of the convection section, the 'bread and butter' of boiler design in that era (at least up through the time of Wagner on the DR) so where you place your 'partition' has consequences. Near the fire has many of the same consequences as in-firebox steam elements: you have high tube temps that abruptly and effectively drop to saturation across the front tubeplate, the rear tubeplate runs at elevated temperature, and the varying pressure in the 'superheater' as steam flow through it changes with throttle and cutoff causes repeated stress in all the joints at the front due to the effective differential with what the saturated boiler contents exert.
Then you move forward to 'someplace in the middle' where Bruce has the Pielock ultimately located -- this doubles the sheet exposure to differential stresses, and leaves you wondering just how to caulk two sets of tube joints in a reasonably-well-populated tube nest. You can imagine the stress profile in a tube allowed to float up in temperature compared to regions almost immediately adjacent to it, the whole thing being exposed to lolog flux as Fry indicated.
If I recall correctly, the Player superheater used a larger surface in the smokebox area to produce the necessary contact with that famously good insulator, steam, to superheat it properly with the combustion gas about to be blown out the stack. This is a nifty idea in principle but didn't seem to produce the effect desired ... I wonder why.
All this puts the Schmidt superheater in much more appropriate perspective. Provided you take reasonable care either to size things properly, and assess drag well enough, and if necessary provide proportional damper control of induced draft in tubes vs. larger populated flues, it makes sense to use a long, expensive-alloy steam pipe inside a larger flue so the same lolog heat that is warming water outside the flue is acting directly on the steam in what is at least half countercurrent action. It is not difficult (at least, today) to make the necessary standoffs to center the tube in the bore and make it self-guiding during service extraction. The late developments that gave us the awful Frankenstein smokebox door arrangement in the NYC Selkirk front end specifically make access to the superheater elements nearly as easy as possible with a minimum of tinkering, and there is no compromise with the tube and flue system itself, whether the boiler is 'sectional partitioned' for other purposes or not.
The only 'unfortunate' consequence of a Schmidt arrangement is that you can't put Besler tubes easily in the same nest, or to put it another way Besler tubes can't also be cleverly designed with heat-recovery elements inside: it defeats the purpose the Besler tubes actually serve (no pun intended).
A problem you haven't explicitly mentioned yet is that the Schmidt arrangement is wildly expensive compared to any firetube or smokebox heater arrangement. Special alloys, cast bends that have to be precision-welded for any kind of service life, design criteria so complex the Superheater Company kept them proprietary, welded refractory-metal shields on the top of the header bends to ameliorate char cutting -- the list is long, and it's unsurprising that when the 'specialty' manufacturer stopped making money the party stopped fairly quickly. In fact it could be (and has been) argued that the type E was already 'over the top' in cost/benefit to railroads at the time (and some of the other 'thermodynamically optimized' superheater-element ideas of the time, such as the 'sine-wave' superheater or elements with inside or outside area enhancement like fins or vanes, were even more expensive for their dubious addition to operating efficiency).
Hoo, boy! Some boiler explosion.
Seriously, folks, was it Overmod who recommended the book Alfred Bruce (1952) The Steam Locomotive in America?
p. 154 "One of the first superheaters seen in the United States was the Pielock type . . . consisted of an internal chamber in about the middle of the boiler barrel through which passed the boiler tubes. The chamber was filled with steam which absobed heat from the exposed tubes, and although superheat up to about 170 to 180 (deg-F) was obtained, the construction was inaccessible and never adopted in the United States."
Maybe me and everyone else are reading too much into one of Porta's drawing reproduced in Wardale's book, but I guess the oddball type of superheater I had asked about was part of steam locomotive history. Yes, it didn't produce a high level of superheat, but maybe Porta thought by truncating the boiler section it could? I now see why this could be a maintenance nightmare -- at least the Schmidt-type superheater elements can be pulled out with the help of a crane, shop workers with strong backs and without smashing fingers, but getting access to the tubes in this tube-and-shell contraption might be quite the challenge.
Maybe this is further vindication of the artistic license exhibited by many on the Model Railroader forum, that there is a "prototype for everything."
Incidentally, here is a rare picture of the shell during fabrication
and here is the result (when Penny Trains tried to fire it up)
M636C... While I can understand this interpetation, a couple of features are of concern.... Why is this completely separated from the "firebox and short boiler" behind it, since tis wasn't a feature of Chapelon's 160A1 which had such a partitioned boiler.
Why is this completely separated from the "firebox and short boiler" behind it, since tis wasn't a feature of Chapelon's 160A1 which had such a partitioned boiler.
The firebox and 'short boiler' are a unit with steam space above. A Chapelon 'partitioned boiler' with through firetubes would also have by necessity a 'steam space' unpopulated by gas tubes in its upper section. This is avoided in the pictured design by allowing the gas flow through the 'boiler' section to expand, in the 'cone-shaped' section, until it passes through the increased number of tubes in the fully-populated 'sectional' boiler as pictured. That is the point of the separation of the two boiler sections by the conical "gas box"; as I said, I think in a practical design the interstage would require a different design from that pictured.
Where is the superheater? An E type through the small tubes in the separate forward boiler?
I see none pictured. Of course, I don't see any steam supply from the boiler to the cylinders, regardless of what could be populating the two "HRSG" partitions, so the only possible solution appears to be some combination of Chapelon-like steam heating around the cylinders (as on one variant of 160 A1) combined with multiple small Wagner throttles operating as fluidic servos (remember Porta used 'Waggoner' when discussing these) close to the ports. I see no dome, and no indication of external piping for one, so I am left with the impression that this drawing shows only a partial set of systems applied to the locomotive, with the valve-gear arrangement and steam supply reserved to another part of the discussion that no longer accompanies the drawing as provided.
I've never claimed to be an expert on Porta's work, but most of his projects were more conventional than that...
This is actually quite familiar, as far as it goes. You will note that quite a bit of previous discussion in threads here (and elsewhere) mentions the diverging double exhaust pictured as being required to accommodate some form of feedwater-heater exchanger in 'the middle'. If Jos Koopmans is correct in his thinking, it would probably be better to use a single larger (angled if desired) chimney in the front end, since the shell diameter in the smokebox clearly permits this; alternatively, since there is no baffling (as in a Master Mechanic front end's longer gas path from the front tubesheet to the entrainment area around the exhaust nozzle jets) and no sort of self-cleaning screening on what is obviously a solid-fueled engine, the long Kylchap-style skirts and vanes are an attempt to help all the gas flows from top to bottom of the extended tube coverage of that front tubesheet have proportional draft induction. (In my opinion he would have needed quite a bit of tinkering with vanes, screens, and other flow elements to get all the gas flows proportional at the front tubesheet to give the proper flow induction from the smaller convection-section outlet into the fully-populated tube section in the first place. The absence of a formal water-jacketed combustion chamber forward of the formal grate area to extend combustion interval for luminous particles is difficult to explain ... but much of this is nitpicking that may have been fully covered in the 'rest' of the material that accompanied the drawing in the '70s.
Porta had far more unusual things than this -- I notice there is little if any optimization of the boiler for GPCS or for 'cyclonic combustion'; he very famously worked out a watertube (more precisely, perhaps, tubular waterwall) firebox for high pressure with, perhaps, even more tube complexity (and difficulty of mechanical internal cleaning) than the Winterthur alternative, that could have been incorporated here.
What did Wardale say? (This time I'm in a different city to my copy?)
This is of more than usual critical importance. My copy is in storage and inaccessible, too. Someone with the book needs to help us quickly.
The problem is that it is ONE dark vertical line, indicating it is the thin partition of a Chapelon-style sectional boiler, with equalized hydrostatic pressure on either side permitting the use of a simple partition between sections.
While I can understand this interpetation, a couple of features are of concern....
What is the gain in separating the two boiler sections by the conical "gas box?"
I've never claimed to be an expert on Porta's work, but most of his projects were more conventional than that.....
Paul Milenkovic I assumed that the vertical line marked a separation of that front section into a front-front section filled with water to form an "economizer" type water preheater, with the back section filled with dry steam to form a different type of superheater.
The problem is that it is ONE dark vertical line, indicating it is the thin partition of a Chapelon-style sectional boiler, with equalized hydrostatic pressure on either side permitting the use of a simple partition between sections. There would be incredible heat loss between 'superheated' steam in the rear partition and the sizable 'wetted' area of tubesheet between tubes (which would limit peak superheat, perhaps dramatically, depending in part on the circulation of steam going through the space of the vessel) and nearly-incredible fabrication difficulties in the highly-stressed partition (which must of course be seal-welded and then kept effectively stress-relieved inside a vessel with tubes welded at the front and rear -- this perhaps being more justified than in a regular steam-space boiler because the tubes acting as tension members populate the sheet evenly, but unjustified if the gas flow into the rear tubesheet is segregated in temperature at any substantial gas mass flow even in conditions of relative 'turndown'. How you deal with any leaks or set up to confirm full-pen integrity of all those right-angle welds is an interesting question, as is diagnosis of progressive cracking or corrosion in all those joints -- remember that desuperheaters in power plants, which are very effective in reducing peak steam temperature, work precisely by sprinkling hot feedwater into the steam...
On the other hand, there have been discussions (including Tuplin's) about the relative efficiency of saturated steam at high pressure if you cut the cylinder losses 'by other means', and this may be one such; I will look carefully at the drawing again when I have time and see what I can deduce from the throttle and front-end arrangements.
Yes, that drawing.
I see a dark vertical line within the front portion of the "sectional boiler" and I don't see any header for a Schmidt superheater. I assumed that the vertical line marked a separation of that front section into a front-front section filled with water to form an "economizer" type water preheater, with the back section filled with dry steam to form a different type of superheater.
Part of what got me thinking along those lines was the 2018 Tom Morrison book describing the "boiler" segmented into multiple secions -- I will look for that book at State Historical Society of Wisconsin in a few days.
Ah, that drawing.
I had a very different interpretation of what that is, in part because it is similar to an idea of my own derived from reading the B&W 'Steam, Its Generation and Use' mid-Fifties edition.
A 'fault' of practical "Stephenson" multi tubular one-pass locomotive boilers is that tubes and flues alike must be kept below the effective 'waterline' (actually a region) rather than the shell being full of tubes and water as in heat recovery exchangers or boilers. (The B&W illustration shows a cylindrical shell that is tilted, with an elevated drum taking off the steam at one end).
Unless I'm mistaken you can see a water-level line in that drawing, across the crown and then the short 'tube bundle' -- if you could see the transverse section that goes with this elevation section I'd bet it would show the bottom of the shell fully populated with tubes as in a saturated-engine boiler, with the steam space above common to both. The combustion gas from the firebox goes through what is effectively a short convection section, then diverges through that cone transition and traverses the 'sectional' boiler part which is kept full of water at all times (and presumably with caring arrangement for quick and positive stripping of any evolved steam or of gases coming out of solution) and therefore can have much larger heat-transfer surface for a given diameter and length of shell than one that must accommodate a mix.
And here I am, a third of the way to Ohio, with my copy of the 'second edition' thoroughly inaccessibly in storage! Is there any way you can scan and link a 'fair use' copy of the diagram at adequate scale to show the construction details?
The first diagram on that page.
Note that there is a very similar looking 2-10-0 with a conventional boiler shown elsewhere.
I wouldn't want to say if this diagram is adequate in any way but it's the one in Wardale's book.
M636C ... Overmod has not addressed the Porta diagram on page 25 of "The Red Devil".
The 'boiler' in this design is only about 1.5 metres between tubeplates and forms an extension to the firebox and combustion chamber and is referred to as an "evaporative tube bundle". This certainly concentrates the boiler function in the radiant heating surface area.
Tom Blasingame at one time had a very similar-sounding idea. The potential issue here is that, with limited uptake in the 'time of flight' of the combustion-gas mass required for a modern locomotive (of adequate peak horsepower to justify the relative inefficiency of this kind of engine), you need extremely good recuperation or bottoming downstream to make use of what is basically 'rejected heat'. I trust the diagram shows these; I have no idea how to visualize what this is:
Forward of this is a conical area which allows the combustion gases to circulate through the whole barrel structure.
although I suspect it would be clear from the diagram itself.
Then, as described the full barrel diameter consists of a fire tube superheater followed by a larger feedwater heater.
Which would logically be like a Chapelon sectional boiler 'in reverse', with the remaining high heat transfer for 'steam generation' (with probable DNB if adequate circulation patterns were not assured, etc., much of which I would trust Porta to understand reasonably correctly) and at least a thermally-effective partition between this and an up-to-saturated (by necessity, think about it) "steaming feedwater heater" which is basically doing what the forward section of most conventional convection boilers probably does by some point in long Wagner-formula tubes. Certainly seems to get around some of the issues with Henderson-style 'sectional flexible boilers' in the latter iteration (where the forward 'hinged section' came from Baldwin optimized specifically as a 'feedwater heater')
Again without seeing the actual diagram, and therefore more than usual potentially talking through my hat, you would need an additional partitioned pure steam space just ahead of the 'rear tubeplate/sheet' communicating to the combustion chamber in the design, with the 'evaporative' bundle being forward of that and in contact with something to evaporate, whether circulated, sprayed on, obtained from a Cunningham-circulator-like arrangement, etc.
In this case I assume that the intermediate tubeplate would in fact be welded to the tubes, since feed water leaking into the superheater would not be acceptable.
"Yes, but..." all the sections including a 'superheater' would (as you noted with the intermediate piping) need to run with equalized pressure, analogically to the situation on a Super-Power engine with front-end throttle which eliminates the ESC's contention that the superheater can constitute a 'separate pressure vessel' necessitating its own set of safeties.
The thought of attempting to actually put those welds in that tubeplate gets worse and worse the more you think about how such an arrangement would have to be fabricated with actual material by actual welders (robotic or otherwise). This is far from unusual with Porta designs. There are also differential-expansion concerns if the tubes are continuous and presumably inserted from the smokebox end as everyone's maintenance practice probably even in the days of Seguin would likely be, but that's another matter to analyze, as it only becomes necessary when you actually make and then stress-relieve all those right-angle full-pen welds in the first place.
This type of superheater was used by Golsdorf in the 1908 210 class 2-6-4 express locomotives. It was described as a steam dryer, "dampftrockener" in German.
There is a contemporary design with a name that I always confuse with the Pielstick engine makers - I think it is 'Pielock' - that I remember as being like this. I can't think of any installation purporting to do superheating on extended convection firetubes that was suvvessful 'enough' to compete with a Schmidt setup in flues in the same shell dimensions as the 'firetube superheater' equipped boiler.
However, since this was added to a conventional boiler, the superheater was at the smokebox end where the heat was insufficient for high temperature superheating ...
It will pay you to look specifically at the physical reasons that the 'heat was insufficient' (there are a number, a couple of which I addressed in the earlier post).
... If located at the firebox end, such a superheater would probably have worked well ...
Perhaps frighteningly well, if one of those 'superheater tubes' were to soften and start failing as it might have an unexpected propensity to do with poor 'indirect' steam circulation as would almost certainly be found in such a design. This would be mitigated -- probably nowhere near enough -- by the pressure being imposed on the OD of the tubes rather than building up hoop stress. Someone could model this with multiphysics to see what the implications might turn out to be.
... but at the expense of a complex boiler design with a number of external connections, between feed heater at the front and boiler at the rear and the superheater in the middle and the throttle (somewhere) and the HP cylinder...
That gets weird on a number of levels.
"Posit" that Fry's general assumption of lolog heat transfer as a function of tube length is correct, and that at each 'stage' along the tube length the actual uptake of "available heat" is actually made comparable to what the effective function says would be available. Superheater at the rear, then convection boiler, then 'feedwater heater' perhaps in partitions approximating part of a more conventional feed train.
Stean from both the presumed optimized circulation in the radiant section and the convection boiler go together into the 'superheater' and thence into some approximation of a top-mounted dry pipe (and steam drying/separating arrangement of sufficient capacity and achieved steam quality at highest demanded mass flow). Now, since the 'superheater' is unlikely to suffer from any of the ;conventional' issues carryover into Schmidt elements poses, except perhaps accrual of boiler-treatment chemicals and the like, we might put some of the steam separation in the 'outflow' from the superheater -- and in my opinion, we now have to consider how the flow 'through' the superheater is going to be managed from the saturated input(s) to the output going to the throttles.
If I were building this, there would be one Wagner throttle as the 'main stop' (approximating the function of a multiple front-end throttle in that sense, or what the conventional dome throttle on engines fitted with 'two throttles' does) and then the individual Wagner throttles close to the valves that Porta advised (you can see them in some of the ACE3000 material). I don't think it is difficult to find both a best 'position' and a good detail design for either one, and the routing of the required 'transfer piping' to get the steam from the rear superheater forward to the cylinders (assuming the cylinders are still in the 'conventional' position for induced draft) is less of an issue in this modern era of stack-train overhead clearances.
I think the concept founders on effective heat transfer and flow in the 'superheater region' and fabrication considerations, but if either the diagram as provided or anyone's ingenuity can deal effectively with these, I'd be interested to see or hear it.
What I am trying to figure out is an arrangement that L. D. Porta had proposed, as shown in a drawing in Wardale's "Red Devil" and some online sources. It is also mentioned in Tom Morrison (2018) "The American Steam Locomotive in the 20th Century" in connection with either superheaters or reheaters between compound expansion stages of some of the earlier U.S. Mallet designs.
What I am talking about is a superheater version of a fire tube boiler instead of the Schmidt superheater that is more like a watertube boiler, except the tubes are filled with steam. It is more along the lines of a "shell-and-tube" heat exchanger.
I have found the discussion interesting.
However, Overmod has not addressed the Porta diagram on page 25 of "The Red Devil".
This is nothing like Chapelon's 160A1 which had two Schmidt type superheaters, one as a reheater between stages of expansion, even if the boiler was partitioned.
The Porta diagram is incomplete in that no external steam pipes are shown which with the unconventional design of boiler leaves a lot to the imagination....
The boiler in this design is only about 1.5 metres between tubeplates and forms an extension to the firebox and combustion chamber and is referred to as an "evaporative tube bundle". This certainly concentrates the boiler function in the radiant heating surface area. Water and steam are confined to the area above this short boiler and firebox. Forward of this is a conical area which allows the combustion gases to circulate through the whole barrel structure.
Then, as described the full barrel diameter consists of a fire tube superheater followed by a larger feed water heater. In this case I assume that the intermediate tubeplate would in fact be welded to the tubes, since feed water leaking into the superheater would not be acceptable.
This was a proposed metre gauge three cylinder compound 2-10-0 and one imagines that it would have worked.
This type of superheater was used by Golsdorf in the 1908 210 class 2-6-4 express locomotives. It was described as a steam dryer, "dampftrockener" in German. However, since this was added to a conventional boiler, the superheater was at the smokebox end where the heat was insufficient for high temperature superheating, and on the later (1911) class 310 a Schmidt superheater was fitted.
If located at the firebox end, such a superheater would probably have worked well, but at the expense of a complex boiler design with a number of external connections, between feed heater at the front and boiler at the rear and the superheater in the middle and the throttle (somewhere) and the HP cylinder...
I have run out of time to address some of the issues here, particularly the ATSF experimentation and the philosophy behind some of it, but there is one thing I want to address before I leave:
Paul MilenkovicThe 160-A1 was a testbed that tried out every combination, but it seems the best arrangement was to feed to HP cylinders with saturated steam that they only partly expanded and then to feed their exhaust into a steam reheater. That and, was it, only jacket the HP cylinders with saturated steam to suppress wall-effect losses and relay on the steam reheating to suppress LP wall-effect losses? Again, he could configure this locomotive to any combination of cylinder jacketing and superheating and reheating, but is the combination I just mentioned the best from both the standpoint of not cooking valve liners and optimum economy down to slow cylinder speeds?
We had some posts about this, one issue being that all the relevant technical information is not only hard to come by but is essentially all in French, with Claude Bersano (who could have explained it) now beyond the veil, and Thierry Stora down with the same general sort of health problem I have (complicated by French socialized-medicine priorities, but we'll not go there). Juniatha explained some of it brilliantly in one of the last threads before she left, but I don't think she said enough about the conclusions made from all the testing (I think much of the actual timeline and 'ultimate conclusion' may be lost in between poor permanent documentation and the near-immediate push to 'demonize' steam in general and Chapelon and particular on SNCF starting in the early Fifties.
As I recall, the interesting part of the testing premise was the idea of using very high degree of superheat with the steam passing over the metal of the HP valve block before it got to the actual valves, substituting for any passive or separately-circulated heating (as in my system) or the use of really good passive insulation as in Wardale's approach. The presumptive idea, and it is far better than any use of superheat 'in the actual Rankine cycle' to heat walls from the inside or heat steam a bit better to stave off nucleate pressure collapse from phase change, was to increase the heating through the mass of the cylinder block itself so the actual temperature of the cylinder wall just inside the .007 or so that actually 'excurses' between admission and exhaust would remain as high as possible -- minimizing the actual heat flux across the internal wall metal, probably by tens of degrees. Amusingly, the last technical reference I could find in English that actually put numbers to this sort of approach was from 1874, which will tell you something about where people thought their priorities were in the various eras since then.
It was difficult enough to figure out that this was the thing going on in the design in the first place, let alone whether later testing proved the practical gains, or the operating flexibility or something else, were not enough to warrant the arrangement. The practical effect -- to heat the cylinder block before there are condensate losses, either near starting or due to normal expansive working -- is still relatively poorly understood or appreciated today.
The Lewty booster was supposed to transfer power hydraulically?
So instead of steam and exhaust connections to the booster (or those auxiliaries), you had inlet and return high-pressure hydraulic hoses? And hydro-static booster motors?
I guess there is experience with that kind of hydraulic transmission of power on garden tractors and some larger units, but would something like that hold up in railroad service.
Paul MilenkovicTom Morrison, by the way, describes booster engines as part of what Lima was trying to sell in their package of Super Power improvements to the steam locomotive. By the end of steam, the railroads were tired of messing around with booster engines, either ordering the Superpower types without boosters or removing them from locomotives that had them.
The best description of what the NYC-designed booster was supposed to do is that it added the functional equivalent of an additional driving axle to a given locomotive design, with little additional steam (or fuel or water) costs. In those days (and on a 'sweetheart' road like NYC that might 'buy invented here' preferentially) that capability was important; with locomotives like the J1 Hudsons on the Great Steel Fleet, the importance rises to near-critical. (Note that this gives the equivalent of a good 4-8-2 in most common areas where that 4-8-2's capacity would be deemed important in the 1920s, without the balance issues with eight-coupled locomotives that hadn't even begun to be solved correctly until near the end of the '20s)
Some proof of this might be determined from the non-scam development of the booster into high-speed reversible (to relieve a stuck engagement without having to stop, not to use boost when backing up) version by the '30s. There was easily a use for 'an axle's worth of extra steam oomph' with air-conditioned consists that had sequential axle generator engagement at various target speeds down low in the acceleration profile, but far higher than a 15mph-or-so cutout.
Frisco at the end had a somewhat different rationale for buying an expensive booster for 2-8-2 'rebuilds' -- but there is little doubt the locomotives so modified kept their boosters until the end. I believe the two I know of still do.
Far more expensive to use a $125,000 or greater (in Bretton Woods dollars, no less) unit as an in-train helper 'in place of a booster' -- although it certainly would have helped 'start any train the steam locomotive could pull' and likely developed a significant part of its power in the roughly 10mph to 30mph range that its traction motors could endure and the booster would be increasingly wasting reserve steam...
The correct answer probably was, and certainly is now, the use of the Lewty booster arrangement or something like it. This puts the 'booster engine', in Lewty's version thermodynamically efficient at triple expansion, with no variable-speed issues and presumably full and effective cutoff control, somewhere on the frame, with hard connections to steam and exhaust, uses it to drive or power many of the auxiliaries (making a plethora of wasteful little steam turbine leeches unnecessary), and allows relatively lightweight but high-torque connection to one or more 'idler' axles on the locomotive or tender arrangement.
It may be highly instructive to read the account in Charles Fryer's book about the testing done in Britain starting in the 1920s to use or incorporate Franklin-style boosters on some British locomotives. The tests were highly successful, but none of the 'improvements' lasted. Part of this might be the same sort of licensing or patent issue that killed any market for Beyer-Garratts in the United States. Fryer takes this subject up in some detail (and, in my opinion, is correct as he does).
Tom Morrison, by the way, describes booster engines as part of what Lima was trying to sell in their package of Super Power improvements to the steam locomotive. By the end of steam, the railroad were tired of messing around with booster engines, either ordering the Superpower types without boosters or removing them from locomotives that had them.
By the 1950s, what did they do when a road locomotive was just shy of the tractive effort to start its train? Behind the steam locomotive, they coupled a Diesel -- we got your "booster, right here!"
Leaving out the question about the superheater, there are two kinds of heat recovery systems that add heat to the feedwater before it adding it to the boiler. Because boiling as well as condensing (whether in a proper condenser or simply exhaust to the atmosphere) takes place at constant temperature for a given pressure, the steam (Rankine) cycle would be closer to an ideal Carnot cycle if you use waste heat to warm the feedwater to near the boiling temperature.
There are two types of such heat recovery devices. What is commonly called a feedwater heater condenses exhaust steam to produce large amounts of heat to warm the feedwater. The "open" type feedwater heater mixes steam with the feedwater (need some way of filtering out the cylinder oil before pumping feedwater back into the boiler otherwise foaming could happen) whereas a "closed" type uses a tube-and-shell heat exchanger that keeps the exhaust steam separate from the incoming water. The exhaust steam injector is an example of a combined closed-type feedwater heater along with a feedwater pumping arrangement.
Feedwater heaters on locomotives can get the water from tender temperature to just shy of the atmospheric boiling point of 212 deg-F. Wardale suggests that when a locomotive is working "hard" at a long cutoff, cylinder release occurs at higher-than-atmospheric pressure and that a check-valve arrangement could be used to supply pressurized steam to a feedwater heater getting a higher feedwater temperature.
An economizer is properly a device using the heat left in the flue gases to get a higher feedwater temperature than what the feedwater heater can reach. Chapelon had such a thing in the form of the baffle plate arrangement semi-isolating water in the front of the boiler from the back. Wardale also proposed this arrangement with the 5AT project that ran out of steam (money) a couple years ago.
It appears that Porta proposed such a thing for his "second generation steam locomotives" that were never build. The drawings of it in Wardale's "Red Devil and Other Tales . . ." suggest three separate drums -- the front drum being an extension of the firebox and being too short to extract all the flue-gas heat according to your above explanation, a middle drum as the alternative superheater arrangement I asked about, and a final drum as the economizer. The economizer and boiler tubes would have to be separated drums in this arrangement because especially the middle drum cannot contain any boiler water in it if it is expected to superheat steam.
I have to return to State Historical society to see of they have Tom Morrison's 2018 book to give you page numbers on descriptions of such a segmented arrangement on early Mallets -- the Santa Fe Mallets are prime suspects. I guess some of the earlier 20th century locomotives and especially Mallets had compound expansion, other-than-Schmidt superheaters, steam reheaters, complications that were dropped by the end-of-steam as being unnecessarily complicated, not effective enough for the trouble, costly to maintain, and so on.
As to Wardale wanting to cool his piston-valve liners with de-superheated steam, I am thinking Chapelon's 160-A1 (2-12-0 wheel arrangement) was on an interesting path with a compound arrangement. The 160-A1 was a testbed that tried out every combination, but it seems the best arrangement was to feed to HP cylinders with saturated steam that they only partly expanded and then to feed their exhaust into a steam reheater. That and, was it, only jacket the HP cylinders with saturated steam to suppress wall-effect losses and relay on the steam reheating to suppress LP wall-effect losses? Again, he could configure this locomotive to any combination of cylinder jacketing and superheating and reheating, but is the combination I just mentioned the best from both the standpoint of not cooking valve liners and optimum economy down to slow cylinder speeds?
I'll take up a few comments 'inline' to keep the context. Please keep the comments going in this thread.
Paul MilenkovicDoes anyone know of alternative arrangements to the widely used Schmidt superheater? This is the kind where the superheater elements are tubes carrying dry steam, which make hairpin turns inside enlarged firetubes in a steam locomotives' boiler. One end of each element is connected to a manifold or "header" receiving steam from the dry pipe of the boiler whereas the other end connects to a second manifold feeding the cylinders. My understanding is that there is a "Type A" where the elements make only one loop whereas the "Type E" has each element making multiple loops into multiple tubes in an effort to get hotter steam.
There are many, both before and after the Schmidt/Superheater Company approach. Santa Fe in particular got the 'bug' for superheat in the first years of the 20th Century, recognizing at least some of the advantages; they had a problem in that they also wanted to recuperate some of the energy in the combustion gas after the convection section, so arranged the superheater passes in the smokebox area. This worked some, but not nearly as well or as proportionally as a good Schmidt arrangement. (This was also before the general recognition of radiant heat uptake for direct steam generation; Santa Fe was also famous for small-fireboxed Mallets (and encouraging Henderson to try peddling equally small-fireboxed quads and quints) and for building the Jacobs-Shupert to the wrong dimensions and circulation. We could go over some of the history, but I think you're focusing the discussion on things that are 'better' than Schmidt.
It is difficult to find an arrangement that works better on locomotives than the arrangement of multiple-pass headers in enlarged flue tubes, with the return bends kept well ahead of, let alone projecting into, the radiant space past the rear tubesheet. The arrangement minimizes weight and consequences of weight distribution, preserves balance on the suspension (or at least the ability to achieve balance with minimal rigging change), and solves some of the issues associated with relatively poor heat uptake from relatively rarefied combustion gas (it is effectively below atmospheric pressure except 'accidentally') to steam, which is a good insulator and likes to 'choke' if in turbulent flow.
What I am trying to figure out is an arrangement that L. D. Porta had proposed, as shown in a drawing in Wardale's "Red Devil" and some online sources. It is also mentioned in Tom Morrison (2018) "The American Steam Locomotive in the 20th Century" in connection with either superheaters or reheaters between compound expansion stages of some of the earlier U.S. Mallet designs ... what I am talking about is a superheater [for a] version of a fire tube boiler ... that is more like a watertube boiler, except the tubes are filled with steam. It is more along the lines of a "shell-and-tube" heat exchanger...
[EDIT - Note to TL;DR and MEGO readers: the following is now something of a digression, and you can skip down a bit to where this specific discussion takes up again...]
I thought this initially seemed like an extreme version of what in marine practice is called a 'steaming economizer' -- the idea of which is to use combustion gas to heat feedwater all the way up to saturation at pressure and then beyond, with the steam generated in the process tapped off and combined with that from the convection-boiler throttle. Something like that would have to be somewhat larger if run on pure steam, as there are two things to consider in its design (proportional heat transfer out of the combustion gas, and proper heat transfer into the steam without 'cooling' it for part of its traversal). Built as such, it would impose additional weight, and additional 'packaging' issues about where to put the arrangement (which has historically been a reason you don't see economizing on locomotives, despite some very clear advantages thermodynamically. Some of the early boilers had additional firetubes that worked in the steam space above the water level -- you can fairly quickly see why these failed to thrive if you calculate the corresponding stresses in the front and rear tubesheets alone under variable firing or steam pressure. If you use a separate shell, you come up against the age-old problem of parallel cylindrical pressure vessels (and all that implies) stacked within restricted loading gage with lagging and so on. All this is avoided very simply and positively in the Schmidt system, and the 'header' system allows very quick removal of any particular section with minimal dead time for the locomotive. (The piping design of a given element was very complicated with proprietary empirical formulae heavily used -- according to Matt Austin, who is about the highest authority we could ask as he had an interest in the specific history, essentially none of these have survived, so we can't duplicate how the Superheater Company performed this service for its clients, but it most certainly did.)
However, after that long digression (which I have now left in only to keep the details up to help 'inform' any subsequent discussion)...
... Porta's arrangement shows three drums. Starting at the firebox, the first drum is the fire tube portion of the boiler. The second drum is the superheater (or reheater on some Mallets receiving exhaust steam from the rear engine before passing it on to the front engine for a second stage of compound expansion). The third drum is the economizer, a cousin to the feedwater heater.
We should stop here for a moment. What you are describing has a very different name, and is analyzed in a very different paradigm. This is Chapelon's 'sectional boiler' as interpreted by Porta, and it is not 'three drums' so much as a single convection-boiler-like shell that is partitioned (with 'leaky tubeplates' where there is supposed to be water mixing -- the partition sheet clearance-drilled so the tubes aren't a tight or 'upset' pressure fit, on the same general principle as the condensers in Holcroft-Anderson) to compose the 'drums' you mention. If you have access to the T1 Trust technical pages, Joe Burgard calculated the length of the 'economizer' feedwater-heating section and hence the theoretical best placement of the partition for it in the existing boiler shell for the specific characteristics of the T1 boiler; it was surprisingly short (at about 3' and change for best theoretical uptake)
The drum of the economizer is filled with water injected to boiler pressure but not heated to the point it boils. The superheater/reheater has the drum filled with pure steam. The actual boiler is partly filled with boiling water with steam space on the top.
If you run the thermodynamics, the arrangement as written is little better than the last evolved generation of smokebox feedwater heaters; it just moves it back to where the lolog heat-transfer from combustion gas is at a higher level. You still have all sorts of steam-flow and blanket insulation problems in how the steam takes up the heat under load, under the variable flow conditions that operation under variable throttle and load produces. You will note that even very late Chapelon designs explicitly preserve Schmidt-style elements (in reasonably enlarged flues) both for main throttle superheat and for compound reheat (see the 160 A 1 for perhaps the best evolved version, as it's clearly determinable from the existing elevation section, and that locomotive used particularly high-grade superheat for its novel cylinder-heating arrangement)
Of course the 'weight penalty' as well as cost of using little more than naked firetubes and flues for the superheater transfer are less than anything involving complex heavy elements and header convolutions serving the inside of those flues. I think that was part of Chapelon's thinking, at least, in proposing that arrangement -- net of the understanding that a substantial part of actual steam generation was being made from the radiant section, so de-optimizing the length of the actual convection section to enhance both the length and the available combustion-gas heat for the superheater section makes better sense than it would to, say, the pre-war Wagner in Germany.
One possible advantage of this arrangement is that you can get an unobstructed path of fire tubes connecting the three drums.
Not only is it 'unobstructed' -- they are in fact continuous firetubes running through the three 'sections'; you can't unbolt one and pull it out as a 'module' for servicing. (A boiler could be designed that way, but it would be vastly heavier, difficult to seal in pressure, and almost ridiculously prone to combustion-gas leaks in service. Be advised that one characteristic of many Porta designs is that they are not designed with typical North American servicing practices in mind...
You can clean or do other maintenance on the tubes without the back-breaking finger-smashing work of extracting those heavy superheater elements.
Also noted for contemporary competitors of the Schmidt approach, 'back in the day'. You might also mention the uninterrupted gas path with no weird induced turbulent flow (utterly incalculable in the days of the Superheater Company, and probably only vaguely predictable with the best nondeterministic modeling and math even now) from complicated multipath elements and their corresponding supports and spacers. (Meaning at least theoretically better flow through downstream Rankine-cycle recuperation a la Franco-Crosti, which you will surely be interested in if the complexities of a practical sectional boiler are intriguing enough to actually build)
In modern practice, you would build specific partially-robotic arrangements to pull and place superheater elements of any practical complexity or construction. In fact you are one of the leading authorities on what would be required, and how it would be constructed, operated and maintained.
Another advantage could be less pressure drop through the steam space.
This might be immaterial, as prospectively you have what is cumulatively a very large 'dry-pipe equivalent' (with carryover separation,etc.) between the two sections, but preserve a front-end throttle arrangement (whether multiple-poppet or Wagner -- not 'Waggoner' as Porta spelled it) far displaced from that transfer porting.
The problem is that flow, and flow stability (which may induce wire-drawing-like effects in the steam as it traverses the complex space) is likely to be determined only empirically, and in almost any conceivable design is wackily more complex than flow through even the most convoluted 'sine-wave' closed Schmidt-style elements. It has the nominal advantage, of course, that there is 'one degree of superheat' at the throttle, regardless of any differentiation of steam temperature in various areas of the 'superheater section' reasonably distant from there. I suppose it is worth modeling in one of the computer multiphysics environments like COMSOL (which I believe has good phase physics for water-to-steam) to see how it stacks up (no pun intended, but now that you mention it...).
A disadvantage could be that this arrangement doesn't get the steam hot enough?
Well, it SURE wouldn't get it as hot as a good Type E (with its four separate passes much of which is in the near-radiant gas flow toward the rear of the flues). But the argument here is more complex (and a bit more nuanced): a commonly-recognized emergent problem in the last decade or so of advanced North American steam-locomotive development is what one account termed 'crazy high superheat' when the locomotive is operated at very high speed for an extended period of time without proportional arrangements like superheater dampers that are used 'correctly'. Armchair thermodynamicists just love this, in principle, usually ignoring the effects of differential thermal expansion and particularly effective tribology on locomotives so affected. (As noted, the actual pressure contribution of high degrees of superheat is relatively small, as the enthalpy added incrementally for each degree of high superheat is likewise comparatively small (see even the rubber bible for the numbers and units of your choice) and so comes right back out with even fairly small expansion... the effect is on reducing both wall-condensation and nucleate condensation effects in the steam as it 'long expands' in economical high-speed operation)
Between the recent Tom Morrison book and the David Wardale book that is a bit older, anyone know anything about this?
Actually, I don't know Tom Morrison, so this is an opportunity in seach of a reader. [EDIT: I missed the paragraph with the cites, above] Post, if you can, a range of pages with synopsis of what he says on this subject, as that will advance the discussion before I can actually find the book and read it.
Wardale is of course famous for understanding that you have to DE-superheat your piston valves on a modern engine -- see the frankly both very good and fascinating arrangements he built on the Red Devil. (As I recall, the engine currently operates without the valve cooling operational, but with the passages for it still incorporated in the cylinder structure -- might want to e-mail him about this specifically.)
Does anyone know of alternative arrangements to the widely used Schmidt superheater? This is the kind where the superheater elements are tubes carrying dry steam, which make hairpin turns inside enlarged firetubes in a steam locomotives' boiler. One end of each element is connected to a manifold or "header" receiving steam from the dry pipe of the boiler whereas the other end connects to a second manifold feeding the cylinders. My understanding is that there is a "Type A" where the elements make only one loop whereas the "Type E" has each element making multiple loops into multiple tubes in an effort to get hotter steam.
Think of something akin to a fire tube boiler. The cylinder is filled with steam and tubes carrying hot gases pass between tube plates at each end of the cylinder. Porta's arrangement shows three drums. Starting at the firebox, the first drum is the fire tube portion of the boiler. The second drum is the superheater (or reheater on some Mallets receiving exhaust steam from the rear engine before passing it on to the front engine for a second stage of compound expansion). The third drum is the economizer, a cousin to the feedwater heater. The drum of the economizer is filled with water injected to boiler pressure but not heated to the point it boils. The superheater/reheater has the drum filled with pure steam. The actual boiler is partly filled with boiling water with steam space on the top.
One possible advantage of this arrangement is that you can get an unobstructed path of fire tubes connecting the three drums. You can clean or do other maintenance on the tubes without the back-breaking finger-smashing work of extracting those heavy superheater elements. Another advantage could be less pressure drop through the steam space. A disadvantage could be that this arrangement doesn't get the steam hot enough?
Our community is FREE to join. To participate you must either login or register for an account.