I'll just toss some things in:

Properly-engineered 4-4-4-4 Duplex, conjugated via Riley Deem's system (as I interpreted it) with Spicer drives incorporated in the cannon boxes of the two drivers adjacent to the rear cylinder block, and a locable Ferguson viscous clutch between. Detented to lock preferentially at 45-degree phasing between front and rear engines.
Slip control via air-over-hydraulic caliper independent, working on cheek plates on the driver like the brakes on the early AEM-7s. Traction control by modulating the same.
All the boiler controls needed to facilitate sliding pressure firing -- you know why.
Snyder preheater, Cunningham circulator, optimized circulation at the top of the water legs. GPCS or torrefied fuel if you want'em. Either proper poppet valves, or a scaled version of David Wardale's articulated valves, with proper steam chest volume and doubled (one admission, one exhaust). Franco-Crosti-style economization with FGR to the air preheater.

Regarding fines in the stoker: even a cursory read of Louis Newton's 'Tale of a Turbine' would disabuse you of the idea. Go with the early-'50s recommendation: coal about 2" size, graded and washed. Dip it in dolomite coating a la some of the 'clean coal' approaches, to flux any glassing in the ash. (Shades of the White Train!)

Use Russell Brown's asynchronous compound idea for the exhaust -- it passes through a good-size turbine driving a traction alternator, putting minimal flow disturbance on the cylinder exhaust. TMs on the tender axles. Lewty booster (hydraulic transmission from a properly-located small triple-expansion engine) for driving the stoker engine, feedwater heater pumps, booster engine, etc.

Insulation via modern 'nano' approach -- about 25 little aluminized-Mylar or Kapton layers, with dry nitrogen in between, the layers held apart by very small 'caltrop'- shaped ceramic spacers.

Concur with Hugh on the welded construction -- laser keyhole welding; HIP components, drivers of course have to be cast in one piece so ... WebSpoke? SCOA-P?
Dynamic balance to within a couple of lb at AAR max diameter speed (510 rpm?) and Cossart rod drive (with the salmon rods) complete with pumped mass to give perfect reciprocating balance with minimum impact on vertical moment. Air-conditioned cab; clear vision forward by providing redundant cameras and scanning devices, including a persistent scanning display to give large angle of lineside vision.


OR: something better.

Modified version of the PRR V1 turbine: two 4-8-0 chassis elephant style, full-flow ceramic turbines, *mechanical* drive, but with Bowes drive and magnetorheological lockup. Two or more speeds, and you're already paying for the gearcase because reverse is via an idler at each turbine pinion -- that gives you high starting TE, flexible band in midrange speed, very high top end, out of one locomotive with no back EMF issue. Perfectly practical to have paired rotors in the turbine design, with HP admission at the center and large exhaust plena at the ends, and the driving pinion in the center like the old Honda F1 engine.

The original had a modified Q2 boiler, most of the problems with which would have been solved with proper welded construction. Circulation, preheat,etc. as above. You will almost certainly want some version of Holcroft-Anderson 'recompression' to get the effective water rate down; 8000+ HP is thirsty work otherwise. You would want better primary suspension, but that is hardly rocket science...


OR

Take a leaf from the team at IAV/enginion AG/Amovis, or Harry Schoell, and build an AC-drive hybrid genset locomotive using ultrasupercritical steam (see [url=http://projekt.beuth-hochschule.de/fileadmin/projekt/sprachen/sprachenpreis/erfolgreiche_beitraege_2007%2F2._Preis_07_-_Zero_Emission_Engine_-_Saskia_Scherfling.pdf[/url]this useful file[/url]. 7250psi at up to 900C superheat, delivery of the supercritical 'liquid phase steam' through what is basically a modified IC engine fuel injector. Use these to drive the MegaGens developed for the ALPS locomotive, and then use... well, any high-speed electric running gear; watch this if you need any reminders.


OR

If you don't care about giving world terrorism a big leg up -- design a locomotive optimized for the Oxford Catalysts cycle, which produces 11 molecules of superheated steam at the correct level to use in a reciprocating locomotive from ONE molecule of methanol, plus 30-35% H202 and clean water (which can be recovered from the engine exhaust). Methanol (or, if you are adventurous, dimethyl ether) is easily produced from plentiful natural gas (or synthesized from renewables); there are natural sources AND 'green' methods of concentrating the H202.

The one catch, of course, is how easy it is to make kilotons of TATP. Rats!!!

I leave aside consideration of designs like those of the 5AT project, or Turbomotive 2 (you will have to look it up via the Wayback Machine, but it's worthwhile...) because those deal with common-sense applications and not 'ultimate' ones.

If ya got questions, ask 'em.

Yes I have a question. What does "Properly Engineered 4-4-4-4 Duplex" mean? Wasn't it properly engineered the first time?

Brian

No, it was not. I won't go into the full litany of complaints, but there were a number of interesting ones. By 1947, for instance, there was lateral motion on ALL FOUR DRIVER AXLES.

The biggest issue imho was the high-speed slipping. The premise of the 'double Atlantic' balancing turned out not to work right, even with the FA cranked up to 4.21 or whatever; I have never had trouble understanding exactly why that was so. What was needed was a conjugating system akin to what Riley Deem proposed for the Q2 (it needed to be re-created from description and first principles, but that proved do-able).

Some of the problems were due to design assumptions that were not problems at design time but rapidly proved so. A whole group of issues centered around the general quality of PRR passenger coal in the late '40s: grate area effectively too small in some cases; crew blackened with low-hanging smoke after a run; that sort of thing.

A properly engineered T1 wouldn't have the valve gear in little OC boxes; it would use something like Franklin type B (as fitted to the 5500) but with dual ThrottleMasters on the two engines (complete with automatic slip positioning) and perhaps with some version of the Long Compression cam (and, in my opinion, modulated Okadees to control compression). Some possibility, at least, of using the type D system (see Kirchhof's patent in the USATC 611 thread) to arrest excessive slipping at any speed. The principal modification, as noted, is to provide a conjugating shaft that keeps either engine from *propagating* a slip.

Don't mistake me -- I love the T1 design, and something like it is almost certainly necessary if you want to reach speeds over 130 mph with two-cylinder engines without disaster. Whether any sane railroad management would run any reciprocating locomotive that fast as part of regular service is another question entirely.

Overmod's comments about the T1 reminded me of the high-speed slip tests performed by the New York Central on a number of locomotives, among them Hudsons and the Niagaras. This was to check out the counterbalancing.

What they did was lay some grease on a stretch of track, and run the locomotive, with train, onto the grease at high speed and high power output. When the locomotive hit that grease, driver rotational speed went into the stratosphere. Reportedly they attained rotational speeds well over 140 mph, which is getting into modern bullet train speeds; I've seen a quote somewhere (might have been second-hand) of rotational speeds up to 165 mph! High speed movie cameras recorded the action (and I wonder if that film is still around somewhere). Even at those speeds there was no lifting or driver bounce from dynamic augment, a remarkable testimony to the art and science of locomotive counterbalancing late in the steam era.

One thing that was a problem then, and would still be a problem now, would be getting the steam into and out of the cylinders fast enough at those speeds, which was the whole rationale behind poppet valves. Even a gas that approaches theoretical perfection as high temperature superheated steam does still has mass, and in most steam engine designs this is complicated by the fact that at least part of the steam has to come into the cylinders, stop, and be reversed to get back out again. That's something relatively minimized in a Uniflow engine, and it's almost completely absent in a turbine.

Fiddling around this thread tickled my brain cells, and made me remember this attempt at ultimate steam traction:

http://en.wikipedia.org/wiki/SR_Leader_class

http://www.semgonline.com/steam/leader_01.html

http://www.bulleidlocos.org.uk

Rotational speed QUOTED by Kiefer for this testing (and he would know!) was 161 mph.
He noted 'minimal lift' at that speed.

It should be remembered from Ralph Johnson's book, published a couple of years before the testing, that the stiffness of the track is important when considering augment. Think of the system as comprising two springs: one a kind of leaf spring in the beam of the rail, and the other the resultant of suspension forces on the wheel under consideration. (BTW, the Hathi Trust has provided an online copy of Johnson's "The Steam Locomotive" for you to read, so I will not quote him at length...) So it may not be surprising that this rotational speed was reached without wheel lift on NYC trackage, while reports of bouncing drivers on, say, E4s (someone out there in the Midwest, might have been C&NW, took the movie film of wheel lift) would also be credible, *even if given the same care in balancing*.


The point often missed in piston expanders is that they are more efficient, as a positive-displacement device, at part load. That is one reason why the S2 was highly efficient at continuous high speed, but lavish with steam waste when starting and in the early stages of acceleration. A key point with turbines is to accelerate them to a speed where 'slip' becomes tolerable, *even if you have not loaded them yet* -- and this is what the Bowes drive on the PRR V1 was supposed to do.

(Incidentally, according to material at the Hagley, PRR took some pains to correct the presumption that turbines developed low starting torque. They produce it just fine; it's that they use high volume flow (and hence, at low speeds, mass flow) to accomplish it, compared to pistons.)

Leader had some intolerable problems -- not least of which is that it was a long, complex locomotive meant to replace ... wait for it ... an M7 tank locomotive. (Shades of the TE-1!) Many of its problems were due to ... I don't have quite the right word for it, "kludging" perhaps, but that would imply it worked ... tinkering with the design in ways that did not improve it. The original designs were single-ended and oil-fired, either of which would have solved most of Leader's expedient service problems. And most of the testing troubles with the running gear, imho (and the failure of the last road test almost certainly) were the result of intentional sabotage.

That does not change the design problems with the locomotive: wheels too tall, inherently weak crank structure, unsealable sleeves, and a host of other problems, most notably including that the performance and thermodynamic advantages of the sleeve-valve system turned out not to mean much (even when not compared with more conventional systems accomplishing the same thing within the same packaging and envelope constraints). Some of the problems with the broken rings, etc. might be due to materials problems in impoverished postwar Britain -- but that is not a direct criticism of the engineering (can you say Meehanite? I knew you could...) and even today there might be difficulty in getting effective oscillation with those thin little ears, sealing the ends of the sleeves, and getting the required plethora of rings to seat and seal reliably without... well, without as much lube as a Merchant Navy would use on average... ;-}

Problem with the external chain drive MIGHT have been solved if double pinions had been used. At it was, the stresses in the main crank axle, already excessive, were made intolerable. OF course, perceived restrictions in the English loading gage were the reason the single chains were used; relieve that issue and most of the other ones become less critical.

As I have said elsewhere, Bulleid loved chain drive because he had seen industrial machinery using it. He appears not to have quite realized the problems inherent with reversing a chain drive, or putting severe periodic loads on it, or of relying on inherent precision *in bidirectional operation* when the chain started to wear or stretch. Apply this knowledge when looking at what worked, and what didn't, and you will have a much better understanding of why things went so wrong with Bulleid valve gear...

Personally, I think Leader might have been made workable if given a (single-ended) very large boiler, and the kind of drive found on the Paget locomotive (a large number of comparatively short-stroke cylinders en bloc between the axles). It's interesting to study exactly why the Paget locomotive was built as it was, particularly how it avoided the need for cranked axles entirely; had it been equipped with workable valve gear instead of the unsealable but thoroughly seizable rotary nastiness it turned out to have, there was a reasonable basis for a workable relatively-low-unsprung-mass locomotive. (And yes, you could have gotten reasonable power into a Leader-size power bogie.)

An interesting exercise is to look at the (much more workable) Irish 'turf burner' detail design, particularly the boiler and the cylinder/valve arrangements.

This thread is an interesting read that beings to mind the V19 1001, a V8 steam motor powered 2-8-2 , of which, there does not seem to be much information to be had outside of a general synopsis. The division of strategy between a modified classical reciprocating engine and that of other propulsion technologies remains an unanswered question as to which is the correct choice. I think there is a general naivety on the subject which is understandable.
Personally I would be pragmatic and follow Porta's more conservative approach of choosing a test bed locomotive for the purposes of research and development, and of all the new designs, the general design principles of the A5 would be my candidate for "new steam" from the ground up perspective in terms of a more conservative approach that primarily deals with more known operating attributes than unknown.
The fact that the British successfully funded a new steam locomotive from the ground up while news of possible new "copies " of old designs circulate, why the A5 design languishes, is a mystery to me and I am sure it's due to my own naivety. Perhaps, I am wrong, but once the ball was rolling toward the fruition of this design, I cannot imagine that the emergence of this new design would generate a great deal of attention...and excitement. Just as much or if not more so than the Tornado.
I think the operating environment is a more complex issue, although the current experimentation with natural gas by Class One's due to market forces leads one into projecting the trending of fuel costs, if one moves beyond the blueprint stage as in the saga of the ACE.
One aspect of this feasibility issue came to mind while watching an old film of a Railroad Fair and one locomotive was emblazoned with "American Railroads" which brought to mind how practical it would be for a scenario of shared costs among Class One's and economies of scale , and instead of all these separately funded steam programs there was one shared program that perhaps would use an improved, from the ground up design, that would have the advantage within it, to explore alternative coal fuel, rather than trying to rebuild and maintain aged examples. Naive? I suspect it is, and the nostalgia factor regarding historic steam would go against it. Yet, the appeal of American "know how" and an annual, national touring locomotive emblazoned with "American Railroads" still somehow appeals to the dreamer in me.

Yes, you can build a nuclear reactor power generating system small enough for the clearance envelope. Alco built one in that size range. Google "Camp Century" for details.

PC

As I recall, the main problem with nuclear powered locomotives is not the size of the reactor, it's the size of the required shielding, which is directly based on the power output.

Of course, I guess you could make the right-of-way a little wider and move the crossbucks back a few yards at all the road crossings and take care of that. ;-)

Boy, this should bring stuff out of the woodwork.

I'm getting the popcorn popped and the recliner ready.

I have had ghastly Internet connectivity problems this morning, or I'd have responded more quickly. I'm delighted to see that this has been moved out of the Railfanning section.

The 19 1001 is covered extensively in Gottwaldt's book on streamlined steam locomotives of the Reichsbahn, and in volume 5 of the Jahrbücher für Eisenbahngeschichte (Karlsruhe, 1972). An interesting precursor, according to Douglas Self, was a partly-built design for the LBE (interestingly, NOT covered in Gottwaldt, although he has a whole chapter dedicated to that road) which is here:

http://www.douglas-self.com/MUSEUM/LOCO ... lubeck.htm

It's useful to see how the progressive changes were made on the motor installation. Both the torque and longitudinal alignment of the motors apparently needed substantial 'beefing up'. The method used to control cutoff on all the motors simultaneously was interesting and unusual -- there was a reference on the Web but I have lost it.

Very high speed with reciprocating power might best be achieved with individual-axle drive, but VERY good slip control needs to be provided. If you thought a duplex was slippery... this will curl your hair. If the steam circuit is internally streamlined for proper 'breathing' at high speed, you will likely have the fast slip propagation observed on other highly-powered locomotives like the T1.

A 90-degree double-acting V2 is inherently in balance, and is thin enough to fit outside the wheels (suspension and frame are inside). Compare this with the French 232 P1, which used individual 3-cylinder motor units, and the 221 TQ that used a large motor at the front of the chassis (shallow-angle V12 if my reports are correct, probably single-acting).

My opinion is that the V-2 motor locomotive was better thought out than the Besler W-1. Aside from the unsprung-mass and lateral-compliance issues ... would you want complex motors hung down between the wheels? A specific point about the Barske-Roosens design was that the motors were modular; that loop at the top was intended for a custom hoist that would make R&R of an individual motor possible in no more than a few minutes. That eliminates putting the locomotive out of service for any motor-related fault; if you use care in the detail design, the motors could be changed out with the engine in steam. The masses balanced nicely (there is a slight obliquity in the balance, but that ought to be inconsequential to suspension performance) and the manifolding lines up more nicely than it would for inside motors all in line.

I am not sure what this "A5" is supposed to be ... do you mean 5AT? That is intentionally not intended as a heavy high-speed locomotive; you would need to modify it with a wide firebox and therefore a trailing truck, and at that point you're better off modernizing something like the 242 A1 than working with the size constraints of the LMS Black Five. I have no hesitation whatsoever in recommending use of the FDCs for locomotive design of any size. I agree that it is jaw-droppingly astounding that no one is underwriting the 5AT group to the extent of building a prototype, but there may be reasons for that I'm not 'privy to'.

There is a 'steam future' for natural gas firing -- but it certainly doesn't involve reciprocating steam; it would be bottoming for something like a GTCC. (I doubt that IC engines burning CNG or LNG would produce enough exhaust heat for significant propulsion steam; you're better off running the auxiliaries or ancillary systems from the bottoming steam source, or perhaps using ORC (organic Rankine cycle, fancy term for non-water vapor-pressure systems) for further energy recovery from exhaust.

You saw the film with the 'American Railroads' and don't have a comment on the locomotive design you saw? In THIS discussion? Have ye no SOUL? ;-}

The scenario of shared costs you describe has been tried. Look at the history of the Bituminous Coal Research effort to design a coal-burning gas turbine in the '40s and '50s. The sticking place for this has always been in two places: (1) where does railroad support properly become the responsibility of locomotive manufacturers or others who would benefit directly from the work; and (2) how do you apportion costs fairly when the member roads may not derive equal benefit from the work product?

Another interesting comparison would be to the USRA locomotive development effort. Yes, it would be nice to have a 'national incentive' to develop better locomotives that any road could use. You can bet your bottom dollar they would not be coal-burning locomotives under our present administration...


Over on the nuclear side: the PM-2a was the subject of a very comprehensive report written by James Barnett in 1961, which is well worth reading if you can find a copy. Enough detail is here, however, to draw appropriate conclusions:

(1) Fuel -- 93% enriched uranium-235 (similar to what Borst was planning...)
(2) Plant weight 800 tons (well, at least it's down from 2500 tons for the SM-1)
(3) 2 MWe may look like a lot, but it's about 2681 horsepower. And that's before you get into motor and control losses... etc.

I'd rather use a nuclear-electric 'battery' on a multi-power locomotive that could also act as a road slug or run from catenary or advanced third rail. Which is a very different thing from saying that I would RECOMMEND using a nuclear-electric battery on a real-world contemporary locomotive...

PCook is right: the way to use nuclear power for railroad use is to provide grid power which is then used for electrification. (Or you could use it for carrier-fuel synthesis or purification of normal locomotive fuels -- q.v.)

Looking up what I could find on proposed nuclear locomotives:

http://brainmindinstrev.blogspot.com/20 ... clear.html

http://books.google.com/books?id=bVMEAA ... ve&f=false

http://bbs.stardestroyer.net/viewtopic. ... 2&t=123883

Interesting that much of the interest in this subject is in the realm of Sci-Fi, Dieselpunk, and things of that nature, not so much in the "serious" rail enthusiast field, even though there was an apparent serious look at this technology. But then again, we have a lot of interesting stuff that actually did get built!

Just for laughs--an idea of what the future would look like for locomotives, ca 1920. The streamlining is one thing--but look at what it's on! I don't think it would have helped the Virginian's triplex that much!

http://www.invisiblethemepark.com/wp-co ... y-1920.jpg

I read with interest some of the comments on the Pennsy T-1 duplex, and have read about their purported "slipperiness" for years. Many years ago I was fortunate enough to meet with Paul Dietz (a Penny fireman in the 40's and 50's). He informed me that the problems crowed about are overblown regarding the T-1 engines . He said the problem lied with heavy handed hoggers who didn't know how to handle them. He said the good engineers never had problems with them, only the "hacks"as he put it. That is straight from someone who operated them. He did mention that his preference was the Q-2 "a miracle hauler" he called them.

FindersKeepers comments about his interview with a Pennsy veteran made me recall an article that used to be online, but sadly no longer is:

Chesapeake & Ohio Tests the PRR T1 - Chesapeake and Ohio Historical Magazine, May 2005 by Stephenson, David R - The C&O test report contains information that is not widely known, and some of it contradicts generally accepted beliefs about the T1

A discussion page on the T1:

http://www.railroad.net/forums/viewtopic.php?t=42536

Anyway, among the comments about the T1 on the C&O, a couple of things are still in my mind. One was that the engines actually ran fairly well on the Chesapeake & Ohio, mostly meeting schedules as required, with one notable exception, a stop at Waynesboro, Va. Apparently this trouble was anticipated; the locomotive was actually loaded heavier than one of C&O's own J3 4-8-4s would have been, and a switching locomotive in the area was put to work as a pusher to get the train moving; loss of time was only about 10 minutes. Acceleration was considered a bit sluggish, but that was blamed on a combination of taller drivers than the C&O used in that particular territory, and perhaps more on the lack of a booster on the trailing truck (reportedly C&O was supposed to get a booster-equipped T1, but that wasn't what was sent). The boiler steamed fine, there were no problems from that source. What was most notable was that the C&O noted there were little or no problems for slipping or adhesion, suggesting two things:

(1) By the fact that the C&O mentioned there was little slippage suggests the T1s were already gaining a reputation for being slippery. Of course, the C&O noted this was not a problem for them. This included a stop or near-stop at the top of grade to meet another train, and it was reported that the engine started and accelerated on this grade, with the train wrapped around curves, with no real problems.

(2) I can't help but wonder if the reason for this sort of performance was that the engine crews of the C&O would have been familiar with divided drive engines in the form of all the Mallet and simple articulated engines the C&O rostered over the years. This would have included something like 250 2-6-6-2s, in territory so rough in places that these engines were used the way other lines used 2-8-0s. Other articulateds on the C&O included the lumbering H-7 2-8-8-2s of the 1920s (which were the first large application of the simple articulated concept), and of course the famous H-8 2-6-6-6s. I do believe this experience with all these variations of Mallet types may have made a difference; on the Pennsy, articulated engines were limited to 0-8-8-0s in transfer service in the Columbus, Oh., area, and to N&W (and some purchased ex-N&W) 2-8-8-2s on the Cumberland Valley line in Pennsylvania. Elsewhere, divided drive engines and their peculiarities (including weight transfer between engines as water surged back and forth in the boiler during acceleration) would have been strangers to Pennsy crews.

I consider that blasphemy.