Trains.com

What steam we haven't seen - relaunch Locked

80132 views
507 replies
1 rating 2 rating 3 rating 4 rating 5 rating
  • Member since
    July 2008
  • 755 posts
What steam we haven't seen - relaunch
Posted by Juniatha on Saturday, December 29, 2012 9:39 AM

Hi all

The original thread was looking for what had been in the development line of the three major American builders in 1949 - along the lines steam had developed so far .

Yet , the thread went 'off-track' so to say and into exploring the unconventional .

Since most of the unconventional concepts would not have retained any resemblance with the classic concept of the steam locomotive but resulted in pretty much the kind of 'box on bogies' appearance of any Diesel or Electric - yet without , in my opinion , standing any chance of reaching competitiveness with Diesel or Electrics - they present a totally different look-out .

Sorry , at present the split-up " Extreeem Steeam !" is a mess - hopefully it will get sorted out ..

Meanwhile , let's get back to the original question :

What further classes of ( conventional , i.e. in line with development so far ) steam locomotive could - realistically - have been expected to appear from the three builders had steam development been continued post 1949 .. to , say , some ten more years ? 

I think , notes posted before on Duplex engines were a likely option .. what further wheel arrangements could have been turned out ?

And ..

Happy New Year !

Juniatha

  • Member since
    August 2010
  • From: Henrico, VA
  • 8,955 posts
Posted by Firelock76 on Saturday, December 29, 2012 10:29 AM

Hi Juniatha!   As far as developements in steam had it continued another ten years, I think the biggest change would have been welded boilers.  The Delaware and Hudson tried welded boilers in the late 1930's and they turned out to be very successful, so there's a good possibility steam builders would have followed suit.  Possibly other changes in manufacturing techniques as well, say making greater use of castings as opposed to machining from steel billets to speed up production.

Different valve systems might have just caught on, say poppet or Capriotti's, as it was they came along just a little too late.

Different wheel arraingements?  I don't know, I think all the practical possibilities had been tried so I don't know just where they could have gone past the 4-8-4 dual-purpose type.  The duplex types tried by the Pennsy turned out to be an answer to a problem that just didn't occur, not that they didn't work well.

The "what if steam lasted"  question kind of reminds me of the question of "what if the jet engine had never been invented" question in aviation circles.  The US Air Force might still be flying P-51 Mustangs!   Wouldn't THAT be something!

  • Member since
    December 2005
  • From: Cardiff, CA
  • 2,930 posts
Posted by erikem on Saturday, December 29, 2012 1:04 PM

The first thing that a railroad would be looking at in a new steam locomotive is how much money will it save to buy a new locomotive vs keeping an older one? Cost savings could come from reducing maintenance, increasing tonnage ratings and reducing fuel/water consumption.

I rather doubt that new steam would have been sold to the western RR's due to lack of water in some cases and long hauls for coal, along with diesels adapting well to both mountain and low grade railroading. Going to larger articulateds would involve investment in longer turntables, so I would rather doubt seeing exotic new wheel arrangements.

What would make sense for the likes of the steam hold-outs (IC, NKP, C&O, N&W) would be a 2-10-4 or 2-10-6 with a welded boiler, poppet valves, all roller bearings, booster and either a Geisl ejector or another improved front end design. In addition, there would be a lot of effort put into making maintenance as easy as possible - along the lines of what the N&W did with their post war locomotives. The idea of the improved Texas type is to haul longer fast freight trains than the existing Berkshires.

I would expect the combustion chamber to be even longer on a new design, if for nothing else than to promote more complete combustion for reducing air pollution.

Another line of development would be detail improvements to the Y6b.

The only other line would have been follow-ons to the Jawn Henry.

- Erik

P.S. Happy New Year to all!

  • Member since
    February 2005
  • From: Vancouver Island, BC
  • 23,330 posts
Posted by selector on Saturday, December 29, 2012 3:28 PM

Most of the recent contributors speak far above me for me to sound anything more than a dilettante.  However, there always seems to be the restrictions in the boiler and throughput for both boiler heating and steam efficiency.  I think this may have been the stemming point to the way the other thread evolved.

So, let us suppose the boiler and firebox configurations remained largely unchanged, including feeder and pre-heat and super-heat systems, syphons, etc.  Do we want an even larger boiler, either in diameter or length, or maybe some of both?   If we don't, how many coupled or driven wheels could be placed under the boiler, but not under the massive firebox needed for ever larger steam demand from larger boilers?   Or, should we keep the boilers smallish, but haul more capacity tendered and provide dual feedwater pumps?

This is my preamble to the question of wheel configurations.  Unless I am mistaken, a pilot truck of at least one axle is essential to help keep a steamer running smoothly down the curves.  Could we have steerable front coupled axles?   Diesels don't seem to need pilot trucks, and some of them move at a frightful clip when necessary.  But, if we need a pilot truck, it limits the number of drivers.  If you want to support a massive firebox, ditto.

I have always wondered if a 'transmission' couldn't be added to a steamer.  I realize that the cut-off provides a measure of steam efficiency, but I have wondered if the length of the main crank had to be fixed.  If we could figure out how to do it, the crank could be turned outward along a track of sorts for greater leverage on start-up.   While hooking up, the crank could also be screwed back up toward the hub so that there is less thrashing, a tighter turn about the center of rotation.  It would add complexity, surely, but also weight that would have to be countered.

I end my musings with the heresy that with steam locomotives, attempting to improve them in a way that retains their sensory and emotional appeal is like trying to make a silk purse out of a sow's ear.

Crandell

PS - I have long felt that a 2-10-4 is and was the epitome of steam success.  Lots of traction, tons of poop from the boiler if it steams well, and they can handle the tracks and vice-versa if their engineering and materials are right for the job.

  • Member since
    August 2010
  • From: Henrico, VA
  • 8,955 posts
Posted by Firelock76 on Saturday, December 29, 2012 4:06 PM

Considering Crandell's comment about a 2-10-4 being the epitomy of steam success, I think that the available wheel arraingments at the end of the steam era were as far as they could probably have gone without going to the ridiculous but then who knows?  Fifty years earlier (1900)  steam locomotive designers thought they'd gone as far as they could, and as we know events certainly proved otherwise. 

As a guide I think in this thread we should pretend the diesel engine itself was never invented.  As it was, it WAS invented, and steam was dooomed no matter what improvements could have been made. 

You know, in 1900 some railroad experts thought electricity was going to doom steam, and of course it did, but not in the way they expected.  

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Saturday, December 29, 2012 6:59 PM
OK, let's use 1950 (Cunningham patent) and 1952 (Smith's instantiation of Franklin type B on ATSF) as the relevant timeframe, as both those systems were essentially available in 1949.

You'd have Snyder preheaters, Cunningham circulators, and some careful applications of internal flow both in the radiant section and in the convection section (probably done through taps on the Cunningham system). Longer-travel valves, either with some flavor of Trofimov or with Wagner bypass a la ATSF. More careful attention to balancing -- rodwork closer to center, better suspension, better compression control (better than just Okadees with accumulator reservoirs) and the like. Zero overbalance, with better side bearings and lateral compliance (Fabreeka having been conclusively utilized in the Centipede tenders) and both the machinery and support for very precise dynamic balancing rods-on and then trimming for rod thrust.

Duplex would have to be conjugated to be practical -- Riley Deem's gears would do this. Not difficult to put analogue to Spicer drive on non-lateral-motion axles and conjugate either with a Ferguson clutch or the Bowes drive. I'd expect type B poppets on any really fast power, but NOT with the intentionally restrictive poppets put on NYC 5500 to 'match' sisters' IHP at lower fuel and water rate. If you're building an updated T1, you might not need Smith's recommended larger grate if all water-rate-reducing mods are implemented.

Far and away the best 'development option' for 1949 would be the PRR V1, which was actually approved for production during WWII and was only not 'proceeded with' because of practical road diesels. This had a Q2 boiler and two Westinghouse direct-drive turbines, each geared to four axles and 48" drivers. The thing that made this practical was the Bowes drive; with it, you have a 'dry' method for matching turbine shaft speed with road speed. Comparing anything contemporary with this -- the only thing that comes close is the S2 above around 85mph. Not difficult to modify the design slightly with '40s tech to address this issue (perhaps with 2-speed final drive and Maybach clutches). Should be mentioned that in 1949 the N&W variant of this, with the 600psi watertube boiler, was still slated to use the V1-style running gear... not full electric drive to all axles in that chassis...

I disagree with the idea of using relatively smaller drivers and better balancing to get better starting adhesion -- on the other hand, the 'sweet spot' for driver diameter could easily be right around 70" for that era., and there are a number of designs, notably the N&W J and A classes, that could easily be improved with flow streamlining, better combustion effectiveness, etc.

I do NOT see high pressure compounding being cost-effective in this era, even before you get to anything more complex than the Smith-style system on Chapelon's 2-10-4 design... which was likely too small for the ways American railroads would use ten-coupled power. I do NOT see turbine-electrics as anything but a boondoggle (no practical all-adhesion drive until later in the '50s, and even then largely driven by truck development for diesel-electrics). I do NOT see the S2/Turbomotive direct turbine design being developed further (if for no other reason than that the V1 already solves all the problems better, and eliminates heavy axle load and restricted number of drive axles immediately)

I'd have expected one exception to compounding: the improvement of the 'booster' system on the Y6b along general Chapelonian lines to equalize effective piston thrust on the two engines, which would have allowed practical operation up to the 40-45mph speed range with reasonable efficiency. That approach might have had a future compared to orthodox simple-expansion articulateds.

I think the superiority of 2-8-4s over 4-8-4s, and 2-6-6-4s and perhaps 2-6-6-6s over Challengers, would have been demonstrated at some point. There's a bit of fudging involved here because dynamic analysis of what that leading truck would have to do at high speed wasn't yet worked out -- but it poses no particular difficulty to implement with '40s means and materials, and wouldn't have involved expensive calculation or numerical methods...


RME
  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Saturday, December 29, 2012 7:06 PM
Crandall: two short points:

Change-speed transmissions on locomotives are difficult and comparatively expensive, because their construction has to be far more robust than just 'scaling up' examples from automotive practice. If you look at the final drives of any of the contemporary turbines, you will get some idea of what's involved. Then can be done, and a couple of my designs did in fact work them out, but the cost-benefit for railroads in our period would almost certainly not have made them worthwhile.

The one exception would... or perhaps 'might' -- have been a Ljungstrom-style balanced double turbine on an S2-style locomotive, where both change-speed and reversing could be introduced at the turbine output shaft. Synchronization is then the 'fun' exercise...


With regard to steerability, a proper 2-wheel lead truck IS essentially radial-steering, provided only that its effective radius and centerpoint give the correct geometry on deflection. The issue then becomes how the restoring force, compliance, etc are managed, especially at higher speeds, to give the 'right' combination of minimized flange force and proper
  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Saturday, December 29, 2012 7:07 PM
Oops... chassis steering. You won't do this just by separating vertical load at truck centerline and restoring forces, although that's a big step forward. Assume 3-axis strut control over axle motion, for example (it was contemporary in French electric locomotive practice, and could easily have been adapted...)
  • Member since
    February 2005
  • From: Vancouver Island, BC
  • 23,330 posts
Posted by selector on Saturday, December 29, 2012 7:13 PM

I should remember how to spell epitome.  My bad.

Thanks for your explanation, Overmod.  I'll continue to watch this thread with a view to learning more.

Crandell.

  • Member since
    July 2004
  • 2,741 posts
Posted by Paul Milenkovic on Saturday, December 29, 2012 7:38 PM

I am just scratching the surface in terms of figuring out "the story" on stability of wheel-on-rail, why steam locomotives needed "pilot wheels" and why Diesels don't.

On the 5AT steam project Web site, I found some discussion of tracking stability of steam locomotives, that most approaches to that were cut-and-try, but that a fellow named Carter from England had written scholarly mathematical and engineering papers on the subject in the 1930's.

If a wheel is forced to roll in not quite the direction it wants to go, there is this effect called creep that allows it to roll that way in exchange for offering some resistance force to being steered that way.  The amount of allowed creep is small, so for any substantial difference from the way to wheel is directed and the curve followed by the rails, the wheelset (wheel pair connected by an axle) will be in a partial slide.

In a conventional steam locomotive, the drivers are held quite stiffly in a rigid frame to facilitate connecting them with the rods to the cylinders.  Given the long rigid wheel base, the drivers will be in a partial slide for being pointed not quite the right way for anything but the most gradual mainline curve.  As such, creep forces do not come into play in stabilizing the tracking of the drivers, and the one paper by Carter I looked at suggests that the entire rigid wheel base can be treated as a single wheelset. 

A solitary wheelset is unstable and will "hunt" or "nose" back in forth in direction. When that wheelset is pointed a little off the track direction, it will roll to one side, placing a larger radius of the taper of one wheel on the outside track, the lesser radius of the taper of the other wheel on the inside track, steering that wheelset in the opposite direction.  Problem is that it overcorrects in steering and will hunt or nose back and forth as it rolls down the rails.

Carter's theory is that a set of steam locomotive drivers on a rigid frame will hunt or nose back and forth as does a solitary wheelset.  If one driver is pointed the right way on a rigid frame of long enough wheelbase for a curve, the other drivers will be slightly misaligned, but for most curves, the misalignment will be beyond what generates creep forces that could stabilize the locomotive and those drivers will be in a partial slide, which does not generate an increase in force for an increase in misalignment in such a way to stabilize the locomotive.

This theory suggests that a rigid-frame locomotive without carrying wheels will nose badly at any kind of speed, and guess what, 0-6-0 or 0-8-0 wheel arrangements are reserved for switch engines or pushers operating at only slow speeds.  I also read in Brian Hollingsworth's book about early express engines in England that also had carrying wheels but with such wheels part of the rigid wheelbase, and guess what, those locomotives were said to nose badly.

How about a 4-6-0, with a pilot-wheel truck and 3 driving axles in a rigid wheelbase?  Carter gives charts of speed ranges where such can be stable up to some critical speed, not only for the usual forward direction but also operated in reverse at speed (as in a suburban-service passenger engine).  It seems that the pilot truck with its rigid frame forms one effective wheelset, the 3 driver axles as a unit forming a second effective wheelset, and if the two wheelsets are allowed to pivot in relation to each other, but if there is some degree of resistance to their relative turning, either through friction of the pilot truck support pad or spring resistance to lateral displacement of the pilot truck from the driver centerline, you can stumble upon a stable locomotive.

So the narrative is that steam locomotive drivers need frequent "turnings" in the machine shop.  This much is true as you probably get flange intererence on a long rigid wheelbase on curves in the absence of "lateral motion devices", and on any substantial curve, the drivers held in a rigid frame are never quite pointed the right way and spend a lot of time in a partial slide.  The other part of the narrative is that the pilot wheels "help steer the drivers into curves", and that part of the narrative may be simply made up in the lack of a theory explaining how railway train wheels work.  Carter shows that the pilot wheels on a 4-6-0 can allow it to travel in a stable manner in reverse for a reasonably high speed.  How do the pilot wheels "guide" the drivers when they are in the trailing position?

Yet another part of the narrative is that some locomotives were known to buck like broncos whereas other steam locomotives were smooth riding up through 3-digit speeds.  I can understand that a railway vehicle without a swing-motion suspension such as on a passenger coach or a Diesel could ride badly, given the tendency of wheelsets and trucks to hunt or nose, but how is it that some steam locomotives (on account of the Pennsy T1 from interviewing engine crews comes to mind) are said to "ride like a Pullman"?

So what is amazing is that some steam locomotive designs rode smoothly, with no shock absorber type dampers that a modern Talgo has a half a dozen at the guided-axle articulated juncture between cars, with no swing hangers as on a passenger car or an EMD Blomberg truck, and they guided down the tracks well with no engineering behind them besides a builder's experience.  Someone needs to figure out a theory to explain this.

My dad had worked on the railway truck hunting problem in the 1970's, and he thought the problem was the fixation on bi-directional running that had become in fashion in modern railroad passenger operations of not turning trains on wyes.  Maybe the trick is an assymmetrical design such as a 4-6-0 Ten Wheeler type, a 4-6-2 Pacific, or the combination of a guide truck with a rigid multi-axle frame such as on a Centipede tender?  Do you suppose that a Centipede tender could be made more stable than a symmetric design such as a pair of 3-axle trucks?  Even for operations in reverse?

I am thinking that there is much that we don't yet know about steam locomotives let alone the modern HSR passenger cars and locomotives.  Maybe something akin to a Centipede tender could lead to a breakthrough in high-speed rail, but for deep physics reasons rather than the superficial explanation of the "pilot wheels guiding the wheels on the rigid frame"? 

If GM "killed the electric car", what am I doing standing next to an EV-1, a half a block from the WSOR tracks?

  • Member since
    December 2005
  • From: Cardiff, CA
  • 2,930 posts
Posted by erikem on Saturday, December 29, 2012 9:43 PM

Paul Milenkovic

My dad had worked on the railway truck hunting problem in the 1970's, and he thought the problem was the fixation on bi-directional running that had become in fashion in modern railroad passenger operations of not turning trains on wyes.  Maybe the trick is an asymmetrical design such as a 4-6-0 Ten Wheeler type, a 4-6-2 Pacific, or the combination of a guide truck with a rigid multi-axle frame such as on a Centipede tender?  Do you suppose that a Centipede tender could be made more stable than a symmetric design such as a pair of 3-axle trucks?  Even for operations in reverse?

Paul,

I'm curious about your having seen any of the literature describing the PRR's testing of the DD1 and GG1 locomotives? The PRR picked both designs after testing them in comparison with other electric locomotive wheel arrangements as the DD1 and later GG1 were easier on the track than competing designs. The PRR engineers attributed the difference to the asymmetric wheel arrangement of the DD1 and similar argument with respect to the GG1 where each half of the locomotive was asymmetric at least in regards to the pivot points.

My guess is the asymmetry would lead to two different hunting resonances, as opposed to two equal hunting resonances with a symmetric design. Note the "symmetric" steam wheel arrangements, specifically the 4-6-4 and 4-8-4 use very different designs for the leading and trailing trucks.

Keep in mind that the 2-6-2 was a popular wheel arrangement for rod type logging locomotives, with the trailing truck added to improve tracking in reverse.

[edit: A 1940's issue of Trains had a "spec sheet" for the Western Maryland 2-8-0's. What caught my eye was the axle loading for the lad truck was half of the axle loading for the drivers, i.e. the WM loco was an 0-8-0 with improved tracking.]

- Erik

P.S. A very interesting set of questions you've asked.

  • Member since
    November 2003
  • From: Rhode Island
  • 2,289 posts
Posted by carnej1 on Sunday, December 30, 2012 5:40 PM

Ironeagle2006

How about a 2-10-10-4 Simple Articulated Call it the Super Texas type.  Think about it the 2-10-4 Texas was considered the Best Fast Heavy Freight mover for a Single Coupled Engine out there.  Now take 2 sets of their Drivers set up for High Speed Running give them a Boiler and Firebox that could handle the Steam Needs and Turn that monster loose.  Bet the Big Boy and Alenghny lovers around here would go Crap we got Slammed. 

 Nope, the Big Boy and Alleghany lovers would say "oh crap, the 2-10-10-4s all got cut up for scrap because of their tendency to derail and destroy track". There really is a limit to how big a locomotive can be, I doubt even a 16 driver version of the Alleghany (i.e a 2-8-8-6) would have been practical.

"I Often Dream of Trains"-From the Album of the Same Name by Robyn Hitchcock

  • Member since
    August 2010
  • From: Henrico, VA
  • 8,955 posts
Posted by Firelock76 on Sunday, December 30, 2012 7:14 PM

This talk of wheel arrangements and diffferent internals is interesting, but I wonder what greater use might of been made of more modern materials.  Say for example, stainless steel for its corrosion resistance in non-boiler areas.  Or aluminum for running boards and cabs.  How about corrosion resistant or corrosion proof linings for boilers?   Pyrometers or other temperature sensors to monitor the firebox to prevent over-firing or under-firing?   Could greater use of roller bearings and automatic forced lubrication have been made?  I  could go on but I think everyone's gotten the picture.

The point is I think the basic machine had gotten as far as it could, what could have been done to make building one easier and extending the service life and maintanance intervals?

  • Member since
    July 2004
  • 2,741 posts
Posted by Paul Milenkovic on Sunday, December 30, 2012 10:32 PM

erikem

Paul,

I'm curious about your having seen any of the literature describing the PRR's testing of the DD1 and GG1 locomotives? The PRR picked both designs after testing them in comparison with other electric locomotive wheel arrangements as the DD1 and later GG1 were easier on the track than competing designs. The PRR engineers attributed the difference to the asymmetric wheel arrangement of the DD1 and similar argument with respect to the GG1 where each half of the locomotive was asymmetric at least in regards to the pivot points.

My guess is the asymmetry would lead to two different hunting resonances, as opposed to two equal hunting resonances with a symmetric design. Note the "symmetric" steam wheel arrangements, specifically the 4-6-4 and 4-8-4 use very different designs for the leading and trailing trucks.

- Erik

P.S. A very interesting set of questions you've asked.

Well, what is a "DD1" but a double 4-4-0 (Pennsy class D) wheel arrangement; a GG1 but a double 4-6-0 (Pennsy class G) wheel arrangement?  That much I had read about prior.

The interesting thing about a GG-1 is that it is two 4-6-0 Ten-Wheeler types connected back-to-back with a kind of kingpin drawbar, with one 4-6-0 at all times going full speed forward and the second 4-6-0 going full speed backward.  And the GG-1 superstructure kind of floats on top of those two sets of 4-6-0 running gear.

Who would have thunk that a Ten Wheeler could go at 100 MPH speed in reverse, with drivers leading and pilot wheels trailing, but it does as half a GG1.  OK, the pair of 4-6-0 running gears are tied together through a kingpin pivot.  I suppose you could do the same thing with a Ten Wheeler steam engine, connect it through such a drawbar to a 0-6-4 "pedestal" tender?  Arrange the coal space for vision from the cab and have a high-speed bi-directional suburban steam engine?

Again, Father thought that the problem of a stable HSR was the railroad insistence on bi-directional running; he asked me rhetorically, "You don't drive your car at 60 MPH in reverse?"  Maybe it doesn't depend as much as what direction you run the train, that the problem is a symmetrical running gear?

I guess I need to take a look for what is written on the PRR testing of the DD1 and GG1.  There may be "gold in them thar hills" to be prospected for stable HSR locomotives and coaches.

If GM "killed the electric car", what am I doing standing next to an EV-1, a half a block from the WSOR tracks?

  • Member since
    February 2003
  • From: Guelph, Ontario
  • 4,819 posts
Posted by Ulrich on Monday, December 31, 2012 8:47 AM

Past 1949 would have probably seen bigger locomotives employing more modern technology and materials. Maybe a new class would have been developed as well... like a "triple articulated"...i..e. a 4 6-6-6-4... greater articluation would result in less stress on curves and would allow somewhat bigger and heavier locomotives.

  • Member since
    December 2012
  • 193 posts
Posted by eagle1030 on Monday, December 31, 2012 8:59 AM

The first thing to run out in the tender is usually water.  I wonder if the western railroads would've followed the example of South African Railways' Class 25 and built condensing locomotives.  They were highly efficient machines, good for 500 miles between water refills.  They were very complex and therefore harder to maintain, but so was Pennsy's S2 turbine and (I know it's not a steam locomotive) UP's coal turbines.  I imagine at least one would've been built as an experiment by a western railroad given enough time.

  • Member since
    March 2016
  • From: Burbank IL (near Clearing)
  • 13,540 posts
Posted by CSSHEGEWISCH on Monday, December 31, 2012 10:03 AM

In the United States, extending range between water stops usually took the form of track pans (not that common), huge tenders (PRR "Coast-to-Coast" tenders) or extra water cisterns behind the tender (N&W and some other roads).  Condensing tenders would tend to be viewed as an unnecessary complexity.

The daily commute is part of everyday life but I get two rides a day out of it. Paul
  • Member since
    December 2005
  • From: Cardiff, CA
  • 2,930 posts
Posted by erikem on Monday, December 31, 2012 1:34 PM

Paul Milenkovic

Again, Father thought that the problem of a stable HSR was the railroad insistence on bi-directional running; he asked me rhetorically, "You don't drive your car at 60 MPH in reverse?"  Maybe it doesn't depend as much as what direction you run the train, that the problem is a symmetrical running gear?

I guess I need to take a look for what is written on the PRR testing of the DD1 and GG1.  There may be "gold in them thar hills" to be prospected for stable HSR locomotives and coaches.

Paul,

Let me know if you find anything interesting with respect to the PRR tests. Middleton mentioned the testing for both the DD1 and GG1 in "When the Steam Railroads Electrified", and the engineers postulating that the asymmetric wheel arrangement led to lower track forces. This was noted in the comparison between the R1 (a 4-8-4 wheel arrangement) and the GG1 (a 4-6+6-4 wheel arrangement). While Northern's have the same wheel arrangement as the R1, the engine trucks and firebox trucks are different designs and mountings.

I've also run across a couple of references about the articulation of the GG1 running gear (and other electric locomotives, e.g. the Little Joe's) reducing hunting as yawing of one of the trucks causes and almost equal and opposite yawing of the other trucks. In addition, the tractive force passed through the articulation acted to keep the trucks in line - as long as the locomotive was pulling, not pushing.

White's "The American Railroad Passenger Car" references testing on high speed trucks by Karl Nystrom of the Milwaukee Road, with comments about wheel shimmy. This included a quote from one of Nystrom's assistants in the March 1945 issue of Railway Mechanical Engineer, page 108.

Asymmetry by itself is unlikely to be a cure-all, as the "Maximum Traction" trucks used on streetcars were not know for good rides at high speeds.

- Erik

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Monday, December 31, 2012 4:33 PM

A little off-topic, but deserving of a response here because of commonalty to steam issues:

 

All you really need to know about PRR frame stability can be seen in the difference between the DD1 and the L5.  Staufer covers this relatively adequately.  Long rigid wheelbase is not preferable!  (Some of this might have been addressed on an L5 chassis with four-wheel lead and trailing trucks, or placement of the motors to give lower polar moment of inertia, but it still isn't anywhere near a high-speed locomotive!)

 

Likewise, the R1 had the high-speed issues of the P5, now with even longer chassis and four axles without optimized lateral compliance.  Not a 'failure' in the sense it was unworkable, just Not The Equal Of An Articulated Underframe.  Note that even the larger engines in the 1942 motive power planning all had articulated underframes (and 428-A size motors) -- the DD2, the "GG2", and the eight-axle pushers/snappers).  Note also how quickly and definitively the P5s were taken out of fast passenger service.

GG1 will run in push as fast as in pull, given a little bit of adjusted compliance at the articulation point between the underframes, and a bit of care with the lateral in the four-wheel truck and bolsters.

The big, big issue with the GG1 at high speed is precisely the issue that was identified: relative lack of braking power with mandatory lightweight consists.  Admittedly with the locomotive pushing, you shift the awful wheel wear to the consist more than the locomotive -- but you're still braking that outsized 242 tons with the independent, and heating up the tires so they come off.  We have modern technology to keep that problem minimized, but then you get into heat soak between rims and spokes and other matters  -- which you  don't have with simpler, lighter designs like the AEM-7 or FlexiFloat where guiding, riding, and steering can be handled independently... and all the adhesive weight is available for traction...

A key issue to remember about the carbody 'floating' on a GG1 is that none of the buff and draft force goes through the carbody at all.  It just floats on its two little pivots and side bearers and transfers the carbody and transformer weight properly to the underframe castings.  Important to remember this when doing the dynamic analysis in 'push mode' as well as understanding where the weight and stiffness in the chassis has to be...

 

The issue of 'steam in reverse' involves a few issues, some of which were solved in practice.  Stability issues from bidirectional Adams trucks was encountered, and addressed, on the German Henschel-Wegmann-Zug tank engines -- one approach is to use air to move the effective pivot point of one or both trucks so they're stable in the direction of running.  There IS a consequence for rod thrust and inertia when running a steam engine'backward' in that the vertical component of those forces expresses vertically UP rather than DOWN on a double-acting engine (model the forces if you don't believe this) which might exacerbate high-speed slipping on engines susceptible to that problem -- which is just about ANY workable reciprocating locomotive of cost-effective dbhp over 120mph or so...  You have increased dust and stone damage to the piston rods, but this can be addressed in the same general way that offroaders keep their shock-absorber pistons clean and intact.  One substantial advantage in practice to 'backward' running is that the smoke deflection becomes relatively unimportant.

You capture some of the importance of integrating the "4-6-0" with its tender via an articulated 'steering' joint immediately between locomotive and tender frames.  Some German power 'pushes' the forward tender truck up as far as possible to get some of this advantage, but I don't believe this was carried through to practical articulation, or even the use of 'spherical' Franklin radial buffering on that power.

Only two pedestal axles would be an unstable configuration.  Remember Dean's experiment with 'no fixed wheelbase'?  Same problem with pedestal... and you don't need the cast-bed underframe on that small a tender anyway.  All axles in a pedestal setup have lateral compliance (via the composite Fabreeka springs between the axleboxes and the coil primaries) and so a long 'rigid wheelbase' is beneficial for guiding and weight-bearing without introducing lateral binding.  Of course, if you are wying an engine with long radial overhang, you might want to put a Bissel truck at the extreme rear of the Centipede tender... note that this is NOT a Delta trailer, as the Lima drawings might have you believe, but something optimized to steering the rear overhang so the rear centipede axles don't pick or climb...  as the UP's experience has often encountered...

And yes, if you know what you're doing, and if you have a proper transmission, you can drive quite happily in reverse at 60mph -- it's just that the steering angles aren't optimized for driving in that direction, and there is no 'self-correcting' In steering motion.  The metaphor doesn't work... but the issues behind the situation there ARE relevant to many of our continuing discussions...

 

RME

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Monday, December 31, 2012 4:44 PM

  Triple articulated would be overkill for that wheel arrangement, and comparatively unstable compared to, say, a simple 2-8-8-4 with an axle-style tender booster.  You would NOT have a 2-6-6-6-4 as it's not dynamically stable to have two hinged 'floating' engines forward of a fixed chassis -- you've got multiple periods of oscillation and the chance to build up some truly nasty resonances, as well as having almost no way to keep all the required steamlines hinged and leakproof.  I'm prepared to say categorically that there's nothing I couldn't do with a 70"-drivered 2-8-8-4 that your x-6-6-6-x would possibly do, at virtually any speed, even assuming that you ran it in 'Allegheny-optimal' speed ranges with more or less constant output -- and there isn't anyplace I can think of where such a thing could have been done cost-effectively in North America.

 

If you are going to Juniatha's version of a Triplex, there is no practical way to drop the firebox between two of the engines, so don't bother -- go straight to the Mallet-Garratt configuration, which would be a 4-6-6-4+4-6-6-4 or 2-6-6-4+4-6-6-2 (you need the extra bearers under the necessary size boiler and firebox).  Easier to implement proportional compounding on such a thing, too.  Not *that* much more involved, particularly with better balancing allowing smaller drivers, to use eight-coupled engines.  As I said before in the other thread, I believe 10-coupled power, either Triplex or super-Garratt, is functional overkill for any service in the '40s or '50s

Obligatory gentle caning:  Remember that a locomotive isn't judged by its thermodynamic efficiency or its dbhp or starting TE alone.  It is a railborne vehicle, operating within relatively restricted loading gage, usually with greatly varying demanded power output.  "Locomotives" that don't do those things well, no matter how colossal they are, will not be successful *in-service* locomotives...

 

RME

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Monday, December 31, 2012 5:02 PM

The water rate is certainly the thing cited in PRR records as killing the V1 turbine after 1946.  The 'other half' of the lesson of the class 25 is that the evaporative capacity of the condenser has to 'scale' not only with mass flow of steam, but also the effectiveness with which the steam RATE of practical condensation can be implemented... not just a given mass flow per hour.  You can easily extrapolate what the condenser area on an 8000-shp locomotive would be.  (It was easy to recognize that the relatively tiny ACE 3000 would run out of practical condensation at somewhere around 110 F ambient temperature, and the thermodynamics truly and and promptly goes into the toilet as soon as the condenser operation degrades...

The practical issues of track pans with high-pressure alloy boilers and 'mandatory' water treatment is worth addressing.  Oddly enough, the system using treated water is on a grade, not 'flat', and requires some careful timing and space between sequential fillings to work -- I do not think any installation of that kind was made prior to the success of the diesel-electric, and the issue was never revisited because the losses involved in track-pan operation with water treatment were just too great in any case I'm familiar with, and perhaps in general.  Water purified and treated for use in even something as low as a 300psi boiler is not free; add the requirement for deoxygenation and the cost goes up even more... as does the possibility of dangerous problem if the onboard treatment fails to work properly.

The A-tank issue bears some careful looking at, as on N&W it wasn't strictly speaking used for 'longer' trips between watering... it allowed elimination of an intermediate stop which prevented gravity-grade' operation in one particular spot on the railroad.  In freight service you'd have to derate the consist in load by the tare and load weight of the auxiliary, and also in length by one or more cars (this is, remember, still the era of 40-foot boxcars) for siding and yard-approach length.  Having said that, there is little question in my mind that separating a 'coast to coast' size tender into coal and water 'modules' would make operating sense for a number of reasons... Europeans who are familiar with the RENFE 4-8-4s will readily recognize that such a setup would have made them much better machines all around...

I will re-introduce here the idea of recompression as a means of increasing the practical water rate.  I invite people to consider what increases and what decreases when this approach is used at various percentages of total mass flow (I'll 'spot' the start of this by noting that at large North American locomotive size it's probably not practical to recompress the whole of the mass flow, even net of all other cost-effective refinements in the Rankine cycle...

 

RME

  • Member since
    February 2012
  • From: CAPE CORAL FLA
  • 511 posts
Posted by thomas81z on Sunday, January 6, 2013 9:41 AM

Firelock76

This talk of wheel arrangements and diffferent internals is interesting, but I wonder what greater use might of been made of more modern materials.  Say for example, stainless steel for its corrosion resistance in non-boiler areas.  Or aluminum for running boards and cabs.  How about corrosion resistant or corrosion proof linings for boilers?   Pyrometers or other temperature sensors to monitor the firebox to prevent over-firing or under-firing?   Could greater use of roller bearings and automatic forced lubrication have been made?  I  could go on but I think everyone's gotten the picture.

The point is I think the basic machine had gotten as far as it could, what could have been done to make building one easier and extending the service life and maintanance intervals?

i have been on the same page for years , GREAT POINTS!!!! wow i have always said that with all the modern materials & sensors we could have  Optimum Performance, with ease , the 40's are over , its the 21st century . its well within our grasp

  • Member since
    August 2010
  • From: Henrico, VA
  • 8,955 posts
Posted by Firelock76 on Sunday, January 6, 2013 9:47 AM

Thanks Thomas, I was wondering if anyone was paying attention.

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Tuesday, January 8, 2013 9:09 PM

See my  notes.

Stainless steel is not particularly well-suited to boiler service -- at least the normal alloys.  You must take particular care with the water treatment, as there are weird corrosion mechanisms in these alloys.

Aluminum for running boards and cabs was a distinctive feature of the T1s; use of light alloys and composites WHERE JUSTIFIED would be a hallmark of 'future' power going into the '50s.  Just be sure not to get carried away and use lightweight construction of the wrong kinds in the running gear!  What worked on the Milwaukee A will NOT scale to higher thrust even in proportion... but the thin deep rods and thin bearings of the later Timken system will.  Nicely.

I think the 'best' protection system for boilers remains some version of electroless plating (a la Manhattan Project) -- you put inflatable balloons in the available spaces to reduce the volume of fluid required (very expensive, very toxic!) and arrange to 'tumble' the boiler in a jig (the same one used to allow all the fabrication and assembly joints to be downhand when made and NDT'd) to give complete coverage of all the nooks and crannies.  This is what I mean in my previous references to 'Parkerizing', but other chemistries could of course be used.

Avoid anything that puts a 'layer' of material on the outside of the flues and tubes; that will act much like scale.  And that is NOT good for integrity or life of the coating... and can damage the thermal-barrier coatings you're using in some zones of the radiant section -- but I digress.

Pyrometers were used on at least one NYC Hudson (in the four corners) to my knowledge, and this might have caught on if steam had lived longer.  Just be sure to avoid the siren call of unattended firing on locomotives -- you cannot do it effectively with '40s or '50s technology, even though all the physical systems are 'thinkable'.  I have covered this issue a number of times already in previous threads.

Big point about automatic forced lubrication: there were late experiments into making this low-loss.  There is a tantalizing mention (in the Encyclopedia of World Railway Locomotives) of an experiment which cross-drilled crankpins and journals as in automobile practice, and provided lube lines either integral or coaxial with the rods.  I leave the solution of how to get oil to this system 'to the reader' as it's a useful thought exercise to address all the design desiderata.

Look at some of the fabrication approaches used on Leader for inspiration... both pro and con.  I have not commented on hydroformed frame sections in lieu of casting as the required technologies were not mature in the period of discussion (although a case could be made for advancement of submerged-arc methods for heavy frame fabrication, if steam power had remained in hot demand.)

Maintenance practice at NYC and N&W goes a long way toward what was desired.  Other approaches, such as dynamic wheel balancing, were brought to an advanced state in Britain and Continental Europe, and could be easily adapted from there.

Keep going... and keep looking for precedents that were actually tried.

RME

  • Member since
    November 2003
  • From: Rhode Island
  • 2,289 posts
Posted by carnej1 on Wednesday, January 9, 2013 11:15 AM

Ulrich

Past 1949 would have probably seen bigger locomotives employing more modern technology and materials. Maybe a new class would have been developed as well... like a "triple articulated"...i..e. a 4 6-6-6-4... greater articluation would result in less stress on curves and would allow somewhat bigger and heavier locomotives.

There wer triple articulateds...called Triplexes both Erie and Virginian experimented with them but found that the boiler ran out of steam at higher throttle settings..they were 2-8-8-8-2s

"I Often Dream of Trains"-From the Album of the Same Name by Robyn Hitchcock

  • Member since
    August 2010
  • From: Henrico, VA
  • 8,955 posts
Posted by Firelock76 on Wednesday, January 9, 2013 7:30 PM

Hello Overmod!

OK, stainless steel is actually illegal for boiler use here in the US, it's subject to cracking due to the heating and cooling cycles a boiler goes through.  My postulated use for stainless would be in non-boiler areas such as the jacketing over the boiler insulation, the "skins" over the cylinder and valve units, hey even the whole tender for that matter.  Possibly even the ancillary piping on the locomotive itself.

As far as a layer of material on the flues or boilers innards to prevent corrosion acting like scale, remember a LITTLE bit of scale wasn't a bad thing, it helped to seal the flues and boiler joints.  Remember, that's a LITTLE bit of scale.

Anyway, the point I was trying to make, and mind you I'm not a mechanical engineer, just an amateur historian, is that in my opinion steam locomotive wheel arraingements had gone just about as far as they could by the 1940's, as far as I'm concerned reaching it's zenith in the 4-8-4 dual purpose locomotive.  I think any further improvements would have come in the manufacturing and upkeep departments.  Of course, we'll never know, will we?

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Wednesday, January 9, 2013 7:31 PM

carnej1

There wer triple articulateds...called Triplexes both Erie and Virginian experimented with them but found that the boiler ran out of steam at higher throttle settings..they were 2-8-8-8-2s

The thing to remember about these (and the other Henderson multiplexes) is that they are NOT three engines under a boiler,  Think of the Triplexes as 2-8-8-0s with motor tenders (as used on Southern and a few other places in this general form) and you'll have a better idea.  If you are going this route (as Juniatha seemed to have done) you want to remember to keep proper steering and buffer action between the motor-tender engine and the rear of the locomotive chassis; I do NOT recommend that this be the 'typical' Mallet double-hinge, and the flat-plate equalizing design used by Bruce et al. would produce waaaaaaaaaaaay too long a vertical-curve rigid 'wheelbase'...
You also have the design compromise of a shallow firebox over drivers, which is unavoidable in a Triplex.  You can go to a modified Meyer arrangement (which would be a 2-8-8-0-with-no-wheels-under-the-firebox articulated to an 0-8-2 tender) and it would be interesting to see how the buffer arrangement and articulation would be handled to prevent 'picking' the forward driver flanges on the tender.
RME
  • Member since
    September 2011
  • 6,449 posts
Posted by MidlandMike on Wednesday, January 9, 2013 7:49 PM

I had a thread on use of stainless steel on diesels.  Only a few early ones had it, and very few after that.  The consensus seemed to be that stainless is to too hard to work with on with something as maintenance intensive as an engine.  There may be more useful alloys for a modern steam loco.

  • Member since
    September 2003
  • 21,669 posts
Posted by Overmod on Friday, January 11, 2013 8:00 PM

Stainless was comapratively 'easy' to fabricate (via Shotwelding) and not *that* difficult to work with, if you understood the relative limitations of the material and the characteristics of work-hardening.  See the Reading Crusader and original NYC Empire State Express J3 (and, I believe, the 'Aeolus' Burlington Hudson design) for examples of how the material could be used on steam locomotives.

Problem is that the material isn't all that much lighter when used for the sorts of application on locomotives where aluminum was notably weight-saving.  Much of the 'point' of using stainless in car construction was that the panels could be corrugated for weight-saving at thinner sheet-metal gauge.  I believe if construction 'comparable' to aluminum's weight were used, the denting problem would be severe (and relatively easily noticed!)

If the hypersonic-aircraft revolution had caught on circa late '50s (with large amounts of brazed honeycomb structure in thin-gauge stainless alloys) there MIGHT have been some use for that material... assuming all the production was costed-down as it essentially had been for the original WS-125/XB-70 by 1960 or so.  Even then, it would be exotic and expensive compared to... aluminum... or one of the aircraft-grade aluminum alloys being developed in our period.

RME

  • Member since
    February 2012
  • From: CAPE CORAL FLA
  • 511 posts
Posted by thomas81z on Thursday, January 17, 2013 1:16 PM

Overmod

Stainless was comapratively 'easy' to fabricate (via Shotwelding) and not *that* difficult to work with, if you understood the relative limitations of the material and the characteristics of work-hardening.  See the Reading Crusader and original NYC Empire State Express J3 (and, I believe, the 'Aeolus' Burlington Hudson design) for examples of how the material could be used on steam locomotives.

Problem is that the material isn't all that much lighter when used for the sorts of application on locomotives where aluminum was notably weight-saving.  Much of the 'point' of using stainless in car construction was that the panels could be corrugated for weight-saving at thinner sheet-metal gauge.  I believe if construction 'comparable' to aluminum's weight were used, the denting problem would be severe (and relatively easily noticed!)

If the hypersonic-aircraft revolution had caught on circa late '50s (with large amounts of brazed honeycomb structure in thin-gauge stainless alloys) there MIGHT have been some use for that material... assuming all the production was costed-down as it essentially had been for the original WS-125/XB-70 by 1960 or so.  Even then, it would be exotic and expensive compared to... aluminum... or one of the aircraft-grade aluminum alloys being developed in our period.

RME

i was also thinking along the lines of graphite based lubricate / like slick 50 or a mirad of other lubricate producers .i mean heat is already delt with inside car motors

Join our Community!

Our community is FREE to join. To participate you must either login or register for an account.

Search the Community

Newsletter Sign-Up

By signing up you may also receive occasional reader surveys and special offers from Trains magazine.Please view our privacy policy