Three-Cylinder Mountain - New Haven R-3a and Southern Pacific Class SP 4-10-2

[Overmod]

M636C

But that was a second three cylinder compound in the USA. It had a curved nameplate on the leading splasher reading "Pennsylvania".

See if you can find a picture, or better, one of the contemporary accounts of construction or service experience in the contemporary trade press.  I'd think Sinclair in particular would mention this, and perhaps cover it in some detail even if short-lived or ultimately unsuccessful.

My notes are put away and inaccessible, and for various google-related reasons I can't find search terms that aren't swamped by T1 references, but someone can probably find a description and reference to the original PRR class T 4-4-0, which among other things had low running boards and splashers, a three-axle "British appearance" tender, and ... wait for it ... 84" drivers.  If I remember rightly this was built as a kind of von Borries compound, where HP was on one side, LP on the other, and as you can imagine some very clever use of a proportional 'starting valve' needed to get the thing up to speed and balance it.

Alternatively, the PRR a bit later (I think by the time of the Columbian Exposition) famously tried an imported de Glehn-du Bousquet 4-4-2, which in theory would have been one of the best approaches to pre-superheat 'compounding' in that great era of the compounding fad in the United States.  The accounts I've read basically say that the locomotive was built to contemporary European standards and some of the working parts were far too lightly constructed for American service; I don't recall reading anywhere in the accounts of the time that there was anything wrong with either the premise of the layout or the ability of contemporary engine crews to operate it effectively -- it being one of the more difficult arrangements to keep 'optimizing' during operation as the crew had two full sets of cutoff that had to be regulated independently.  (Not for nothing did the French have 'mecaniciens' instead of just hogheads on their compound locomotives!)  Of course this is on the 'opposite' side of a three-cylinder engine by being a balanced four-cylinder design. 

For no reason other than perhaps autodidactic sampling bias, I thought that most of the practical use of compounding in the early 20th Century was 'balanced compounding' using an equal number of HP and LP cylinders, which in practice with a quartered engine would be a four-cylinder design (as on the ATSF, or Vauclain type 2, or Cole compounding, to mention three potentially overlapping search references).  The Reading was the great domain of the three-cylinder express locomotive in the pre-WW1 years and as I recall at least one of these was a compound, albeit briefly.

[Overmod]

timz

Steamlocomotive.com is even wronger than that -- it says two high-pressure cyl and one low-pressure. BLW 60000 was the only compound 4-10-2 in the US -- the UP and SP engines were all simple.

I no longer know quite what to believe, especially hearing that someone thinks that two HP cylinders of the stated dimensions could exhaust into one LP cylinder of stated dimensions in any practical sense of the word 'economy'.

A more practical* approach to compounding is the version used on the Triplexes as designed, where one high-pressure cylinder exhausted into two LPs (the high-pressure engine was in the 'middle', with one cylinder exhausting forward and then providing the boiler draft, and the other exhausting to the rear and thence through an escape-pipe minimizing the back pressure there).

As I recall and vastly simplifying without notes, the 'optimal' expansion ratio with reasonable limitation of heat loss being taken is something around 1:2.3 for a Smith compound, which is IIRC what Baldwin 60000 is supposed to be -- a practical modern design would probably use explicit reheat as well as intermediate modulated steam injection (a better version of N&W's 'booster valve') to equalize the effective thrust from the three pistons at any given steady-state operation.  (In my opinion the greatest failing of traditional compounding schemes was not the "lack of expected power from the LP cylinders" but the performance and maintenance implications of lower effective thrust with a different risetime from the LP side per revolution working against the HP characteristics -- see Vauclain type 1 compounds for a fairly dramatic set of troubles so caused.) 

The whole premise of Alco three-cylinder engines was to divide the drive and permit larger horsepower without increasing the mass of the reciprocating parts so needed.  Much of what made this look attractive was the Gresley adaptation of Holcroft's conjugated valve gear for four-cylinder simples, which used a fairly simple linkage of pinned levers to derive the motion for a center cylinder's valve (with the valve rod facing forward instead of back) in a fashion easily adjusted for necessary angle, different size, or other characters of that cylinder.  There are some characteristics of Gresley gear that would not 'give good service' over an extended service period without more careful maintenance than typical shop crews might provide.  Note that it would be highly difficult, and even more difficult to adjust or maintain, a Gresley gear to provide the very, very necessary independent cutoff control for a LP cylinder vs. what the reverse provides for the outside HP cylinders, even if anyone were so stupid as to duplicate Webb's whopper of a misdesign.

In case anyone needs a good laugh, the Webb compound started by using two comparatively tiny cylinders for HP and one relatively colossal one (an account I've read terms this 'dustbin-sized' which is appropriate in the senses intended) in the center.  To compound the felony (pun intended) the HP cylinders drive on one axle, the LP on another, and there is no mechanical connection or even conjugation between the two.  Now, it would be difficult to find something even stupider to add to that, but Webb certainly managed; he did not see any need for a third set of expensive valve gear to drive that center cylinder, which after all would take only steam from the HP with its determinate cutoff settings, so the original version had only a slip eccentric that reversed 'itself' according to the direction the locomotive started moving under HP propulsion.  This was an obvious improvement over trying to 'simple' an engine with so disparate a cylinder dimension, or just leaving the LP out of near-starting TE with the engine in reverse, and it worked great until you, say, backed the engine down on a train and then (as was almost incredibly likely, if you think about it even a fraction of a second) slipped the HP axle when starting -- this would result in the HP happily spinning 'ahead' and the LP obligately spinning up on the exhaust mass flow -- the opposite way as the slip eccentric wouldn't have been pulled to forward motion by the locomotive actually moving forward yet.  One could not recover from this quite as elegantly as a problem with Winans valve gear, which required a starting bar to be inserted rather than some ad hoc reaching under the locomotive with something like a fire tool and pushing on things to set the LP in the right direction.  A very good thing Mr. Webb knew Mr. Moon so well, I think.  (At least a good thing for Mr. Webb, if not for the stockholders' interest or the folks who actually had to operate the railway...)

Apparently the later Teutonic 2-2-2-0s were somewhat better, and undeniably economical when running in steady state ... but as noted they did not last too long, once Webb was no longer dictator over the mechanical department, and no particular surprise. 

To my knowledge (which is almost entirely derived from a quick reading of 'Three Barrels of Steam') there were no compound experiments tried with the Overland types on SP.  In my opinion the UP Nines were an expansion on the 4-10-2 engines, and of course those were all obligate simples, so I find it hard to believe UP tried compounding either.  Certainly the production engines were all simple, and perhaps the problem is that we have armchair folks seeing two different cylinder dimensions and unthinkingly concluding 'compound' from that.

*Perhaps 'less impractical' has more appropriate semantic meaning...

[Jones1945]

M636C

The Webb componds of the London and North Western Railway, one of which was purchased by the Pennsylvania and was operated for some time, had two (small) outside high pressure cylinders feeding a single Low Pressure cylinder between the frames. The cylinder was similar in diameter to the boiler...

But that was a second three cylinder compound in the USA. It had a curved nameplate on the leading splasher reading "Pennsylvania".

Peter 

Thanks a lot, Peter. Details of the "PRR Webb Compound 1320" (a LNWR Dreadnought Class): https://www.deviantart.com/rlkitterman/art/PRR-Webb-Compound-1320-Pennsylvania-550259422

"The LNWR Dreadnought class was a class of 40 passenger three-cylinder compound 2-2-2-0 locomotives designed by F. W. Webb for the London and North Western Railway......"

https://en.wikipedia.org/wiki/LNWR_Dreadnought_Class

 

[M636C]

Deggesty

 

 

timz

Steamlocomotive.com is even wronger than that -- it says two high-pressure cyl and one low-pressure. BLW 60000 was the only compound 4-10-2 in the US -- the UP and SP engines were all simple.

 

 

 

 

Huh?! So the two high pressure cylinders exhausted into the single low pressure cylinder? What do they say were the comparative sizes of the cylinders?

 

 

The Webb componds of the London and North Western Railway, one of which was purchased by the Pennsylvania and was operated for some time, had two (small) outside high pressure cylinders feeding a single Low Pressure cylinder between the frames. The cylinder was similar in diameter to the boiler...

But that was a second three cylinder compound in the USA. It had a curved nameplate on the leading splasher reading "Pennsylvania".

Peter 

[Deggesty]

timz

Steamlocomotive.com is even wronger than that -- it says two high-pressure cyl and one low-pressure. BLW 60000 was the only compound 4-10-2 in the US -- the UP and SP engines were all simple.

 

Huh?! So the two high pressure cylinders exhausted into the single low pressure cylinder? What do they say were the comparative sizes of the cylinders?

[Jones1945]

timz

Steamlocomotive.com is even wronger than that -- it says two high-pressure cyl and one low-pressure. BLW 60000 was the only compound 4-10-2 in the US -- the UP and SP engines were all simple. 

Cheers! I am gonna inform them of the mistake after finishing David Paulides' book. 

[timz]

Steamlocomotive.com is even wronger than that -- it says two high-pressure cyl and one low-pressure. BLW 60000 was the only compound 4-10-2 in the US -- the UP and SP engines were all simple.

[Jones1945]

timz

Did they really say the SP 4-10-2s were compounds?

The caption below the image is probably created by the California State Railroad Museum. According to steam locomotive dot com, the SP 4-10-2s had one high-pressure cylinder 25 X 32, two low-pressure cylinders 25 X 28. A video of the SP-1s with sound: 

[timz]

The Trains table says the IC 4-6-4 had cyl 24-1/2 x 30, drivers 73-1/2, 265 psi, so its nominal TE was 55186 lb.

Did they really say the SP 4-10-2s were compounds?

[Jones1945]

I found these photos of SP SP-1 from the archive:

"Celebration for arrival of the first 3-cylinder compound 4-10-2 class of locomotive, 5000 series. It was the first 4-10-2 locomotive built for the SP. The view is looking east. Both ends of the Roundhouse are visible."

 

"Celebration for arrival of a new 3-cylinder compound 4-10-2 class of locomotive, 5000 series."

 

 

"The No. 5004 was placed into service July 10, 1925 and scrapped February 18, 1953. It was a 3 cylinder 4-10-2 freight locomotive built by the American Locomotive Company."

 From California State Railroad Museum (For fair use here)

[Jones1945]

Overmod

...The SP engines are the ones that were like 'having an expensive mistress at every division point'.  If you have an interest in them, you need to find and read "Three Barrels of Steam" -- a quick look on the Internet today shows copies available for $30...  

Thanks a lot, Overmod, I found a "copy" much cheaper than $30.  Yes, the NH R-3a have their own page on steam locomotive dot com but it is not mentioned on their page for the three-cylinder engines. According to the page you mentioned, these mountain type "could run like a scared cat, and pull an awful train at high speed. They could take 5000 tons up the Shore Line while an R-2a might take 3500 tons at best......"

Flintlock76

...Out of curiousity I went on steamlocomotive.com's locobase and found those IC 2600's had a tractive effort of a whopping 78,540 pounds.  If you're looking for the most powerful 4-8-2's that's got to be it...  

Impressive! When I searching IC's 2600s on the web, Illinois Central No. 1 appeared in one of the results, the only Hudson type in North America built (rebuilt from a Berkshire) for freight service which had a TE of 68,288 lbf! Oh my! 

 

https://www.deviantart.com/edjack14/art/Illinois-Central-No-1-696255073

 

[timz]

IC 2600s had two cyl 28 x 30, 275 psi, drivers 70, so nominal TE 78000+ pounds

Some IC 2500s had two cyl 30 x 30, 240 psi, drivers 70, so about same TE (this from the table in 10/48 Trains)

DRGW 3-cyl 4-8-2 had about 75000 lb nominal TE

NH 3-cyl had 71100 (and NH Power says NH never converted them to 2-cyl). Cyl 22 x 30, 265 psi, drivers 69

[Flintlock76]

A "Mistress in every division point?"  Is that like Zack the Engineer with a wife in every town along the line?

https://www.youtube.com/watch?v=5m_7WLW_h1o  

Back to the New Haven.  According to George Drury's book on American steam the NH's three cylinder 4-8-2's kept the third cylinder right up to their post-WW2 retirement, even though those third cylinders were a maintanance headache.

George doesn't say who had the most powerful Mountain types, just the heaviest,  Illinois Central's 2600-2619 at 423,893 pounds and the lightest, Alaska Railroad's 802 at 268,000 pounds.

I checked my Alfred Bruce book and interestingly he refers to the 4-10-2 type as the "Overland" type, built originally for the Union Pacific in 1925 and with three cylinders.  The Southern Pacific 4-10-2's came a bit later and were heavier than UP's.  Also, Bruce says 60 4-10-2's were built in total, and none as two-cylinder engines.  

Out of curiousity I went on steamlocomotive.com's locobase and found those IC 2600's had a tractive effort of a whopping 78,540 pounds.  If you're looking for the most powerful 4-8-2's that's got to be it.  

[timz]

No one knows which US 4-8-2 was most powerful -- NY Central? But we know all about nominal TE, and the NH engines weren't anything special. Guess the DRGW 1600s had top TE for 3-cyl 4-8-2s? Don't recall if the 2-cyl IC 4-8-2s beat them.

Did the NH engines really keep their third cylinder?

[Overmod]

We have posts about the glorious joys of the three-cylinder engine revolution (simple, to cut augment from heavy running gear, with Alco; compound, with higher boiler pressure and watertube firebox, with Baldwin) and the reasons for its effective demise even with cheap Depression resources on most railroads.

I have no idea why the New Haven engine is not mentioned; it was one of the most significant of them all. Did you try their page on "Mountain locomotives of the New York, New Haven & Hartford"?

One of the delightful S. Kip Farrington books has an account of a trip on one of these at 65mph (probably on the Speed Witch, although I don't remember if that was mentioned.  I suspect that money that could have gone toward converting them wasn't available at the time, say, LV was doing theirs, and a bit later every cent was being applied to the Alco DL109s which were seen as more flexible than any NH steam... someone with more knowledge of sad New Haven financial prospects in its latter years will have better details.

The SP engines are the ones that were like 'having an expensive mistress at every division point'.  If you have an interest in them, you need to find and read "Three Barrels of Steam" -- a quick look on the Internet today shows copies available for $30.  

[Jones1945]

Hello all,

I saw a listing on the "Bay" of a good looking New Haven R-3a in HO scale, a three-cylinder 4-8-2s manufactured by Alco. The front end of it reminds me of the Pacific of CNJ. I note that it is not listed on the three-cylinder steam engine page of steamlocomotive.com, so I presume the R-3a must be something long forgotten by the railfan.

I thought PRR M1b was the most powerful "Mountain" type steam engine in the world in terms of tractive effort until I heard about the New Haven R-3a, though PRR M1s were two-cylinder engine while the R-3a had three. I wonder if there were any interesting stories about the NH R-3a? Some sources say New Haven wanted to convert them into the two-cylinder engine but dropped the idea due to a financial concern, therefore they keep them running with 3 cylinders until 1951.

Another Alco product that caught my attention recently is the Southern Pacific "Southern Pacific" Class 4-10-2 three-cylinder engine. There is probably only one post on the Trains forum discussed the SPs of SP in the past. There were not many steam engine types that had the privilege of being given a name which was the same as the railroad they were serving. I wonder what was the reason they were given the "Southern Pacific" class name? Was it because they were one of the most powerful rigid frame steam engines in SP's fleet when built? Thanks!

NH R-3a #3555

SP SP-2 #5021

SP-1 when new

https://www.railwaywondersoftheworld.com/southern_pacific_locos.html