Admittedly, I never paid much attention to railroading prior to my interest in model railroading (the past year and a half or so). I got a copy of George Drury’s Guide toNorth American Steam Locomotives for reference and I find myself reading it like a novel.
What amazes me is a common thread throughout the book - how many locomotives just didn’t work out as intended or just didn’t work at all. The pressure in today’s business world to “hit a home run” every time seems absolutely opposite of the attitude back in the day.
Am I reading this right or did the successes far outweigh the failures? Things like (I’m paraphrasing) “26 locomotives of x type were delivered but were scrapped 2 years later and replaced by y,” seems like a recipe for disaster.
Some railroads designed and built their own steamers. They had hits and misses…the PRR comes to mind. Still, their hits were sufficient that, during the time when they were doing well economically, the PRR had a solid run. Later, they began to fool around with the Q, S, and T types, and only managed to put the T1 into production in quantities and with a real purpose…too late because they had to scrap them just a few years later…four or five, if I recall, and buy diesels.
If I have it right, the NP was offered a trial of the first 4-8-4 from whichever builder (ALCO?) that stood them up…on spec…and the rest is history. I think that some roads were impressed with some designs and bought locos that were ultimately rendered marginal for what was required of them in just a few months of change.
I’ve never seen any evidence that yesterday’s business world was low pressure or even less pressure than today. Railroads in particular are ultra-high pressure cultures, and the contrast to non-railroad organizations I’ve worked in are quite remarkable. The only work cultures I’ve experienced that I have felt comfortable in and compare well to the railroader obsession with 7/24 hard work and getting it right at all costs are the U.S. Marines and combat brigades of the U.S. Army. I suspect that U.S. Navy warships have similar cultures but I have no direct experience to measure that. But your ordinary for-profit corporation? They think that us railroaders are workaholics that were expelled from charm school.
This handbook is a very useful reference source but no means should it be used as a quantitative analysis of steam locomotive design failure rates. The overwhelming preponderance of steam locomotive designs were just fine for their intended task. What you’re seeing is an artifact of the author’s personal obsession with iden
Since steam locomotives were pretty much custom designed to a certain railroad’s spec’s and needs, if a design turned out to be a bowser, was there any resale value in being able to sell the locomotive to another railroad?
Other than purchases during war emergency, sales of castoffs to short lines, and transfers between lines with common ownership, steam locomotive transfers from one railroad to another were uncommon. Examples of the first are wartime purchases by SP, Santa Fe, D&RGW, and UP. Examples of the third are between CB&Q and C&S/FW&D.
One exception is the resale of FEC 4-8-2s, but apparently those were pretty good locomotives no longer needed by a railroad with a big traffic reduction.
Castoff sounds a little more charitable than bowser. If a railroad found itself in need of ‘casting off’ a locomotive, did it get sent to the scrapper, or simply sent off to some "out of sight, out of mind’ part of the system?
If it was sent to another part of the system, it wasn’t really “gone,” was it? Indeed there were many locomotives cascaded to lower-density, lower-tonnage lines, but they weren’t necessarily poor designs, only no longer sufficiently high in productivity for the high-density routes. On capacity limited lines, maximizing train lengths and tonnages matter greatly, whereas on low-density lines there’s much less economic reason to fully tonnage every train, and often not the tonnage presented daily to require that.
Most of the locomotives sold to short lines were fine locomotives, just no longer needed by Class Is, generally because they were too light, too slow, too costly to operate, or all of the above. The short lines were very good at avoiding the dogs.
The junk no one wanted was often pushed into dead lines awaiting traffic upturns, and after awhile when it appeared it would never emerge it was scrapped.
I suppose I can agree with that. With attention span being what it is today, businesses, in many instances, can bury failures or spin them into some alternate outcome. Or we can hire consultants and blame them when things don’t work out. It’s kind of hard to hide a 200,000 lb locomotive.
I must do some more reading. This has become a very interesting subject for me.
IIRC, some (if they had the facilities) might re-use the parts of a locomotive that did work into something they could use - perhaps changing the wheel arrangement.
I fully agree with these comments. The vast majority of new steam locomotives built after 1920 or possibly even earlier conformed to standards of design that had been proven to be successful. Whether built by ALCO, Baldwin, Lima or in the railroads’ own shops (PRR at Altoona, IC at Paducah, T&P at Marshall, Frisco at Springfield, etc.) there were very few locomotives that failed to perform well in the actual service for which they were intended. Those few that did not live up to expectations were either rebuilt or assigned to other duty for which they were better suited.
Sometimes locomotives were built in anticipation of a need that never materialized. For example, the DL&W bought, IIRC, five 4-8-2’s to handle the heavier passenger trains they expected as a result of continued long distance traffic growth. Along comes the Depression and instead of increasing, passenger traffic actually declines. These light Mountains were not f
I’d push the date for successful locomotives back to 1870. The 4-4-0s, 2-6-0s, and 2-8-0s of the 1870-1910 period were almost uniformly good. The exceptions between 1870 and 1920 were the compounds (except for Mallet type), three-cylinder locomotives, and various underboilered articulateds. The compounds, a fetish in the 1890s and early 1900s, were quickly simpled as the fuel savings did not equal the maintenance escalation. Some of the 3-cylinder locomotives were modified to 2-cylinder, and most of the rest were scrapped earlier than would be expected had they been 2-cylinder when built – or should have been scrapped earlier. The 3-cylinder locomotives are real head-scratchers: by that time people should have known better.
One might throw nickel-steel boilers into this group.
Some of the most successful standardized types of steam locomotives were designs created by the USRA, considering it was a federal agency, this seems counterintuitive. The PRR late steam experimentation was a anomaly as most in house designs were very carefully evolved up to that point and the PRR was very conservative in that regard. The best book I have read in that regard is “Black Gold. Black Diamonds.” Most steam locomotives had very long lives. Experimentation was not always a dead end…compounded steam was not popular here but was elsewhere. Many late steam developments were not imported over here from Europe. Baldwin experimented with alot of diesel designs that did not work or were not widely adapted…dual fuel…modular engine packages…are some of them. The Q1 was not a success but the Q2 was… it could have been improved but was too late in the game.
This was another common practice (simpling) I noticed throughout the book. Was the compound an idea that was just ahead of its time technologically, or would it have not worked in any time period? My modern-day logic says it’s a good idea but that’s my inexperience coming through.
Compounding was effective in certain applications, two railroads bought articulated (mallet) compounds as the last (C&O) or nearly last (N&W) steam locomotives. It is hard to build a locomotive that produces high horsepower and also uses compounding.
Regarding locomotive failures, a newer example of a less than successful design was the Alco designed NYC (B&A) A-1 Berkshire, within its limitations the NYC made them work, but the other purchasers of locomotives built to the same design (IC, AT&SF, MP) quickly rebuilt them to a different locomotive. In the MP’s case they reused the firebox, frame, and trailing truck and rebuilt them into successful 4-8-4s. The AMC designed Berkshire (NKP, C&O, PM, W&LE, RF&P) was more successful, it came out after the limitations of the Alco design had become apparant.
I think the prime reason so many compounds were converted to simple expansion was to boost their tractive effort. On the negative side this resulted in an decrease in the thermal efficiency of the locomotive.
The KCS had seventeen Class G-1 and G-2 2-8-8-0’s which were originally all compounds when delivered from Brooks and Schenectady in 1924. Ten of these remained in service as compounds their entire lives. The other seven were rebuilt to simple expansion during WWII. All were scrapped in the 1947-53 time period. As compounds their tractive effort was 122,000 lbs and as simples 147,220 lbs, an increase of 20.7%. This is probably representative of the relative TE’s of a compound and simple version of the same locomotive.
Compounding was done to increase thermal efficiency by utilizing the power remaining in the steam exhausted from the high pressure cylinders. The technology was well advanced and compound locomotives saw many years of highly successful service. Mr. Hadid is correct in mentioning that the compounds were more costly to maintain but IMHO the objective of increased TE was more of a factor in the decision to convert those that became simples. I have no idea as to the numbers but would be interested to learn how many compound locomotives were simplified and how many remained as compounds for their entire lives.
Compounding worked well for Mallet articulated locos, which only had two cylinders driving each set of drivers, but only on those railroads which operated Mallets within their proper low speed environment. OTOH, several railroads bought high speed simple articulateds, then used them in low speed drag service for which a Mallet would have been a better choice. N&W, which had the Y’s for drags and the A’s for manifests - and assigned them to appropriate trains - got it right.
The compounds most usually ‘simpled’ were those designed by Vauclain (Baldwin,) Cole (Alco) and others which had both high and low pressure cylinders driving the same set of drivers. The increased complexity of additional cylinders, and the asymmetric stresses caused by unequal cylinder thrust, cost much more in maintenance than the improved thermal efficiency saved in fuel.
Across the big puddle, our French brethren had great success with compound locomotives - but their engineers were trained to handle the much more complex controls fitted to their locomotives and could make far better use of the principle.
N&W Y6 locomotives had Intercepting valves to allow HP Steam to be admitted to the LP cylinders for starting and low-speed operation. I think another factor in the success of locomotives like the Y6 versus earlier Compounds is a greater understanding on a scientific basis of the priciples and requirements. Things like test plants and dynamometer cars had put facts and figures at the disposal of designers who would have received a University Education versus, earlier people who would have learned solely by apprenticeship.
Kind of my point above - maybe it took a while for everyone to catch up with the technology and by then there were just better alternatives.
I can remember way back in the early days of IBM PC’s - we had this wonderful piece of technology but no real applications to run on it. I was running IBM 370/148’s and 4331’s at the time (yea, I’m somewhat of a relic). Our IBM CE’s laughed at the PC and really weren’t worried because they figured it wasn’t going anywhere…hmmmm.