Union Pacific 4-12-2 Locomotives in the USA

Class UP-1 to UP-5 (Locobase 290)

Data from UP 10-1936 Locomotive Diagrams supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange; "Union Pacific Type Locomotive", Railway Age, Volume 80, No 24 (15 May 1926), pp. 1295-1300; "A Powerful Three-Cylinder Locomotive for the Union Pacific Ry", Railway and Locomotive Engineering, Volume XXXIX [39], No 5 (May 1926), pp 123-129; and "Three-Cylinder Locomotive", Brotherhood of Locomotive Firemen and Enginemen's Magazine, Volume 81, No 1 (July 1926), pp. 55-61. See also Alfred W Bruce,The Steam Locomotive in America - Its development in the twentieth century (New York: W W Norton, 1952), pp 302-303); and "The Collection: Union Pacific 9000", on the website of the Southern California Chapter, Railway and Locomotive Historical Society at [][] . Page last revised 1 July 2003, last accessed 2 January 2020. (Thanks to Chris Hohl for his 9 May 2015 email commenting on tender weights and capacities.)

Brooks works numbers were 66544 in March 1926, 67024-67034 in August, 67035-67037 in September, 67581-67584 in June 1928, 67585-67595 in July. Schenectady followed with works numbers 67596-67603 in July 1928, 67944-67949 in 1929 (month undetermined), 67950-67968 in July 1929, 68490-68504 in July 1930, 68505-68514 in August 1930. (NB: With the exception of two Mikado tanks for logging service and a snowplow, these were the last locomotives turned out at Brooks' Dunkirk Works.

Railway Age's detailed appraisal of the then-new design reported the railroad's desire to have an engine that would pull the same tonnage as their 2-8-8-0s at speeds then achievable by the the 2-10-2s and 4-10-2s then in service. (See Locobases 87, 16218, 6590, 16219, 6591 for the 2-10-2s; 13994 for the 4-10-2s; and 7466, 16520, 310, 7467 for the 2-8-8-0s.) Together with the RA, both the R&LE and BLF&E reports contain several pages of description and diagrams and together provide a comprehensive profile of the 4-12-2.

"In other words", the article said, "it was desired to have an increase in permissible speed from 20 mph (32 kph) to 40 mph (64 kph) and an increase in the average speed over the district of from 12 mph (19 kph) to better than 20 mph."

Choosing a three-cylinder layout allowed the locomotive to deliver its power to two driving axles, effecting "a bettter distribution of stresses over the whole frame structure''' ...[allowing] the use of six coupled axles together with comparatively high speed and greater power."

Valve motion had limited cutoff. The firebox's enormous heating surface included 62 sq ft (5.76 sq m) from five arch tubes and a long combustion chamber. BLF&E Journal observed that part of that combustion chamber was created by the Gaines wall placed just behind the last driving axle. Thus the effective grate was reduced, but overall direct heating service was preserved. Another benefit, added the Journal, of the layout was "the long distance from front tube sheet to cylinder center [which] should improve the draft condition by evening up the pull on the upper and lower flues". Unusually, both tubes and flues measured 3 1/2" in diameter.

Further thermal encouragement took the form of a Worthington 5-SA feed water heater. The SC Chapter of the R&LHS account noted that the 5-SA replaced the originally fitted 4 1/2 BL. "Changing to the 5-SA," the Chapter's description wrote, "allowed the weight of the components to be distributed along the length of the locomotive and and not concentrated in one spot and on one side."

UP-3s also substituted the American Throttle Company's front-end throttle at the superheater in place of the original throttles housed in the steam dome. The BLF&E's Journal helpfully provided a listing of the "special equipment" incorporated in this landmark locomotive:

Among the special equipment with which this locomotive is supplied is the following: Draftac spark arrester;

Franklin Precision Type "F" reverse gear;

Worthington feed water heater;

New York air pumps, brake and signal equipment;

Franklin adjustable wedges;

Elvin stoker;

Vapor Car Heating Co. steam heat equipment;

Vissering sander;

Detroit flange lubricator;

Nathan force feed lubricator;

Nathan hydrostatic lubricator;

Pyle National headlight equipment;

Hancock feed water equipment;

Franklin driving box lubricators;

The Superheater Company's Type "E" superheater;

Franklin Unit Safety drawbar;

Hancock injector equipment.

16-degree curves demanded flexibility, which the 9000's leading bogie accomplished by permitting it to traverse 6 1/4" (159 mm) to either side. Moreover, both the lead and rear driving axles used lateral motion devices. Indeed, stated the BLF&E Journal, the three castings that made up the front truck required no bolts and the truck was "perfectly equalized" without the usual equalizers" and its design permitted "the use of very long easy riding springs" (This may account, at least in part, for the 4-12-2's renowed ability to handle 60 mph speeds.)

The two outside cylinders had 32-in strokes; the inside cylinder was driven by Gresley conjugated gear and had a 31-in stroke. In a rare (for North America) instance of divided drive, the center cylinder drove the second driving axle, the outside two turned the third axle. Notice the very rare instance of identical tube and flue diameters. Although the center motion gave some maintenance problems and the fixed wheelbase was the longest ever mounted, these "Union Pacifics" were considered a great success.

Bruce notes with satisfaction that the locomotives had been delivered as "slow-speed units" with a restriction of 35 mph (56 kph). "However," Bruce comments,"because of the long wheelbase the engines were very stable riding [sic], and speeds of 60 mph or over were very frequently reported." They also handled 2 % grades and 6-8 degree curves despite their extreme length.

No other railroads built such long-wheelbase engines, however. Another unusual note about this class was that the highest axle loading was found on the trailing truck under the firebox and not one of the driving axles.

Note on the cylinder and valve gear layout: Based on correspondence with John E Bush and examination of photos in the latter's book, Wes Barris ([], viewed 30 March 2004) has concluded that the first axle was in fact not cranked. Instead, the spacing between the first two axles was 18" greater than was true of the other four. He notes that Barry Koeb, an R&LHS member responsible for the UP 9000 has been inside that engine's frame and saw no crank.

"The Cowboy" (permstev@cableone.net as of 30 March 2004) creatds new files for the virtual railroad simulator Trainz Railroad Simulator 2004. In his notes on the UP 9000 series he added quite a bit of additional information. The center cylinder and piston rod were raked at 8 degrees above the horizontal to clear the first axle. He continued: "This inclination resulted in a uneven positioning of the intervals for the 3 cylinders. When looking at the locomotive from the right side, the intervals are as follows rotating in a clockwise manner: 112 Degrees to the middle pin, then 128 Degrees to the left pin finally 120 degrees back to the right pin. This positioning gave the locomotive a very distinctive and uneven sounding exhaust."

TheCowboy says that the friction bearings in the Gresley gear wore quickly and that eight of the class were converted to a "third-link" motion layout in which a second Walschaert gear was added to the right side. This additional link transmitted motion to the inside cylinder through a rocker arm. Later 9000s had roller bearings on their Gresley gear and were not so modified.

He also notes that the air compressors on the early 9000s, whose smokebox location gave the class such a distinctive look, were easier to maintain when they were moved to the sides of the boiler. The result gave rise to the name "Bald-Faced Nines".

Brooks's first sets of tenders, built in March 1926 for the UP-1 and UP-2 batches, were equipped with an Elvin stoker, carried 15,000 US gallons (56,775 litres) of water and 21 tons (19.1 metric tons) of coal. Light weight came to 122,500 lb (55,565 kg), loaded weight reached 287,000 lb (130,181 kg). The basic structure remained unchanged in the later tenders, which had Duplex D-3 or D-1 stokers, but loaded weight and capacity increased to the levels shown in the specs.

Stoker issues began with the UP-1 and UP-2 batches (9000-9014) equipped with "an Elvin Mechanical 'paddle' Stoker, which also quickly proved to be inadequate" reported the Chapter description. UP-3s--starting with road number 9015-- were the first delivered with Duplex D-3 tower stokers, but they "were noisy and took up significant space in the small cab". Ultimately, Duplex D-1s replaced the D-3s beginning with their delivery in UP-5, the last batch of ten delivered in July 1930..