4-6-2 "Pacific" Steam Locomotives in the USA

Class K-1 (Locobase 7607)

Data from table in June 1906 American Engineer and Railroad Journal (AERJ) and from Erie 1907 Locomotive Diagram supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange.

These large-boilered Pacifics were the largest class of 4-6-2s built for the Erie. Alco built three (works numbers 30712-30714) in 1905, Rogers 41 (works 38578-39724) in 1906, and Baldwin 15 (works 32917-32929, 32937-32939, 32948-32950) in 1908.

Delivered with Stephenson link motion actuating inside slide valves, K1s were soon rebuilt with Baker gear and Schmidt superheaters; see Locobase 455

Class K-1 - superheated (Locobase 355)

Data from Erie 1930 Locomotive Diagram (noted as revised "Feb 1, 1927") supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange.

Delivered with Stephenson link motion actuating inside slide valves, K1s were rebuilt in 1912-1917 with Baker valve gear and outside piston valves. At that point, the renovators removed 17 tubes, reducing the tubes' heating surface. They also modified the firebox so that, including the 26.5 sq ft of arch tubes, its area increased 15.9% to 226 sq ft. And the makeover allowed an increase in boiler pressure to 215 psi and tractive effort to 32,540 lb. That increase was later retracted and the 1948 diagrams show 200 psi.

Class K-2 (Locobase 356)

Data from Erie 4 - 1928 locomotive diagram book supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange. See also "Pacific Type of Locomotives for the Erie Railroad," Railway and Locomotive Engineering, Vol 26, No 6 (June 1913), p. 210. Lima works numbers were 1230-1234 in 1913. ( Thanks to Chris Hohl for the valve gear ID.)

Five locomotives built by Lima in 1913 introduced mechanical stokers to the Erie Pacific. The firebox heating surface included 28 sq ft (2.6 sq m) of arch tubes. They were Lima's first rod-driven road locomotives (as opposed to switchers, e.g.) to be ordered by a trunk line.

The railroad press acknowledged the company's audacity. R&LE commented: "The new corporation ...bids fair to take a leading place as builders of high speed locomotives suitable for passenger or fast freight service."

And Lima was attracting notice for the quality of its work as well:"A notable feature in the construction of these locomotives is the excellent materials or which the various parts are constructed ar well as the marked superiority of the various appliances and attachments with which the locomotives are equipped."

10 Alco-built sisters (K2a; 1917; 2905-2914) had similar power dimensions; see Locobase 6671.

Class K-2-A (Locobase 6671)

Data from Erie 4 - 1928 locomotive diagram book supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange. Works numbers were 57521-57530. (Thanks to Chris Hohl for the valve gear ID.)

These locomotives were the ten sisters to the Lima-built K-2s described in Locobase 356.

The data as built were very close to the Limas and fit the reported differences well, so Locobase has taken the opportunity afforded by a much later diagram book to show the changes to the heating surfaces wrought by decades of operation and rebuilding. By now,

The firebox heating surface included 27 sq ft of arch tubes. These follow-ons to the K-2 (Locobase 356) had more superheater area as a percentage of total heating surface, but were otherwise similar to the 1913 locomotives.

Class K-3 (Locobase 128)

Data from Erie 4 - 1928 locomotive diagram book supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange, supplemented by data from reproduction of 1913 Alco Bulletin 1016 on Richard Leonard's [] (accessed 16 June 2006). See also long article "Experimental 4-6-2 Type Locomotive," American Engineer, Volume 86, No. 1 (January 1912), pp. 5-12; and "American Locomotive Company's New Pacific Type Locomotive", Railway and Locomotive Engineering, Volume XXIV [24], No 2 (February 1912), p. 50-51; and a very useful Steam Engine Reverse Lever: What Does It Do?" at [],4770579, last accessed 26 March 2022. Many thanks to Chris Hohl for his 22 September 2017 email reporting unlikely boiler pressure values for 177 entries. A Locobase macro caused the error .) Alco works number was 50000.

Originally built as an experiment by Alco-Schenectady. The length of the boiler tubes is exceptional for a Pacific of the time and highlights the size of the boiler. Locobase finds that the firebox heating surface area is a surprisingly small percentage of the whole, but observes that the great length of the tubes and flues are the principal reason. The 1913 table reveals that 28 sq ft (2.60 sq m) of the firebox heating surface consisted of arch tubes; it also credits the superheater with 897 sq ft (83.33 sq m), 25 more than the Erie book. Piston valves measured a healthy 14" (356 mm) in diameter.

The 50000 was also the first locomotive to use vanadium cast steel for the castings. (These were lined with vanadium cast iron bushings.) The American Vanadium Company claimed in American Vanadium Facts (February 1912, no page) that "use of this material together with the outside steam pipes has not only saved about 4,000 lbs in weight, but results in a decided simplification and improvment of the casting." Other components using vanadium steel included driving wheel centers, frames, rods, crankpins, piston rods, valve motion parts [e.g., link supports, link yoke, and link yoke bracket]."

The L&RE report focused on Alco's use of a screw reverse gear in place of the usual lever. (aka Johnson bar): "The screw reverse gear can be easily arranged to give about eleven times the leverage obtained with the reverse leveer as usually proportioned." Spinning the wheel allowed a complete reversal of motion in 4-6 seconds, the author noted, "probably faster than can be done with the ordinary reverse lever." The engineer needed to get into position "and well braced before he can exert sufficient muscular force to throw the reverse lever."

It seems that screw reverse gears never quite overtook lever reverse. Instead power reverse gear accomplished the same effort reduction by using a much shorter (12-18 inches/305-457 mm) lever to change the settings. (See the very useful discussion

Another innovation, this one critical to a time-saving proper alignment of the Walschaert valve gear, was the use of a self-centering guide for the valve stem, which could be "erected, taken down and replaced without any lining up, at the same time insuring [sic] that the valve stem guide is absolutely in line with the piston valve chamber."

Interesting note (dated August 1940) on later Erie diagrams specifically states that the weights do not include the 7,000-lb (3,175 kg) Worthington Type BL feedwater heater or the engine end of the Duplex stoker, which weighed 5,000 lb (2,268 kg). Later tenders held 9,000 US gallons (34,065 litres) of water and 20 tons (18.2 metrtic tons) of coal and weighed 197,000 lb (89,358 kg).

Drury (1993) comments that this engine lasted until 1950, considerably longer than most experiments.

Class K-4 (Locobase 358)

Data from November 1913 American Locomotive Company Bulletin 1016 on Pacifics found in [] and Erie 4 - 1928 locomotive diagram book supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange. See also DeGolyer, Volume 48, pp. 215+ and Volume 50, pp. 196+. Alco works numbers were 53451-53160 in 1913, 55853-55862 in 1916; Baldwin works numbers were 40812-40821, 40858-40859, 40893-40895, 40910-40911, 40935-40937 in November 1913; and 41640-41641, 41655, 41664 in August 1914.

Locobase reproduces the data from the Alco Bulletin as the starting point for this design. Firebox heating surface area included 27 sq ft (2.5 sq m) of arch tubes in the Alcos; Baldwin's firebox area measured 235 sq ft (21.8 sq m) of which 32 sq ft (3 sq m) were arch tubes. Steam admission to the cylinders came through 14" diameter piston valves.

Weight on the drivers varied slightly. One of the Baldwin specification pages (217 in Vol 48) has a very detailed list of of the component suppliers for this locomotive. It shows the web of suppliers that supported what was in 1913 the key transportation mode in the United States.

If the later Erie diagram books are a good guide (and wouldn't they be?), the Erie tweaked the boilers of the entire class to gain more heating surface. See Locobase 14495 for the result.

Class K-4 - modified (Locobase 14495)

Data from Erie 4 - 1928 and 4 - 1941 locomotive diagram books supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange. See also See also Chuck LaFever, "Farm Boy to Fireman", Steam Glory 2 Lost World of the Great Locomotives, Classic Trains Special Edition No. 5 2007. p32.

In Locobase 358, Locobase shows the large K-4 class as delivered in the 'teens. The new locomotives all had 261 two-inch (52.4 mm) tubes and 36 5 3/8" (137 mm) flues. Sometime after the engines entered service, they were all modified to the proportions shown in this entry. The flues each gained an eighth-inch, probably to accommodate a different Type A flue-element design, and two more small tubes joined the others in the boiler. Boiler pressure also increased by 10 psi.

After this makeover, the class was set for the rest of its long career on the Erie. LaFever's reminiscence about the Erie steam power he worked included the following about this design: "Hand-fired, she was a sweet steamer, rode like a Pullman, and had plenty of power for the typical train on this flat, fairly straight line. Of all the hand-fired switchers and road engines on which I worked, I like the trim K-4 the best."

Class K-4-A/K-4-B (Locobase 6672)

Data from Erie 4 - 1928 locomotive diagram book supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange. See also DeGolyer, Vol 69, pp. 52+. Works numbers were 56305-56310 in March 1923, 56479-56482 in May 1925.

Baldwin added these ten locomotives to the K-4 design several years after the earlier engines (described in Locobase 358). The basic power dimensions stayed the same: 14" (356 mm) piston valves fed steam to the same cylinder volume at the same 200 psi and the locomotives rolled on the same 69" drivers.

In the meantime, however, arch tubes had been replaced in many fireboxes by thermic syphons, so the firebox heating surface in these Eries comprised 14 sq ft (1.3 sq m) of arch tubes, but 58 sq ft (5.4 sq m) of syphons. Note, too, the high degree of superheat and the rearrangement of tube and flue dimensions. As delivered, weight on the drivers was measured at 178,180 lb (80,821 kg) and engine weight came in at 285,410 lb (129,460 kg).

Although they were intended for fast freight and were delivered with 69" (1,753 mm) drivers as in the K-4, their obvious value as passenger engines led to their being fitted with 75" drivers and cylinders bored out 1/4" more (to 641 mm) six years later and it's this version--redesignated K-4-B and trailing a much bigger tender--that is shown in the specs. In addition to the big tender, changes included a 4-ton gain in adhesion weight and a related increase of almost 12 tons in engine weight. In addition to the Duplex Type D16 stoker and Precision power reverse gear, the K-4-Bs also had a Worthington Type-BL feed water heater.

Locobase correspondent David Mainey added some colorful detail in a September 2012 email: "[T]he K-4-Bs seemed to be a no-nonsense engine, i.e., a loud

barking exhaust compared to K-1s and even the K-4s. I remember standing

next to 2748 leaving Glen Rock [New Jersey] with eleven commuter cars with a very

impressive sound. I miss that!"

A very similar overhaul of the design occurred with the Erie's K-5 USRA Heavy Pacifics; see Locobase 145 for the K-5, Locobase 9241 for the K-5-A.

Class K-5 (USRA) (Locobase 145)

Data from Erie 4 - 1928 and 7-1949-1 locomotive diagram book supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange. See also DeGolyer, Vol 69, pp. 68+. (Thanks to Chris Hohl for the valve gear ID and for his 22 September 2017 email reporting unlikely boiler pressure values for 177 entries. A Locobase macro caused the error .)

Firebox had combustion chamber. Intended to be the standardized Pacific heavy passenger express locomotive (Locobase 173) for the government-run USRA beginning in 1918, but only Erie operated any. These were built by Alco-Richmond (works numbers 59320-59329) and Baldwin (51802-51965).

As delivered, they had 28 sq ft (2.60 sq m) of arch tubes in the firebox, but the Erie later renovated the furnace along the lines of the K-4-A (Locobase 6672) and K-5-A (Locobase 9241) engines of the mid-1920s. In the process, two of the arch tubes were replaced, leaving 14 sq ft (1.30 sq m), but 71 sq ft (6.60 sq m) of thermic syphons were added. The tender also grew as it carried 4 tons more coal or 20 tons in all.

Class K-5-A (Locobase 9241)

Data from Erie 7-1949-1 locomotive diagram book supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange. See also DeGolyer, Volume 69, pp. 68+. (Thanks to Chris Hohl for his 9 May 2015 email and spreadsheet outlining differences in tender capacities and weights.) Works numbers were 56316-56318 in March 1923; 56416-56417, 56422-56424 in April; and 56536-56537 in May.

Based on the K-5 Heavy Pacifics delivered a few years earlier (Locobase 145), these engines had the same grate area and the 14" (356 mm) piston valves were no larger. But Locobase finds a discrepancy between the Erie locomotive diagrams in the six editions of diagram books that he can consult and the Baldwin specifications. Baldwin's superheater area is the same as several other similar Erie Pacific classes and is the number given in the specifications. Erie's diagrams consistently show the superheater area as 1,315 sq ft (122.15 sq m), even though the relevant flue length and count are identical to the other engines. The 1933 diagram has a mysterious "20" inked on the diagram near the tube length, but even that wouldn't account for all of the increase.

(Locobase has been advised by other researchers that such errors--if that is what this is--crept into locomotive diagrams over the years and often went undetected. Moreover, Locobase found an 80 sq ft spread in reported supeheater areas for 28 classes of locomotives that had 45 flues of 5 1/2" diameter that measured 19 feet in length.)

As in the K-4-As (Locobase 6672), thermic syphons accounted for 78 sq ft (7.25 sq m) of the direct heating surface while arch tubes contributed another 14 sq ft (1.3 sq m) and the combustion chamber added 66 sq ft (6.13 sq m) more. The 1923 tenders were marked for 10,000 US gallons of water on the rear ends; coal capacity came to 16 tons and the tender's loaded weight was 191,400 lb (86,818 kg). Stenciled on the front end of the tank was a possibly unique "Save Coal." As Locobase's specs show, these Pacifics pulled much heavier tenders later in their careers.

Class K-5-B (Locobase 16082)

Data from Erie 7-1949-1 locomotive diagram book supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange. See also DeGolyer, Volume 76, pp. 1+.. (Thanks to Chris Hohl for his 9 May 2015 email and spreadsheet giving the details for this lone K-5-B.) Works number was 59259 in May 1926.

Three years after the K-5-As (Locobase 9241) left Eddystone for the Erie, the railroad ordered one more that was to be based on the Uniflow cylinder design concept. Already tested on Erie's N-3 2-8-2 #3199, the Uniflow idea is described at that Locobase entry 78. A look at Baldwin's specs suggests, however, that unsatisfactory performance from the 3199 led to a redesign with conventional cylinders.

The design duplicated the K-5-As except for the squaring of the cylinder dimensions to 28 x 28. Thermic syphons accounted for 78 sq ft (7.25 sq m) of the direct heating surface while arch tubes contributed another 14 sq ft (1.3 sq m) and the combustion chamber added 66 sq ft (6.13 sq m) more.

In August 1940, the 2960 was modified to K-5-A configuration.