Atchison, Topeka & Santa Fe 2-8-2 "Mikado" Locomotives in the USA

Class 3100 (Locobase 5442)

Data from AT&SF 7 -1920 locomotive diagrams supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange. See also DeGolyer, Volume 46, pp. 174 and data from table in May 1916 issue of Railway Mechanical Engineer (RME). Information about how many, when, and by whom this class was supplied from [] (also visited 6 Feb 2006), the Santa Fe All-Time Steam Roster maintained by Evan Werkema. Works numbers were 39767-39768, 39803-39807, 39846-39856, 39888-39898 in May 1913.

Among the first of the Atchison, Topeka & Santa Fe Mikados, these were free-steamers by the looks of the numbers.

In terms of weight, the Santa Fe engines were about at the median for US 2-8-2s of the pre-USRA era. Relatively large boilers and 16" (406 mm) piston valves supplied the steam. Part of the firebox heating surface included four arch tubes comprising 25.4 sq ft (2.36 sq m). Locobase 5498-5499 show the 3129 and 3160 classes that built on Baldwin's experience producing this design.

Later in their working careers, the class was converted to oil-burning and supplied by a tender now carrying 3,300 US gallons (12,491 litres) of oil and weighing 166,370 lb (75,464 kg). By then, boiler pressure had been reset to 200 psi (13.8 bar).

Virtually all of the class lasted until the 1950s with the first to be scrapped (3114) going in July 1949 and 3116 broken up in December 1949. 3102 and 3123 were the last to go in June 1954.

Class 3129 (Locobase 5498)

Data from [] April 2003).

Data also from AT&SF 7 -1920 locomotive diagrams supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange and DeGolyer, Volume 61, pp. 31+. Information about how many, when, and by whom this class was supplied from [] (also visited 6 Feb 2006), the Santa Fe All-Time Steam Roster maintained by Evan Werkema. Works numbers were 43064, 43091-43095 in March 1916; 43138-43139, 43154-43155, 43186-43189, 43263-43266 in April; 43292-43299, 43352-43353, 43397-43398 in May.

Similar to the 3100 class that immediately preceded them (Locobase 5442). The heating surface area of the relatively short but wide firebox included 25.4 sq ft (2.36 sq m) in four arch tubes. Instead of the usual Von Boden oil-burner, the railroad specified a Booth system. The 200-psi boiler increased pulling power, sent steam to the superheater through a 7" (179 mm) stainless steel dry pipe, and capacious 16" (406 mm) diameter piston valves supplied steam to the cylinders. These had been modified from the earlier 3100 design.

The Baldwin specs included a 10 January 1916 letter from M H Haig, Mechanical Engineer from the Santa Fe's Topeka office, that spelled out how three selected engines (the first, 15th, and 30th) would be weighed. The engines were to be under steam and at the working pressure, coal on the grate of the coal-burners, and two gauges of water in the glass. All three selections would be weighed twice each. As it was, the railroad's diagram book shows a slight underrun of the 228,000 lb adhesion weight estimate, a 3,000 lb overrun for the engines, and a 2,000 lb overrun of the tender's weight.

A year later a still larger design of Mikados entered Santa Fe service; see Locobase 5499.

Like the 3100s, these engines saw out steam, retiring in 1948-1954.

Class 3160 (Locobase 5499)

Data from [] April 2003) and AT&SF 9 - 1920 Locobmotive Diagrams supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange. See also "Santa Fe 2-8-2 Type Locomotive", Railway Mechanical Engineer, Volume 92, No 2 (February 1918), pp. 83.

All of Drawing Room Specification Book Number 61 of the DeGolyer Library's Baldwin collection is devoted to Santa Fe locomotive orders in the late 19teens and 1920s. The 3160 class specs run from page 31 to 224 and cover seven batches. (Anyone wishing to build a Santa Fe engine delivered by Baldwin in these years could do so from these specs, assuming one had the foundry, components suppliers, overhead cranes, floor space, and the work force to do so.)

(Many thanks to Chris Hohl for his 9 May 2015 email and spreadsheet laying out in detail the various capacities and weights of the different tenders used by this class.)

Works numbers were

1917

October 46676, 46728, 46818-46819; November 46961-46965; December 47281, 47389, 47488

1918

January 47571, 47689-47691, 47743; February 47794, 47863, 47915-47917, 47964; March 48024-48025, 48077, 48146-48147; April 48263-48266, 48349-48350; May 48670-48673, 48730; June 48896-48898, 48948, 49082-49083; July 49163-49164, 49292-49294, 49389-49391; August 49470-49471, 49540, 49619; September 49923; October 50128, 50206, 50209; November 50549; December 50886-50889, 50990

1919

January 51028=51029, 51078, 51116-51117, 51150-51151; February 51231, 51252-51253, 51303-51304, 51380-51382, 51437-51439; March 51476, 51511-51513, 51554, 51575, 51580-51583, 51626; April 51659

1920

May 53272; June 53299-53301, 53318-53319, 53330-53331, 53350-53351, 53364; July 53376-53380, 53398-53399, 53430-53434, 53461-53464; August 53504-53505, 53526.

Preceded by the 1913 3100s (Locobase 5442) and the 1916 3129s (Locobase 5498), this class increased cylinder volume to compensate for taller drivers. Specified to have a boiler setting of 190 psi (13.1 bar), the class soon operated at 200 psi.

Firebox heating surface area included 34.2 sq ft (3.18 sq m) in four arch tubes space 16 1/4" (413 mm) apart; coal-burners were fitted with Duplex stokers. For all of the batches, all of the heating surface areas were calculated to the nearest 1/10 square feet: tubes 3,067.8 sq ft, flues 1,279.6 ft, fire box 232.4 sq ft, and superheater 1,085.6 sq ft.

All were equipped with Ragonnet Type B power reversing gear. 3160-3187 had Baker gear; 3188-3257 had Walschaert. Both types actuated 15" (381 mm) piston valves, which were an inch smaller than those in the earlier classes.

One clue suggesting why the railroad and builder switched to Walschaert after the first batch comes in the Baldwin specs' supplement paragraph 447: "Pilliod Co. to be warned to allow ample clearance in valve gear as RR [railroad] has had some trouble due to insufficient clearance especially with gear connecting rod."

Another watchword concerned the tires. Baldwin understood that the driver and tender wheel tires needed to meet ASTM specifications that included an acceptable variation among driver sets of 3/32". But the Santa Fe's Charles Riddell of the Chicago office noted in his 8 November 1919 letter to Baldwin's Vice President G Greenough that the railroad's Mr Purcell, who ruled on such matters, was "of the opinion that builders have instructions to turn both driving and truck tires when mounted on axles to uniform diameters, as to pairs." Riddell added that Purcell was considering whether to insist that all drivers in the the same set (i.e., all eight as opposed to each axle's pair) be turned to the same diameters. Given the force of axle loadings and tractive power on these tires, Purcell's position doesn't seem unreasonable.

As the data shows, these Mikes were considerable larger than the 3129s that immediately preceded them. If, as Drury (1993) notes, the 3129s were akin to the USRA's light Mikados (Locobase 40), then these were the Heavies (Locobase 41). They came in two subclasses -- the last 30 apparently being strictly coal burners while the first 98 were a mixture of coal- and oil-burning engines.

As late as the specs for 3202-3217, drafted 29 May 1917, the axle loading target was set at 58,500 lb (26,535 kg), but a inked correction dated 21 April 1918 reset the goal to 60,000 lb (27,216 kg). Supplement paragraph 36 suggests one (of many) reasons why this was of such concern when it prescribes the arrangement of equalizers according to the axle loading.

Locobase uses the limit-busting axle loadings laid out in the 1920 Santa Fe diagrams, which held true for all but the last 30. This set, ordered well after the Armistice on 11 February 1920, featured a sterner requirement in supplement paragaph 17: "Locomotives will not be accepted by the Railway Co. if the weight per each axle varies more than 500 pounds above or below that specified." The specs called once again for 240,000 lb on the drivers, 325,000 lb (147,418 kg) for the engine. The 1920 diagram for this set shows a maximum axle loading (first and last driving axles) of 60,450 lb (27,420 kg).. Engine weight came in at 328,920 lb (149,196 kg).

These engines were delivered with cast iron cylinders and later refitted with cast steel types.

By 1941, however, the entire class's maximum axle loading had risen to 67,000 lb (30,391 kg); see Locobase 16083.

Volume 61's page 143, part of the 9 July 1917 specs for the batch road numbered 3218-3227 includes supplement paragraphs 81-82, which describe in lengthy (but oddly fascinating) detail just how the finished locomotive is to be cleaned up for delivery.

Chris Hohl's parsing of the tender assignments reports the following combinations:

Road numbers 3189-3201; #3213-3287 #3160-3188; #3202-3212

Water capacity 12,000 US gal (45,420 litres) 15,000 US gal (56,775 litres)

Fuel 16 tons (14.55 mt) coal 20 tons (18.18 mt) coal

17.5 tons (15.91 mt) coal with coal boards

4,000 gal (15,140 litres) oil 5,000 gal (18,925 litres) oil

Loaded weight 232,700 lb (105,551 kg) 277,00 lb (125,645 kg)

235,700 lb (106,912 kg)

236,600 lb (107,320 kg) 298,600 lb (135,443 kg)

NB: In supplement paragraph 575 of the specs for the first batch (p. 79, Volume 61), the loaded tender weight as delivered is given as 242,000 lb (109,769 kg),

The class was big in both senses and performed throughout the system until 1950-1956. Also, the Santa Fe was sufficiently pleased with the design to go back to Baldwin for 101 more after the USRA era ended; see Locobase 2.

Class 3160 - updated (Locobase 16083)

Data from ATSF 1953 Locobmotive Diagrams supplied in August 2013 by Allen Stanley from his extensive Rail Data Exchange. See also E D Worley, Iron Horses of the Santa Fe (Dallas: Southwest Railroad Historical Society, 1976), pp. 186-191.

Locobase 5499 describes in detail the Baldwin-built Mikados delivered from 1917-1920. Almost two decades later, Santa Fe diagrams would show how much a main-line design might change to meet the demands. The basic power dimensions remained the same, tube count dropped by only nine, the fire box was essentially untouched. But adhesion weight rose by five tons, engine weight by a lesser amount. Tender weight for the oil-burner increased by almost 20 tons for the coal burners, 30 tons for the oil-burners.

There were other, more subtle differences. The engine now had an Elesco feed water heater, two sand domes ahead of the steam dome (the bell relocated ahead of the first dome), and an extended cab. The running gear's main rod was 8" longer (up to 119"/3.03 m). In both old and new configurations, the third axle took the main rod's thrust. In the 'teens, the journal bearing measured 12" x 20"; in the 1940s, that same bearing measured 12" x 12".

At least one engine--3192- was tested with a Worthington feed water heater in the late 1940s.

Their survival until the end of steam in the early 1950s testifies to the qualities described by E D as "well-balanced, free-steaming, capable machines".

Class 4000 (Locobase 15202)

Data from tables in 1930 Locomotive Cyclopedia. See also the Kansas Memory archive of Santa Fe diagrams at [], last accessed 18 May 2013 and DeGolyer, Vol 70, pp. 316+.

Works numbers were

1921

June 54860-54862; July 54894-54896, 54903-54908, 54915-54916; August 54937

1923

May 56510-56511, 56571; July 56697-56703, 56819-56823; August 56992-56994; September 57086-57087, 57115-57122; October 57242-57244, 57269-57271, 57318-57328

Looks to have been based on the USRA Mikado design (Locobase 41), but according to Wes Barris --[], accessed 21 June 2006 -- these were repeats of the 1917 3160 class (Locobase 5499). And in fact the design had the same firebox and only minor tweaks to the heating surface, which increased the superheat percentage. All of this group were oil burners.

A few of the earlier engines had their fireboxes retrofitted with thermic syphons that added 70 sq ft (6.5 sq m) to their heating surface areas. These were 4028, 4037, 4038, 4056, 4059.

The last 40 of the class were produced with more superheater area and fewer small tubes; see Locobase 2.

Class 4060 (Locobase 2)

Data from tables in 1930 Locomotive Cyclopedia. See also the Kansas Memory archive of Santa Fe diagrams at [], last accessed 18 May 2013 and DeGolyer, Volume 70, pp. 316-490. (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 .)

Works numbers were

1924

March 57695-57698; April 57735-57737, 57742-57754; June 57815-57819, 57850

1926

July 59317-59318, 59339, 59349-59356; August 59398-59401.

Following the delivery of the first 61 engines in this class (Locobase 15202), Baldwin began producing a slightly modified design that had two more superheater flues (and more elements) and eight fewer tubes. All had Elesco feed water heaters and 15" (381 mm) piston valves. A relatively short engine wheelbase combined with a relatively tall boiler gave the class a stocky look.

Twenty-five--4060-4069 and 4086-4100--were oil burners and had the tender weights and capacities shown. According to the Santa Fe diagrams, weight distribution differed considerably between those engines burning oil and those fueled with coal. Oil burners put 11,000 lb (4,990 kg) more on the drivers, but coal burners (16 tons/14.5 metric tons of bituminous) had so much more weight on the two non-powered axles (6,250 lb/2,835 kg on the leading truck, 13,050 lb/5,919 kg on the trailing trucks) that their loaded weight reached 344,800 lb (156,399 kg).

Class 885 (Locobase 8249)

Data from AT&SF 7 -1920 locomotive diagrams supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange. See also DeGolyer, Volume 25, p. 32. Information about how many, when, and by whom this class was supplied from [] (also visited 6 Feb 2006), the Santa Fe All-Time Steam Roster maintained by Evan Werkema. Works numbers were 21247, 21278, 21284-21285, 21323-21324 in November 1902; 21368-21369, 21403, 21419 in December; 21440, 21450, 21490, 21539, 21552 in January 1903.

Baldwin's spec shows that this class was to be numbered 900-914. Before delivery, however, a note in the specs changed the range to 885-899 to make room for the class 900 Santa Fes.

At the same time Baldwin constructed the first-ever 2-10-2s for any railroad (Locobase 463) as tandem compounds, the builder produced these very similar balanced four-cylinder compound designs, in which the high-pressure cylinders lay inside of the frame and the low-pressure cylinders outside.

Locobase doesn't know which way the emulation flowed, but notes that these Mikes (which were early examples of that arrangement as well) had more tubes stuffed into the same boiler barrel and consequently more heating surface. In fact, other than the single Decapod shown in Locobase 5364, this class had the most tubes measuring 2 1/4" in outside diameter of any locomotive ever built.

Such a big engine couldn't be left a compound. Moreover, it needed superheat. So some time later the Santa Fe simpled and superheated this class; see Locobase 8250.

Class 885 - superheated (Locobase 8250)

Data from AT&SF 7 -1920 locomotive diagrams supplied in May 2005 by Allen Stanley from his extensive Rail Data Exchange. Information about how many, when, and by whom this class was supplied from [] (also visited 6 Feb 2006), the Santa Fe All-Time Steam Roster maintained by Evan Werkema.

Locobase 8249 described the 885s in their saturated, compound-working original state. It wasn't too long before the Santa Fe was simplifying the layout in conversions that occurred in 1908-1909.

Not too long after that they installed a superheater. It's odd that so packed a boiler would yield 230 small tubes to allow for 40 large ones, but the total loss in heating surface wasn't extreme. The Baker gear originally required a rocker arm to compensate for the raked HP cylinders, but the later installation of standard cylinders allowed deletion of the rocker arm.

With the exception of the 895, sold in January 1947 to the Apache Railway as their 400, and 896 and 891 (scrapped in August 1947 and November 1949, respectively), the long-lived class operated into the 1950s. 898 was the last to be scrapped in September 1955.