London & North Western Articulated Locomotives in Great_Britain

Class 687 "Fore and Aft" (Locobase 21114)

Data from "Compound Tank Locomotive (Webb's Patent), London & Northwestern [sic] Railway", Railroad Gazette, Volume 19 (23 September 1887) , pp. 614-615, 620; and "Second Compound 2-2-2-2 Tank", Passenger Locomotives on the London & North Western Railway Society at [], last accessed 11 February 2023. See also "2-2-2-2T: Second compound tank: 1885-, in steamindex at [], last accessed 11 February 2023. (Thanks to Alexander Blessing for his 5 February 2023 email supplying a detailed spreadsheet entry on this locomotive from which Locobase took the data.)

Francis Webb experimented with an unusual method of compounding his divided-drive system beginning with the 1881 Experiment, a passenger engine (Locobase 21113). Although he built a class of 30 such 2-2-20s, he continued to explore his approach and put this tank locomotive in commuter service in 1885.

Like the Experiment, he set up two high-pressure cylinders outside and a much larger one inside to compound the effort by reusing the exhausting steam from the HP cylinders. 687's compounding ratio complied with preferred practice by varying the stroke. The HP strokes traveled 8" (203 mm) less in each cylinder than the stroke in the LP cylinder. It was a difference that he never repeated and was seldom adopted by anyone else.

Railroad Gazette reported that the HP valve travel and cutoffs could be set independent of the LP cylinder, allowing the HP cylinders to shut off live steam early in the stroke while the LP cylinder worked in full gear. This allowed the LP cylinder to take the most advantage of the exhausted steam.

The variable strokes and settings likely created the difficulty Ernest L Ahrons described in his Locomotive and train working in the latter part of the nineteenth century. Volume 1. Steamindex quotes an excerpt, which reads in part:

" "687, a 2-double-2-2 arrangement built in 1885. Had the London and North Western tank engines had names like the tender engines, he would have respectfully suggested the name "Fore and Aft" for 687".

After noting that the express engines' surging motion damped out at higher speed, the excerpt continued:

"Now the "Fore and Aft" had it much worse than the "Dreadnoughts"; and as this engine was for a time on the District line (Willesden and Mansion House), and consequently was continually starting, the result can be better imagined than described. I had some experiences of it myself-on one occasion when leaving Victoria (Underground), a carriage full of passengers were swinging backwards and forwards after the manner of a University "eight." [rowing crew]. I am afraid that the "Fore and Aft" was the cause of much bad language".

This was the only Webb compound that used different strokes for his HP and LP cylinders.

Class Dreadnought (Locobase 2174)

Data from James Dredge (ed.), "Section V Historical Locomotives and Rolling Stock -- 47 -- The Exhibit of the London & North Western Railway '", in A Record of the Transportation Exhibits at the World's Columbian Exposition (New York: John Wiley & Sons, 1894), pp. 174-176; Daniel Kinnear Clark,The Steam Engine--Half-Volume IV (London: Blackie & Son, Ltd, pp. 609-610. See also Hollingsworth, Steam Passenger Locomotives, (1982); "The Webb Compound Locomotive, Marchionesse of Stafford", American Journal of Railway Appliances", Volume V [5], No. 8 (10 October 1888), pp. 421-422. F W Webb, "Development of the Crewe System of Compounding Locomotives," Engineering Review, Volume 2, (20 June 1895), pp. 391-393; and W A Tuplin, The Steam Locomotive: Its Form and Function (New York: Charles Scribner's Sons, 1974), (Thanks to Alexander Blessing for his 28 December 2019 email supplying the Dreadnoughts' wheelbases and tender coal and water capacity.)

Citing the Engineer (London), the AJRA's report gives details of the Dreadnought's performance, noting that the lead engine had amassed 15,477 miles (24,918 km) between September 1884 and 20 March 1885. The rake on the trains between Crewe and Euston averaged twelve carriages each. Virtually all of the tubes in the boiler measured 1 7/8" (48 mm), but a desire to reduce tube-plate cracking led to reducing the diameter of the top and side rows of tubes to 1 1/2" (39 mm). The reduction alse limited the draft throught these tubes.

Hollingsworth (1982) explains why these relatively incompetent locomotives were built: "One of the problems was that Webb [their designer] was an autocrat and anyone who suggested that his beloved compounds were less than perfect was regarded as questioning his superior officer's judgement and hence offering his resignation." Against that harsh review is Steamindex's "Francis [Frank] William Webb" summary at [] (last accessed 28 December 2019.)

Acknowledging that the three-cylinder compounds were not very successful. Their divided drive is described by others as sometimes working against itself and in any case offering too little tractive effort on starting. But steamindex reports: " Provided the high- and low-pressure motions were in harmony, a Webb three-cylinder compound would start and accelerate a train of equal tonnage far more rapidly than would a contemporary simple engine."

The review discusses several biographers and other analyses and argues that Webbs very significant achievements across the board (locomotive design, factory management, etc) stand up to any other British engineer.

Like the Experiments, these engines were retired as soon as decently possible with the last departing by 1907.

Note: Tuplin (1963) gives a nominal tractive effort of 21,500 lb, commenting that it is "that of the simple engine produced by removing the high-pressure cylinders." The number given in the specifications is based on a 1924 ARA formula.

Class Experiment (Locobase 21113)

Data from Francis Webb, "On Compound Locomotive Engines",Proceedings of the Institute of Mechanical Engineers, (July 1883), pp. 438-462. See also "Argus" (pseudonym) , "Compound Locomotives", Engineering, Volume 40 (11 September 1885), pp. 251-252. (Thanks to Alexander Blessing for his 5 February 2023 email supplying a detailed spreadsheet entry on this locomotive from which Locobase took the data.)

When Francis Webb delivered this seminal description of his compounding system, the first engine in this class had already amassed 100,000 miles (161, 000 km) in service. It headed up the Irish and Scottish limited mail trains and covered 319 miles (514 km) each day. Webb claimed that the Irish mail runs reduced train mile consumption of coal from 34.6 lb (9.75 kg/km) in a standard ftwo-axle express engine to 26.6 lb (7.49 kg/km)/.

Its cylinder arrangement was distinctly Webb's. He placed the two high-pressure cylinders outside under the running board between the front truck and the first coupled axle. The connecting rods drove the second coupled axle. Dividing the drive, he argued resulted in "less grinding action round curves." If so desired, one could even use drivers of different sizes on the two axles. (Locobase hasn't found many if any such setups on steam locomotives.) No coupling rods and the three-point cylinder layout meant smoother running at high speeds.

The single low-pressure cylinder sat on the centerline under the smokebox and drove the leading coupled axle. A relief valve on the low-pressure cylinder restricted its boiler pressure to 75 psi (5.17 bar).

Webb's presentation elicited considerable comment (pp. 446-462). His compounding ratio seemed low, but, despite the apparent complication of machinery, several agreed that eliminating coupling rods between drivers would prove beneficial.

In August 1885, "Brazilian", a pseudonymous author of a short question to Engineering's letter section asking for more information about the actual advantages of compounding ignited a firestorm of replies. The entire thread bears reading (and patience) , but "Argus's" long and detailed reply on 2 September 1885 that appeared in the 11 September 1885 issue of Engineering serves to lay out the case against. One writer responded to Argus's criticisms by saying that if Argus persisted in taking up 3 columns of print (which Locobase can attest was a lot of words), he should be willing to reveal his name. Engineer, in its own 2 October rebuttal argued that compounding's value should not be examined solely in terms of savings of fuel and water.

Although mostly supportive and usually appreciative of the effort Webb put into testing the idea, British railways use relatively few compounds, few of which repeated the two HP, one large LP cylinder concept.

Class Experiment/Compound (Locobase 2173)

Data from Francis W Webb, "On Compound Locomotive Engines", Proceedings - Institution of Mechanical Engineers July 1883 (London: , 1883), pp. 438-445. See also Frederick Colyer, A treatise on modern steam engines and boilers: Including Land, Locomotive, and Marine Engines and Boilers (London: E & F N Spon, 1886), pp. 87-89; and F W Webb, "Development of the Crewe System of Compounding Locomotives," Engineering Review, Volume 2, (20 June 1895), pp. 391-393; and Data from W A Tuplin, The Steam Locomotive: Its Form and Function (New York: Charles Scribner's Sons, 1974),

(Thanks to Alexander Blessing for pointing out the original 11 1/2" dimension of the HP cylinder.) Crewe works numbers were 2500 for the prototype in 1882, 2625-2628 in February-April 1883, 2669-2673 in July, 2734 in January 1884, 2735-2740 in February, 2741-2743 in March, 2744 in May, 2745-2748 in June, and 2749-2753 in July.

FW Webb set his sights on a three-cylinder compound arrangement soon after he took over locomotive-design duties at the London & North Western. This "Experiment" offered the unusual arrangement of two high-pressure cylinders exhausting into one low-pressure one. The HP cylinders drove the trailing axle, the LP -- "a great dustbin" according to Hollingsworth (1982) -- powered the lead axle. Each set had its own valve gear. OS Nock (RWC III, pl 103) comments that this "'double-single'...was hoped to have the freedom of a 'single-wheeler' with the power of a coupled engine."

The original Experiment was tested in 1882 using 11 1/2" (292 mm) HP cylinders. Webb told the IME meeting in July 1883 that the Experiment had headed up the Scottish and Irish limited mails for almost 100,000 miles, averaging 319 miles per day. He claimed that the engine had shown itself to be "very steady running, which is no doubt due to the arrangement of the cylinders; the engine being practically balanced, and having no coupling-rods, is enabled to run at very high speeds.".

Only a few changes seemed necessary to . HP cylinder diameter increased by 1 1/2" (39 mm) and valve ports were increased from 1 1/2" x 14" (39 x 356 mm) to 2" x 16" (50.8 x 406 mm). Webb's high opinion of the resulting production variant hasn't stood up well. Most historians agree the class was not a success. Hollingsworth notes they "were not especially economical and were bad starters." Production ended in 1884 and the last was retired early in 1905.

Note: Tuplin (1963) gives a nominal tractive effort of 13,500 lb, commenting that it is "that of the simple engine produced by removing the high-pressure cylinders." The number given in the specifications is based on a 1924 ARA formula.

Class Greater Britain (Locobase 2177)

Data from "The Webb System of Compounding Locomotives", Engineering, Volume 57 (11 May 1894), p. 610-611. See also Gustav Reder, The World of Steam Locomotives (1974), pl 238; F W Webb, "Development of the Crewe System of Compounding Locomotives," Engineering Review, Volume 2, (20 June 1895), pp. 391-393; Charles S Lake, "Special Features in the Design of Locomotive Boilers and Fireboxes," The Engineering Magazine, Volume 26, No. 1 (October 1903), p. 71; and W A Tuplin, The Steam Locomotive: Its Form and Function (New York: Charles Scribner's Sons, 1974),

Unsuccessful elaboration on the Teutonics using a Stephenson link motion on the outside with a "slip-eccentric" link between the two outside HP cylinders and the single LP cylinder under the smokebox. Reder noted that this class had an intermediate combustion chamber in the boiler such that the tubes ran 5' 10" (1.78 m) from the firebox to the chamber then 10' 1"(3.07 m) from the chamber to the smoke box. Combustion chamber area measured 39.1 sq ft (3.63 sq m) and tube area within the chamber came to 506.2 sq ft (47.03 sq m).

Lake detailed the actual layout of this divided boiler:

"The tubes between the firebox and the combustion chamber, are of copper, 2 l/8 inch diameter outside and 6 feet over all in length. "

This set were horizontally laid. But "those which are situated between the forward tube plate of the combustion chamber and the smokebox are inclined in a downward direction towards the smokebox, the total fall amounting to 6 inches in the length of 10 feet 1 inch which separates the two tube plates."

Lake totted up the areas afforded by this apparatus: "These forward tubes ...are of steel and provide 875.0 square feet of heating-surface area, which together with the 506.2 square feet of the rearmost tubes, the 39.1 square feet of the combustion chamber, and 120.6 square feet which is derived from the firebox, gives a grand total of 1540.9 square feet (143.15 sq m), an unusually large allowance at the time when the engines were built."

Reder assessed the result decades later: "But since the exhaust gases had cooled down too much before entering this intermediate chamber, the arrangement did not live up to its purpose." Reder also reports that the "driver's footplate had so little room that the fireman had to stand on the swaying running board between the engine and the tender while firing."

Note: Tuplin gave a nominal tractive effort of 19,000 lb, commenting that it is "that of the simple engine produced by removing the high-pressure cylinders." The number given in the specifications is based on a 1924 ARA formula.

Class John Hick (Locobase 2178)

See also F W Webb, "Development of the Crewe System of Compounding Locomotives," Engineering Review, Volume 2, (20 June 1895), pp. 391-393; and Data from W A Tuplin, The Steam Locomotive: Its Form and Function (New York: Charles Scribner's Sons, 1974),

Slightly larger Greater Britains using a Stephenson link motion on the outside with a "slip-eccentric" link between the two outside HP cylinders and the single LP cylinder under the smokebox. Boiler pressure is quite a bit higher and the driver diameter is lower because these engines were meant to operate between Crewe and Carlisle. Such changes did little to improve an unimpressive design and Reder (1974) notes that these engines wound up on local branch lines.

Note: Tuplin gave a nominal tractive effort of 24,500 lb, commenting that it is "that of the simple engine produced by removing the high-pressure cylinders." The number given in the specifications is based on a 1924 ARA formula.

Class Teutonic (Locobase 2176)

Data from Hollingsworth (1982) as amended and supplemented by R L Rowley, letter in "Correspondence", The Locomotive Magazine Illustrated, Volume 4, No 41 (April 1899), p. 64. See also F W Webb, "Development of the Crewe System of Compounding Locomotives," Engineering Review, Volume 2, (20 June 1895), pp. 391-393; and W A Tuplin, The Steam Locomotive: Its Form and Function (New York: Charles Scribner's Sons, 1974), (Thanks to Alexander Blessing for his 28 April 2019 email correcting the wheel arrangement shown in the specs.)

Unlike Webb's other divided-drive compounds, the Teutonics were reasonably successful and put up an impressive performance in the then-famous "Race to the North" in 1895. Hollingsworth (1982) notes that the divided drive could result in the embarrassing predicament of letting steam into the cylinders and having one set of drivers rotate one way and the other set rotate the other. Part of the problem was the use of "slip-eccentric" link between the two outside HP cylinders and the single LP cylinder under the smokebox.

Note: Tuplin (1963) gives a nominal tractive effort of 19,000 lb, commenting that it is "that of the simple engine produced by removing the high-pressure cylinders." The number given in the specifications is based on a 1924 ARA formula.