Giants of Steam

by Jonathan Glancey

  Alfred Belpaire also had a widespread influence on international steam locomotive design practice. Born in Ostend, he was trained in Paris at the École Centrale des Arts et Manufactures, and went on to become chief mechanical engineer of the Belgian railways, a position he held from 1864 to 1884. His distinctive square-topped fire-boxes, first introduced in 1864, were fitted to nearly all Belgian railway locomotives over the next thirty years. Their principal advantage over round-topped fire-boxes was the larger water and steam space above the crown of the inner fire-box, the zone in which steam generation is most intense. Curiously, a change of regime after Belpaire saw the Belgian railways reverse the square-topped fire-box policy and order round-topped locomotives from Neilson, Reid of Glasgow, which were virtually identical to the two-cylinder express passenger Dunalastair III class 4-4-0s designed by John Farquharson McIntosh, chief mechanical engineer of the Caledonian Railway. In fact, between 1898 and 1913 hundreds of Scottish locomotives were put to work on the Belgian railways.

  Continental order was restored by Jean-Baptiste Flamme, chief mechanical engineer from 1904 to 1914, who designed two extremely powerful four-cylinder, superheated, simple-expansion classes, the type 10 Pacific and the type 36 2-10-0. These brought Belgium early on into the realm of what the Americans would call super-power steam. Both types had round-topped fire-boxes, and large grates of 53.8 sq ft. With tractive efforts of 44,000 lb and 60,000 lb, the Pacifics and the 2-10-0s were well equipped to work heavy trains on the steep Brussels to Luxembourg main line, with its ruling gradient of 1-in-62 – so much so that the type 10s remained the principal passenger locomotive on this route until its electrification in 1956.

  The type 10 Pacific, of which fifty-eight were built, had an unusual outline, with the smoke-box set well back behind decking plates designed to be lifted in order to provide easy access to the inside cylinders and motion below. In 1934, a Flamme Pacific was exchanged with the French Nord railway for one of the latter’s Chapelon compound Pacifics. The French locomotive showed fuel economies of about 35 per cent. It was not surprising, then, that in 1938 the type 10s were ‘Chapelonized’ with improved steam circuit and double Kylchap exhaust; these modifications increased their maximum power by 20 per cent, with an appreciable reduction in fuel consumption.

  The type 36 2-10-0, of which 136 were built, had meanwhile greatly impressed a Lancashire and Yorkshire Railway delegation visiting Belgium in 1911. An attempt was made to get a similar locomotive – the type 36 was the most powerful European freight locomotive – into production as late as 1922; its size and weight, however, were unacceptable to the railway’s civil engineer.

  In 1932, Raoul Notesse, the newly appointed locomotive design engineer, working for chief mechanical engineer Fernand Legein, prepared designs for a four-cylinder Pacific intended to be faster and more powerful than the Flamme type 10s. Notesse visited Britain the following year to study Stanier’s new Princess Royal four-cylinder Pacific. The result was the type 1 – the heaviest of European Pacifics, at 126 tons – a class of thirty-five engines with 2 m (6 ft 6 in) driving wheels, boiler pressure of 256 psi, and large double-admission piston valves. The type 1s developed a maximum of 3,400 ihp at 100 kph (62 mph), making them, along with Stanier’s Coronation class, the most powerful simple-expansion European Pacifics.

  Then, in 1939 the Belgian railways introduced what was at the time the fastest steam-hauled train in the world, covering the 92.9 kilometres (57.7 miles) from Brussels to Bruges, en route to Ostend, in 46 minutes each way, start to stop, at an average speed of 121.3 kph (75.3 mph). To work these trains, Notesse designed the type 12 streamlined Atlantics, elegant dark-green machines, weighing 89 tons. To minimize disturbing forces at high speed, the two cylinders were located inside the frames, but with external drive to the valve gear; four of the Atlantics had Walschaerts valve gear, one had Dabeg rotary-cam valves, and one Caprotti valve gear. With very large steam-chests to optimize steam flow, double blast-pipe and chimney, 256 psi boiler pressure, 2.1 m (6 ft 10¾ in) driving wheels, and a grate of 39.8 sq ft, these were fast, free-steaming, and powerful locomotives. On test, they developed 2,500 ihp and one reached 165 kph (102.5 mph). The high-speed train was usually loaded with just three or four coaches, weighing up to 220 tons, and limited to 150 kph (93 mph).

  Sadly, the German invasion of Belgium in 1940 put an end to the high-speed trains, although the type 12s continued to run on heavier express trains and semi-fasts until the last of them was withdrawn in 1962. (No. 12.004 was preserved and has been in steam since, notably for the 150th anniversary of the Belgian railways, in 1985.) At the time of the invasion, Notesse was working on two very large and powerful three-cylinder types, a 4-8-4 passenger and 2-10-2 freight locomotive, the former designed for 5,000 ihp. With massive 2.1 m (6 ft 11 in) diameter boilers, triple blast-pipes and chimneys, and mechanical stokers, they promised a great deal. But the Second World War put a stop to these notable designs and Notesse, who came to Britain with the Belgian government-in-exile, was killed by a V-2 ballistic missile at Harrow-on-the-Hill in July 1944. For all its efficacy and apparent innocence, the steam locomotive and its designers, whether in Czechoslovakia, the Soviet Union, China, South Africa, or Belgium, have never been free from the depredations of politics and war.

  CHAPTER 6

  EXPERIMENTATION

  Reinventing the Rocket

  The photograph is telling. Standing on the buffer beam of No. 10000, an experimental high-pressure, four-cylinder 4-6-4, at the buffer stops of King’s Cross station are the two crews in soot-stained overalls who have just brought the non-stop Flying Scotsman express up from Edinburgh. The date is 31 July 1930. The platform clock shows the time as 18.25. The train had arrived ten minutes earlier, exactly on time. The men look exhausted. Standing on the platform to their right beside the whale-like, battleship-grey locomotive is its designer, Nigel Gresley. He looks intensely serious. It is hard not to think that the chief mechanical engineer and crews were simply relieved that No. 10000 had run the 393 miles from Scotland without incident and to schedule. Not that the 1930 timing was especially demanding; with eight and a quarter hours allowed from Edinburgh Waverly, the Flying Scotsman trotted and occasionally cantered, but never had to gallop, much less fly. To steam non-stop over such a distance, however, was no mean achievement.

  Doubtless, the men cheered up after a wash-up and a few beers. As for Gresley, what might he have been thinking? Did this unusual locomotive at King’s Cross point the way to the future, or was it nothing more than an engineering sidetrack? From its inception at the beginning of the nineteenth century, engineers had experimented with the steam locomotive. Although the Stephensons had established the fundamentals of a design that was to endure until the end of regular main-line steam on the world’s railways, there was always a belief that somehow the low thermal efficiency of the steam railway locomotive could be radically improved. The Union Pacific’s Big Boys and the Chesapeake and Ohio’s Alleghenies were to be hugely powerful, while Chapelon’s 240Ps and 141Ps were to set new standards of thermal efficiency – but were locomotive engineers missing a trick?

  Gresley was certainly willing to experiment, and No. 10000, the one and only member of the LNER W1 class, was his most radical design. Completed in November 1929, No. 10000 was taken out of service in October 1936. It had spent more than half of its time in the workshops. The following year, it reappeared as a conventional, three-cylinder simple-expansion locomotive, a 4-6-4 version, with a larger P2 boiler, of Gresley’s magnificent streamlined A4 Pacifics which had first entered service in 1935. What is clear from the rebuilding is that, ultimately, Gresley had decided that any marginal fuel savings were more than offset by the increased servicing and maintenance costs, which ruled out any multiplication of the W1 type.

  The engine itself took five years to design and build, and had emerged from Doncaster in late 1929 from behind a veil of the greatest secrecy; this was why No. 10000 was known as the ‘Hush-Hush’ locomotive, although
some said the nickname was inspired by the engine’s exceptionally soft exhaust. The heart of this big express engine was its high-pressure water-tube boiler. Designed originally for marine work, the water-tube boiler differs fundamentally from a fire-tube boiler. In the latter, heat produced by combustion in the fire-box is transferred as convected heat through the fire-tubes in the water barrel. In the water-tube boiler, water is heated and converted into steam in vertical banks of tubes linking lower and upper water drums. By the time Gresley designed No. 10000, water-tube boilers had been at work in factories, ships, and power stations for several decades. Indeed, Francis Webb, the inventive chief mechanical engineer of the LNWR, had employed water-tube boilers in a pair of tiny 0-4-0 tank engines operating at Crewe works as early as 1875.

  The advantage of the water-tube over the fire-tube boiler lies in its ability to work at a much higher pressure. With their high working pressures and multiple-stage expansion of steam, marine steam engines tended to be far more efficient than steam railway locomotives. The difficulty Gresley faced was fitting an efficient water-tube boiler to a locomotive within the tight limitations of the British loading gauge.

  Some of the only large-scale locomotives at work in the 1920s with water-tube boilers, or hybrid water-tube and fire-tube boilers, were four experimental engines built by the Delaware and Hudson Railway in the United States, where the loading gauge was far more generous. Gresley followed the design of these with great interest. The first, in 1924, was a monstrous two-cylinder compound 2-8-0 built by Alco, which looked as if a factory boiler had been lowered on to on a locomotive chassis. Numbered 1400 and named Horatio Allen, this 174 ton engine – the world’s largest 2-8-0 – was equipped with a 350 psi water-tube boiler (at a time when 200 psi was the norm for conventional boilers), designed by the Delaware and Hudson’s consultant engineer, John E. Muhlfeld. Water-tube side walls with top and bottom drums formed the fire-box, while the boiler barrel itself was of a conventional superheated fire-tube design. With Young valve gear, which allowed a maximum cut-off of an exceptionally high 90 per cent, a booster engine on the two axles at the rear of the tender, high adhesive weight, and a tractive effort, with the booster, of 102,560 lb (simple) or 88,620 lb (compound operation), Horatio Allen was prodigiously strong. She was recorded hauling 3,217 US tons at 26.5 mph up a 1-in-200 gradient with a drawbar thermal efficiency of 8.72 per cent, twice that of a conventional contemporary US steam locomotive. This prototype, however, did require frequent and specialist servicing and maintenance.

  Two more two-cylinder compound 2-8-0s with similar hybrid boilers – 1401 John B. Jervis, with a boiler pressure of 400 psi, and 1402 James Archibald, with a 500 psi boiler pressure – followed in 1927 and 1930. The Delaware and Hudson went further in 1933 with 1403 L. F. Loree, named after the railway’s long-serving president. This 190 ton triple-expansion 4-8-0 was an impressive machine, although it looked rather top-heavy. It was also complex. Steam pressed to 500 psi was passed to a 20 × 32 in high-pressure cylinder located under the right side of the cab, and then to a medium-pressure 27½ × 32 in cylinder under the left side; it was then exhausted to two 33 × 32 in low-pressure cylinders positioned between the bogie wheels at the front of the locomotive. Connecting rods from the fore and aft cylinders drove the second pair of driving wheels, with steam distribution through rotary-cam poppet valves. A six-wheeled booster working at 500 psi replaced the rear truck of the tender. L. F. Loree duly proved able to haul a 6,103 ton freight train up a 1-in-200 gradient, although at no more than walking pace. This, though, was quite some achievement, especially as her maximum drawbar thermal efficiency was 12.5 per cent – although one Delaware and Hudson employee was overheard to say, with some exaggeration: ‘Every time we sent her out, a machine shop had to go with her.’

  There were two further American experiments with high-pressure steam. The first was a 350 psi, three-cylinder compound 4-10-2 with a water-tube boiler, built by Baldwin in 1926. It produced an impressive 4,515 ihp on test at the Pennsylvania Railroad’s Altoona test plant – the limit of the test plant and not that of the locomotive – and gave generally excellent results, running one hundred thousand miles over the territories of six US railroads. Mechanical engineers, however, disliked the fact that the locomotive needed special water-tube descaling treatment; they prized simplicity, and no matter how effective she was, the 4-10-2 needed special treatment. The second was the New York Central Railroad’s HS-1a, No. 800, a handsome 850 psi, three-cylinder compound 4-8-4, built by Alco in 1931. This had much the same complex Schmidt-Henschel water-tube boiler as the LMS’s high-pressure 4-6-0 Fury. The locomotive did little main-line work and was soon relegated to shunting at the New York Central’s Selkirk yard, a duty for which she was unsuited. She was scrapped in 1939.

  If anything, Gresley’s plans were even more ambitious. His 450 psi, water-tube compound 4-6-4 was intended for high-speed and long-distance express passenger work. It was intended to be a more efficient version of the LNER’s fleet-footed A1 Pacifics. Development took place over five years. Working closely with the shipbuilder and philanthropist Harold Yarrow of Yarrow Shipbuilders, based at Scotstoun on the river Clyde, the Canadian locomotive engineer Lawford H. Fry developed a water-tube boiler suitable for one of Gresley’s Pacifics. Initial drawings, however, showed an A1 with a boiler so long that there was precious little room for a cab. As a result, the Pacific’s frame was stretched to a 4-6-4 arrangement, although because the rear axles were independent of one another, No. 10000 was really a 4-6-2-2.

  The locomotive’s streamlined external cladding was devised by O. V. S. Bulleid and William Ernest Dalby, professor of civil and mechanical engineering at Imperial College, London. An early expert in applied aerodynamics, Dalby started work aged fourteen as an apprentice machinist at the Stratford works of the Great Eastern Railway. At Crewe with the LNWR from 1884, he studied part-time for a BSc degree in engineering at London University, before helping to set up the new engineering department at Cambridge University in 1891.

  The ‘Hush-Hush’ locomotive was not a failure; equally, it was not a complete success, particularly as servicing and maintenance costs were appreciably higher than for a standard Pacific. A major problem with poor steaming, however, was overcome with Chapelon’s help; in June 1935, No. 10000 was fitted with a double Kylchap exhaust and a separate low-pressure superheater, eliminating the condensation formed in the low-pressure cylinders which had hampered performance. Thus modified, it produced 1,701 dbhp at 57 mph and could meet mile-a-minute schedules with 450 ton trains between York and Darlington. However, with the introduction of the A4 Pacifics that year, giving 10 to 15 per cent more power than the extremely able A3s, Gresley felt unable to justify further work on No. 10000, especially with the arrival at the LNER in 1936 of a new general manager with the mind of an accountant.

  There had been plans to name No. 10000 British Enterprise and, perhaps, for the engine to be the prototype of an extended class of water-tube compounds. Instead, in 1937 the locomotive was transformed into a three-cylinder 4-6-4 along the lines of a Gresley A4 and ran in this guise until 1 June 1959, when she was withdrawn and scrapped. The sole member of the W1 class, and now numbered 60700, the former ‘Hush-Hush’ locomotive was very nearly named Pegasus, but because non-standard British Railways locomotives were culled at the end of the 1950s, the name plates cast for her were never fitted. The water-tube boiler, meanwhile, steamed on until 1965 under a cover of industrial bricks at Darlington, where No. 10000 had been assembled, used for pressure-testing and space-heating. The locomotive’s eight-wheeled corridor tender runs on today behind the preserved A4 60009 Union of South Africa.

  Meanwhile, the truly furious Fury emerged from the North British Locomotive Company in Glasgow a few weeks after No. 10000 made its debut at Darlington. This semi-compound variation on the three-cylinder Royal Scot 4-6-0s of 1927 was rooted in German rather than American research. Dr Robert Hartmann, engineering director of the Schmidt Superheated Steam Company
, had devised a special type of hybrid water-tube boiler in a bid to increase heat and efficiency. The idea was to raise steam in banks of vertical water-tubes, forming the fire-box walls, to top drums in a closed primary circuit of distilled water at very high pressure, between 1,400 and 1,800 psi. This steam was transferred via tubes to raise pressure at 900 psi in a further drum above the fire-box. Finally, now mixed with steam generated at 250 psi in the fire-tube boiler barrel, it was fed to one 11½ × 26 in high-pressure cylinder and, after being re-superheated, on to a pair of 18 × 26 in low-pressure cylinders. The aim was to reduce fuel and water consumption. The LMS was intrigued. On 15 December 1928, it ordered its first high-pressure locomotive. The boiler was built by the Schmidt Superheated Steam Company, with advice from Henschel in Kassel, while the fire-tube boiler barrel was from the North British Locomotive Company. No. 6399 was ready for testing on 30 January 1930.

  Faith in this daring and rather terrifying-looking machine was severely tested less than a fortnight later. Passing through Carstairs station on its first main-line run, Fury exploded – or at least a high-pressure fire-box tube did. The ensuing rush of steam through the fire-hole door at a pressure of 1,000 psi and more mortally wounded Louis Schofield, the Superheated Steam Company’s representative, who died the following day in hospital in Edinburgh. Fury’s boiler was repaired and further tests were made under Stanier in 1932 and 1933; the results were poor. There was, though, a happy ending of sorts. In 1935, she was transformed by Tom Coleman, under the direction of William Stanier, into 6170 British Legion, prototype of the rebuilt Royal Scots, a class of freshly invigorated express passenger locomotives completed in 1955. These tapered-boiler, three-cylinder machines proved to be among the most efficient of all British steam locomotives.