Giants of Steam

by Jonathan Glancey

As John van Riemsdijk, former keeper of mechanical engineering at the Science Museum in South Kensington, pointed out in Compound Locomotives (1994), Chapelon was never more or less during the war than an apolitical technician and he would have kept his thoughts to himself, although after the German defeats in Russia and North Africa in 1942–3 he did contact a Resistance group. As Van Riemsdijk recalled: ‘I had some contacts with SNCF personnel when I worked in the Special Operations Executive and followed these up in early 1946, when I formed lasting friendships with railwaymen of various grades. The conclusion to be drawn from all the sources is that, apart from a politically motivated minority, the prime loyalty was to France but some collaboration with the Germans was necessary to protect French lives. Chapelon continued working on post-war locomotive design and advising on motive power matters. He was a backroom boy largely uninvolved in politics. However, both he and Marc de Caso, like numerous others, were investigated after the war.’

  Whatever professional and personal jealousies were involved in these purification committees, Louis Armand put a stop to all such witch-hunting when he was appointed assistant general manager of the SNCF in 1949. A railway mechanical engineer and a genuine Resistance hero, Armand had formed the Resistance-Fer (Railway Resistance) in February 1943 when he was chief mechanical engineer of the SNCF. Denounced by collaborators, he was about to be sent to a concentration camp in Germany in July 1944, but was saved by the allied bombing of the Paris to Strasbourg line, and then by the Swedish Red Cross and the timely arrival of French and American troops. He, if anyone, would have known who the real collaborators were.

  Today, Armand is best known to steam enthusiasts for his Traitement intégral Armand, a system of controlled chemical water treatment to eliminate boiler scaling and corrosion, which was applied to all SNCF steam locomotives, reducing boiler maintenance costs by between 85 and 90 per cent and cutting fuel consumption by 7.5 per cent as a result of improved heat transfer within the boiler. In fact, Armand’s treatment, together with a preventive maintenance system introduced by Joseph Poissonier of the Nord railway, enabled the SNCF’s fleet of 9,000 steam locomotives in 1955 to average 250,000 kilometres (156,000 miles) between failures, a degree of reliability never achieved in Britain with any form of traction.

  As for Marc de Caso, he was not only Chapelon’s friend and ally but also the one other truly notable French steam locomotive engineer of the super-power steam era. Joining the Nord railway in 1921, his first job was to renovate a fleet of ex-Prussian State Railways P8 class 4-6-0s, locomotives given to France as part of Germany’s post-war reparations, and in deplorable condition. He moved on to design all-steel monocoque coaches for the Nord railway – the first to enter quantity production in Europe – before improving the valve mechanism of the Nord railway’s Super Pacifics, which he did very successfully.

  His first locomotives were the powerful 5.1200 class four-cylinder compound 2-10-0s of 1933 featuring Super Pacific boilers and a very large steam-flow circuit. With an official top speed of 105 kph (65 mph), these heavy freight and mineral train locomotives were also often put to work on heavy passenger trains making frequent stops. Capable of 3,000 ihp, these highly successful machines continued in production in SNCF days, classified as 150Ps.

  In 1933, de Caso became locomotive design engineer of the inter-railway central design office, where he continued his design work on his remarkable 4-6-4s for the Nord railway, with 290 psi boilers, 54 sq ft fire-grates, mechanical stokers, robust frames, and large steam-flow circuits. Three three-cylinder simple-expansion 4-6-4s and five four-cylinder compound 4-6-4s were built from 1941, the latter developing a maximum of 4,500 ihp. They were so successful, and impressive, that the SNCF’s north region proposed building a further twenty improved de Caso 4-6-4s in 1950, but this was stopped by the ministry of public works and the region was allocated 241Ps instead.

  Meanwhile, Chapelon’s own masterpiece and, arguably, the finest steam locomotive yet built, was to emerge from the embers of the war. Projected in 1938, authorized in 1942, and finally completed in 1946, this was 242A1, a three-cylinder compound 4-8-4, a prototype for the new standard types the French engineer hoped to build to meet traffic needs up to 1970. This was not the fastest or the biggest or the most powerful steam locomotive in the world; but it was extremely efficient, and one of the very few steam locomotives that forced the design engineers of the latest electric engines to think again.

  Chapelon had first proposed this development when he was given the opportunity to rebuild an État railway three-cylinder simple-expansion 4-8-2 which had fractured a cylinder. There was, though, no chance of building a prototype express passenger locomotive during the war. So 242A1 was not revealed until 18 May 1946. Built at the forges of Ateliers de la Marine et d’Homécourt, near Saint-Étienne, this locomotive was the stuff of painstaking research, mechanical inspiration, and – or so it must have seemed to those who experienced the solitary French 4-8-4 in action – something like sorcery. Weighing 148 tons, 242A1 employed three, rather than four, cylinders. This was a major departure for Chapelon. The idea was not so much for simplicity’s sake as to create room for much thicker inside driving cranks and wider axle boxes; while these components (in thinner and lighter form) gave satisfactory results in the 240Ps and 141Ps at up to 4,000 ihp, Chapelon believed they would need to be much stronger for the higher power output expected of the 4-8-4.

  The appearance of the locomotive was that of a scaled-up 240P. It looked long and racy, especially with its small cab which made the driver seem like the jockey of a particularly strong and lithe racehorse. With its triple Kylchap exhaust and chimney, thrusting conical smoke-box, high running boards, huge ‘elephant-ear’ smoke deflectors, massive outside low-pressure cylinders, and close-coupled driving wheels, it seemed like one of those futuristic French steam locomotives bursting out of the brilliant 1930s posters by graphic artists like Adolphe Cassandre. It looked every last centimetre the dashing and heroic, if tragically short, part it was to play in the story of the steam locomotive.

  The driver’s job was certainly not one for the inexperienced. There were two regulator handles, instead of the one found on conventional steam locomotives. On starting, the driver would open the low-pressure regulator to get the train moving, followed by the high-pressure regulator. Full compound working would begin at about 15 mph, when the driver would close the low-pressure regulator, and as 242A1 accelerated he would adjust the two cut-offs for the high-and low-pressure cylinders so as to give equal power. On production engines, the low-pressure regulator would have been retained for starting but the ratio between high- and low-pressure cut-offs would have been achieved automatically.

  There was no question, though, that Chapelon had done something truly exceptional with 242A1. Here was a European steam locomotive as powerful as many American engines 50 per cent heavier. Here was a 148 ton 4-8-4 that used approximately 20 per cent less coal than a 105 ton Stanier Coronation – one of Britain’s finest – needed to produce a proportionate amount of power. Here was the first European steam locomotive able to maintain a constant 4,000 dbhp. Fire-box and boiler – pressed to 290 psi – produced steam at a continuous rate of 52,240 lb per hour (a Coronation would produce a maximum of 40,000 lb per hour on test). The engine rode well up to its specially permitted maximum of 150 kph (93 mph); it was normally limited to 130 kph (81 mph). With high and reliable power on tap, sure-footed starting with the heaviest trains on the steepest gradients, rapid acceleration, and the ability to run at the normal speed limit uphill and down, except over the most severe gradients, 242A1 was able to equal the performance of existing SNCF electric locomotives.

  On trial between 1946 and 1948 on the ex-PLM Ligne impériale, 242A1 bettered the performance capacity of the planned 144 ton 2D2.9100 class electrics, which were designed to go into service between Paris and Dijon in 1949. SNCF officials could only gawp in astonishment as, on 14 October 1948, the Chapelon locomotive tackled an eighteen-coach Paris–Lyon–Nice
express weighing 861 tons, storming over the steep and curving gradients north-west of Dijon without falling below 101 kph (63 mph). The 31.2 kilometre (19.4 mile) section between Les Laumes and Blaisy was covered in 18 minutes and 17 seconds, despite slowing down for permanent-way works. At 117 kph (73 mph) the engine generated 5,500 ihp, with the single high-pressure cylinder producing 1,920 ihp and the two low-pressure cylinders 3,580 ihp.

  An immediate consequence of this was the order to uprate the maximum one-hour rated power of the new Ligne impériale electrics from 4,000 hp to 4,950 hp – there was no way that management could allow a steam locomotive to outperform their electrics even before they had gone into service. Here was that very rare case of a steam locomotive influencing the design of electrics. However, due to its much lower frictional resistance, a 2-Do-2 with 4,950 hp at the motor shafts developed 4,500 dbhp at 75 mph; at the same speed, 242A1 gave about 4,400 dbhp from the 5,500 ihp generated in her cylinders.

  On 12 September 1952, and with the brilliant Argentine locomotive engineer Livio Dante Porta on the footplate, 242A1 worked a Paris to Brest express under the electric wires to Le Mans. Loaded with twenty cars, weighing 810 tons, Chapelon’s three-cylinder compound averaged 116 kph (72 mph), undercutting the new electric schedule. From the first time it turned its eight 1.95 m (6 ft 4¾ in) driving wheels, there had been something quite unparalleled about the performance of this peerless locomotive. Imagine what it must have been like to watch the green giant lift a 742 ton train east from Lisieux, on 23 April 1947, up a snaking 1-in-125 gradient on a blustery day with a 40 kph (25 mph) side-wind, accelerating to 80.5 kph (55 mph) in just under 4 kilometres (2.5 miles). Truly, nothing like this had been seen or experienced before in Europe. And, true to Chapelon form, the 4-8-4 went about its work with great brio.

  John van Riemsdijk later summarized the capability of this great machine: ‘On those main lines with formidable gradients, such as Blaisy, between Paris and Dijon and the Lisieux bank between Paris and Caen, 100 kph would be sustained. On the well-known climbs to the summit at Survilliers on the Nord main line, this locomotive was able to run freely with 700 or 800 tons at 75 mph, which was the normal speed limit. In fact, gradients almost ceased to exist for practical purposes because the speeds of the trains did not drop appreciably when this locomotive was pulling them.’

  Allocated to the west region, 242A1 was based at Le Mans. When it pulled Paris to Brest expresses over the 412 kilometres (256 miles) from Le Mans, where it took over from electrics, the locomotive gained up to forty minutes in recovery time, with trains of up to 850 tons, on schedules designed for the smaller 4,000 ihp 141P class 2-8-2s. As French engine crews were paid bonuses for regaining time, as well as for saving fuel, 242A1 was greatly in demand at Le Mans depot, although there was some degree of rivalry over who should be allowed to drive such a boost to the pay-packet. With the final cancellation of Chapelon’s high-powered locomotive designs in 1951, the wholly reliable 242A1 was withdrawn prematurely in 1954, and finally scrapped in 1961 – despite an offer to buy the locomotive made by Dr Armin Glazer, a Swiss enthusiast, who received a reply a year later informing him that the locomotive no longer existed. There have long been dark murmurings to the effect that the engine was not saved for posterity because it was an embarrassment to the SNCF modernizers. Why else would the most powerful steam locomotive ever built outside the United States, and one of the most economical of all steam locomotives, have been cut up? As it is, enthusiasts from around the world would pay the SNCF handsomely to see a reconstruction of the greatest steam locomotive of all in action once again.

  With its unsurpassed performance, 242A1 should have been the genesis of a new era of SNCF steam development. On 7 June 1944, the day after D-Day, Louis Armand, then chief mechanical engineer of the west region, attended a meeting to discuss future motive power policy. Assuming – though not lightly – that the Germans were finally on the run, it was time to think ahead once again. While the electrification of key main lines was not in doubt – Paris to Lyon had been agreed in 1942 – there would still be a fairly lengthy period of changeover from steam to electric and a need to improve the speed and frequency of services on routes scheduled for electrification in the more distant future. To meet this need, Chapelon and Chan worked up proposals for a high-powered series of modular locomotives, this time combining the best in modern American construction and practice with the best in French thermodynamics.

  Chapelon discussed these designs for the first time at an SNCF-authorized lecture that he gave on 14 December 1945 to members of the Association Française des Amis des Chemins de Fer (AFAC) at the Conservatoire des Arts et Métiers in Paris. I wish I could have been there. Outline drawings revealed four types of three-cylinder compound to be produced from standardized components. These machines, intended to be capable of developing 5,000 ihp continuously, with a maximum of 6,000 ihp, comprised: a 4-8-4 for heavy express duties, designed to haul 950 ton rapides up 1-in-200 gradients at 120 kph (74.5 mph), with a maximum speed of 140 kph (87mph); a 4-6-4 for ultra-rapides, capable of 140 kph (87 mph) up 1-in-200, and a maximum speed of 200 kph (124.5 mph); a 2-8-4 for heavy mixed-traffic use, capable of 90 kph (56 mph) up 1-in-200 pulling 1,200 tons, with a maximum speed of 120 kph (74.5 mph); and a 2-10-4 for heavy freight, capable of 70 kph (43.5 mph) up 1-in-200 with 2,000 tons, and a maximum speed of 110 kph (68 mph).

  Artists’ impressions reveal a good-looking, modern group of engines, of elegant outline, with generous and fully enclosed cabs, and a fine sense of artistry as well as technical excellence. The locomotives would feature single-piece integral frames, roller bearings fitted throughout, and automatic lubrication. The large boilers, similar to that of 242A1, would be pressed to 320 psi, but with longer, mechanically fired, 64.5 sq ft grates. Willoteaux double admission and exhaust piston valves would ensure an unrestricted flow of steam in and out of the three cylinders.

  Chapelon also showed designs for two complementary projects. One was for a one-man-operated, oil-fired, four-cylinder compound 4-6-4 tender-fourgon, a streamlined tank engine with built-in luggage and guard’s compartments. All four cylinders were to have been mounted externally in side-by-side pairs. This intriguing machine, derived from a Paris–Orléans project of 1936, would have provided fast services on main-line stopping trains and on secondary and tertiary cross-country lines. The tank locomotives would have replaced what were, at the time, rather noisy and uncomfortable diesel-railcars. A 2,000 ihp engine, with a top speed of 130 kph (81 mph), it would have accelerated a 100 ton, three-car train from rest to 120 kph (74.5 mph) in two minutes, revolutionizing the kind of service provided by slower diesel-railcars. The proposal was strongly supported by the southeast region’s chief mechanical engineer, Marcel Japiot. Chapelon’s figures, always erring on the side of caution, can be accepted as fact. Throughout his career, he calculated to remarkably small margins of error; figures that appeared on paper were equalled, and usually bettered, when the locomotives were completed and out on the track.

  The other project was for a streamlined, 3500 ihp, mechanically fired, four-cylinder compound 4-6-0, using a slightly shortened 240P boiler at 290 psi, which was to have rocketed 350 ton trains-drapeaux (prestige trains) from Paris to Marseilles and Paris to Strasbourg at an average speed of up to 144 kph (89 mph), with a maximum speed of 200 kph (124.5 mph). Unfortunately, post-war track conditions ruled out such maximum speeds and the 230P, as it was classified, was cancelled. This dashing machine was, however, to influence the design by David Wardale, a modern British steam locomotive engineer, of the proposed 200 kph 5AT class 4-6-0 (see Chapter 7) half a century later. The dream – far from idle – of a steam locomotive that can cruise at 200 kph was ignited by Chapelon; it has yet to dissipate. Chapelon himself believed in higher speeds still for steam traction, outlining proposals for a possible further generation of three-cylinder triple-expansion locomotives designed for up to 250 kph (156 mph).

  A start was made at the Compagnie des Ateliers et F
orges de la Loire, Saint-Chamond, on the frames and cylinders of what was to be the first of a hundred 152P class 2-10-4s, but in 1947 SNCF management, following government instructions to reduce consumption of coking coal in order to conserve supplies for the steel industry, declared the suspension of new steam development in favour of electrification. Four years later, electrification of the trunk route between Valenciennes and Thionville, for which the 152Ps had been intended, prompted their final cancellation. Substantial compensation had to be paid to the builders.

  Chapelon succeeded Georges Chan as chief locomotive design engineer of the SNCF in 1949. While construction of new steam locomotives was ruled out from 1951, Chapelon was able to make major improvements to a number of locomotives, including the 141Rs and 241Ps. The latter were a class of express passenger four-cylinder compound 4-8-2s, of which thirty-five were built between 1948 and 1952 for service on the former PLM main line from Paris to Lyon and Marseilles, and the Nord railway line from Paris to Lille and the Belgian border. Developed from the rebuilt 241.C1, the PLM prototype four-cylinder compound 4-8-2, the 241Ps, once given the Chapelon treatment, were capable of producing a maximum 4,300 ihp. When displaced by electrics on the PLM main line, they worked on the État main lines from Paris to Le Mans, Brest, and Nantes, where they powered expresses of 800 to 900 tons at average speeds of 100 kph (62 mph) until 1969. In 1956, two pooled 241Ps from Lyon Mouche depot each averaged 19,250 miles in thirty-one days pulling the heaviest and fastest expresses to Marseilles, without incident. These were the highest monthly mileages recorded by European steam locomotives.

  Chapelon retired from the SNCF in 1953. He had already taken up an appointment as consulting engineer for GELSA (Groupement d’Exportation de Locomotives en Sud-Amérique), for which he had designed twenty-four simple-expansion 4-8-4s and sixty-six 2-8-4s for service on metre-gauge railways in Brazil, built between 1950 and 1953. In spring 1951, Chapelon led a mission to Brazil to see the first of the new locomotives in action; he journeyed on to Argentina where he met his most fervent disciple, the engineer Livio Dante Porta, who was described by Chapelon as ‘eminent amongst those who believe in the further development of the steam locomotive’.