Giants of Steam

by Jonathan Glancey

  Got that? Although such complex and frustrating episodes were rare, they soon became the stuff of railway lore. Webb, although in many ways a great engineer, never got his many compounds – Experiments, Dreadnoughts, and Teutonics – to run half as well as his lighter and older simple-expansion engines. The legendary performance of 790 Hardwicke, Webb’s bantamweight Precedent class two-cylinder 2-4-0, in the 1895 railway races from London to Aberdeen was quite outstanding. On 22 August 1895, Hardwicke sprinted the 141 miles over Shap from Crewe to Carlisle in 126 minutes, at an unprecedented average speed of 67.1 mph, a record that stood throughout the steam era. Hardwicke’s three-coach train might have been light, but so was the locomotive, at 35 tons. Its tractive effort was 10,918 lb. One also has to imagine what it must have been like for the crew on the flimsily protected footplate of this fiery machine as it rode down Shap at up to 90 mph.

  Webb’s compounds, although not as fast, could in fact be very reliable. The Teutonic 1304 Jeanie Deans worked the heavy 14.00 Glasgow ‘corridor’ dining-car express from Euston as far as Crewe, and the corresponding train back, almost every day for eight years. They were controversial, however, providing grist to the anti-compound mill and giving compounding a bad name in Britain from which it never quite recovered. Although the forty-five Midland Railway compounds built from 1902 to 1909, and the 195 built by the LMS from 1924 to 1932, were successful and economical engines, they were not sufficiently powerful to make significant improvements to heavy main-line services on the west coast route to Scotland. They were relatively old-fashioned in front-end design for machines built in the mid-1920s, by which time the Collett Castle class and Gresley A1s were in service.

  In France, the first compound locomotive, a two-cylinder 0-4-2 tank engine – a slip of a thing – was built by Anatole Mallet for the Bayonne– Biarritz railway in 1876. Two key factors were at work to make the compound a success. The first was coal. France was not endowed with such huge reserves of coal as was Britain, and much of the coal that was mined, mostly in the north of the country, was small, dusty stuff, sometimes with a low calorific value. This meant that coal had to be imported, at great expense. One way of reducing consumption on the railways was to adopt compounding, where the softer exhaust beat enabled small coal to be burned more efficiently. To the GWR fireman brought up on glistening lumps of high-grade Welsh coal, the French offering would have seemed very poor. If the supply of decent-sized coal for locomotive use was one problem the French railways faced, the other was the way in which most of the main lines had been laid with relatively light rails, compared with many in Britain. This meant that locomotives had to be relatively light too, putting as little stress on the track bed as possible. Partly because of this, a national speed limit of 120 kph (74.5 mph) was ordained for the railways during the reign of Napoleon III. When it was revoked in the 1930s, the new limit was just 130 kph (81 mph).

  If, then, an express train was going to live up to its name and average a mile-a-minute between the two world wars, it had to be able to accelerate rapidly, and it had to maintain high speeds uphill because there was little opportunity for racing downhill. The express locomotive would also have to run as smoothly as possible in order to spare the track and to burn as little costly coal as possible. Once these conditions are understood, the rise and rise of the compound steam locomotive in France becomes not just comprehensible, but almost a matter of principle. For many French steam locomotive designers, compounding was as much a national religion as it was a form of engineering.

  Because many early compounds required two sets of valve gear and controls to regulate steam supply to the high- and low-pressure cylinders, they were often more complex machines to drive than their simple-expansion counterparts. In France, drivers – mécaniciens – spent two years in the workshops as a part of their training and were instructed in the theory and practice of working steam locomotives. In this regard, they were a breed apart from drivers in other countries, whose job was primarily to drive or fire locomotives and not to be overly concerned with the niceties of thermodynamics and the intricacies of mechanical engineering. There are many stories – some apocryphal, others true – of mécaniciens spending their holidays at locomotive works to see their compound Pacific through its annual general overhaul, while for many years in France individual main-line express locomotives were allocated to specific crews.

  This devotion to duty, or obsession, is well caught both in Émile Zola’s 1890 novel La Bête Humaine and in Jean Renoir’s classic adaptation of this disturbing psychological thriller for the cinema in 1938. In the story, Jacques Lantier, a mécanicien on the Paris–Le Havre line, is as obsessed with his locomotive as he is with the femme fatale and stationmaster’s wife, Severine. The sequences on the footplate of a compound Pacific are as compelling as the story itself.

  Equally, Pacific Senator: A Train Driver’s Life (1984), by the Pacific mécanicien Marcel Péroche, is an extraordinary book which helps one to understand the French love affair with a form of locomotive that could belong to no other country. Speaking of Pacifics that were pooled and not assigned to individual crews, Péroche writes: ‘ On les appelait les “putains” parce qu’elles n’avaient pas de mécaniciens titulaires ’ (‘They were called “whores” because they didn’t have regularly assigned drivers’). This compelling book, which is particularly good on how French railway workers coped with the German occupation during the Second World War, captures the bloody-minded yet critically intelligent spirit of the men who helped to make André Chapelon’s and other outstanding compound locomotives so very effective.

  ‘When I took leave,’ Péroche writes, ‘I made sure my engine was in the hands of a conscientious driver so as to be sure to find her in good shape when I got back. With today’s way of thinking, no doubt I’d be classed as a fool, but I can tell you that I often spent my rest day tinkering about on my engine. I used to get to the depot around nine and wash down the wheels and connecting rods with boiling water. Then I’d clean out the axle boxes and clear the grease nipples so that the oil could get through. It was the old school. Time was of no matter since we lived for the railway. For us, looking after our locomotive was a pleasure in itself.’

  To watch a highly polished, dark-green Chapelon Pacific start a boat train for Paris away from Calais Maritime was a wonder. Almost silently, save for a piercing high-pitched screech from its whistle, the complex engine would move its heavy train out of the station in one smooth movement. Even at speed, with little variation uphill and down, the exhaust was silent and, in warm months, no more than a heat haze from the chimney. These locomotives ran like silent sewing machines and it was hard to gauge just how very hard they might be working. They were the product of a designer who was little short of a genius and was backed up by highly trained engine crews and maintenance staff.

  André Chapelon was born on 26 October 1892 in Saint-Paul-en-Cornillon, a small village, of which his father was mayor, dominated by a medieval castle and set high above the Loire in the Massif Central, some forty miles south-west of Lyon. His great-great-grandfather, James Jackson, pioneered the crucible-melting process for steel making when he came to France from England in 1814, establishing a steelworks in Saint-Étienne. By 1850, the Jackson family steelworks produced half of all the steel in the Saint-Étienne valley. In 1857, Jackson’s granddaughter, Elizabeth Esther, married Antoine Chapelon.

  As if the landscape were not sufficient distraction, the young Chapelon would spend as much time as possible watching the trains of the PLM, the Paris–Lyon–Mediterranean railway, running close by between Le Puy and Saint-Étienne. His first footplate ride, in 1903, at the age of eleven, was on board a venerable 4000 class PLM 0-8-0, with its tall chimney, huge dome, and weatherboard cab. Chapelon was in love with steam from as far back as he could possibly remember. A woman he had fallen in love with married another man while he was away on service during the First World War. After that, he appears to have given himself over wholly to a lifelong passion f
or steam.

  What is regrettable is the fact that Chapelon’s life’s work was badly interrupted when he was in full creative flow by the Second World War, and then by the SNCF’s decision to put a stop to steam development in 1951. Chapelon hoped to continue building mainline French steam until 1970, by which time, he believed, electric traction would be able to make a quantum leap over steam in terms of high-powered performance. Was he wrong? The TGV, an electric train that has truly revolutionized travel by rail, started running in regular service in 1981. One of these striking, prognathous machines has reached 574.8 kph (357.2 mph), while regular services have been scheduled, from Lorraine to Champagne-Ardenne, at average speeds of 279.4 kph (173.6 mph), or very nearly 3 miles per minute. This is just what Chapelon, still very much alive when TGV development began in the early 1970s, had wanted. Before French steam development came to what must have felt like an emergency stop, he had designed express passenger locomotives that would have run comfortably over long distances at 200 kph (125 mph), and he was thinking ahead to locomotives capable of 250 kph (155 mph). These were not idealistic pipe dreams; they could have been realized.

  When Chapelon was eight years old, one of a pair of highly successful four-cylinder compound locomotives was put on show at the 1900 Paris World Fair. These Atlantics had been built at Mulhouse the previous year to the designs of Alfred de Glehn, the English-born chief engineer of the Société Alsacienne de Constructions Mécaniques, then in German Alsace, and Gaston du Bousquet, chief mechanical engineer of the Nord railway. They were impressive machines, light on the track and yet fast and powerful for the day; trains of 350 tons could be run over considerable distances at 120 kph (74.5 mph) and, on test, these non-superheated Atlantics achieved up to 1,444 ihp. It was one of these Atlantics that Churchward bought for testing on the GWR, followed by two more of the slightly larger Paris–Orléans version. One other small, if hugely significant, exhibit on show in Paris in 1900 was a revolutionary form of engine running on peanut oil; its inventor was the French-born Rudolf Diesel.

  While du Bousquet was making a great success with compounding on the Nord railway, in Germany Wilhelm ‘Hot Steam Willy’ Schmidt, a mechanical engineer working in Kassel, was developing a practical superheater for steam locomotives. This device reheated saturated steam produced in the boiler until it was all but dry and very hot – upwards of 350 °C (662 °F). Superheated steam was so effective that it substantially raised the thermal efficiency and power output – by up to 25 per cent – of many of the early locomotives to which it was applied. The first, in 1898, was a Prussian State Railways two-cylinder 4-4-0. With compounding and superheating, twentieth-century steam locomotive engineers had two very powerful tools to hand, and Chapelon was to make highly effective use of both.

  Chapelon himself took great interest in the work of both du Bousquet on the Nord railway and Churchward on the GWR, even before he had passed the necessary exams to enter the École Centrale des Arts et Manufactures. Founded in 1829, this was the leading French university in the field of practical engineering. But, with the outbreak of the First World War, on 28 July 1914, Chapelon, an artillery officer cadet, was soon in action at the front, observing artillery fire. During the Somme offensive of June and July 1916, he was an aerial observer, spending up to sixteen hours a day high over the battlefields in the basket of a hot-air balloon. This was a perilous position and on a single day five French balloons, with his friends aboard, were destroyed around Chapelon’s – his was the only one to survive unscathed.

  Chapelon went on to become a staff officer with the 106th Heavy Artillery Section, based in the fortress at Verdun. His biographer, Colonel Hugh Rogers, recalls Chapelon’s great gift of accurate range prediction, achieved through a mixture of imagination and meticulous research:

  Here Chapelon used his off-duty time in writing a paper setting out, with examples, his method of fire . . . His Colonel submitted his treatise to the Artillery Improvement Centre. It was examined there by the mathematician Émile Borel, who could find no fault in the system. A practical demonstration was then given by the Artillery Training Centre at Mailly camp. During a practice carried out with 75 mm guns the average point of impact with fifty shots practically coincided with the target. As a result, Chapelon’s new system was embodied in Artillery Regulations.

  In April 1919, a decorated Chapelon returned to the École Centrale and on graduating in 1921 he joined the PLM. (It was, as we will see, some of those in the engineering department of the PLM component of the future SNCF who later did much to undermine Chapelon, and even attempted to destroy his legacy.) In 1921, the PLM was one of six major French railways; the others were the Nord, Midi, Est, Ouest, and Paris–Orléans railways. They had been formed through the government-led merger of the nation’s railways in 1852. The state took control of the Ouest railway in 1908, and had overall control of the railways during the First World War. Suffering immense reconstruction problems in the wake of the war, and then the Great Depression, the SNCF was formed on 1 January 1938, two years after the return of the leftist Popular Front government in 1936. The government owned 51 per cent of this new national corporation, and the railway shareholders 49 per cent. The six major pre-SNCF railways became regions, and continued to develop along lines of their own, although locomotive and rolling stock policy was now to be centrally controlled. The state finally took full control of the SNCF in 1983, when the last private shareholdings were bought out.

  In the early 1920s there was little locomotive development going on at the PLM to stretch a young man of Chapelon’s abilities, nor, in the conservative engineering climate that prevailed there, were his theories on how to improve the flow and use of steam in locomotives likely to be well received – so he left to work with the Société Industrielle des Téléphones in the autumn of 1924. The realization, however, that he might now be separated from working with steam forever hit him quickly and hard. With the help of Louis Lacoin, his professor of thermodynamics at the École Centrale, Chapelon was back with the railways in January 1925, now in the research and development section of the Paris–Orléans railway.

  This was the challenge Chapelon needed. In 1923, the Paris–Orléans had decided to electrify its main line from Paris to Bordeaux and Toulouse, and not to order any further steam locomotives. However, with the widespread introduction of heavier, all-steel coaches, the performance capacity of the railway’s 189 compound Pacifics was becoming inadequate, and the solution was clearly to increase the performance of the existing steam stock. The Paris–Orléans railway was also interested in fuel economy. The upshot was that Chapelon’s design chief, Paul Billet, gave the young man his head. Chapelon took 3566, one of the Paris–Orléans’s Pacifics, built in 1913 and known by shed staff as ‘Le Choléra’, and transformed it, after careful analysis, from an adequate, if unloved, 1,450 dbhp express engine into a 2,700 dbhp machine. Perhaps not surprisingly, news of Chapelon’s remarkable achievement was met, at first, with disbelief on both sides of the Atlantic.

  It was only when trials began in November 1929 between Orleans and Bordeaux on the main line from Paris, almost exactly a century after Rocket’s winning performance at Rainhill, that the apparent miracle that Chapelon had performed was proved to be no illusion. On 4 April 1930, 3566 ran the 70.1 miles start to stop from Poitiers to Angoulême with a train of 362 tons at an average speed of 70 mph. Once up to the line limit of 120 kph (74.5 mph) the compound remained at that speed uphill and down, the speedometer needle only moving when the train had to slow down to stop – it was almost as if the engine were being supplied by a constant electric current. Shortly afterwards, 3566 worked a 567 ton train between Poitiers and Angoulême at an all but incredible average speed of 67 mph, again without exceeding the 75 mph line limit.

  What Chapelon had given 3566 was an unrestricted flow of high-temperature steam. His great achievement was in the application of holistic thermodynamics to the steam locomotive. Intuitively, he saw the steam engine as an organism, and eve
n before techniques existed to see inside the steam locomotive under power, Chapelon had thought through the best way for steam to flow through the circuit of boiler, superheater, cylinders, and exhaust for optimum effect. Every attempt was made to create a steady flow of steam superheated to a very high temperature. Eddies and vortices of boiler gas flow needed to be stabilized and controlled. The whole cycle had to be as seamless as possible.

  To achieve this, and to produce such spectacular results, Chapelon made the following improvements to 3566. The cross-sectional area through the steam-flow circuit, from regulator through to superheater, was doubled, reducing pressure drop and power loss by 75 per cent. The steam-chest volume was quadrupled, as was the steam-port area through the large-diameter poppet valves. Steam temperature was raised from 300 °C to 400 °C. A Kylchap exhaust, patented in 1926, provided a strong and uniform exhaust across all of the boiler tubes in the smoke-box, greatly improving combustion efficiency while decreasing back-pressure in the cylinders. The Kylchap exhaust system was based on the Kylala mixing device invented by the Finnish locomotive inspector Kyösti Kylälä. Kylälä had been looking, primarily, for an effective form of spark arrester – Finnish locomotives often burned wood – but found that his invention also saved fuel. Chapelon’s eye fell on research from anywhere in the world that would help him improve the steam locomotive. Eventually some four thousand locomotives worldwide – including two thousand in France and two hundred in Britain – were fitted with Kylchap exhausts, greatly improving power while reducing fuel consumption.

  A test run on 24 March 1931 with 3566 and a train with guests drawn from the rival French companies showed the locomotive averaging 65 mph between stops and gaining many minutes of lost time in the process, without exceeding 75 mph. When the special drew in to Bordeaux, Marcel Bloch, the Paris–Orléans’s chief engineer for rolling stock and workshops, turned to his young locomotive engineer and said, ‘Chapelon, vous avez fait quelque chose’ (‘You have certainly done something’).