Giants of Steam

by Jonathan Glancey

  In 1916, a year after he had designed a lacklustre four-cylinder Pacific which stayed on the drawing board – it was little more than a stretched Ivatt Atlantic – Gresley read a series of articles, complete with detailed drawings, in Engineering magazine about the Pennsylvania Railroad’s hugely impressive class K4 Pacific, built by Alco from 1914. By 1928, 425 of these competent locomotives were at work. Developed over many more years, they became ever more powerful – 4,325 ihp with poppet valves, or 35 per cent more than the first of the class – and fast.

  The class A1 Pacific Great Northern owed something to the K4s, something to Churchward, and something, of course, to Henry Alfred Ivatt, Gresley’s predecessor as locomotive engineer of the GNR. But it was also very much his own machine, the locomotive that saw him ascend the ladder to his seat in the pantheon of the world’s great steam locomotive engineers. With the design of the A1s, Gresley had had the full support of his management. As the GNR was about to be merged into one of the Big Four privately owned railway companies in 1923 – an intermediate stage on the iron road towards the nationalization of Britain’s railways in 1948 – the company was keen to bow out in style. In the event it was, through Gresley, to stamp its mark on the locomotives of the new LNER right up to the great man’s untimely death in 1941 and beyond, since Gresley’s locomotives continued to run the crack express passenger trains of British Railways’ Eastern, North-Eastern, and Scottish regions as late as 1966.

  When the LNER was formed, the new board of directors offered the job of chief mechanical engineer to the most senior locomotive engineer from among the constituent railways brought into its capacious fold. This was John George Robinson, chief mechanical engineer from 1900 to 1922 of the Great Central Railway, for which he had designed three excellent classes of locomotives: the class 11B ‘Jersey Lily’ Atlantics of 1903 (their curvaceous good looks were likened to those of Lillie Langtry, the beautiful actress who caught the roving eye of Edward, Prince of Wales), the class 11E and 11F Director 4-4-0s of 1913 and 1920, and the class 8K 2-8-0s of 1911, a freight locomotive of which an extra 521 were built for the Railway Operating Division of the Royal Engineers for service in France, Belgium, the Middle East, and Persia during the First World War, and which continued in front-line freight service with British Railways until 1966.

  Feeling himself too old, at sixty-seven, for a role that he felt should rightly belong to a dynamic and much younger engineer, Robinson politely refused the job, suggesting Gresley in his place. Inevitably, this rather upset Vincent Raven, who had been chief mechanical engineer of the North Eastern Railway – the upper English half of the east coast main line from London to Scotland – from 1910. Raven had unveiled a Pacific at almost exactly the same time as Gresley. The two, both with three-cylinder drive, were pitted against one another in tests and, perhaps inevitably, the Gresley A1 was chosen as the standard-bearer for GNR locomotive design. This made sense, for even while the Raven locomotives were decent machines, they looked Edwardian, or even Victorian, whereas Gresley’s mechanical racehorse looked like a modern express passenger locomotive should: long, lithe, taut, and very much all of a piece. From the words ‘Right away, driver’, the A1 was one of those designs that could only be spoilt if anything was added to it – spoilt, too, if anything was taken away from it.

  Not only did the A1 look just right, but the idea of naming these beautiful, modern, apple-green engines after racehorses was also truly inspired. The third of the class, 4472, was dubbed Flying Scotsman to match the 10.00 King’s Cross to Edinburgh express service. First run in 1862, this had been known as the Special Scotch Express, but was renamed the Flying Scotsman by the LNER in 1924. That was the year of the British Empire Exhibition, held at Wembley, alongside the tracks of the Metropolitan Railway and the former Great Central Railway. There, the LNER displayed Flying Scotsman along with new coaches from the Flying Scotsman train. Gresley’s imperious new Pacific stood beside the GWR’s new, and considerably smaller, Collett four-cylinder 4-6-0, 4073 Caerphilly Castle.

  Flying Scotsman weighed over 92 tons, while Caerphilly Castle was 13 tons lighter. Flying Scotsman was 70 feet long, Caerphilly Castle 65 feet. With a grate area of 41.25 sq ft, Flying Scotsman’s fire-box was much bigger than Caerphilly Castle’s, at 29.36 sq ft. The evaporative heating surface inside Flying Scotsman’s boiler and fire-box added up to 2,930 sq ft; that of Caerphilly Castle totalled just 1,963 sq ft. Despite these statistics, stacked overwhelmingly, or so it seemed, in the A1’s favour, a notice in front of Caerphilly Castle stated, quite categorically, that the GWR engine was Britain’s most powerful express passenger locomotive. The basis for this claim was the fact that Caerphilly Castle boasted a tractive effort of 31,625 lb, compared with Flying Scotsman’s 29,835 lb. Tractive effort is calculated as the square of the diameter of the piston, multiplied by the piston stroke, multiplied by 85 per cent of the maximum boiler pressure, divided by the diameter of the driving wheels. The final figure, based on the dimensions of a two-cylinder locomotive, should be increased by 50 per cent for a three-cylinder engine and 100 per cent for a four-cylinder engine. (The vast majority of steam locomotives have been built with two cylinders.)

  What the figures for tractive effort show is the maximum (or theoretical) pulling force a given locomotive can exert on starting. It might be compared to the figures given for torque for internal combustion engines. While horsepower is the combination of force and speed, torque is the pulling power needed to get a car moving from rest, or from low speeds in high gears. Until the 1930s, locomotive engineers tended to rate the power of locomotives in terms of tractive effort alone, as this enabled the operating department to calculate the train weight a locomotive could start and could haul on varying gradients. However, this did not necessarily relate to how fast and powerful a locomotive was when running. While a high maximum tractive effort is essential for a goods locomotive, which is asked to start very heavy loads from rest, it is not required for a passenger locomotive called on to run light trains at high speeds. A better measure of performance capacity at speed is horsepower output. As speed and haulage capacity began to matter more and more, for both passenger and freight trains, so, from the 1930s, locomotive engineers increasingly evaluated horsepower outputs. For steam locomotives, the two most important measures are indicated horsepower (ihp), the power generated in the locomotive’s cylinders, and drawbar horsepower (dbhp), the power a locomotive can exert to pull a train after the power needed to move itself against friction and air resistance has been extracted.

  Generations of enthusiasts – in Britain and abroad – believed that the most powerful locomotives were those with the highest tractive effort. But this was not always true, and so visitors to the British Empire Exhibition would perhaps have been right to question the GWR’s claim for Caerphilly Castle. In any case, the matter was put to the test the following year when the GWR and LNER agreed to a trial of strength. In late April 1925, 4474 Victor Wild was packed off with an LNER crew to run the GWR’s principal express, the Cornish Riviera Limited, a demanding non-stop run from Paddington to Plymouth, leaving London with a full load weighing about 530 tons. Crew and locomotive did well. Victor Wild proved able to maintain the very tight schedule. However, the GWR team, with 4074 Caldicott Castle in the spotlight, was able to knock fifteen minutes off the schedule on a single run with a 380 ton train. More than this, Caldicott Castle was burning considerably less coal, with an average saving of 6 lb per mile. Over the 225.5 miles from Paddington to Plymouth, this would add up to more than half a ton. Given that Caldicott Castle carried just 6 tons of coal in its small, six-wheeled tender, this was both a considerable saving in what was becoming, even at that time, an increasingly expensive fuel, and an appreciable saving over the fuel consumption of Victor Wild. To be fair, the LNER crew had been instructed to adhere to the timetable and not to try to run ahead of it, while the long wheelbase of Victor Wild had to be run at lower speeds downhill due to the restrictive alignment of GWR curves. As a result
, Victor Wild had to work harder uphill, burning more coal in the process. Following the trials, the GWR improved its curves and Gresley improved his Pacifics.

  To the astonishment of the LNER management – and of enthusiasts – much the same thing happened when 4073 Caerphilly Castle ran test trains of up to 480 tons from King’s Cross to Doncaster and back. Refusing to slip its driving wheels – which, at 6 ft 8½ in, were a half-inch larger in diameter than those of the rival A1, 2545 Diamond Jubilee – as it lifted its heavy trains up the greasy rails through Gas Works Tunnel immediately beyond the platform ends at King’s Cross, the GWR engine was easily master of the east coast main line, while burning 6 per cent less coal.

  Gresley might have designed locomotives that appeared to be more modern and powerful than the GWR Castle class, but what he had yet fully to appreciate was that, years earlier, Churchward had understood much about the need for a free-steaming boiler – which both the A1s and Castles possessed – matched to a valve design that allowed the maximum expansive use of steam and its rapid ingress and exit from cylinders, to give a free flow of steam through the locomotive’s operating cycle. Despite their old-fashioned looks, the Castles were, in this respect, the more advanced of the two designs.

  A part, though, of what made Gresley a great engineer was that he learned the lessons from Churchward quickly. The new class A3 Super Pacifics which emerged from Doncaster from 1928 onwards had boilers pressed to 220 psi, rather than the 180 psi of the A1s. Like many engineers, Gresley was persuaded that boiler scaling increased at pressures above 180 psi. However, his new chief chemist, the forceful T. Henry Turner, pointed out that chemical water treatment would overcome this, and its introduction by the LNER in 1928 enabled Gresley to use higher pressures.

  The A3s were fitted with improved valve settings. With long-lap, long-travel valves – standard on Churchward designs – steam could now flow freely through the valves into and out of the cylinders and do more work by expansion at short cut-off. Before the advent of long-lap, long-travel valves, many British locomotives were unnecessarily choked, unable to channel steam freely enough through the cylinders to allow them to run rapidly and economically. There had been notable exceptions, especially in the designs of Patrick Stirling and Henry Ivatt for the GNR and William Adams for the London & South-Western Railway. The ideal, as Churchward and Chapelon demonstrated and explained, for a steam locomotive at speed was for the driver to open the regulator fully whenever possible and to cut off steam to the cylinders as early as possible. In 1925, Victor Wild and Diamond Jubilee were being worked at speed uphill with cut-offs of up to 50 per cent with the regulator partially opened. This was a bit like driving a fast car in third gear with the foot down on the throttle, rather than changing up to fourth or fifth and going easy on the gas.

  Bert Spencer, Gresley’s chief technical assistant, was very much instrumental in persuading the chief mechanical engineer of the advantages of Swindon- and French-style valve settings. The result paid off in the Pacifics: the A3s built between 1928 and 1935 were magnificent machines, while the A1s were modified with improved valve settings and, where necessary, new A3 boilers. It was not that the A1s were poor performers, but they were heavier on coal and water and not as efficient as they might have been.

  Significantly, Gresley’s use of the term ‘Super Pacific’, borrowed from the French Nord railway, coincided with the development in the United States of what was known as ‘super power’ at the Lima works in Ohio. The A-1 class 2-8-4 built by William E. Woodard in 1925 was, along with the work of Chapelon in France, one of the key turning points in the development of the steam locomotive, leading to a new generation of machines that were to give diesels and electrics a very good run for their money before the management tide turned against steam in Britain and the USA after the Second World War. Woodard described super power as ‘horsepower at speed’, the goal of most American locomotive engineers from 1925 until the end of steam construction in the USA in 1955.

  The LNER’s A1s, however, had been built to haul heavy expresses at moderate speed. Gresley intended them to pull 600 ton trains – 30 per cent heavier than before – at average speeds of 50 mph. Even by the standards of 1922 this was slow. There was, though, a reason. From 1895, after a summer of hair-raising railway races to the North, when the east and west coast railways from London to Scotland attempted to go ever faster on rival runs from King’s Cross and Euston to Aberdeen, the companies had agreed to a truce, with a leisurely timing of 8 hours 15 minutes – an average speed of 48 mph – from Edinburgh to King’s Cross. So there was no imperative to design for speed. Instead, express trains became ever more comfortable and, in consequence, very heavy indeed.

  In May 1928, the LNER made the headlines with the introduction of the non-stop Flying Scotsman service. The first train was headed by 4472 Flying Scotsman. There was no need to hurry, the express ambling down to Edinburgh (in Britain, trains go ‘up’ to London and ‘down’ to all other destinations) in 8 hours 15 minutes. It was no mean feat, though, to maintain what was, at 393 miles, the world’s longest non-stop run, on a daily basis. It was possible, in summer, partly because the engines could pick up water at speed from six troughs along the route, and partly because Gresley’s new design for a corridor tender allowed crews to change halfway through the journey as the locomotive thundered on. Significantly, from the point of view of speed, the winter timetable included five intermediate stops within the same time schedule.

  The relief crew of the non-stop train would ‘ride on the cushions’ in a reserved compartment in the leading coach, chatting, reading, staring out of the window, or snoozing until they were needed. Then driver and fireman would make their way through a corridor connection and into an 18 inch wide, 5 foot high passage inside the tender, leading to the glare of the rolling footplate. This was such a sensible invention that it seems odd that no other railway adopted it. The experience, meanwhile, of huddling through one of these Gresley corridors and out into the cab of an A1, A3, or A4 Pacific is one that few fairgrounds can rival in terms of sheer thrill and drama. I was lucky to experience it once, when riding behind the streamlined A4 Pacific 60009 Union of South Africa at 75 mph. The comparison my excited mind made at the time – this is what I scrawled in my notebook – was with the experience of first negotiating the claustrophobic corridors and anterooms of the Doge’s Palace in Venice, before experiencing the inexplicably shocking scale and splendour of the vast and opulent Sala Maggiore at the heart of the building. The footplate of a big and powerful steam locomotive at speed is never less than an operatic spectacle.

  A revolution in speed, though, was about to begin which would indeed lift the steam locomotive into a new reality of super power, and Gresley was at its forefront. The spark was ignited by the launch of high-speed inter-city diesel and steam services in Germany and long-distance streamlined diesels in the United States. These events made railway management in Britain at long last think seriously about the rise of new forms of competition, both internal and external. Traditional heavy steam express trains were being challenged by lightweight, high-speed diesel and steam trains, at the very same time as the first inter-city air services were taking off and the wealthy were able to buy fast cars which could purr elegantly and comfortably across long distances.

  A step change in speed in railway travel was deemed essential – by the LNER at least – and it was this, and the need to accelerate existing heavy main-line trains, that drove the design of the steam locomotive to unprecedented heights. In Britain, Gresley was very much the steam locomotive engineer of the moment. In 1934, when the first high-speed tests were made on the LNER between London and Leeds, Gresley was fifty-eight years old. Within the next four years, he gave the British steam locomotive a new lease of life, proving that it was easily possible for a steam locomotive from Doncaster to take on Germany’s revolutionary new express diesels – and to win very convincingly indeed, especially in terms of passenger comfort. He was to be dead within seven year
s.

  Gresley was a big man in every sense. Large-framed, 6 ft 2 in tall, commanding and direct – friendly and good-humoured, too – he was born the fifth son of the Rev. Nigel Gresley, rector of Netherseal (or Nethereseale), in Derbyshire, and Joanne Beatrice, née Wilson, in June 1876. The family was aristocratic, although both its major homes, Drakelowe Hall and Netherseal Hall, were demolished in the 1930s. The future designer of Flying Scotsman, Mallard, and Green Arrow could trace his ancestry back to soldiers who fought with William the Conqueror at the Battle of Hastings and, through Hugh de Calvacamp, archbishop of Rouen in the tenth century, to the Viking warriors who settled in Normandy in the late ninth century.

  St Peter’s, Netherseal, in the churchyard of which Gresley is buried (his tomb was restored in 2009 by the Gresley Society, with the help of readers of Steam Railway magazine), is a modest thirteenth-century building with a fifteenth-century tower. It was largely rebuilt in 1877, while Nigel was learning to walk, by Arthur Blomfield, the church architect who employed the young Thomas Hardy (an architect before he found fame as a novelist and poet) and who, daringly, built a house for himself in concrete in East Sheen in southwest London. There is a simple Gothic bench in the church made by the young Gresley.

  The Gresleys lived comfortably, with a staff that included a Swiss nursemaid who taught Nigel to say his prayers in French before he could say them in English. From Marlborough School, Gresley was taken on as a premium apprentice at the LNWR’s Crewe works in 1893 under the aegis of the autocratic and eccentric, if extremely able, Francis Webb. At a dinner held in Crewe in 1919 by former apprentices and pupils, Gresley, according to Geoffrey Hughes, one of his biographers, sang a duet with Sir John Aspinall, the retired chief mechanical engineer and general manager of the Lancashire and Yorkshire Railway, who, years before, had worked for Webb at Crewe, and who had employed Gresley in his drawing office at Horwich in 1898. To the tune of ‘The Holy City’, a popular hymn written by Frederick E. Weatherby and scored by Stephen Adams (the pen name of the musician Michael Maybrick) in 1892, the engineers crooned the ‘Crewe Steam Shed Song’: