GT3: a good try - but it got there too late
In the third and final part of his survey of British Railways’ gas turbine locomotives, PHILIP HAIGH looks at English Electric’s stylish yet dated GT3
When John Hughes arrived at English Electric in 1947 he sat down to consider the limitations of locomotives. They had to be heavy enough to produce enough adhesion to haul their trains but not so heavy that they contributed an excessive proportion of their train’s overall weight. If they were then they’d expend most of their fuel shifting themselves rather than their cargoes of goods or passengers and this would not be efficient.
Steam remained king and, Hughes wrote, was limited in power to around 1,500hp as a result of the size of locomotive fireboxes and the physical ability of firemen to shovel sufficient coal on a continuous basis.
Yet in 1947, there was a hint of the future with the London, Midland and Scottish Railway building two main line diesel locomotives in Derby. They each packed 1,600hp. Given the weight and space of diesel engines, generators and traction motors, Hughes thought this was the maximum likely to fit within the space and axle loads available.
He also considered diesel-electric locomotives to be expensive and complicated when compared with steam and would require railways to add electrical engineering workshops to their repair organisation.
These early diesel locomotives were heavy. Hughes reckoned a pair of them would be 122ft long and weigh 240 tons. When hauling a 16-coach, 500-ton train, they would comprise 33% of their train’s total weight. In contrast, a gas turbine locomotive of 2,700hp hauling the same coaches would be 20% of the total weight.
For a locomotive, Hughes argued that a gas turbine would be a third of the weight of a diesel engine of similar power. Electrical transmission was also heavy. Direct mechanical transmission would cut weight by 24%, he claimed at a time when English Electric was considering a gas turbineelectric locomotive. This was the type the Great Western Railway had ordered in 1946 from Brown Boveri (18000, RAIL 825) and Metropolitan-Vickers (18100, RAIL 834).
Mechanical transmission would be simpler to maintain than its electrical equivalent. Gas turbines had fewer moving parts than diesel engines and should need less maintenance.
Hughes wrote: “Recent investigation into the gas turbine electric locomotive show that within the British loading gauge and with the present electrical practice, anything more than 2,000hp from one vehicle can be achieved only with great difficulty.”
He argued: “The gas turbine direct driver locomotive has a transmission which is only 20% of the weight and 20% of the cost of the electrical transmission it replaces. The cost of the whole locomotive will be even more favourable to the direct drive because the vehicle itself is lighter and smaller than is the case with electrical transmission.”
In the conclusion to his 14-page paper, Hughes wrote: “The gas turbine mechanically
connected to the driving wheels through reduction gearing is the only application in which the features of high power, and light weight of the prime mover are reflected in a locomotive of general simplicity and low first cost. In this form it is the natural successor of the steam locomotive and will, when satisfactorily developed, take the principal place in the world’s railway transport.”
He had in mind a locomotive with a 4-8-4 arrangement with four coupled axles driven by a gas turbine and two twin-axle bogies at either end of the locomotive. The 2,700hp machine would weigh 90 tons and be 50ft long, less than half of a pair of 1,600hp diesel electrics and 12ft and 30t under a 2,700hp gas turbine-electric locomotive.
English Electric must have been convinced by the arguments of its newly appointed senior design engineer with Hughes later saying that his proposal had fallen on receptive ears at English Electric and that the British Transport Commission had said it was a promising proposal.
Former Rolls-Royce colleague Bill Allen produced design sketches of a locomotive that combined style and grace. One was centrecab 4-8-4 bearing the name Sir George Nelson; another was a 4-6-0 that drew on a mix of American styling and the grace of Sir Nigel Gresley’s pre-war Silver Jubilee.
Sketches were the easy part. Hughes and English Electric needed a suitable gas turbine and a suitable locomotive in which to place it. The 1948 locomotive exchange trials provided a rich source of information about the speed and torque needed for several different lines. From this, Hughes concluded that brake horsepower at the turbine shaft would need to be around 3,000hp with a transmission efficiency high at 94%.
In adopting a 4-6-0 wheel arrangement similar to a steam locomotive, Hughes could design a single cab which removed the need for a corridor inside the locomotive, creating more space. He could also place the air-intake at the locomotive’s front where it would be simple to design the necessarily large passages. It also reduced the chance of hot exhaust gases re-entering the turbine.
In balancing a need for the turbine centre line to be as low as possible to allow a heat exchanger and combustion chambers to sit above, and be as high as possible to allow for a gearbox, a compromise gave 5ft 9in driving wheels.
It was the turbine, heat exchanger and gearbox that comprised the bulk of GT3. There was little more to it but both took considerable design effort. Hughes developed what EE would brand its EM27L gas turbine, although he generally called this component the ‘engine’ reserving the term ‘turbine’ for those specific parts of it.
GT3 took filtered air and compressed it to 60psi before feeding it into a heat exchanger where it was warmed by exhaust gases, passing from there to two combustion chambers. Now a mixture of hot air and combustion gases, it passed through a turbine - the charging turbine - which directly drove to the compressor. A second turbine on a separate shaft - the power turbine - transferred remaining energy via a drive shaft and reduction gearbox to the centre axle and via coupling rods to the other two axles.
The compressor drives lubricating and fuel pumps, cooling fan and a generator to charge the batteries that started the engine via a 60hp motor.
In connecting the power turbine to the wheels, it had to be designed to produce the correct characteristics for rail operations. Hughes wanted a maximum speed of 90mph and a starting tractive effort of around 40,000lb. His preliminary studies showed for an optimum running speed of 50-60mph, the turbine would need a standstill torque of 5,600lb running at 5,400rpm.
Turbine models were tested in low-speed
Next morning the locomotive was towed outside and the engine again started in order to turn the driving wheels on rollers set in the track, which had previously been used for setting the valve events of steam locomotives. John Hughes, Chief Designer, Gas Turbine Locomotive Projects, English Electric
wind tunnels but when incorporated into the engine, running at 5,000rpm, Hughes found them unsuitable for locomotive use although they gave the high starting torque needed. They were unsuitable because the turbine was unable to cope with rise in gas flow as it speed increased which could cause the compressor to surge. This led to changes in the compressor design and a decision to switch the two-stage turbine in favour of one of three stages, running at 5,000rpm.
Fuel ignition within the combustion chambers came from a spark plug in each that lit an initial supply of propane which in turn lit gas oil jets allowing the propane supply to be automatically cut.
GT3 used a plate heat exchanger. In a paper he read to the Institution of Locomotive Engineers (ILocoE) in April 1962, Hughes described this as the largest single item in the locomotive and “possibly the most complex both in its design problems and manufacture”. However, he added that it was the solution to the high fuel consumption of the simple gas turbine.
The engine drove a shaft connected directly to the locomotive’s gearbox. Within it there was a splined shaft always in mesh with the turbine shaft and which could be engaged with either forward or reverse gear pinions.
Engagement needed a brake to briefly stop the turbine and this brake could itself rotate slightly to ensure the splines aligned before engaging.
The gearbox contained three sets of pinions and gear wheels given an overall reduction of 20.3 to 1 so that a turbine shaft speed of 9,000rpm gave a driving wheel speed of 443rpm (equating to a road speed of 90mph). Between the wheels and the output of the gearbox was a final drive that needed to allow for the driving axle to move relative to the gearbox as the driving wheels reacted to the track. This final drive used the principle of the Oldham joint but made the sliding surfaces of that joint cylindrical to allow it to cope with the driving axle tilting.
Around the time that British Railways was commissioning its locomotive testing plant at Rugby, the ILocoE organised a visit to Swindon’s test plant. Here Hughes bumped into Roland Bond who was the British Transport Commission’s chief mechanical engineer. As Hughes later wrote: “There is no doubt that this meeting was of decisive importance in the history of the project, as it brought about interest and encouragement at a high level in the British Transport Commission.” It led to a meeting chaired by Bond’s assistant, ES Cox, on January 27 1955 which agreed the basis of GT3’s trial running, including the use of Rugby.
Hughes didn’t yet have a locomotive but he met by chance Vulcan Foundry’s works manager at Rugby. This resulted in GT3 being built by Vulcan with its engine coming from EE. Vulcan assigned the works number 11657 to the locomotive and its frames were sufficiently complete to merit an official photograph on February 21 1957.
EE sent the turbine north to Newton-leWillows and Hughes records: “The main centre of activity was now at Vulcan Foundry where in the main erecting shop, now empty of the steam locomotives for so long built there, and in the process of being adapted for the diesels yet to come, the rudiments of a turbine locomotive were rapidly taking shape.
“It was in this great empty shop late one night that the engine was first brought to life coupled to driving wheels. Next morning the locomotive was towed outside and the engine again started in order to turn the driving wheels on rollers set in the track, which had previously been used for setting the valve events of steam locomotives.”
It was now time to take GT3 to Rugby. It had no cladding so the ‘naked’ locomotive was sheeted over and prepared for the 104-mile trip that was scheduled for Sunday May 19
1957. It reached only Acton Bridge, 13 miles away, before there was a snatch on the drawbar and a puff of smoke from the gearbox breather pipe. There was severe overheating in the roller bearing cage of the forward bevel gear. The train crept forward into a siding, the driving rods were removed and the driving axle jacked and packed clear of the rail following a call for help to Vulcan Foundry.
Back at the works, an examination of the gearbox revealed that the oil was too cold to flow properly. For GT3’s second attempt to reach Rugby it was stripped of excess equipment and travelled with its driving wheels lifted.
GT3 was the first non-steam locomotive tested at Rugby. It was not the finished product, Rugby’s test would help iron out problems. And problems there were, particularly with the oil system which used one grade for both turbine and gearbox and could be too viscous when cold.
Rugby flushed out turbine problems with rotor blades rubbing against stator blades. This forced the shortening of the rotor blades in the compressor which meant that full power was no longer available. This work meant removing the engine and returning it to EE. Another fault saw oil leak past a bearing into the stream of combustion gas where it caused several minor fires.
With the locomotive held on Rugby’s rolling road, tests could determine power, tractive effort and fuel consumption. They took place between 10mph and 90mph. Maximum power was limited by the compressor speed and the maximum temperature of the gas entering the turbine. Although the maximum design speed was 8,250rpm, tests never reached over 8,100rpm because of the shortened blades. This difference seems small but the effect was considerable because actual power output varies very nearly as the cube of the charging set speed. After refit, the speed was around 7,700rpm so power was cut considerably.
Under Rugby Superintending Engineer DR Carling, a test report dated July 31 1958 commented: “The characteristics are such that the locomotive will show to best advantage if it is employed on duties involving sustained high power and mainly running between 40 and 70mph, such as heavy night trains with sleeping cars, fitted freight trains and such.”
In addition GT3 made short moving tests alongside the Rugby-Peterborough line, forming what was surely a strange train. Behind an unclad GT3 came a four-wheel tank wagon containing fuel and then a pair of ‘8F’ steam locomotives as load, one in steam to provide braking if needed. The track used was steep (1-in-44) and sharply curved ( 5.5 chains), providing a stiff test.
The characteristics are such that the locomotive will show to best advantage if it is employed on duties involving sustained high power and mainly running between 40 and 70mph, such as heavy night trains with sleeping cars, fitted freight trains and such. DR Carling, Rugby Superintending Engineer
This first test took place on a damp morning. “The engine was started, the throttle opened and the brake released. The engine revolutions were building up rapidly and a twinge of doubt no sooner arose than it was dispelled by the realisation that the train was on the move… That morning at Rugby, in the anxiety to make a good climb, rather more power had been applied than was needed, and the startled onlookers - used to the slow passage of an ex-LNW [locomotive] with a couple of coal trucks, saw a dramatically fast climb,” Hughes wrote. Roland Bond visited one of these tests on June 10 1958.
A supplementary note on November 22 1957 addressed GT3’s fuel consumption. It revealed GT3 consumed 0.715 pounds per horsepower hour (lb/hphr) when running at 60mph on 2,360 drawbar hp (effectively full power). At half-power (1,180hp) specific fuel consumption rose to 0.99. A diesel-electric at 60mph on full power of 1,325hp consumed 0.565lb/hphr. The note explained: “The specific rate of the gas turbine is nearly 27% greater but it is capable of 78% greater output.”
It showed that to have best effect, a gas turbine locomotive such as GT3 needed to be worked hard. Running at less than full power was inefficient.
Tests produced a graph that showed the relative performance of GT3 against Southern Region 10202 (a 2,000hp diesel-electric locomotive). GT3 took a minute to reach 20mph, 10202 took half the time, 0-60mph just over 4½mins for GT3 and just under 4½mins for 10202. Both took just over seven minutes to reach 70mph. The difference was that GT3 had 19 coaches (762.0t) behind it and 10202 had 12 ( 524.8t). Quite a difference and enough to show the performance advantage that GT3 held.
For Hughes and the EE team, the challenge after Rugby was to complete the locomotive and its tender that would contain a trainheating boiler and tanks for fuel and water. This tender would include a corridor and connection to allow crew to move between footplate and train. Vulcan squeezed it into a works now busy building AC electric locomotives for BR.
He also needed a base near some lightly used lines to allow for on-track tests. BR suggested Whitchurch where the motive power depot had closed but could be made available and from where a quiet line ran via Malpas to Tattenhall Junction on the CreweChester line.
BR and EE met at Whitchurch to discuss arrangements. They agreed to have the local electricity board boost the old MPD’s supply to allow for immersion heaters in GT3’s oil tanks. Hughes also arranged for heaters for staff comfort, assuring BR they would not run at the same time as the immersions. EE would also hire a camping coach from BR, to be housed in No 4 road of the old shed. The pit in No 2 road would be dug out for GT3.
They met in Crewe on October 10 1960 and talked about tentative arrangements. With GT3 expected to be ready in early November, BR and EE proposed that it run under its own power to Whitchurch and then conduct
light engine and loaded tests between there and Tattenhall Junction and then a 200-mile loaded run that would include testing the steam-heating boiler. The plan was then to put GT3 into revenue earning service on the former Great Central line between Nottingham and Marylebone.
BR and EE refined the plan to see GT3 leave Vulcan Works at 0430 on Sunday November 27 with light engine tests starting the following day and lasting until December 21 when GT3 would run to Crewe Diesel Shop for stabling over Christmas. Eight-coach tests would start on January 2 1961 for a week and then GT3 would run with eight coaches to Llandudno Junction and back in the week starting January 9.
Reality took a difference course. As Hughes related in his ILocoE paper in 1962, there were various defects on the reversing system, mainly to do with the power turbine brake. There were also problems with starting the turbine as a result of a small leak which affected a pressure reducing valve.
BR rescheduled the test runs for March 1961. GT3 ran successfully five trips between Whitchurch and Llandudno Junction with 12 coaches. Starting problems followed and then three further trips with the final one loaded to 16 coaches.
GT3 then moved to Leicester which was near EE’s gas turbine headquarters at Whetstone and worked daily test trains between Leicester and Woodford Halse while drivers were trained. This lasted for three weeks but a recurrence of power turbine brake problems prevented planned runs to Marylebone. With the problems fixed, 13 runs followed.
In May 1961, GT3 was on display in Marylebone goods yard as one of the exhibits marking the ILocoE’s 50th anniversary. It kept company with the final steam locomotive BR built, 92220 Evening Star, as well as ‘Deltic’ D9003, ‘Warship’ D867 Zenith, 25kV electric E3059 and others.
Further tests followed, including on the steeply graded West Coast Main Line over Shap. These tests included runs with a dynamometer car in December 1961. The first was planned for December 13 but flat batteries prevented GT3 starting. December 14’s test produced some results but was riddled with problems. Even with the controller fully open, the compressor would only spin some 500rpm short of its maximum, so reducing power. This was the result of a fuel spill valve operating to limit fuel when turbine inlet gas temperature approached its maximum 777°C.
There were also unforeseen stops for adjustments and for adverse signals which delayed the 15-coach test train. This resulted in the ascent to Shap summit being made as quickly as possibly. GT3 reached a balancing speed of 20mph, short of the 38mph it should have been capable of.
The return run from Carlisle started an hour late after a fire caused by a lubricating pump overheating. GT3 was then unable to start its train on the 1-in-131 rising gradient at Upperby Bridge Junction and needed banking assistance.
With GT3 short of up to 400hp and wide variations in compressor speed, these tests were disappointing. BR report concludes: “In view of these undesirable features, it is considered that further dynamometer tests would not be justified until the control system has been sufficiently developed to produce a reliable output.”
Earlier Shap runs had produced better performances. With 12 coaches, it had
The performance of GT3 is just what I should have expected from our Rugby tests. I think it would give a ‘Deltic’ a good run for its money on a suitably timed run and could easily beat any other class of locomotives in the country. DR Carling writing to Hughes on January 2 1962
breasted the summit at 42mph (and at 34mph with 15 coaches). In comparison a ‘Duchess’ class steam locomotive managed 37mph with 11 on and an English Electric Type 4 (Class 40) managed 25mph with 12.
Despite the poor dynamometer car runs, Hughes told the ILocoE that GT3 had managed 10 coaches on its first Shap run, 12 on the second day of tests and 15 on the third. “The locomotive showed itself the master of this load over this difficult route and thereafter eight consecutive trips were worked with a 12-coach train,” he said.
He and GT3 were too late. 1961 was not 1947. Back then, Hughes was arguing that a gas turbine with mechanical transmission provided a weight and power advantage over diesel-electrics. By 1961 as the 90mph GT3 worked over Shap, 100mph diesel-electrics ran on the East Coast Main Line. These were English Electric’s ‘Deltics’. They packed 3,300hp against GT3’s 2,700hp and weighed 100 tons. GT3 weighed 80t but needed its 44t tender to carry fuel and its steam-heating boiler. This gave a total weight of 124t. GT3 was slower, weaker and heavier than the latest diesel-electrics.
It was the end of the line. Former Rugby engineer DR Carling wrote to Hughes on January 2 1962 from Utrecht: “The performance of GT3 is just what I should have expected from our Rugby tests. I think it would give a ‘Deltic’ a good run for its money on a suitably timed run and could easily beat any other class of locomotives in the country (electrics being hors de concours if not hors de
combat)… I do hope EE don’t abandon the project just when it has proved its worth: I hope they are not as broke as that, though I know things aren’t too good just now.”
Renowned photographer Derek Cross wrote on May 19 1963: “I am very sorry to hear that B/Rlys are not interested. Incidentally if you can give me any information as to what she is doing at present I would be most grateful, especially if I might be able to get any more photos of her for my collection.”
There was no information. GT3 was not to run again despite discussions with BR in autumn 1961 about introducing it to revenue traffic. Hughes told the ILocoE in 1962: “Had its withdrawal for reasons of economy been foreseen then the running programme carried out in eleven months of 1961 would have been conducted differently, minor troubles would have been lived with in order to have run a bigger mileage. As it was the work was conducted with the conviction that all defects brought to light should be remedied in order to achieve the best possible record on entry into regular service.”
EE took the locomotive back in 1962 and stuck it under a tarpaulin. It was sent to Thomas Ward’s scrapyard in Salford in 1966.
Hughes addressed the Rugby Engineering Society on February 19 1964. His conclusions have echoes of today’s problems.
“The very high capital cost of electrification has restricted it to routes of high traffic density or particular operating requirements or difficulties and it seems certain that the majority of the world’s route mileage will be operated by independently powered locomotive at least for the remainder of this century. It also seems pertinent to point out that continued expansion of so costly a system is very much at the mercy of fiscal policies.
“Of outstanding importance has been the realisation of the high power requirements for the modern railway and whereas in this country in 1947 a diesel locomotive of 1,600hp would be considered for main line duties, now it is 2,750hp and shortly will be 4,000hp.”
Hughes acknowledged the change in railway operating that had come with the development of DMUs in the 1950s and the rise of lightweight, high-speed diesel engines combined with hydraulic transmissions (the Western Region practice in ‘Warships’ and Westerns’). At the same time, passenger numbers were falling and trains becoming shorter than the 15-coach behemoths that stretched BR’s steam fleet.
His 1947 reference to pairs of diesel locomotives was not entirely wrong. Never mind the ‘Deltics’ (they contained two engines after all), BR used pairs of 2,700hp English Electric Class 50s on its trains over Shap in the 1970s and when it introduced High Speed Trains in 1976 they consisted of a 2,250hp diesel power car at each end of a fixed formation train. Rising passengers numbers are now pushing operators back towards longer trains. Today’s London-Glasgow trains over Shap consist of 11-car Class 390 electric multiple units with 7,980hp and weighing 567 tonnes.
Of GT3, Hughes rightly said: “A project such as this locomotive not only requires a sustained effort by the organisation concerned with its design, construction and development, but also a belief in its future by a possible user to carry it forward.”
About the author Philip Haigh, Contributing Writer
Philip was formerly Business Editor at RAIL, leaving in September 2013 after spending 16 years with the title. He has a background in engineering, and is now a freelance writer and regular contributor to RAIL.