Living Museum of the Great Western Railway

Home » Other Articles » Going Loco Index » Going Loco - October 2021 - Going Loco - October 2021

Going Loco - October 2021


Oh The Horror! Part 2...

As if the last lot wasn't enough... The sort of engines that make you want to go and hide in the cupboard are back. With a guest blogger too no less! Your friend and mine, Mr Frank Dumbleton – GWS official photographer and contributor of the vast majority of the images to Going Loco – has put down the camera and picked up the virtual pen to continue to terrify traction lovers everywhere. Remember folks, stick to the path...*

Continuing the Hallowe'en theme from last week, we have some more horrible engines. The first is broad gauge 2-4-0 four cylinder compound No 8 which enjoyed an extremely short working life in its original manifestation. It may well have contributed to the GWR's suspicious view of compound locomotives in the following years.

The broad gauge tandem compound 2-4-0 No 8, the subject of this horror story.

To understand the background to No 8's explosive debut in passenger traffic, we turn to E L Ahrons' seminal work, Locomotive & Train Working in the Latter Part of the Nineteenth Century which first of all sets the scene:

Three ungainly 6 ft. coupled broad gauge engines were stationed at Swindon for piloting purposes. They were the Acheron, Phlegethon, and Hawk. These engines had 6 ft. coupled wheels, and 16 in. by 24 in. cylinders, and were known on the line as the Avonside class. The three engines in question were the last broad gauge coupled tender engines, other than the much later ‘Convertibles’, to be built, and they came out from Swindon in 1865-6, having been constructed to the designs of Mr Joseph Armstrong. There had been 26 of them, of which 20 were built by the Avonside Engine Co. of Bristol, Swindon being responsible for the other six. Of these, ten had been converted into 2-4-0 saddle-tank engines with 5 ft. wheels, leaving 16 of the original tender engines at work.

They were extremely ugly awkward-looking engines, and the appearance of the Phlegethon was not improved owing to its old-fashioned ‘rolled top’ chimney having a bit out of the top on one side, and a dent in the other. The Acheron was more presentable, as it had been fitted with one of Mr. Dean's standard chimneys. It is just within the bounds of possibility that 60 miles per hour might have been ‘pumped’ out of them.

The role of these three engines was for two of them to stand pilot at Swindon on alternate days in charge of the two oldest broad gauge drivers, on which job they put in a mileage of about 1½ per day, and did it well. The third was retained as a spare engine to be requisitioned when either of the other two became exhausted. Of course, if one of the 8 ft. ‘singles’ had to give up its train, then the pilots had reluctantly to take a hand, and occasionally go down to Bristol, but I cannot recollect a single instance of their services being required between Swindon and London, and this, it may be added, speaks volumes for the excellent condition in which the main line 8 ft. engines were kept.

Suddenly, early in 1886, a bolt fell from the blue on to the peaceful Swindon pilots, and they were put down to work a regular daily train. The manner in which this came about was as follows. The 3 o'clock up train from Bristol had left Cornwall in the morning at what was probably the most convenient hour for the ordinary traveller. Moreover, it carried third class passengers, and being sandwiched between the up Dutchman and up Zulu, into both of which the third class passenger was forbidden to set his foot, the train in question gradually became the up West of England express of the day, and assumed proportions as to load which were utterly beyond the powers of one 8 ft. ‘single’.

Until about the end of 1885, the London engine which had worked down with the 5.30 a.m. newspaper train to Bristol, returned with the 3 o'clock ‘up’ from the latter place, and invariably had to be piloted as far as Swindon, generally by a second 8 ft. ‘single’ from Bristol shed, and sometimes by coupled engines of the ‘Avonside’ and ‘2015’ classes, for on several occasions the train had ‘stuck’ in the Box tunnel.

The No 8 which was nominally a rebuild of the compound No 8, but possible only the driving wheels were re-used.

The load was usually nine, and sometimes as many as eleven eight-wheeled coaches, and it took the two big-wheeled engines all their time to climb the two-mile rise of 1 in 100 through the tunnel, out of which they usually emerged at an almost imperceptible crawl. Then it occurred to the running department that perhaps a coupled engine would be more suitable for this train, and it was decided that the London engine should come off the down newspaper train at Swindon, and that one of the Swindon coupled pilot engines should take its place as far as Bristol, in order to work the 3 o'clock back to Swindon. But, even after this change, it generally happened that Acheron, Phlegethon and Co. had to requisition the services of a Bristol pilot, so that matters were only improved to a slight extent, and the once placid countenances of the Swindon pilot drivers began to wear a very worried look.

It was at this period, the spring of 1886, that the little known four cylinder broad gauge compound engine No. 8 was built. No. 8 bore this number in place of a name, for it was constructed on the ‘convertible system’, i.e., by changing the axles it could be converted to a standard gauge engine. The broad gauge engines of the Great Western proper had names only but no numbers. Those taken over from the Bristol and Exeter Railway had the Great Western Railway numbers 2001 to 2095, but never possessed names, and the South Devon stock, which became Great Western Railway 2096 to 2180, had both names and numbers. A considerable number of these names were the same as those of the Great Western engines proper, so that there were many cases of two engines having the same name. ‘Convertible’ engines were numbered with the narrow or standard gauge stock and were nameless.

No. 8 had six frame plates, of which the inside and middle ones would correspond to the inside and outside frames of a standard gauge double-framed engine, and in fact when the engine came to be converted they would become such. In a broad gauge engine the wheels were placed outside both frames, and in addition No. 8 had a third pair of frames which provided outside bearings for the leading pair of wheels, as well as a deep angle plate upon which the side platform rested.

The chief feature, however, was in the four cylinders, which were arranged inside in tandem fashion, with the larger low-pressure ones in front. The high-pressure cylinders, 14 in. diameter by 21 in. stroke, had the usual form of piston rods screwed into crossheads, but in place of front tail rods connected to the low-pressure pistons, there was no direct rod connection between the pistons of the high-pressure and the low-pressure cylinders. In place of the more usual tandem arrangement, the low-pressure cylinders, 22 in. diameter by 21 in. stroke, were arranged each with two piston rods, which were screwed into the pistons at the sides instead of in the centre. These rods issued from glands in the back castings and passing outside the high-pressure cylinders were connected to the same crossheads as the high-pressure pistons, so that each crosshead formed the connection for three piston rods.

The arrangement was similar to one used for 20 eight-wheels-coupled compound coal engines designed by the late M. du Bousquet for the Northern Railway of France. In other respects No. 8 looked like an ordinary 2-4-0 express engine. The coupled wheels were 7 ft. in diameter, and the boiler had 1,258 sq. ft. of heating surface, and carried 180 lbs. pressure. The weight on the coupled wheels was 31¾ tons, and the total weight was 45¼ tons in working order.

Had No. 8 had a longer existence it is probable she would have been used exclusively for the 3 p.m. from Bristol, which was in sad need of a more powerful engine. But she was an unfortunate machine, and things happened to her. She ran but a few hundred miles, and as I was on the footplate during most of her trips, I have a vivid recollection of these happenings.

To begin with she was tried as a light engine for a number of short runs between Swindon and Dauntsey, during which she seemed to behave fairly well except for one or two minor faults such as might occur in any new engine. These having been rectified, it was decided to give her a run on a passenger train, and for a start, the piloting of the celebrated 3 p.m. from Bristol, as between Bath and Swindon, was chosen.

The Acheron had gone down to Bristol with the down newspaper train as usual, and when old Mr. Jones, her driver, had hooked on to the return 3 p.m. and sent out the daily ‘S.O.S.’ signal for piloting help, he was instructed to proceed to Bath, where he would find the new compound engine waiting to help him on to Swindon. At Bath No. 8 was attached in front of the Acheron, to the great disgust of Mr. Jones, who scented trouble. There were plenty of us to help him out, for besides the trial driver and fireman, the chargeman fitter and two other individuals were on the footplate. Moreover, Mr. Jones's instructions were to give the Acheron only sufficient steam to keep his own engine going, and to leave the work of pulling the train to No. 8.

For about 5 miles all went well, until we reached a point about 10 yards outside the western end of Box Tunnel, when two loud explosions were heard, and amidst the roar of escaping steam we entered the tunnel. It was pitch dark, and a rain of fragments of cast iron mixed with large gun-metal nuts was projected against the roof of the tunnel, from which they rebounded like shrapnel on to the footplate. The position was distinctly uncomfortable; no one could tell what had happened, but the driver dare not shut off steam or the train would have ‘stuck’ on the 1 in 100 rise, for had this occurred all the efforts of Mr. Jones and the Acheron would have been useless.

As a matter of fact, No. 8 had smashed to bits three of her four pistons and cylinder covers, the fragments and nuts from which had broken the ports and been blown out of the chimney. Luckily no one was hurt, but one or two of us were hit on the shoulders by gun-metal nuts. The fourth piston – one of the high-pressure ones – was fortunately intact, and by keeping steam on we just managed to crawl out of the Corsham end of the tunnel, the Acheron keeping the train moving. At Chippenham we were pulled off by a goods engine that happened to be there, and after taking down the motion, we were ignominiously towed home to Swindon about supper time.

As the train emerged from Box Tunnel I happened to look back at the other engine, and to this day can recall to mind the pained expression depicted on the features of Acheron's driver, as he hung over the side of his cab. Mr. Jones was, as a rule, a very taciturn individual, and on this occasion thought a good deal more than he said, but I was afterwards told that ‘Jones on compound engines’ was worth hearing if only on account of its extreme pungency of expression.

No. 8 was repaired and stronger pistons were put in, but the same thing happened again. For some reason it appeared that unequal expansion of the piston rods caused excessive strains in the piston heads. Afterwards the cylinders were taken out and No. 8 lay for some years on a pit in the erecting shop with the North Star and Lord of the Isles as neighbours. The new standard gauge No. 8 Gooch of 1894 contained nothing of the original engine except, perhaps, the 7 ft. wheels.

With much of the decorative brasswork removed, the cab widened to cover the rear driving wheel splasher, and renumbered 4172, Gooch has lost much of her original elegance.

Thank you, Mr Ahrons, for that most entertaining narration of a locomotive horror story. The No 8 Gooch that was nominally a rebuild of the compound No 8 was one of just four 4-4-0s that were built in 1894 known as the Armstrong class. The others were No 7 Armstrong, No 14 Charles Saunders and No 16 Brunel. They vie with the more numerous Dean Single 4-2-2s for the title of the most beautiful locomotives ever built. Unfortunately their careers aesthetically were marked by decline and fall after they lost the original boilers with large brass domes in favour of standard tapered versions. Fortunately you can visit the Museum and Archive at Didcot Railway Centre and see the magnificent 7¼ in gauge model of No 8 Gooch in full original glory, bequeathed to us by Sir Hugh Ford.

The model of No 8 Gooch in the museum at Didcot is a reminder of how elegant the locomotive was when originally built.

Talking about locomotive rebuilds, the additions hung on Dean Goods 0-6-0s for wartime service overseas did nothing to improve their appearance. No 2466, converted to War Department No 179, had pannier tanks, condensing apparatus, Westinghouse air pump and “Whites” water lifter in this photo issued on 25 March 1942.

The Dean Goods hung with all the military modifications in preparation for overseas war service in 1942.

Thanks Frank! A further spine chilling set of engineering faux pas to chill us to the bone. Again, it has to be said that we only know that some of the things tried here didn't work was because somebody tried them. It is easy to sit back in our I.T. driven world where an aircraft can take flight in virtual space before the first piece of metal has been cut and say “They should have known”. But look at it like this, even something as relatively recent as Concorde was entirely designed and developed on paper. It took tens of thousands of drawings that all had to work in concert with each other to make that beautiful white bird a reality. Imagine if you spent a week designing a part, only to find that the person on the next drawing table over had designed the bit that went next to it and because of what he did, your bit won't fit. We use the phrase ‘back to the drawing board’ for a very good reason...

*Bonus points for remembering the 80's horror film this quote comes from!


Oh The Horror! …

In recognition of the family fun Halloween events we have coming up, I thought I had better put in a word on some real horror stories. Nightmares made of metal and rolling on the iron road to the nether world. Basically, it's an excuse to take a look at some really awful engines!

The first horrible haulers were made flesh at the very dawn of the G.W.R. Isambard Kingdom Brunel is an engineering colossus. The mighty structures he built are still wonders of the Industrial Age. A national treasure. However, he has a dark secret. Us G.W.R. fans only talk about this in hushed tones. Don't spread this about ok, but he was RUBBISH at designing steam locomotives. Well, he didn't design the locomotives exactly. He did however put such strict specifications on the designs that the poor engineers trying to meet them were totally hamstrung.

Model of Thunderer displayed in the Museum & Archive at Didcot Railway Centre

There were a whole host of horror stories that were caused by this but the two worst offenders have to be Thunderer and Hurricane. This diabolical pair were built by R & W Hawthorn & Co. in 1838 and if there was a bad idea to be put on a steam locomotive, then this pair had it. At the front was the ‘locomotive’ bit. Thunderer had an 0-4-0 design with 6’ wheels and in order to reduce the piston speeds reached at the theoretical high velocity it was supposed to attain, there was a set of gearing in between. Hurricane's 2-2-2 ‘engine’ was even more extreme. It carried the largest set of driving wheels ever fitted to a steam locomotive at a whopping 10’ or just over 3m in diameter!* Proceeding to the stern of these mutant beasts we next come across the boiler. Yep, the boiler wasn't on the engine. Instead of the normal way of doing things, it was on its own 6 wheeled frame. Of course it was. Then came the tender. Because if you can have two vehicles in a loco, then why not have three?

It's easy to spot the seemingly schoolboy errors here. Firstly, in 1838 there was a lack of an effective technology to have flexible steam tight pipes between the boiler ‘vehicle’ and the ‘engine’. Not a great start. Then we have to talk about the fact that the driver is at one end of this cavalcade and the fireman is at the other. Communication? Good luck, with that... The biggest issue though is one of adhesive weight. The fact that the boiler wasn't over the driving wheels meant that while the engines could have all the power in the world (they didn't) it had no way of putting that power down to the track. Some reports say that they were “Barely capable of pulling themselves along the track”. While other wits claimed that they were “More a procession than a locomotive!” Needless to say they didn't last long. Both were withdrawn in 1839 after doing about 10,000 miles each.

There is a scorpion like sting in the tail of this tale. It is claimed - with very little in the way of corroborating evidence - that Hurricane did the amazing thing in its short career and broke the 100mph barrier while on test. The Locomotive magazine recorded in 1901 that:7 “It is stated this monstrosity ran the 22½ miles to Taplow in 16 minutes, on which occasion it was driven by Richard Wilkinson of the S&DR, and was fired by John Thompson; but it is safe to conclude that this fabulous speed was never attained as a start-to-stop speed of 84 miles per hour would mean an intermediate rate of somewhere up in the 90s at least.”

However unlikely, if it DID do ‘the ton’, it predates No. 3440 City of Truro's claim by 65 years and that made by No. 4472 / 60103 Flying Scotsman by nearly a century...

Model of Hurricane displayed in the Museum & Archive at Didcot Railway Centre

Hurricane was involved in a tragic accident soon after it was delivered, which resulted in the death of W D Field, aged 19, who was a pupil of Dr Dionysius Lardner. Dr Lardner, described by O S Nock as a ‘supreme purveyor of scientific bunkum’, had been commissioned to carry out trials for a report on the newly-opened Great Western Railway. The line had opened between Paddington and Maidenhead in June that year and was not performing well. Shareholders began to express concern in a hostile manner, and Brunel suggested that the directors seek the opinion of other engineers on the GWR system, such as the broad gauge, the type of track and the design of rolling stock. One of the engineers was Nicholas Wood, who employed Dr Dionysius Lardner to assist him.

On Thursday 25 October 1838, according to a report in The Times newspaper, Lardner's assistants were measuring the deflection of rails near Wormwood Scrubs as the 5pm train from Paddington passed over them. At the same time an ‘experimental’ train hauled by Hurricane, consisting of seven trucks with two tons of iron in each, approached from Paddington running on the ‘wrong’ line. Field was adjusting the trackside equipment and was inexplicably so close to Hurricane's train that it ran over him and he died of his injuries almost immediately.

This drawing of No 9 as built was published in The Locomotive magazine, January 1940 edition

We will now spin the clock forward to the reign of William Dean at Swindon. This next engine is semi-mythical. There is no known photograph of it in its original form. The drawings - if they do still exist - are buried in the vast, unsorted pile at York Railway Museum. It's a bit like the steam equivalent of the Sasquatch.** Unlike Bigfoot though, there is enough evidence, both anecdotal and documentary, to prove that it did, briefly, exist. This is the legend of that scourge of the Western Rails - Number Nine.

The idea was to produce a large and powerful tank locomotive (for the late 1870s / early 1880s at least!) that would be able to pull fast passenger trains. The design was quite radical. The engine was a single wheeler - having just one, large driving wheel. But, it also had 4 carrying wheels either side of this. There has been some past conjecture as to just how these were attached to the frames but the latest thinking is that it was via two bogies. This sort of ‘articulated’ 4-2-4 wheel arrangement was weird enough but it is also reported that the water tanks went the whole way down the side of the engine, past the smokebox at the front. It also had Joy valve gear on the outside of the engine which drove the valves of the inside cylinders via rocking levers. Modern reconstructions make this behemoth look really imposing. Brilliant.

The fabulous Bristol & Exeter Railway broad gauge 4-2-4 tank engines could have influenced William Dean's choice of wheel arrangement for his short-lived express tank engine

If it wasn't for just one, tiny, tiny detail...

It couldn't stay on the track.

At all.


David Joy - the inventor of the valve gear - went to see this amazing beast at Swindon. He says in his diaries that it “tumbled over the turntable going out of the shed and stayed there [and was] covered with a tarpaulin.” No matter what the engineers did, the design was simply flawed. It never made it out of the gates of Swindon Works. The surging of the water in those enormous tanks coupled with inadequate support from the bogies is one possibility proposed. Whatever the truth is, the whole affair was covered up by the G.W.R. No. 9, along possibly with the partially started components of a second 4-2-4 were rebuilt. No. 9 became a particularly handsome 2-2-2 tender engine of the Queen Class and as such served until the early 20th century. Photographs of this DO exist and it looks so innocent. A beautiful machine with a dark past...

No 9 after the first conversion to a 2-2-2, with the Joy valve gear eccentrics still outside the driving wheels

The last of our horror stories for today is really one akin to the famous Victor Hugo novel The Hunchback of Notre Dame. It wasn't a bad locomotive per se - it was just really ugly! Ugly is not a word often used to describe the products of Swindon but there have been a few. This particular offender is actually a series of 10 different machines built in 1899. They are known today by their nickname, the Krugers. The name comes from Paul Kruger - one of the leaders of the Afrikaner side of the then current Boer War. To the British public of the day, he held a similar position as Kaiser Wilhelm did in WWI or Adolf Hitler did in WWII. Any chance to ridicule and belittle them was taken.

No 9 approaching Acton, after the second conversion as a 2-2-2 and given the name Victoria

They were highly experimental machines. The first was No. 2601 which was a 4-6-0 that had outside frames. This means that the bearings on which the engine rested were mounted between the wheel and the cranks that the coupling rods attached to. Nothing particularly horrible there but where it really starts looking weird is with the boiler. They were very high mounted with a straight sided firebox. This alone would have caused it to look top heavy but the addition of a sand dome*** on top of that just made it look both top heavy and lumpy! She also had an unusual volute type spring arrangement to her suspension.

The second type was broadly similar but built to the 2-6-0 or Mogul design. These were numbered Nos. 2602 - 2610. While they were fairly successful machines, they were nothing more than stepping stones on the way to the Churchward design philosophy that was to come in the early 20th century. No. 2601 was withdrawn in 1904 and the last of the Moguls went by January of 1907. Many of their components were the same as used in the then new Aberdare Class 2-6-0 freight locomotive so it is highly likely that a recycling of parts happened here too.

Kruger 2-6-0 No 2602

It is easy for us to sit back in our ivory towers and laugh and criticise these bold experiments in steam traction as having being foolish or ridiculous. We do have the benefit of hindsight however. As is often said, hindsight is 20/20. The great engineers of the past did not have access to computer modelling or indeed the experience of those that came before them. That's what being a pioneer is all about at the end of the day. They had to work empirically. Learning by doing and making mistakes. I will leave you with two bits of wisdom on this subject that will hopefully put this blog in perspective.

Firstly, as the old saying goes: ‘Those that never made a mistake never made anything.’ Secondly there is the quote from Sir Issac Newton. The one that is so good, they wrapped it around the edge of the standard £2 coin. It is that we are all ‘Standing on the shoulders of giants’. Where would the great George Jackson Churchward have been without his pioneering forebears?

*There was another broad gauge engine with 10’ diameter driving wheels built for the G.W.R. This was the equally hamstrung but far more conventional 2-2-2 Ajax built by Mather, Dixon & Co. also in 1838.

**Your blogger's nickname on shed. I can't think why...

***The sensible idea being that the heat of the boiler keeps the sand dry. It's very common outside the U.K. but as we have quite the restriction on how large we can build our engines, it never caught on here.


Why is That Engine, That Shape?

It's been a while since I did a back to basics type blog where we explain some of the fundamentals of steam locomotive engineering. So, with that in mind, let's take a look at why steam engines are the shape they are. Steam locomotives come in all sorts of different sizes and shapes. The size and shape depends upon a whole heap of different constraints. The smallest engine in the collection is No. 5 Shannon at just 15 tons. The largest is No. 6023 King Edward II at a little under 140 tons. Why the huge disparity?


The smallest, Wantage Tramway No 5 Shannon and 2, The largest, No 6023 King Edward II

The first thing to say is that we need to look at the railway that the locomotive is to run on. Taking the two examples, the King has ten wheels on the locomotive and 6 wheels on the tender to spread that 140 tons on the track. That means that each of the driving wheel pairs is pushing down with 22½ tons. On Shannon, there are only 2 wheel sets but each of those is pushing down with just 7½ tons. The King is therefore restricted to lines where the bridges, track and other structures are designed to take that kind of weight. Shannon is lighter than some of coaches we have in the collection(!) so, she could traverse even the lightest of railway structures.*

Six driving wheels, No 5900 Hinderton Hall

You can change the downward force that a locomotive applies to the track. If you look at the Halls, the Saint and the 28XX class, you will see that they are about the same size. However, No. 3822 is able to travel on much lighter laid track than the other machines. Although the engines all have ten wheels, No. 3822 spreads its weight over a wider area as it is a 2-8-0. The 4-6-0s don't do this as well. The more driving wheels you have, the better the spread of weight is. There are the same number of axles but each axle carries more weight, thus it being lighter on the track overall. The distance between the first and last driving wheel is called the coupled wheelbase. The shorter this distance is, the tighter the curve it can go around. Also a good clue to the type of railway an engine is supposed to run on.

Eight driving wheels, No 3822

Let's take a look at those wheels too. They come in a vast range of sizes. The smallest in our 1932 running shed belong, again, to Shannon at just 2’ 11” diameter whereas the largest are on the Castles and Saint** at 6’ 8½” diameter. The Halls have wheels that are 6’ in diameter and the 28XX class and 72XX class have 4’ 7½” diameter wheels. So, what's that about? Well, think of wheel size as extra gearing. If you have a small wheel, mark the point where it touched the rail on it, roll it through one revolution and measure the distance it has travelled, you will find it is quite small. This is because the circumference is small. Do the same thing with a larger wheel and it goes further in one revolution. The larger the wheel the further it goes in one revolution. If you can get a large wheel moving, then you can go quite quick!

Mixed traffic at work, No 4958 Priory Hall with a down fast goods climbing the 1 in 36 portion of Dainton Bank, Devon. Photograph by John Ashman FRPS

Great big driving wheels, No 4079 Pendennis Castle and No 2999 Lady of Legend share 6ft 8½in drivers

The trick however, is getting it moving. Larger wheels are bigger, heavier and take more effort to turn. A smaller wheel is far easier to get moving. The best illustration of this is the difference between No. 3822 and the Halls. From the running plate up, they are very similar.*** Broadly speaking, they have the same boiler, same boiler pressure, the same running gear, cylinders and so on. No. 3822 has a smaller 3,500 gallon tender but apart from that, much the same. The wheels are where the big change is. The Halls' 6’ wheels mean they can go reasonably quick, 80 - 90 mph, whereas No. 3822 will top out much sooner in the low to mid 40 mph range. But look at the power they can apply to a train. The Halls are at 27,285lbs tractive effort but No. 3822 can pull with 35,380lbs.

So, by the size of the wheels we can tell what a locomotive does for a living. No. 3822 has small wheels so it's not going fast but it can move a large load doing so. This is a heavy freight machine. The Halls with a 6’ wheel however can pull a variety of services including freight and passenger but with not as much power as No. 3822. These are therefore mixed traffic engines. The Castles have great big driving wheels so are designed for going fast. Express passenger trains are the job for them.

We can also tell a lot by the size of a locomotive. The bigger the boiler, the larger the reserve of steam. Therefore, the engine with a bigger boiler will be needing to put out large amounts of power for long periods of time. A good comparison here is the Large Prairies like No. 4144 with the Halls. Both did similar work - mixed traffic - but there is a marked difference in their size. Halls are required to go longer distances than No. 4144. They bigger boilers and therefore a bigger reserve of steam. No. 4144 can do very similar work with a smaller range and a smaller reserve of steam but is able to travel on a greater range of different routes as it isn't as heavy as the Halls.

Large Prairie, No 4144, good for shorter distance work

Small wheels and small boiler means you probably have smaller cylinders so less power but more route availability. The smaller branch lines will be open to these engines. A Small Prairie, 14XX class or a Pannier Tank would be in this category. The Panniers also bridge the gap between lighter trains / branch line work and shunting. This is where the smallest engines like Shannon, Trojan, Bonnie Prince Charlie and No. 1338 worked.

The number of cylinders is a good indicator of purpose too. Most steam engines in the U.K. were two cylinder machines. This means that you don't get easily into a situation where the pistons are stuck at the ends of their travel, making starting difficult. You CAN do this with a two cylinder engine - it's just far more unlikely! It also reduces complexity. The disadvantage comes when you are going really fast.

Two large pistons thrusting backwards and forwards makes the front of the engine swing left to right. This is known as hunting but at slow speeds it feels like the engine is walking! No. 5322 is famous for its ability to walk off with a load. At high speed this is not good for crews and locomotives alike. To make power delivery smoother, we need to make the cylinders smaller. This makes the masses moving backwards and forward less. To keep the same high power outputs needed to maintain high speed, we add more cylinders. That is why the Castles and Kings have 4 cylinders. It's the same principle applied to a V8 internal combustion engine.

The final big distinction is where you put your coal and water. Larger engines need more coal and water - that much is obvious. However, there comes a limit where putting coal and water on the engine itself, as in tank engine designs, just isn't any longer practical. The limit for U.K. is somewhere about the size and shape of No. 7202. They had to make this one longer to get all the coal and water on board.**** This 90 ton monster is, with two sisters, the largest tank loco in preservation in Britain. If the boiler gets bigger than theirs, the coal and water are evicted - to live in a separate vehicle known as a tender.

The magnificent 72XX class 2-8-2T at work. No 7252 is hauling exchange traffic for the Southern Railway, crossing from relief to main line east of Reading General station prior to running on Southern metals. Photographed in October 1946 by John Ashman FRPS

There are advantages to being a tank engine however. Most of them are quite happy going either backwards or forwards. A tender locomotive has a limit to the speed it can attain in reverse. 45mph is often quoted. This is because if the tender is between the loco and the train, its running qualities are really good but pushed out ahead of the engine, not so much...

Well, that will help you understand the anatomy of steam engines everywhere. These points are fairly universal and will enable you to tell fairly accurately what a locomotive was born to do. Much like the Darwinian theory of biological evolution, steam locomotives followed a similar path. They started as tiny single cylinder machines and evolved into a bewildering variety of different types. They filled every conceivable niche. Sadly, like many types of fauna today, they are only now commonly found in steam zoos. The main railway environment now being the province of far more generalist diesels, electrics and multiple units. The rats and seagulls are doing well but the pandas? Not so much. The ‘biodiversity’ of the modern railway is sadly lacking in comparison to the steam era. I think there's an ecological message in there somewhere too...

*The G.W.R. used a coloured disc, painted on the cab side sheets that tells you what lines the engine can go down. Check out the July 2020 Blog ‘Decoding an Engine’ for more details.

**The ones on our broad gauge replicas Firefly and Iron Duke are bigger but I'll keep it to standard gauge today.

***By coincidence both Halls have 3 row superheaters, so are not quite as similar as the standard Hall would be but the comparison still stands.

****That sounds like a blog challenge. Hmmmm...


Trials and Tribulations - Part 2

There are two more important locomotive trials that the G.W.R. were involved in during the steam age. The first trial for today is the one that took place after the G.W.R. / L.N.E.R. trials that our very own Pendennis Castle was involved in.

This was with the London, Midland & Scottish Railway (L.M.S.). Having observed the goings on between the L.N.E.R. and the G.W.R., they wanted their chance to see what the Castle could do up close. The L.M.S. comprised several railways, grouped in the early 1920s, one of which was the L.N.W.R. The very railway that had been host to the Star in the 1900s. The L.N.W.R. seemed to take on board very little from the experience but the then chief engineer of the L.M.S., Sir Henry Fowler, definitely was interested and ready to learn.

Shame the L.M.S. top brass wasn't...

They were still hidebound by the policies of one of their constituents, the Midland railway. They favoured a small engine policy - the exact same policy that had so troubled the L.N.W.R. The operating department of the railway was against such powerful 4 cylinder beasts and eschewed pacific type machines. Somehow, not now easily fathomed, the authorities were persuaded to allow a G.W.R. Castle to visit their line and it was the beginning of the end for small loco policies!

No 5000 Launceston Castle photographed in workshop grey when brand new in September 1926, with the larger 4,000 gallon tender.

The engine in question was No. 5000 Launceston Castle. She worked between London Euston and Crewe initially and then to Carlisle, during September and October 1926. The LMS's new service - The Royal Scot - was not going well in the hands of their existing fleet. No. 5000 just romped away with it, keeping to time and doing so in the customary efficient manner. This time, a dynamometer coach* was used and again showed that No. 5000 was producing all the power with less fuel consumed. The tests were not done in the same way as the previous ones were handled. For a start, there were sections of the road that required the engine to slow down as to not strike platforms with her wide cylinders. There was also a number of signal checks and so on that delayed the trains still further. There were also occasions where the gravity sanding that had so helped Pendennis Castle was a problem as on exposed sections of the line, the sand was being blown away by the wind before it could hit the rail.

Launceston Castle arriving at Euston during the 1926 exchange. This photograph was printed in The Railway Magazine.

Despite all these challenges, Launceston Castle, driven and fired by No. 4079's Young and Pierce, still covered itself in glory. The trains were dealt with efficiently. The loads, which steadily got heavier as the five week test progressed, were taken with ease and the ability for the Castle to recover time when made late was remarkable. This caused the L.M.S. to sit up and take notes. Their locomotive designs started to improve and were really solidified by the headhunting of William Stanier - the second in command at Swindon - to become their new Chief Mechanical Engineer. The L.M.S. did send a locomotive to the G.W.R. It was a Midland Compound 4-4-0. Let's just say that No. 1047 was put on work befitting its lower tractive effort and even so, it didn't do well...

Midland Compound No 1047 departing from Paddington with the 1.15pm train to Bristol Temple Meads. This photograph was printed in The Railway Magazine.

It would be twenty years, a Second World War and the nationalisation of the railways before another set of locomotive exchanges took place. Given the fact that the newly formed British Railways (B.R.) had amalgamated the design philosophies of 4 separate companies, all with their own needs and histories they wondered if it would be possible to set down a standard set of locomotives that could work over the entirety of the U.K. network. A very Churchward way of thinking perhaps?

City of Bradford at Paddington station during the 1948 exchange, photograph by Ben Brooksbank.

The idea came about therefore to exchange the locomotives of all the previous companies on as many different lines as possible to see just what this standard design might look like and what features have the best advantage to the locomotives. The first decision was to not help the G.W.R. machines. Unsurprisingly, B.R. decided to run all the tests with the locomotives burning the more common Yorkshire hard coal. The G.W.R. had always used a softer, more potent Welsh steam coal. Given that the other three had been making huge leaps forward in their locomotive technology, this wasn't going to be easy. There were different groups of engines tested against each other. In the passenger group was the G.W.R. King class 4-6-0s, the L.N.E.R. A4 class 4-6-2s, the S.R. Merchant Navy class 4-6-2s and the L.M.S. fielded both their Rebuilt Royal Scot class 4-6-0s and their Duchess class 4-6-2s. The mixed traffic types were represented by the S.R. West Country 4-6-2s, the L.M.S. Black Five 4-6-0s, the L.N.E.R. B1 class 4-6-0s and the G.W.R. Hall class 4-6-0s. Finally, the heavy freight selection was from the L.M.S. 8F class 2-8-0s, the L.N.E.R. O1 class 2-8-0s , the War Department 2-8-0 and 2-10-0 designs and the G.W.R. 28XX class 2-8-0s.

King Henry VI departing from King's Cross during the 1948 exchange, photograph by Ben Brooksbank.

The problem with the inclusion of the G.W.R. designs was that due to the extra space around the track caused by its broad gauge origins, their designs were very wide and there were several routes over which they could not travel. No. 6018 King Henry VI did perform on the L.N.E.R. but sadly it was unable to take part in many of the other tests as it just didn't fit. This was the same for the Halls and the 28XXs and this therefore made their inclusion in the trials somewhat difficult. The performance of the King on the L.N.E.R. was surprisingly not as good as most people expected. It seemed that her crew struggled with the hard coal where No. 4079 had excelled. There were other disappointments for other regions too.

King Henry VI on the Eastern Region during the 1948 exchange.

The A4s recorded no less than 3 separate failures. The Black Fives seemed to struggle on their home turf and all in all, the results showed what was probably blindingly obvious to all and sundry. For the lines the engines were built for, they were generally ideally suited! Those Duchesses designed by that nice Mr Stanier did rather well though...

Returning to Paddington with the 8.30am from Plymouth, Mallard failed at Savernake with a hot bearing. This photograph shows the locomotive at Savernake.

The majority of these trials are amazing examples of just how advanced Swindon technology had become under the leadership of Churchward. It did take the other railways a long time to catch up, but catch up they did as 1948 clearly showed. This often leads to the old chestnut that critics pull out about the so called ‘stagnation’ of locomotive design on the Western. It's true that development had slowed a great deal by WWII and beyond. What we as steam fans tend to forget is that this was a commercial organisation. They had designs that worked for them, were efficient using the fuel available to them and they did everything that was asked of them. They weren't trying to sell those designs to anyone else either. The engines were tools for doing a job on their railway.

By the end of the steam era, the double chimneys and higher degrees of superheating showed that there was still room for improvements in the designs too. I wonder how the running qualities of the Kings would have changed had they not been withdrawn before the programme to fit them with roller bearings had been applied. Looking at the work of such steam luminaries as André Chapelon and Livio Dante Porta, efforts such as streamlining the steam passages and so on could have helped still further. Sadly, the burning of coal became the burning of diesel and the Swindon steam age came to an end. Given all that, why on Earth would a company with shareholders or a nationalised industry, both with an economic bottom line, pour money into massive locomotive development? B.R. built Halls and Castles into the early 1950s after all and were only stopped from building another batch of 28XXs when the designs for the new fangled 9F 2-10-0s were put forward instead. They can't have been all that bad...

Following Mallard's failure, the 8.30am from Plymouth was taken to Paddington by the Reading station pilot, No 4920 Dumbleton Hall.

*A coach filled with recording equipment to measure all aspects of a locomotive's performance.


Trials and Tribulations - Part 1

Last time we caught up* with Pendennis Castle's trials on the L.N.E.R. Whilst it wasn't a whitewash by any means, it did show the superiority of G.W.R. locomotive designs of the time. This however wasn't the only trial of its kind that G.W.R. designed engines participated in. There are a few more for us look at. Let's go back in time to see the first.

Cue swirly visual effect and possibly a DeLorean at 88 mph...**


This page of photographs of Polar Star at Euston during the 1910 trials was published in Great Western Railway Magazine, November 1910 edition And L.N.W.R. No 1471 Worcestershire departing from Paddington during the 1910 trials

The first big locomotive exchange for the G.W.R. was in 1910. The London & North Western Railway (L.N.W.R.) had seen the end of the long reign of their Chief Mechanical Engineer, Francis Webb, from 1871 to 1903. He had a lot of radical ideas but his engines were never the biggest or most powerful. This left the L.N.W.R. with the problem that many of their trains were being double headed (two engines on the front). This has an obvious cost implication. His immediate successor was the short to reign (1903-1909) George Whale. He took this problem by the horns and developed the Precursor Class 4-4-0s introduced in 1904 and the Experiment Class 4-6-0s in 1905. These revolutionised express workings on the L.N.W.R.

George Whale's new order at the L.N.W.R., No 1737 Viscount of the Precursor class 4-4-0s introduced in 1904 photographed at Euston. After Francis Webb's reign these were the first new non-compound express passenger design for the L.N.W.R. for 30 years

Whale's successor was C. J. Bowden-Cooke and seeing the improvements made by the last chap, he decided to see where his Crewe Works products stood when compared to those of its neighbours. The first trial comparison was with the Great Northern Railway's large boiler Ivatt Atlantic 4-4-2s versus the Precursor class. Not much of the timings from these trials survives but the efficiency results show that the engines were fairly evenly matched. The Caledonian Railway allowed a comparison between the Experiments and their Cardean class 4-6-0s designed by John McIntosh. This revealed a remarkably similar result to the previous tests.

Earle Marsh's I3 4-4-2T No 21 hauling the Sunny South Express from Willesden Junction. The train is composed of L.N.W.R. carriages

The most remarkable trial was that of the interchange between the L.N.W.R. and the London, Brighton & South Coast Railway. This involved no less a machine than an express passenger tank locomotive! Their chief engineer, Earle Marsh, had designed a series of these 4-4-2T engines. They went in series through the I1 (not very good) to the I2 (also not very good) and just when you think it's going to be three tries and three failures, he brought out the I3. The previous two types were poor steamers and had 5’ 6” driving wheels. The I3 had two key improvements. Firstly, it had enormous 6’ 9” diameter driving wheels*** and secondly, it was superheated. These engines would pull the ‘Sunny South Express’ between Willesden Junction and Brighton. The load was usually about 235 tons and the I3 only had a 2,100 gallon capacity in its water tanks. Running at about 52 mph was the order of the day and there were no water troughs on that route. When you realise that they went non-stop to Brighton without refilling their tanks, their abilities start to become quite remarkable. They also did the whole return trip on just one bunker full of coal. That was just 3¼ tons of fuel. Worked out in efficiency terms, they did the round trip of 264 miles burning just 27lbs of coal per mile. I3 No. 23's performance on the L.N.W.R. was also startling. Although the power and timekeeping was within the grasp of the L.N.W.R. machines, the efficiency with which the work was carried out by No. 23 was certainly not.

Earle Marsh's I3 4-4-2T No 22 hauling the Southern Belle between London and Brighton. A beautiful train of matched Pullman cars, it was introduced in November 1908, the same year that the locomotive was built

By 1910, The L.N.W.R. had organised the final and for us, the most significant of the trials. The G.W.R. agreed a series of exchanges with their latest Star class locomotives. George Jackson Churchward was at the height of his powers when he designed the Stars. They deserve a blog of their own really but the basics are that at the end of the development of the Saint class 2 cylinder 4-6-0s. The French locomotives that Churchward had purchased to compare with the Saints had two features he was interested in. They were compounds and they had 4 cylinders. He had already discounted compounding as not viable for the G.W.R. There was therefore, clearly one more trick to perform with the 4-6-0 design - adding 2 more cylinders! The design became a fantastic success and was the basis for both the Castles and the Kings.

Given Swindon's dominance of the locomotive engineering world at the time, this was always going to be a tall order for the L.N.W.R. The G.W.R. locomotives were seemingly fast and powerful but their lack of participation in trials of any kind meant that the true nature of their prowess against other railways' products was only estimated. The announcement that the Stars were going up against the L.N.W.R. Experiment Class was therefore a bit of a surprise and keenly watched. No. 4005 Polar Star was sent to the L.N.W.R. to be compared to No. 1455 Herefordshire and No. 1471 Worcestershire was sent to the G.W.R. to be compared with now preserved No. 4003 Lode Star.

This photograph of Worcestershire at Paddington during the locomotive exchange was published in Great Western Railway Magazine, November 1910 edition. The loco looks positively antediluvian compared with the sleek GWR designs of the period

Worcestershire passing Subway Junction after departure from Paddington with the 11.50am, this time with the dynamometer car on the train

The comparison was never going to go well for the L.N.W.R. TheExperimentwas a 2 cylinder machine with a boiler pressure of 175psi and a tractive effort of 18,615lbs. This wasn’t bad by the standards of the day but still pale in comparison when compared with the Star. This design has 4 cylinders which were driven by a boiler pressure set to 225psi and a tractive effort of 25,090lbs. It showed from the start on the G.W.R. No matter what the crew of Worcestershire did, as the tasks set before them got progressively harder and closer to the everyday running of the Stars, they began to lag further and further behind. On one trip they were 33 minutes down on the booked time. Another hassle for them was that their water scoop wasn't performing in the wider G.W.R. troughs. This meant extra water stops which compounded the issues. Another trip saw them take a Paddington to Torbay express. Even on the easy running up to Reading, the engine was struggling and the 180 minute booked time to Exeter was extended to 193½ minutes. This wasn't the fault of the crew - far from it - they did remarkably well to do what they did but their steed was simply out of its league.


Polar Star hauling the 10.00am Euston to Glasgow Central on Bushey Troughs 20 August 1910, photograph from the LCGB Ken Nunn collection

Polar Star too was having a good time on the L.N.W.R. Given that the schedules and loads were designed to suit the Experiment class, the Star and her crew found the work easy. The amazing thing being that not only did it not have any issues doing the work, No. 4005 did it while generating far more power and with appreciably less fuel consumed. To put it in perspective, the cost of looking after a locomotive at the time for the L.N.W.R. was £975 per engine, per year. It might be assumed that the operational costs to the G.W.R. were far higher but they weren't. Per loco, per annum the figure was only slightly increased to £1,003. The result of this expenditure was a major advantage in power and efficiency and the opportunity to pull faster, heavier trains and the increase revenue that came with it. It is even more remarkable that despite this evidence being laid before them, the L.N.W.R. seemed to get nothing from the trial. It took a long time for their designs to change. A far cry from how the L.N.E.R. responded in 1925 to Pendennis Castle.

Well, that's enough for today. My cup of tea isn't going to drink itself and there are some biscuits with my name on them. Metaphorically at least. Tune in next time and we will see about going forward in time to just after Pendennis Castle's performance on the L.N.E.R. That's when the L.M.S. knocked on the door...

P.S. A bit of an oldie but a goodie situation here. The data for this blog comes from the book ‘The Locomotive Exchanges, 1870 - 1948’ by Cecil J. Allen, published by Ian Allan in about 1949 or 1950 as far as I can make out. It's a cracking read if you can get a copy...

*If, indeed, learning about events of around a century ago can be described as ‘catching up’!

**Simply the ONLY way to travel back in time! However, it nearly wasn't. The original idea the filmmakers had was to build the time machine out of a refrigerator...

***This is ½” LARGER than those of a G.W.R. Castle!


« Back To Going Loco Index

Didcot Railway Centre Newsletter

Stay up to date with events and what's going on at Didcot Railway Centre.
You may unsubscribe at any time. We do not share your data with 3rd parties.


Make A Donation

Click To Donate