Living Museum of the Great Western Railway

Home » BLOGS - Going Loco

Going Loco

BLOG - A closer look at our collection of historic locomotives

With a collection of locomotives dating from Victorian times to the 1960s, there's plenty to discover.


Latest Blogs



An Injection of Pure Genius

(2914 Saint Augustine - steam coming from the injector overflow)

Warning: There is some massively over-simplified physics here. Whilst I appreciate that some people can tell you chapter and verse, this is a blog for a general audience. Just the basics is enough...

Steam engines need water. Water in constant supply while in steam in fact. The big problem with getting water into a boiler is that the boiler is already at pressure. You therefore have to overcome that pressure in order to get the water in. So you are already fighting an uphill battle so to speak. Early 19th century locomotives did the obvious thing and connected a water pump to the engine somewhere. The big disadvantage here was that these water pumps were connected to the mechanism that made the wheels turn. You can only therefore put water into the boiler when the wheels are revolving. It wasn’t uncommon in that pioneering age for locos to be disconnected from the train until just before the departure time so they could run up and down in order to fill the boiler. There is a multitude of drawbacks here and it clearly wasn’t a satisfactory solution.

(Injector from 6023 King Edward II)

The solution came along as late as 1852. This is where a Frenchman with the easily remembered name of Baptiste Jules Henri Jacques Giffard. Let’s call him Henri shall we?! Born in Paris in 1825, he made many contributions to engineering including being the first to fly in a powered controlled flying machine. His Giffard Dirigible (airship) first flew in 1852 and was powered by a 3hp steam engine. Rather terrifyingly, the gas bag was filled with ultra flammable hydrogen. With a steam engine. That puts out all kinds of hot gasses, sparks and flames.


I think I’ll wait for the Wright Brothers to come along thanks.

The invention we remember him best for at Didcot however is his steam injector. This is an amazing piece of equipment that seems to work via magic as far as most people are concerned.* It seems to do the impossible in that it uses the steam to overcome the pressure of the same steam. The Giffard injector also does it with the absolute minimum of moving parts too. Sounds too good to be true. But let’s have a look.

It is described as a fluid dynamic pump. That means it is using the properties of the fluids and the energy bound up within them to provide the pumping action. No matter which way you cut it, it sounds complex, so let’s break an injector down to its four most basic components. There is the body of the injector, the big metal outside casting. Inside there are three cones arranged thusly: 


(Left: Drivers side injector on 3738, Right: Boiler Feed Injector Diagram)

From left to right you have the steam cone, the combining cone and the delivery cone. The first is the steam cone and as you might expect, this is hooked up to the steam supply from the boiler. Steam goes in from the wide end of the cone and exits from the narrow end of the cone. This causes the Venturi Effect** to come into play. If you restrict a pipe, it has the effect of taking the pressure energy in the working fluid (the steam) and turning it into velocity. To the layman, it speeds up. This is why your garden hose without a nozzle on the end seems to run out in a fairly lacklustre manner but with the nozzle on the end you can squirt it clear across the garden.

This jet of steam then goes into the combining cone. As it is now going quite fast, it reduces the air pressure around it and pulls the water along for the ride. It literally sucks it into the combining cone. Whilst travelling through the combining cone, the steam and water are, er, combined. The steam therefore turns back to water. Up until now it all seems quite reasonable but all we have is a fast moving jet of water. It shouldn’t be able to do more work and overcome the energy of the steam that caused it to happen in the first place right?

(Drawing of a Restarting injector from the Drawing Archive)

Well, no. There is a little bit more energy that is stolen from somewhere else and it’s this that makes the difference. You were probably taught at school that there about the three states of matter. A solid, a liquid, and a gas.*** You also know that if you heat up ice it becomes water and if you heat that water high enough it becomes steam. In order to make that change between one state and another, there is a little extra energy needed and this is known as latent heat energy. As the steam turns back to water, this energy is released.

This gives our steam water combination the last kick of energy. As it goes through the delivery cone, the reverse Venturi effect happens, it slows down. Physics tells us that energy cannot be created or destroyed and it has to go somewhere. The energy is in fact returned to the water as pressure energy. This is higher than the original boiler pressure and it can overcome it and force its way past the one-way valves into the boiler.

Utter genius.

Let this sink in. A man in the middle of the 19th Century thought this thing up. In his head. Henri Giffard truly was a genius (a most overused word today) and developments such as the self restarting injector that GWR locos use and the exhaust injector (that recycles some of the energy from the exhaust steam) only made it better as time went on. It has its limitations. They don’t work well with hot water for example. In fact, No. 4079 Pendennis Castle had to be fitted with special injectors to enable it to operate safely in the Australian desert. We also always run with two operational injectors at the start of the day in case one fails. Even so, they are nothing short of brilliant.

Oh, and one more thing - they are anything up to 98% energy efficient...

(Saint Catherine 2918 Pad c. 1908 - Steam coming from injector overflow)

*It was Arthur C. Clarke who said “Any sufficiently advanced technology is indistinguishable from magic."

** Named after Italian genius Giovanni Battista Venturi. Amongst other things he was a physicist, diplomat, man of letters and a historian.

*** There are actually more but who’s got time for plasma or a Bose - Einstein condensate? You people have got other things to do.


Return of the Champion - Chapter 2

Well, lots of jobs still to do on the mighty 4079 ‘Pendennis Castle’, but progress is being made. Let’s take a look in the history books before again opening the workshop doors and taking another of our exclusive peeks. So, what makes a Castle a Castle? Well, that all starts way back in the 1920s. Let’s hop in the time machine...

Star class 4014 Knight of the Bath at Old Oak Common, 25 June 1932 - photo Ken Nunn

Picture the scene. It’s 1923 and the craftsmen at Swindon are building the first batch of Castle Class Locomotives. These are essentially an enlarged Star Class loco and this is so true that the Star drawings have merely been drawn on with red pencil to show the changes! A larger cab with such luxuries as side window and a roof that has at least a hope of keeping the rain off the crew, slightly larger cylinders and a larger boiler.

The Star was an excellent machine in its own right; the four-cylinder express passenger design that was developed by the GWR was a winner from the start. It had a great deal of power for its size, was highly efficient and made a huge impression on the London & North Western Railway when an exchange trial saw one absolutely trounce them!

Star class 4005 Polar Star hauling the 10 am Euston to Glasgow Central on 20 August 1910 passing over Bushey troughs during the exchange trial with the LNWR - photo Ken Nunn

The Castle boiler was in fact, something of a compromise. Collett has intended to use the Standard 7 boiler as was fitted to the 47XX class 2-8-0 fast freight machines. This was vetoed by the civil engineering department, citing the wear and tear the extra weight would cause. So, forever pragmatic, Collett produced a slightly smaller, slightly lighter version of the boiler and with little thought to poetry, dubbed it the Standard 8. This was an absolute winner.

Postcard of Pendennis Castle when new in February 1924

Now let’s look at some of the work that's been undertaken to ensure 4079 returns to steam . . .

There were a number of jobs needed when we took on the job of bringing No. 4079 back to life, one of the biggest was the cylinder block. There were two main issues. Firstly, the valve and cylinder liners were all in a very poor condition. Replacement was really the only option. This is a big job on a Castle as it has 4 cylinders and therefore 8 valve liners. Tiresome but not in any way new ground.

Machining one of the new cylinder liners

The second job was a bit more of a concern. The cylinders on a Castle are made up of 4 sections. The two outer cylinders and valves, the inner cylinder and valve block that also has the front section of the smokebox saddle. The final section comprises of the rear of the smokebox saddle and the two exhaust pipes from the outer cylinders that lead up into the smokebox. On top of this is a 2 to 1 manifold that then concentrates the blast straight up the chimney.

The problem was that this flange where the manifold bolted was broken. The whole top half was snapped off. This was a case of good news – bad news: the cast iron welding company said they most certainly could repair it but would only guarantee the repair if they could take the part to their works. For those that don’t know, to weld cast iron you need to heat it up before doing it. Not easy if it’s attached to the frames of a locomotive...

This necessitated us taking it out. The bolts that hold it in are what is known as fitted. This means that the plain section of the shaft of the bolt is machined to a very tight tolerance to fit exactly into the hole it goes in. They do not give up their hold easily. We had to start by making a jig to get these bolts out. Which had to fit into all sorts of inaccessible places where Swindon saw fit to secrete a bolt(!). It took a day. It was hooked up to a crane and we tried to lift it but the fit was so good that it took a whole day of very careful jacking, tapping, cajoling and anything else you can think of to get the thing out.

Smokebox saddle awaiting boiler replacement. The temporary wooden cover to prevent debris falling into the valves while work is going on

It went away for repair and when it came back, in the greatest tradition of the Haynes manual, replacement is the reverse of removal so another solid day saw is slowly push it back in and a further day got all the bolts back in...

Recent efforts on the project have been really pleasing. Pete Gransden and Ali Matthews have been working solidly on the boiler to prep it for its examinations by the inspector. This really is just a whole heap of tidying up and the boiler is now watertight. Well done to them!

4079's boiler sits on the boiler wagon in Didcot's lifting shop - November 2020

The last of the serious work on the cylinder block - the grinding in of the pressure relief valves on the ends of the cylinders is now complete and they just now await the new cylinder drain cock pipes and some paint here and there to complete the front end to the top of the running plate.

Not long now...

All the best,

Drew and the 4079 Team.

Celebrating 4079’s return to traffic is amongst the projects included in the recently-launched Diamond Jubilee Fund – learn more and see how you can help.


The Iron Warhorse

5322 in France, 3 October 1918

As it’s early November, thoughts naturally turn to the season of remembrance. We are privileged to have a very special resident at Didcot who is quite literally a World War 1 veteran. This is none other than the sole surviving GWR Churchward Mogul No. 5322. The engine that went to war.

One of 5322's sister locomotives, 5319, on active duty in France

The Churchward Moguls started as an afterthought and became a legend. Eventually 342 were built making it the most numerous single class of GWR tender locomotive. The Mogul (the name of the engine’s 2-6-0 wheel arrangement) wasn’t on the list of standard locomotives that George Jackson Churchward was designing for the GWR at the beginning of the 20th Century. Types like the Star, Saint and 28XX 2-8-0s had all been planned for but it quickly became clear that there was a sort of ‘gap in the market’ type situation. This was for a smaller, lighter tender locomotive capable of being a go anywhere and do anything sort of machine.

This was actually a really good test for Churchward’s standardisation program. Could a whole new locomotive class to meet a whole new set of demands be built from the set of standard parts that Churchward and his engineers had developed? Well, the short answer is yes. The long answer is that, according to legend, only 11 extra drawings were required to do so. If this is true, you need to realise that some of those were general arrangements (the drawings that shows where all the bits go), so the actual number of new parts drawings would be lower.


5322 in ROD khaki livery

The 1911 prototype engine was basically a 2-6-0 tender version of the 3150 class large Prairie (2-6-2) tank engine. Although Harold Holcroft (the project’s chief draughtsman) claimed that they only noticed the massive similarity once the project was finished, it seems that Churchward was keen to point it out when selling the design to the GWR board of directors. The Moguls had a standard 200 psi No. 4 boiler, 5’ 8” driving wheels, 18 1/2” diameter cylinders and this gave them a tractive effort (pulling power) of 25,670lbs. The tender was almost always a Churchward 3,500 gallon version. As might be expected of the majority of standard GWR machines from Churchward’s stable, the Moguls became firm favourites and were maids of all work. Freight, passenger, the lot. This doesn’t however explain what our No. 5322 has to do with World War 1...

In 1917, the call went out to the railway companies of the U.K. to supply some more locomotives to serve behind the lines of the Western Front in order to keep the men and war materiel flowing. The railways were to supply a total of 160 machines and they were all to be of the 8 coupled variety. They should all have 8 small driving wheels which was pretty much standard for British 20th Century heavy freight machines. The problem as far as the GWR was concerned was that all of the 2-8-0 type locos they had and could build were desperately needed for coal traffic. Don’t forget that the Royal Navy was powered by coal at this point and the threat of the German U-Boat submarine fleet was potentially devastating to supplies for both the war and the U.K. itself. The GWR therefore decided to send the mighty Moguls instead. Its General Manager, Mr Frank Potter, explained that: “The Great Western type of 2-6-0 engines is in point of power and efficiency practically equal to other Companies 0-8-0 engines”. Quite the claim - but amazingly true. The government were eventually convinced to release the materials to build 20 new Moguls. Technically therefore these machines were built for the Railway Operating Division (R.O.D. - the railway section of the British Army at the time) and not the GWR although the company officially recorded them as being ‘on loan’. They were turned out at a rate of 5 per month and of the 20, 11 were conscripted into R.O.D. service. For the record, they are Nos. 5319 to 5330 but for some reason not No. 5327. Perhaps it had a problem on completion that could not be quickly fixed and therefore No. 5330 was substituted instead?

5322 in ROD black livery with period re-enactors at a special World War I commemorative event

The engines received the R.O.D. livery where the tender was emblazoned with the letters R.O.D. And then the loco number. The colour is not confirmed. The official colour for R.O.D. Engines was black but a number of references in the GWR paperwork denoted that these machines were painted in khaki. No. 5322 has appeared in both versions in preservation as a result! They also had a modification that our Mogul still carries to this day. If you take a look at where the cab sides meet the cab roof, you will see two hooks. A portion of the rear strengthening rib in this area has also been crudely chiseled away. This is on both sides and the mystery as to what these were for was solved when a deactivated short magazine Lee Enfield .303 rifle - the standard weapon of the British Army at the time - was found to easily clip into place. One for the driver, one for the fireman. It’s the only steam engine that we have with integral gun racks...

The mystery of the extra hooks revealed with a decomissioned Lee Enfield rifle

These locos were sent to the Western Front from Portsmouth Harbour in two trips, one on the 20th August and the second on the 21 September 1917. They were immediately pressed into service in the North of France. They worked army supply trains generally from Calais to the railheads around Hazebrouck. They were worked hard but rose to the challenge, gaining an excellent reputation with the R.O.D. crews. It can’t have hurt to have these relatively small, efficient, powerful locos that were capable of a reasonable turn of speed when required. There is a fantastic recollection of Mr C. E. R. Sherrington (then an officer in the R.O.D.) who saw No. 5322 in France in 1918 here.

The locos served in France until 1919 and are recorded as being returned to Swindon in March and April of that year. Amazingly, given the intensity and horror of that conflict, they all made of back to Blighty but not unscathed. It is said that at least No. 5325 came back with war wounds, receiving bullet holes after coming under enemy fire. The engines were all back in service by the beginning of June and went on to have regular careers on the railway. The last of these old soldiers to succumb to the march of the diesels was our very own No. 5322 when she was withdrawn on 24 April 1964. She was sent to Barry Scrapyard in South Wales from where she was rescued but that’s a tale for another time.

We are obviously very proud to look after such an important piece of history as No. 5322. The human veterans of this conflict are now all gone and it is down to the objects, places and the documents to remind us of what happened. Our ‘old soldier’ stands as our memorial to WWI. It is sad to think that World War Two is also fast leaving living memory and soon the last of that generation’s voices will fall silent too. It is so important for us to keep telling the stories. To remind ourselves and subsequent generations of the huge sacrifices and heroism that are the foundations of the freedoms we enjoy today. This is our duty as the custodians of this small part of WWI history. A very minor burden when you consider what has been and continues to be sacrificed by the brave men and women of our armed forces and their families...

“They grow not old as we that are left grow old. Age shall not weary them, nor the years condemn. At the going down of the sun, and in the morning, we will remember them.”


P.S. While we are usually asking for donations for our projects, it is only appropriate that with today’s blog, we highlight the many organisations that look after our forces veterans and their families. If you can, please make a small donation to one of the many organisations taking care of them after they have done so much to take care of us.

The Royal British Legion

Help for Heroes

SSAFA - The Armed Forces Charity

5322 complete with wreath of poppies on Remembrance Day - 9  Nov 2008



Spooky Shunting Spectacular

It’s alive I tell you - ALIVE!

Sorry, went all horror movie on you there.

With it being the season for such things I though it would fun for you to hear about the spooky goings on at Didcot Railway Centre. As you might know, there’s a Phantom loose at the railway centre and it’s really weird. It doesn’t eat coal, it only sips at water occasionally, seems to burn a funny flavour of oil and it growls around the place making the occasional high patched scream.

What am I talking about?

Well, nothing supernatural obviously. Nope - I’m talking about our trusty workhorse No. 08 604 Phantom. While we are all about the steam locos, we do have a few resident diesels and more often than not, when something needs moving, we summon the mighty Phantom from her slumber. The British Railways Class 08 diesel Shunter that she is an example of is one of those almost unrecognised British success stories and has a history going way back to the 1930s.

The London, Midland and Scottish Railway (LMS) built the first of the 0-6-0 DE shunter’s that were to be the first step on the road to the Class 08. They were built between 1934 and 1936. Another batch of a very similar type was built during WW2. Given the huge success of the LMS design, BR engaged the company English Electric to begin to provide some more of their diesel engine and generator sets as well as the traction motors. The new locomotives to put them in were built between 1952 and 1962 at no less than 5 of BR’s works*.

The 350 horsepower engine is a 6KT. In English Electric’s system, this means it has 6 cylinders, is of Type K and is set up for use in rail Traction. This diesel engine turns not a gearbox like a car but an electrical generator and this supplies electrical power to motors connected via some gears to the wheels. In a seeming nod to the steam locos they replaced, their driving wheels are coupled together with rods, at least giving the outward look of what has come before. These locos aren’t fast. The two versions of the gearing had either a top speed of 15mph in the Class 08 or the slightly uprated Class 09 which hit the heady heights of 20mph...

What the 08s lacked in speed, they most definitely made up for with pulling power. To put it in to terms so that we can compare it with a steam locomotive, it has a tractive effort of 35,000lbs of force. The Class 8750 0-6-0 pannier tank locomotives** that they replaced on the Western had just 22,515lbs T.E. They are even stronger than a GWR Castle which is just under 32,000lbs. Just don’t expect it to pull the Cheltenham Flyer at 80mph...

They slowly took over their assigned shunting duties from steam. This is the moving around and rearranging of railway vehicles into different formations. They did this so well that eventually 996 Class 08s were built. Along with all the other variations on this design, the grand total is around 1,200 individual machines. The way the railways have changed over the years has meant that the need for shunting has declined greatly. The vast majority of both freight and passenger trains are now of what is known as fixed formation - they stay coupled together as a set and are only rarely disconnected. However, there are still occasions where vehicles need reorganising and amazingly the Class 08 is still doing the job in the 21st Century. Although the fleet is now dwindling, there are less than 100 still in front line service, they are still there, doing the job!

Our example was built by BR at Derby in 1959. She was originally numbered D3771 - the D prefix being used while steam locomotives were still around to denote a diesel. The renumbering scheme brought about by T.O.P.S. (the Total Operations Processing System) saw the class designated as Class 08 (the first two digits of her number) and then followed by her individual designation 604. She wound up at Tyseley under the care of a certain Mr Simon Grego (one of our volunteers) and one of the things he did was to be party to repainting her in to a pseudo GWR colour scheme and added the name Phantom.*** Later, she managed to shear one of her crank pins and the cost of the repair, lack of work and her age lead to her being withdrawn. Thankfully she was subsequently purchased for preservation. The then owner needed to have somewhere for the locomotive to live and Simon came to the rescue in suggesting Didcot as the ideal home.

Upon arrival at Didcot, the damage was repaired and a fair amount of work including rewiring in places was undertaken to reanimate the Phantom. She then kept doing what she has always done - shunting. So successful was she at her new preservation role that when the loan agreement ended, she was purchased and now has a permanent home with us at Didcot. She has had a number of liveries over the years including black, War Department black (with the wonderful number WD 40!), BR lined green and most recently, Phantom has received the famous BR rail blue livery, enabling the centre to tell another part of the story of the railways in the former GWR area.

Love them or not, you can’t help but admire the Class 08 350hp diesel shunter and it’s derivatives. There are no less than about 80 that are now classed as preserved making it the most numerous class in British preservation by some margin. They are a classic design and a quiet triumph of British engineering. Railway history, the U.K. preservation scene and indeed Didcot Railway Centre wouldn’t be the same without them. Even if us steam guys still can’t work out where the coal goes.

Nearly 90 years after those first few LMS examples rolled out the factory gates, these unsung heroes are still doing their unglamorous job. You might say that their presence still haunts the railway to this day...

*These were Horwich, Doncaster, Derby, Darlington and Crew.

**Like our residents Nos. 3650 and 3738.

*** She had a fellow classmate that was also put into a steam era livery, this was No. 08 601 which was given the similarly spooky name Spectre and turned out in an LMS style paint scheme.


Another Tender Behind

Yes, I went there with the title.


Anyway, we had a look at the tenders used up to the start of the Churchward era last time and we saw the birth of the standardisation era.* The amazing thing about this was that there were always more engines than there were tenders. Tenders were easier to overhaul and even had their own works - ‘J’ shop at Swindon - where they were maintained. When a loco went in for overhaul therefore, the tender was separated off and went for assessment and work there. When an engine was completed, the very next available tender of the correct type was then coupled up. It goes some way to explaining why, especially on tender locomotives that had long service lives like our own No. 5322 and No. 4079 Pendennis Castle, why photographs of them show quite a bit of variation in their partner vehicles.

The Churchward 3,500 gallon version that we began to describe last time became a mainstay of many classes and was widely produced. There were a whole heap of updates to it over the years. The changes to the side frames were mentioned last time but the tank wasn’t. The coal flare or fender (both terms are used) got progressively longer over time until it reached its final form as shown on all the other 3,500 gallon tenders in our collection. You can tell the difference in the age of the tanks. The earlier ones have flush riveting and the later ones (post 1917) have snap head rivets (the domed kind) which are easier, cheaper and faster to do. There was a war on... No. 7808 Cookham Manor has the earlier flush riveted tank and both Nos. 5322 and 3822 have the later snap head style rivets.

The next type of tender represented at Didcot is the 4,000 gallon. There are two original types of this tender with third type being replicated. There was a first try at building 4,000 gallon tenders at Swindon in the late Dean / Early Churchward period but the instigation of the water trough network meant that only 20 were constructed between 1900 and 1904. As the bigger 4-6-0s got, well, bigger, a renewed interest in larger tenders was shown by Collett in 1926. After this time, the 4,000 gallon tender eventually became the norm for engines such as the Castles, Kings, Halls, 47XXs and Stars.

The more numerous 4,000 gallon tender in our collection is the Collett version. This takes the previous GWR design and extends the tank sides upwards to increase the water capacity. The coal fender, despite still retaining some of the design cues of the previous vehicles, now goes all the way down both sides and around the back too. This meant that rather pleasingly, the cab sides of the big 4-6-0s lined up with the tender tops and gave a much more pleasing ‘flow’ to their overall design. They originally had two water fillers (one either side at the back) as the platform water cranes were not tall enough to reach right over to the centre position. Upgrades to the columns meant that a central filler was soon adopted as standard.

The other 4,000 gallon design present in the collection is the Hawksworth version. This is a very different beast to all previous tenders. Although in silhouette it could pass as a Collett version, 5 seconds of inspection shows it clearly isn’t! The sides are absolutely flat. Not because they have been flush riveted but because they are welded. Another wartime expedience. There is no rolled over top edge, no other shaping at all in fact! The first 30 were oddities as they were built to be 6” wider than any other GWR tender. This is because these tenders were used exclusively with the (then) new 10XX County Class 4-6-0s. They had a wider cabs and so these tenders were made wider to meet them. They became extinct when the last of the County Class engines they served was scrapped in the 1960s. Of course, the new No. 1014 will need one of these and it is under construction in Didcot’s Locomotive works.

There was however a standard width (8’ for those that need to know) version that was built and used by initially the late era Halls and Castles although, once they were in the tender pool, they were used by all the locos that were eligible for them. The one we have at Didcot is rather appropriately matched to our Modified Hall No. 6998 Burton Agnes Hall. As a Hawksworth locomotive that was actually built in 1949 (a British Rail Loco from beginning to end), this paints exactly the right picture! There is some possibility that this tender is the only fully original Hawksworth tender left as the vast majority (if not all? - corrections on this point welcome!) of the other preserved examples have had new tanks made and fitted.

That gets us to the end of the GWR tender story that we can tell at Didcot. There are a few preserved examples of the Collett 3,000 gallon version about but not in our collection. It wouldn’t be right however to end without mentioning some of the real oddball tenders that, while not preserved, crop up from time to time in period photographs.

The obvious first place to go is to mention the few tenders that were converted fas part of the Post WW2 oil firing (using oil burners instead of coal) experiment. This quickly resulted in the realisation of two things. 1)That oil firing and GWR copper fireboxes weren’t a great mix and 2)that a ‘financially iffy’ post WW2 Great Britain had nowhere near the kind of finances to power its locomotives on expensive oil...

As well as the oil firing experiment there was also a 4,000 gallon unit built with an aluminium alloy tank - presumably to mitigate corrosion but this possibly also fell into the ‘well expensive’ side of things and was not repeated.

There were two 8 wheel GWR tenders. The first was built to accompany the GWR’s only Pacific (4-6-2) locomotive No. 111 The Great Bear. Presumably, this elongated 3,500 gallon unit with two four wheeled bogies underneath was thought necessary on visual grounds alone. The then standard Churchward one would have looked a little diminutive alongside a whopping great loco!

The second 8 wheeler was as an experiment by Collett in 1931. This one looks for all the world like a really convincing photoshop job when it crops up. It used all the same components as the regular Collett 4,000 gallon version but just had more wheels. it wasn’t repeated but as it used all standard parts in the main, it lasted pretty much until the end of steam on the Western Region.

From the sublime to the ridiculous. The GWR also built a few short tenders for use on tender locomotives that operated in areas where there were a lot of short turntables. The 80 vehicles were phased out as larger turntables were installed and can best be described as looking cute.

Awwww, bless...

There were also self - weighing tenders that were filled with all sorts of measuring gear so that coal consumption of locomotives on test could be measured.

The final word on tenders probably has to go to the ex. R.O.D. 2-8-0 tenders that, after the WW1 war surplus engines they were built for were scrapped, entered into normal circulation for some classes. One example of this was the 2251 Class ‘Collett Goods’ engines - a diminutive 0-6-0 freight loco. This meant that there was a forty three ton loco pulling a forty SEVEN ton tender! Talk about the tail wagging the dog - that engine really did have a tender behind.

That’s it, I can’t do that joke any more.

I’m off...

*Disclaimer: Again - I KNOW that I am really skimming the surface here. Just pointing out the basics. There is a LOT more to it than this blog can or indeed should hold. If you need more, (and why not?!) there is an excellent ongoing study here which really gets into the nitty - gritty.


A Tender Behind!

I know.

The title.

Couldn’t resist it.

So, tenders. There is always a whole bunch of stuff to be said about the locomotives that pull them but the tenders seem to get a bit of a raw deal. These fascinating vehicles have their own unique and varied history and deserve a little of the limelight as well. So, in the first of a two-parter, pull on the overalls and pick up the inspection lamp and we will go and have a look!*

The vast majority of early locomotives were tender engines. Simply because it’s more convenient to carry your coal and water supply on a separate vehicle rather than on the main chassis of the locomotive itself. Indeed, Trevithick’s early 19th Century prototypes were tender engines. The basics of these early vehicles are wonderfully illustrated by Stephenson’s Rocket. A simple wooden 4-wheeled truck with a flat area that you can put the fuel on and a large barrel, turned on it’s side to put the water in. Simpler times...

Replica Fire Fly in action on the Broad Gauge demonstration line

Looking forward to the earliest (albeit replica) versions in the Didcot collection, we come to the broad gauge versions. You can see the development between that of Rocket to Firefly and Iron Duke. The size of the tank is much larger and is of a riveted construction. The engines were consuming a lot more water as they became more powerful. They also were burning more fuel, hence the general increase in size. The basic shape is one that went through the whole of GWR tender development - in fact many of the design cues on Firefly can be traced through to almost the last GWR tenders built. Features that made it through include the rolled over top edge to the tank sides - although this was radically altered over the years - and the six axle configuration.

Features to note here include the sandwich type construction of the frames. This is an early railway technique where to ensure the strength of the materials, two plates of metal are sandwiched either side of a fairly serious lump of timber. The tank is a fair bit narrower than the frames with a lip around the outside. Some times this lip had the leaf springs above them and sometimes the springs were in / on the frames below. Some even had a small hutch type structure on them in which the train’s guard sat. It must have been exciting to ride there! Unbelievably this isn’t the most scary thing about the originals. That is that the handbrake on these vehicles were quite often the ONLY brake on the train.

Oh, and the brake blocks were wooden.

And on one side of the tender only...

Iron Duke class ‘Lightning’ showing the ‘iron coffin’ on back of the tender where a travelling porter sat looking backwards to watch the train for any signs of a mishap

The rivetted construction of the tender water tanks can be clearly seen on the replica Iron Duke; pictured here during a tender repaint.

We have a bit of a gap in the preservation story here as the oldest preserved GWR standard gauge tender locomotive** is the National Collection’s wonderful Dean Goods No. 2516. The 2,500 gallon tender preserved with this loco is noteworthy on our journey. It is a product of the William Dean era at Swindon and tenders of this era were often constructed with coal rails. These are exactly what they sound like - a metal cage structure that is used to hold the coal in! Later on these were replaced with the now familiar GWR style ‘flare’ section of flat steel plate and the short example shown on this vehicle is typical of the era. Again, many of the features from the broad gauge follow through including the tender tank being significantly narrower than the frames but times were a changing.

Early stage of construction of 2999’s tender tank. Note the well for extra water capacity between the wheels

Our next example comes in the form of the recreation of our Saint class Locomotive No. 2999 Lady of Legend. The Churchward era brought a whole new idea to Swindon - Standardisation. As far as was possible Mr Churchward tried to make as many components as interchangeable. Tenders are an obvious place to look. They are a separate vehicle attached to the locomotive after all! Tender No. 1719 was originally built in 1906 but oddly it’s frames were later replaced with a more modern set and the tank was rebuilt back to its original version for the Saint Project so its a bit of a mix. Probably not much there that dates back to 1906 any more but technically it is registered as having been built in the right time period so, we’ll take it! For serious nerd points, it’s the angles of the plate work where the axle boxes and springs are mounted that give the game away. Narrow plate work = early, wider plate work = late.

2999 and tender part way through (re)construction

The tank has been recreated to look as much like the original as possible and is a huge credit to Peter Gransden and Ali Matthews who built it. It might look like there are no rivets holding it together but there are. HUNDREDS of them. These early tenders were all flush riveted. The process is to drill a countersunk hole on the front. Insert your red hot rivet from behind the plate and then flatten the protruding stem into the countersink. Then you have to grind it flat to blend in with the sides. Imagine the hours of work in this job alone...

2999’s water tank with baffles fitted. Ali Matthews, who built it, standing in the tank

The tank now is almost as wide as the frames leaving just a small lip around the edge. A rolled top is in evidence as is the flare - in its original shorter Dean style but scaled up to fit the new larger 3,500 gallon design. To increase the water capacity, there is a well which protrudes down through the frames. The filler is in the centre, at the rear. The other important feature here is the step structure at the front of the tender. In order to make these tenders truly interchangeable with locomotives of different heights, there were different designs of this shovelling plate to suit specific classes.

2999's tender with the coal sheets fitted

Like the Dean Goods, No. 2999’s tender has its own vacuum brake system and the handbrake for the engine is on the left or fireman’s side of the cab. The other large handle is to operate the water scoop - a subject of a previous blog for you to discover! There are two smaller handles, either side of the coal space. These are the water valves and start the water flowing from the tender to the injectors to be pushed from there into the boiler. There is also a small vacuum valve here that can be turned to allow air into the tender brake system.

The finished locomotive - complete with the all-important tender

Well, that’s enough for now - in part two we will look at the tender designs of Mr Collett and Mr Hawksworth as well as a few of the oddities produced down the years. Like much of the engineering on our locos - they can talk to you and tell you a story if you know how to read it. Your next lesson in how to speak tender will be next week...

*Disclaimer: I KNOW that I am really skimming the surface here. Just pointing out the basics. There is a LOT more to it than this blog can or indeed should hold. If you need more, there is an excellent ongoing study here which really gets into the nitty - gritty.

**Some parts of the Bluebell Railway’s ‘Dukedog’ No. 9017 possibly come from earlier GWR tender locos but as this is only the cab and the boiler it’s not a pure survival like the Dean Goods. The tender is a later version too...


Hydrostatic Hyperbole?

It’s time we talked about hydrostatic displacement lubrication. Please don’t run away - this is really quite interesting! In fact this a fascinating piece of equipment that is vital to the operation of the majority of our steam locomotives but it is something that the average visitor never really realises the significance of. So, let’s warm up the spanners and take a dive into the works shall we?

Hydrostatic displacement lubrication relies on a very simple scientific principle that almost everyone knows. Oil floats on water. This is due to density. The measurement of density is its “mass per unit volume”. Do not be frightened by the sudden outbreak of physics. Let us explain. You may have heard of the old trick question:

“What is heavier - a ton of bricks or a ton of feathers?”

The answer is of course that they both weigh the same (1 ton! Who got that wrong?) but there will be a fundamental difference in the size of each of those loads. In our imaginary ton of bricks there will only be relatively few bricks compared to the probably tens of thousands of feathers needed to make the same weight, or more accurately, mass. A much bigger load. So there is a lot more of the stuff that makes up bricks in a brick than there is of the stuff that makes up feathers in a feather. We measure the amount of stuff a substance by measuring its mass (physics uses kilograms these days). We then define how much space the object is taking up. This is its volume (using centimetres cubed for solids and millilitres for liquids*). Divide the mass by the volume and you get the density of the mass per unit volume. See? It’s not so scary! So, for a millilitre of water there is more of the stuff that makes up water than the amount stuff that makes up oil in a millilitre of oil. That is why the oil floats on the water. It is less dense.

So what has this impromptu physics lesson got to do with steam engines and lubricators? When we last spoke about lubrication, we found out that there were two types of oil. The thinner general lubricating oil and the thicker steam oil. It’s the thicker stuff, that looks a bit like green golden syrup, that we need to deliver to the moving parts where the hostile steam filled environment exists. This is the regulator, valves and cylinders at the front of the engine. In the cab of an engine there is a big brass container of the stuff and it has pipes going into it and pipes coming out of it.

The pipes coming in do so via a strange thing, usually mounted on the cab roof that either looks like a heater bar or a ring off a cooker. They are not heaters - think about where the firebox is... They are in fact condensers. The steam from the boiler is turned back into water by having to travel through this long, winding and uninsulated pipe. It however, remains at boiler pressure. This water is fed into the big brass container and because oil floats on water, the water is at the bottom and the oil is on top. There are a number of galleries or passages that the oil is forced through when the steam pressure is applied and the amount of water in the container increases.

The galleries are also connected to the glass tubes at the front of the brass container. There can be as many as five glasses on GWR engines. Each glass tube has a tap near it and a cone type structure inside it. As the water comes into the container and pushes on the oil, some of the oil is forced up through these cones. As it is working, the oil will drip upwards through the sight glass. The tap controls how fast the drips go a though the cone and therefore how fast the oil is delivered to the front end.

You will also notice that the some of the steam pipes go through a valve that is connected to the regulator lever (the big red ‘accelerator’ control, via a mechanical linkage. This is called the jockey valve and means that the oil isn’t wasted as it only flows when the regulator is open. It does give us a problem though. What if we are going along but the steam is shut off - what we call ‘coasting’? We might coast for a long distance and that means no oil flows and that is not great for the valves and cylinders...

Well, when you move a regulator, you notice that there is a little bit of what feels like slack in the mechanism. The handle moves a bit and then sort of ‘bites’ and opens the regulator valve. When they want to coast, a GWR engine driver will shut the regulator lever right down but then they do a strange thing and lift it up again just slightly. This is enough movement to open the jockey valve but not open the regulator valve - clever stuff, right?!

When the pressure is on the oil it is forced from the lubricator to the front end. You can trace the pipes from the sight glasses to where they disappear under the cladding (the thin metal outer case that holds the insulation on the boiler). These pipes then travel under the lagging to keep things warm. The hotter this thick oil is, the easier it flows. These pipes come up from under the lagging at the front end of the boiler, goes through a series of shut off taps under a small cover** and then dive into the smokebox. These oil pipes then deliver the oil into the steam flow coming from the boiler. It passes through an atomiser and it is carried from there, in the steam, to the moving parts. The number of sight glasses was updated over the years and if you see a locomotive with just one cover on the driver’s side of the smokebox it has a 3 glass lubricator, if it has a second, smaller one on the fireman’s side them it has a 5 glass lubricator. Nerdy fact that one!***


So, if this system is so good, why don’t all of the GWR locomotives have a hydrostatic displacement lubricator? Well, this system is really great as it gives a very fine degree of control to the driver over the oil supply. For most of the first half of the 20th Century, when the GWR fitted superheaters to their engines they did so using just a low degree of superheat. There were a maximum of just two rows of superheater elements in the boiler and the lower resulting steam temperatures involved give you a bit more leeway in oil control. However, when the poor post war coal meant that some locos began to be fitted with three and even four row superheaters (like our own No. 6023 King Edward II), there was a growing concern that the drivers might not supply enough oil to match the increased temperatures that their steads now operated at. As a result, mechanical lubricators were fitted to the running plates of the engines and as these are driven by the locomotive mechanism as it moves, the rate of oil flow is taken out of the hands of the driver completely.

The drivers were not all behind this decision. It was quite common for a GWR driver, when faced with an engine that was running in a sluggish manner, to turn up the level of oil flow to the front end and this could result in a marked improvement in the running qualities of the machine. Indeed, when the first two post war Castle Class engines were built with three row superheaters, they were fitted with hydrostatic displacement lubricators. They were regarded as powerful and free running but when they and the subsequent post war Castles were fitted with mechanical lubrication, their reputation with some crews were not as good. The debate about the merits of both systems carry on to this day.

Wherever the truth lies, the hydrostatic displacement lubricator remains a both a singularly very clever and mostly unsung piece of locomotive engineering. So, the next time you are at Didcot, have a look and see it you can see one of these fine pieces of equipment. Ask our drivers about them and take a moment to marvel at it. It’s really quite amazing. Let’s hear it for the hydrostatic displacement lubricator!

PS: Your blogger was asked about his personal mass per unit volume while writing this. He declined to comment...

*You might remember from your science lessons that 1ml of a liquid is the same volume as 1 cubic centimetre. While we don’t use the metric system a great deal in the works, this is really handy for doing physics!

**For some unknown reason, the myth is that the little blister style covers for the oil pipes and valves are there to hide the superheater. No - a superheater is a massive thing. It bolts to the front of the boiler and fills a reasonable proportion of the smokebox and the boiler tubes. It does not need covering. It’s inside. Pre-covered. The little covers are just there to tidy up the look of oil pipes and valves. This is the second time we have mentioned this but this myth is really persistent. We are here to both entertain and educate!

***This argument is null and void when you have an engine with a mechanical lubricator. It’s a big box thing on the running plate with loads of pipes on - you can’t miss it!



The Return of the Champion - Chapter 1

It’s time to come clean. One of your blog writers here has another duty at Didcot Railway Centre. For a number of years, I have had the unbelievable privilege of looking after one of the Great Western Society’s most precious and famous residents. None other than No. 4079 Pendennis Castle. As you may or may not know, we repatriated the locomotive from Australia in 2000* and in 2001 we slowly took the engine to pieces, intending it to be restored to running order. Yours truly showed up a short time after and once I earned my laurels, I was given the job of looking after the engine. It’s been a long road, a great deal of work has been done already and there is light now at the end of the tunnel (pun intended). For the first of these semi - regular updates, we will take a look at the at some of what has been done already and what the latest updates are.

Photo taken on 7 August 2018

The whole engine was basically complete upon return but also had a few ‘extras’. There were a number of Australian fittings including electric lamp brackets, air brake fittings and most notably what seemed like half of the iron red sand of the Pilbara! It was only relatively recently that the last of this was ejected into the bin, discovered as a thin red haze on the last few untouched bits...

The engine has been right down to the frames. It has meant that we have peered into every corner of this mighty machine and it has needed a lot of work to get us to where we are today. The boiler - quite often the bete noir of many loco restoration project - was pretty good. No. 4079 was fed a diet of de-mineralised water upon her arrival in the Pilbara. Rio Tinto needed it to prevent the desert water from ruining their mining equipment. Also, when they finished running her, the Australian crew emptied her boiler out. The desert heat did the rest.

Inside Pistons

It required some work on the stays - the crown stays in particular - seam and foundation ring rivets and of course a complete re-tube. The frames and mechanical items were sadly a much different story. The structure of the engine and the tender under the cab was pretty much life expired. As this forms the drag box (the bit that connects the two vehicles together and transmits the pulling force of the loco to the train), we were keen that this was a good as possible. A total replacement of the metal here was undertaken with just the main frames and two complex (and thankfully reusable!) pressings remaining. Her front buffer beam was also twisted and a great deal of effort was expended in straightening this out too.

4079 Drag Box

There isn’t much mechanically that hasn’t been overhauled. The wear in the bearing surfaces in some cases was twice the scrapping size. Oil and red desert sand mixing together make terrible lubricant but amazing grinding paste... All the whitemetal bearings have been refreshed or renewed and the cylinder and valve liners are also new with pistons and valves machines to suit.

Valve Liner

A number of features required a bit of reversion back to her original specification. The injectors that No. 4079 went to Australia with don’t like hot water and as you can imagine, a tender tank of water in the desert gets pretty warm, pretty quickly. They were replaced with some injectors that were far less susceptible to this problem. We replaced them with two GWR live steam units. Not entirely authentic (she should have a live and an exhaust injector**) but more so than the Gresham and Craven units as supplied!

4079 Crank Axle

The other ‘start from scratch’ area was the cylinder drain cocks. These little valves are on the bottom of the cylinders and steam chest. These are opened as an engine stands still and remain open for the first couple of revolutions of the wheels to drain the water out of the cylinders. Hence the big whooosh when a steam engine moves off. Water is incompressible unlike steam and if you have too much of it in there your cylinders go bang in a really expensive way. There are 11 of these and they had all been replaced with steam operating versions from an unknown source before she left the U.K. in the 1970s. The GWR ones are operated via a single lever and a mechanical linkage. Now, casting, machining and grinding in these valves are all technical tasks. Getting all 11 to line up and open together took the patience requirement a stage further...

Cylinder Liners

We had the majority of the work on the frames and boiler completed and we reunited them in order to make sure that we had all the bits we needed. The loco had been in bits for a number of years and this we felt was important to ensure that there were no delays that might unnecessarily burn up vital boiler ticket time. A lot of very fiddly and unexciting but very vital jobs got sorted in this time. Just before the lockdown of 2020 started, we had decided to make the final push towards removing the boiler for completion and commissioning.

Connecting Rod

We had everything to the boiler cladding released and ready to come off. Then it’s just a case of the cab roof and boiler out! And that’s when lockdown hit.

For various medical reasons concerning my other half, I was promptly locked into my flat for the next 4 months...

Thankfully, my great friends in the loco works were able to come back to Didcot a lot sooner than I was. They took the cladding sheets and cab roof off and then lifted the boiler, placing it on BR WELTROL No. 901002. Work has now commenced on doing the last few jobs and making it fully pressure tight. I finally made it back to Didcot on the 1st August.

Team fitting Drain Cocks

Despite the fact that the workforce size has been limited, progress has been great. As soon as there is something more to report, I’ll write another blog with a little engine history, a few jobs we have done before and where we are at the moment.

Best Wishes,

Drew and the 4079 Team.

Photo taken 28 November 1970

*Sounds like yours truly has a few Castle history blogs to write here...

**I’ll do a blog post on the way injectors work - it’s REALLY clever!



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