How were steam engines designed/drawn up/constructed?

[Peter C]

Hello
I've found myself reading more and more about steam engines recently, particularly those built at Swindon by the GWR. Whilst reading through various books by O. S. Nock, Cecil J. Allen and others, I've learnt a bit about what different bits of engines do, and how engines can be made to perform better by a driver, but I haven't learnt much about how the engines were actually designed (by this I mean how they were originally devised and drawn up by the likes of Collett, Churchward, Hawksworth, etc.). Hopefully people on here can answer my questions!

• When designing a steam engine, such as the "Castles", for example, how would Collett have worked out the best sizes for all of the different bits of an engine to make them the best he could? Were there equations and calculations which would give one the necessary measurements, or was there some sort of proven relation between various things, so once one was designed, the others went along with it? Or was it a case of trial and error?
• When engines were being designed, what were the various factors involved in what it looked like? How much of a role did style and public relations play in the design of an engine?
• How did engines go from being ideas to being drawings and then to being built? I've seen pictures of a drawing room at Swindon, showing sets of tables with people all drawing various things - what was this for, and was it to make copies of things, or was it to speed up the process of making one set of drawings, e.g. someone does cylinders, someone does the smokebox, etc.? And when these drawings were made and finished, how did the real thing get built? I know there were construction lines at Swindon but how do you start making something which has never been made before?

I know this is quite a lot, and I bet I've got several things wrong in terms of terminology and all sorts! I know very little, if nothing about this stuff, so answers which are as easy to understand as possible would be preferred

Thanks,

-Peter

 

[Ken H]

Hello
I've found myself reading more and more about steam engines recently, particularly those built at Swindon by the GWR. Whilst reading through various books by O. S. Nock, Cecil J. Allen and others, I've learnt a bit about what different bits of engines do, and how engines can be made to perform better by a driver, but I haven't learnt much about how the engines were actually designed (by this I mean how they were originally devised and drawn up by the likes of Collett, Churchward, Hawksworth, etc.). Hopefully people on here can answer my questions!

• When designing a steam engine, such as the "Castles", for example, how would Collett have worked out the best sizes for all of the different bits of an engine to make them the best he could? Were there equations and calculations which would give one the necessary measurements, or was there some sort of proven relation between various things, so once one was designed, the others went along with it? Or was it a case of trial and error?
• When engines were being designed, what were the various factors involved in what it looked like? How much of a role did style and public relations play in the design of an engine?
• How did engines go from being ideas to being drawings and then to being built? I've seen pictures of a drawing room at Swindon, showing sets of tables with people all drawing various things - what was this for, and was it to make copies of things, or was it to speed up the process of making one set of drawings, e.g. someone does cylinders, someone does the smokebox, etc.? And when these drawings were made and finished, how did the real thing get built? I know there were construction lines at Swindon but how do you start making something which has never been made before?

I know this is quite a lot, and I bet I've got several things wrong in terms of terminology and all sorts! I know very little, if nothing about this stuff, so answers which are as easy to understand as possible would be preferred

Thanks,

-Peter

This is a nice book. I actually met Eric but he was sadly very frail.
But he gives good detail of his work

Amazon.co.uk: Buying Choices: Trained by Sir Nigel Gresley

 

[341o2]

You tube video link building a steam engine

Basically built in bits, some from plate, some from castings, then assembled

 

[pdeaves]

When designing a steam engine, such as the "Castles", for example, how would Collett have worked out the best sizes for all of the different bits

Collet, Churchward, et al would not have done very much, at all, of the design/calculations. They were the 'top men' in their department and set policy, oversaw what went on, etc. They had a team of draughtsmen/designers to do the actual work. No doubt they reviewed designs as they went on but would have done very little of the 'actual work', just authorise the final product for construction.

 

[Ken H]

Hello
I've found myself reading more and more about steam engines recently, particularly those built at Swindon by the GWR. Whilst reading through various books by O. S. Nock, Cecil J. Allen and others, I've learnt a bit about what different bits of engines do, and how engines can be made to perform better by a driver, but I haven't learnt much about how the engines were actually designed (by this I mean how they were originally devised and drawn up by the likes of Collett, Churchward, Hawksworth, etc.). Hopefully people on here can answer my questions!

• When designing a steam engine, such as the "Castles", for example, how would Collett have worked out the best sizes for all of the different bits of an engine to make them the best he could? Were there equations and calculations which would give one the necessary measurements, or was there some sort of proven relation between various things, so once one was designed, the others went along with it? Or was it a case of trial and error?
• When engines were being designed, what were the various factors involved in what it looked like? How much of a role did style and public relations play in the design of an engine?
• How did engines go from being ideas to being drawings and then to being built? I've seen pictures of a drawing room at Swindon, showing sets of tables with people all drawing various things - what was this for, and was it to make copies of things, or was it to speed up the process of making one set of drawings, e.g. someone does cylinders, someone does the smokebox, etc.? And when these drawings were made and finished, how did the real thing get built? I know there were construction lines at Swindon but how do you start making something which has never been made before?

I know this is quite a lot, and I bet I've got several things wrong in terms of terminology and all sorts! I know very little, if nothing about this stuff, so answers which are as easy to understand as possible would be preferred

Thanks,

-Peter

There would have been a catalogue of parts. So buffers, injectors. If possible an old design would have been reused.
Then the injector man would have been given a specification and told to design an injector for the engine. He would be able to calculate the weight from the drawing.
Some parts may have been bought in, and the suppliers drawings would be in the set, with weights and other attributes
Other draughtsmen would have placed all the parts onto the main drawing and the weight distribution calculated.
possibly several stages. so a drawing for the tubes. Then the drawing for the boiler - 1 with and one without the tubes.

The cheif designer would manage the process, and also oversee the design so it worked as a whole. But he would have staff to design the parts.

At manufacture, a department would make the tubes, another make the bare boiler. Then perhaps a third department would fit the tubes and other bits to make a complete boiler as a sub assembly.
Then the boiler would be added to the frame.

There would have been a bill of materials. The resources to make the tubes would be on that, including labour, machine shop time and the raw materials etc.
A works order would be for 50 tubes. A clerk would allocate the resources to the order. The tubes would be made. Then the clerk would reduce the stock on hand for the consumables in the stock system (back flushing), and update the stock for the finished tubes.
The stock system would have attributes for each item, including weight and cost. So hen you raise a works order you know the cost of the finished item.

Later, a works order would be raised for a finished boiler. so the same system of allocation and stock update would occur.

Of course all this would be paper based. Today we use computers for managing manufacturing. And today, many components would be bought in, rather than everything being made in 1 factory.

 

[edwin_m]

I don't think anyone would have been copying drawings. I can just about remember at the start of my career when things were drawn by hand on what looked like tracing paper, and could be copied by putting them on top of some special paper then passing them through a machine with a UV tube inside. I think this was quite an old technique. The resulting copy was blue except where there were lines on the master, presumably the origin of the term "blueprint". The originals would have been filed away and the prints given to the people doing more detailed designs of components to show how they fitted together, and ultimately to those making the components themselves. For cast components a skilled toolmaker would produce a replica out of wood, which was then pressed into some special sand to make a mould for the molten metal. Things like screw threads would have been machined afterwards, again according to the drawings.

 

[Peter C]

You tube video link building a steam engine

Basically built in bits, some from plate, some from castings, then assembled

That's a really interesting video - a lot of detail about the different parts of the process.

Collet, Churchward, et al would not have done very much, at all, of the design/calculations. They were the 'top men' in their department and set policy, oversaw what went on, etc. They had a team of draughtsmen/designers to do the actual work. No doubt they reviewed designs as they went on but would have done very little of the 'actual work', just authorise the final product for construction.

I wondered if it could have been that. That was one of my thoughts when I saw the photos of the drawing rooms at Swindon

This is a nice book. I actually met Eric but he was sadly very frail.
But he gives good detail of his work

Amazon.co.uk: Buying Choices: Trained by Sir Nigel Gresley

Looks interesting. I'll look into getting it and see what it has to offer!

There would have been a catalogue of parts. So buffers, injectors. If possible an old design would have been reused.
Then the injector man would have been given a specification and told to design an injector for the engine. He would be able to calculate the weight from the drawing.
Some parts may have been bought in, and the suppliers drawings would be in the set, with weights and other attributes
Other draughtsmen would have placed all the parts onto the main drawing and the weight distribution calculated.
possibly several stages. so a drawing for the tubes. Then the drawing for the boiler - 1 with and one without the tubes.

The cheif designer would manage the process, and also oversee the design so it worked as a whole. But he would have staff to design the parts.

At manufacture, a department would make the tubes, another make the bare boiler. Then perhaps a third department would fit the tubes and other bits to make a complete boiler as a sub assembly.
Then the boiler would be added to the frame.

There would have been a bill of materials. The resources to make the tubes would be on that, including labour, machine shop time and the raw materials etc.
A works order would be for 50 tubes. A clerk would allocate the resources to the order. The tubes would be made. Then the clerk would reduce the stock on hand for the consumables in the stock system (back flushing), and update the stock for the finished tubes.
The stock system would have attributes for each item, including weight and cost. So hen you raise a works order you know the cost of the finished item.

Later, a works order would be raised for a finished boiler. so the same system of allocation and stock update would occur.

Of course all this would be paper based. Today we use computers for managing manufacturing. And today, many components would be bought in, rather than everything being made in 1 factory.

That makes a lot of sense - thanks for so much detail.
Having different people who specialise in certain things seems more sensible than having one person design everything, and using as many existing parts as possible is also very clever.

I don't think anyone would have been copying drawings. I can just about remember at the start of my career when things were drawn by hand on what looked like tracing paper, and could be copied by putting them on top of some special paper then passing them through a machine with a UV tube inside. I think this was quite an old technique. The resulting copy was blue except where there were lines on the master, presumably the origin of the term "blueprint". The originals would have been filed away and the prints given to the people doing more detailed designs of components to show how they fitted together, and ultimately to those making the components themselves. For cast components a skilled toolmaker would produce a replica out of wood, which was then pressed into some special sand to make a mould for the molten metal. Things like screw threads would have been machined afterwards, again according to the drawings.

Thanks for the explanation
So the various things would have been designed by various people specialising in those things. The LMS video linked further upthread describes that as well.

Thanks to everyone so far for their amazing responses!

-Peter

 

[John Webb]

The 'blueprint' system used ammonia as the developer, I recall, to bring up the print on the sensitised paper. Can't remember the exact chemical process.

Swindon in particular was noted for large-scale commonality of parts, from bolts to boilers, for many different loco classes. This kept the costs down. Several other companies followed suit, including the LMS when they persuaded Stanier to join them from the GWR.

 

[Peter C]

Swindon in particular was noted for large-scale commonality of parts, from bolts to boilers, for many different loco classes. This kept the costs down. Several other companies followed suit, including the LMS when they persuaded Stanier to join them from the GWR.

That was one of the things I remember from one of the books I read. A fairly small selection of boilers could be repurposed from one class to another with small changes if needed I assume.

One of the books was Great Western Railway GWR Stars, Castles and Kings by O. S. Nock (omnibus edition, parts 1 and 2), and I've found it to be very interesting indeed. Especially the detail about 111 The Great Bear. I seem to have quite an interest in that loco!

-Peter

 

[Cowley]

That was one of the things I remember from one of the books I read. A fairly small selection of boilers could be repurposed from one class to another with small changes if needed I assume.

One of the books was Great Western Railway GWR Stars, Castles and Kings by O. S. Nock (omnibus edition, parts 1 and 2), and I've found it to be very interesting indeed. Especially the detail about 111 The Great Bear. I seem to have quite an interest in that loco!

-Peter

Re the standardisation of parts - it’s worth having a look at the amount of replicas that are being recreated out of bits of The Barry ten (as well as other Swindon spares) to recreate long lost classes like the 47xx and 68xx ‘Grange’ amongst others.

 

[webbfan]

The 'blueprint' system used ammonia as the developer, I recall, to bring up the print on the sensitised paper. Can't remember the exact chemical process.

Swindon in particular was noted for large-scale commonality of parts, from bolts to boilers, for many different loco classes. This kept the costs down. Several other companies followed suit, including the LMS when they persuaded Stanier to join them from the GWR.

Ramsbottom was very insistent on standardization when he took over the complete LNWR. Webb followed same pattern, sometimes it appears as though he only built one engine as there were so many major components shared between the classes.

 

[coppercapped]

If you can find a copy Harold Holcroft's book 'Locomotive Adventure', buy it! It gives a lot of background to drawing office (and other) work at Swindon under Churchward just before the First World War. Holcroft was the man who designed the GWR's 2-6-0 Moguls, styled the curves in the running plates so typical of GWR locomotives and also did the maths behind Gresley's conjugated valve gear; this Wikipedia article gives and outline of his life.

I can thoroughly recommend the book - it was first published in, if I recall correctly, in the early 1960s when Holcroft had already been retired for fifteen years or so. I bought a first edition and treasure it still.

The Amazon link is here, but there is no stock at the moment.

 

[Peter C]

Re the standardisation of parts - it’s worth having a look at the amount of replicas that are being recreated out of bits of The Barry ten (as well as other Swindon spares) to recreate long lost classes like the 47xx and 68xx ‘Grange’ amongst others.

I've definitely read about that as well. I remember hearing a lot about the "Saint" at Didcot (2999 Lady of Legend) being made from (I think) the frames from a "Hall" (Maindy Hall?) and it proves how you can take one set of parts and re-purpose them for various classes of engine.

If you can find a copy Harold Holcroft's book 'Locomotive Adventure', buy it! It gives a lot of background to drawing office (and other) work at Swindon under Churchward just before the First World War. Holcroft was the man who designed the GWR's 2-6-0 Moguls, styled the curves in the running plates so typical of GWR locomotives and also did the maths behind Gresley's conjugated valve gear; this Wikipedia article gives and outline of his life.

I can thoroughly recommend the book - it was first published in, if I recall correctly, in the early 1960s when Holcroft had already been retired for fifteen years or so. I bought a first edition and treasure it still.

The Amazon link is here, but there is no stock at the moment.

Sounds interesting - I'll definitely look into it. I think Holcroft was mentioned in one or two of my other books. I've realised that I don't actually know much about the work that went on at Swindon so it will definitely be a useful book for learning how things were done!

-Peter

 

[pdeaves]

it proves how you can take one set of parts and re-purpose them for various classes of engine.

The West Somerset Railway (I think) even has a 'non-existent' locomotive; a 'might-have-been' assembled from various standard GWR bits in a configuration never used on the 'big railway'. Someone with more knowledge can tell you which engine it is.

 

[Cowley]

The West Somerset Railway (I think) even has a 'non-existent' locomotive; a 'might-have-been' assembled from various standard GWR bits in a configuration never used on the 'big railway'. Someone with more knowledge can tell you which engine it is.

That’s right, Mogul 9351 is made out of Prairie 5193 (the number is an anagram), I think it might have been because the rear frame section was completely rotten? It looks like a mini 43xx, although to my eyes it looks slightly odd.

 

[Taunton]

As the question is about Swindon, it is interesting to compare CMEs there. The "Chief" typically had two offices, one at the works (Swindon) and one at company headquarters (Paddington). They did tend to be the top person for budgets, manpower, strategy, etc, but also had different approaches to the job.

Churchward normally worked at Swindon and apparently appeared in the drawing most days, always with questions and comments. He knew just what the answers were to just about everything, but would give an outline of advice and then let people get on with it, reviewing it the next day.

Collett more commonly worked at Paddington, spoke with the other top officials and directors, wrote policy notes, etc. He was seen as far more distant.

Whereas Churchward's designs were seen as revolutionary, Colletts were evolutionary. Which suited the style of the time for the GWR.

The Chief Designer in the drawing office was commonly the No 2 in the organisation, even they would have a group of direct admin staff etc. Notable is the split that existed between design and works. Commonly locomotives were designed, then it was a decision point where they were to be built, whether internally at the works, or by one of the many outside contractors. Sometimes a class would be built by both, or by two separate contractors, and it was really not possible to see any difference between them. Even the GWR had considerable numbers of their locomotives built by outside companies, dependent on relative pricing, works load, etc - most internal works capacity was taken up with periodic overhauls rather than new build; when locos were effectively taken to pieces and built back up again at overhaul, with the parts refurbished or replaced, there wasn't a lot of difference. If you had a 63xx on your Taunton to Barnstaple local, only the most informed knew that 6390 had been built by Robert Stephenson in Darlington, and not at Swindon.

 

[30907]

Seconding the reference to Holcroft's book - volume 2 as well, though by then he was on the SR. I'm biased

 

[coppercapped]

Having different people who specialise in certain things seems more sensible than having one person design everything, and using as many existing parts as possible is also very clever.

The concept of the standardisation of parts dates back a long, long way. Think crossbows and cannon and the Roman legions - they had standardised parts such as fencing panels for building temporary (and some not so temporary) fortresses.

More directly connected to railways, Sir Marc Brunel built a series of machines at Portsmouth to manufacture pulley blocks for the Navy in about 1801 or 2. The Navy needed some hundred thousand a year and manual methods couldn't keep up. The world's first transfer machine tool - it did one process and passed the block on to the next step and the concept is now used in the volume manufacture of internal combustion engines the world over and many other industrial processes such as semiconductor chip manufacture.

There is nothing new under the sun...!

 

[Peter C]

That’s right, Mogul 9351 is made out of Prairie 5193 (the number is an anagram), I think it might have been because the rear frame section was completely rotten? It looks like a mini 43xx, although to my eyes it looks slightly odd.

That engine looks really quite odd - yes! I thought it looked a bit like (if you squint/take your glasses off/look the other way) an LNER B17. But that's probably just me!

As the question is about Swindon, it is interesting to compare CMEs there. The "Chief" typically had two offices, one at the works (Swindon) and one at company headquarters (Paddington). They did tend to be the top person for budgets, manpower, strategy, etc, but also had different approaches to the job.

Churchward normally worked at Swindon and apparently appeared in the drawing most days, always with questions and comments. He knew just what the answers were to just about everything, but would give an outline of advice and then let people get on with it, reviewing it the next day.

Collett more commonly worked at Paddington, spoke with the other top officials and directors, wrote policy notes, etc. He was seen as far more distant.

Whereas Churchward's designs were seen as revolutionary, Colletts were evolutionary. Which suited the style of the time for the GWR.

The Chief Designer in the drawing office was commonly the No 2 in the organisation, even they would have a group of direct admin staff etc. Notable is the split that existed between design and works. Commonly locomotives were designed, then it was a decision point where they were to be built, whether internally at the works, or by one of the many outside contractors. Sometimes a class would be built by both, or by two separate contractors, and it was really not possible to see any difference between them. Even the GWR had considerable numbers of their locomotives built by outside companies, dependent on relative pricing, works load, etc - most internal works capacity was taken up with periodic overhauls rather than new build; when locos were effectively taken to pieces and built back up again at overhaul, with the parts refurbished or replaced, there wasn't a lot of difference. If you had a 63xx on your Taunton to Barnstaple local, only the most informed knew that 6390 had been built by Robert Stephenson in Darlington, and not at Swindon.

A very detailed explanation - I think I understand the system a bit more now. As with every company, different people will have different ways of working and the GWR was no different. The book by O. S. Nock I mentioned upthread goes into quite a bit of detail with regards to the evolution of GWR designs and you can see how the "Stars" led to the "Castles" and then the "Kings", both through technical information as each one was more powerful than the last, and also through style as the engines all share the same basic principles of GWR engine design, which goes back to the standardisation and re-purposing of parts and designs.

The concept of the standardisation of parts dates back a long, long way. Think crossbows and cannon and the Roman legions - they had standardised parts such as fencing panels for building temporary (and some not so temporary) fortresses.

More directly connected to railways, Sir Marc Brunel built a series of machines at Portsmouth to manufacture pulley blocks for the Navy in about 1801 or 2. The Navy needed some hundred thousand a year and manual methods couldn't keep up. The world's first transfer machine tool it did one process and passed the block on to the next step - as now used in the volume manufacture of internal combustion engines the world over.

There is nothing new under the sun...!

A very good point! Also one of the world's first production lines I assume from the description? I've heard of production lines devised during the Industrial Revolution, particularly the Garrett traction engine works in Leiston, Suffolk: they were one of the world's first production lines apparently, but they used people to make things.

-Peter

 

[Ken H]

I would assume as time went on the understanding of metals and how to make them stronger made improvements in design possible. I dont know if welding was ever good enough for steam engines, but that would be a great advance over riveting, not least in weight.

 

[Peter C]

I would assume as time went on the understanding of metals and how to make them stronger made improvements in design possible. I dont know if welding was ever good enough for steam engines, but that would be a great advance over riveting, not least in weight.

I think 60163 Tornado was largely constructed using welding - although whether the technology existed to make a full engine using welding during the steam era I don't know. I have a feeling the LMS might have done quite a bit with welding, but I've got no clue why that sprang to mind when I saw your comment!

-Peter

 

[silverfoxcc]

Building a 5in live steam loco has given me some insight on what goes on. After all as Grandpa Pig forcibly reminds people 'It is NOT a toy train, it is a miniature locomotive. Whilst most things can be scaled, and with 3D printing and CNC some quality stuff can be made
One thing that you cannot do is scale nature, so we run with boilers up to 100lb psi, which are small versions of the big boys. everything else as mentioned can be scaled from works drawings ,readily available at the NRM if you book a day. One thing with this scale is to chose a loco and get as much info as possible, and do not take the published drawing by others as being gospel. I found out the hard way as i am now on my 4th B17 GE Tender and have seen three other variants!
But getting back to to OP question. Original it was i would imagine all done 'on the back of a fag packet' and guesswork, and as deigners progressed they began to know the li,in designer based on what the CME wanted. There is a good example of this ( And NO opening up of the Gresley/Thompson saga) When Thompson had decided on his A1/1 variant and his A2 designs ,he was getting towards his retirement and Peppercorn was the 'shoo-in' The drawing office started working on the Peppercorn A1, and one day Thompson paid a visit to them. The new drawings were hurriedly hidden under others! It took nearly 150 years of developments,and locos were built to suit a specific purpose. As Ron White of Colour-rail used to point out While Gresley was bringing out the A4 Collett was working on the 14xx class!! ( Ron hated the GWR) So i would assume that they learnt by the mistakes and improved where possible

 

[Grumpy Git]

Have a read of the Wiki page for Andre Chapelon, who was one of the first (if not the first) railway engineer to use science rather than trial and error.

 

[Bevan Price]

Re the standardisation of parts - it’s worth having a look at the amount of replicas that are being recreated out of bits of The Barry ten (as well as other Swindon spares) to recreate long lost classes like the 47xx and 68xx ‘Grange’ amongst others.

Yes - (post Churchward) the GWR just had a few standard designs of loco boliers. Even the BR 82xxx 2-6-2 tanks had what was basically a Swindon - designed bolier.

Early steam loco design was largely a matter of trial and error - find out what worked, and try to improve it. Churchward's GWR team was amongst the first to realise that smooth flow of steam from boiler to valve gear, cylinders & exhaust gave the best performance and greatest efficiency. Stanier followed the same principles when he moved from GWR to LMSR, but needed to alter some combustion arrangements as the Welsh coal (used by GWR) behaved a bit differently to the (mainly Midlands) coal used by LMSR.

 

[Cowley]

Yes - (post Churchward) the GWR just had a few standard designs of loco boliers. Even the BR 82xxx 2-6-2 tanks had what was basically a Swindon - designed bolier.

Early steam loco design was largely a matter of trial and error - find out what worked, and try to improve it. Churchward's GWR team was amongst the first to realise that smooth flow of steam from boiler to valve gear, cylinders & exhaust gave the best performance and greatest efficiency. Stanier followed the same principles when he moved from GWR to LMSR, but needed to alter some combustion arrangements as the Welsh coal (used by GWR) behaved a bit differently to the (mainly Midlands) coal used by LMSR.

With the team building the 47xx planning to use the boiler from an 8F, I’ve always found it interesting that the dimensions are so close that they’d be able to do that.
It made me wonder if other companies boilers were interchangeable with each other?

 

[webbfan]

Quite good insight into the casual/accidental rather than scientific nature of design can be gained from E.A. Langridge "Under 10 CMEs", Oakwood Press..

 

[341o2]

I would assume as time went on the understanding of metals and how to make them stronger made improvements in design possible. I dont know if welding was ever good enough for steam engines, but that would be a great advance over riveting, not least in weight.

Bulleid was an advocate of welding, it was used extensively in construction of his pacifics at least

 

[coppercapped]

Have a read of the Wiki page for Andre Chapelon, who was one of the first (if not the first) railway engineer to use science rather than trial and error.

I think that is a little unfair on earlier engineers...! Chapelon certainly had the benefit of modern test methods but dynamometer coaches and experimental testing generally goes back to the very early days of railways.

One shouldn't underestimate the effects of the development and use of steam power from the 18th century onwards on scientific knowledge. From being a philosophical debate on the underlying concepts of heat and temperature and alchemy, the need to understand how heat engines worked led to the development of the whole science of thermodynamics which in turn drove practically all of current scientific knowledge. The very word 'thermo-dynamics' was coined by William Thomson, later Lord Kelvin, in the 1850s and statistical thermodynamics was developed by James Clark Maxwell sometime later (one of my heroes!) which led directly to radio and electronics and all that stuff.

 

[Clarence Yard]

If you are interested in GWR M&EE organisation, workshop practice and locomotives, "Swindon Steam 1921-51" by K.J.Cook is definitely the book to get. Cook is one of the unsung heroes of GWR steam and his book shows you what went on at Swindon during that period, including welding in sections of fireboxes. The GWR had a reputation for being conservative but in their workshop practices they were quite advanced. This enabled to locos to run good periods between shops.

"Raising Steam on the LMS" by A.F.Cook, one of the RCTS LMS books, is probably the easiest to read book on boiler design issues and development but if you are interested in the design, development, modification and maintenance of LNER Pacifics then "East Coast Pacifics at Work" and "Top Shed", both by (my old boss) P.N.Townend are pretty good, written by a man who really knew his locos.

The standardisation that the GWR followed was fairly impressive. It allowed for reduced maintenance times and costs of repair - in 1950 WR had about 715 Standard No 1 boilers for the Stars, Saints, Halls, Granges and 28XX locos. The boilers were regularly swapped on overhaul (apart from minor work, a boiler usually takes longer to repair than the engine) and by having a common type, the boiler float could be kept to a reasonable minimum. As the fleet reduced, you could gradually run down your stock - by 1958 the total had come down to about 627 for 576 locos.

The value of this approach can be seen in other ways. When City of Truro was restored to traffic in 1957, it's existing old boiler was junked and 1921 built Standard 4 boiler 6025 (from 3185) was given an A repair (retube and partial restay) and put on City of Truro. Even since 1940, Boiler 6025 had been on 4 locomotives before it was put on 3440 in 1957!

 

[bspahh]

If you can find a copy Harold Holcroft's book 'Locomotive Adventure', buy it! It gives a lot of background to drawing office (and other) work at Swindon under Churchward just before the First World War. Holcroft was the man who designed the GWR's 2-6-0 Moguls, styled the curves in the running plates so typical of GWR locomotives and also did the maths behind Gresley's conjugated valve gear; this Wikipedia article gives and outline of his life.

I can thoroughly recommend the book - it was first published in, if I recall correctly, in the early 1960s when Holcroft had already been retired for fifteen years or so. I bought a first edition and treasure it still.

The Amazon link is here, but there is no stock at the moment.

https://www.steamindex.com/people/holcroft.htm has a lot of information on Holcroft, with links to papers in engineering journals, and an email from his grandson with quotes from him.