steam loco engineering tollarances

[18B]

While diesel locos ultimate went for a component exchange maintenance cycle, would it be too far to say that while Steam age locomotive engineering tolerances were perfectly fine for their day, each engine was pretty much an individual and many of its constituent parts were bespoke to it alone thus preventing component exchange maintenance? .    

[Killian keane]

Components were exchanged like mad! Just look at the quarry hunslets, they swapped virtually everything, (class variations permitting) to such a point as identification becomes a nighmare in a headache! So safe to say a very high standard of engineering, now numbering of components was (is) commonplace mind

[The Bigbee Line]

Reading books it appears that when the Southern Railway was formed. The SE&CR and LSWR workshops had standard "oversize" dimensions for major components. The LBSC seemed to have a make it fit policy. When Brighton locos went to Ashford, the works staff pulled their hair out. So the answer is yes there was standardisation of sizes.

[D854_Tiger]

While diesel locos ultimate went for a component exchange maintenance cycle, would it be too far to say that while Steam age locomotive engineering tolerances were perfectly fine for their day, each engine was pretty much an individual and many of its constituent parts were bespoke to it alone thus preventing component exchange maintenance? .    

 

It may have been the case very early on but by around the 1920s railways like the GWR designed most of their locomotives around the concept of interchangeability.

 

I believe it was the case that many Castle class interchanged boilers with their sister locomotive and even other classes.

 

Know I know you have be very careful getting into debates like this, where a degree of tribalism can be involved, but I do remember reading somewhere that the GWR, in particular, lead the way in engineering standards.

 

So much so, that it was claimed their locomotives were going into overhaul with tighter tolerances than many other railways sent their locomotives out of overhaul with.

 

Whether that was true or not and whether that contributed to the GWR locomotive generally performing far better in the loco exchange trails, I know not.

 

But it does raise the rather interesting question, nowadays, when all restored locomotives benefit from the highest most modern engineering standards are we seeing greater levels of performance, for certain types, than were ever experienced back in the day..

 

Are we only now seeing the true potential of some of those designs, in a way that those performances might have been considered game changing when they were in daily service.

[The Bigbee Line]

One class of diesel had members built in Romania. Some components were hand made (axlebox covers for example) and could not be swapped about!

[D854_Tiger]

One class of diesel had members built in Romania. Some components were hand made (axlebox covers for example) and could not be swapped about!

 

Rumour has it that some of those components didn't even fit the locomotives they were intended for that well, let alone others of the same class.

[asmay2002]

While diesel locos ultimate went for a component exchange maintenance cycle, would it be too far to say that while Steam age locomotive engineering tolerances were perfectly fine for their day, each engine was pretty much an individual and many of its constituent parts were bespoke to it alone thus preventing component exchange maintenance? .    

This is completely wrong.  Component exchange was practicsed from the 1840s at least.  Daniel Gooch supplied patterns to some loco manufacturers to ensure that members of the same class had interchangeable parts.  It doesn't matter whether the tolerances are small or large - as long as the parts are within the tolerances they can be swapped between locos.

[10000]

While diesel locos ultimate went for a component exchange maintenance cycle, would it be too far to say that while Steam age locomotive engineering tolerances were perfectly fine for their day, each engine was pretty much an individual and many of its constituent parts were bespoke to it alone thus preventing component exchange maintenance? .    

 

 

That all time favourite (of some people ) Flying Scotsman is a prime example of running with parts from another loco. Duchess Of Hamilton carried at least 5 different boilers during its years on LMS/BR

 

Post nationalisation the level of part swapping gradually increased in some areas. Crewe works in latter years when multiple locos from the same class where in didn't worry about rebuilding with the same parts, it was a case of what was picked up first.

Towards the very end you could often see locos where the smoke box number did not match the cab side.

[pH]

Towards the very end you could often see locos where the smoke box number did not match the cab side.

I can honestly say I did not see a single occurrence of that. Now, that may have been because I didn't think of checking they were the same - seeing either one of them was enough for me to note the number. Any examples, preferably with pictures?

[Phil Bullock]

Have a look at a GWR loco the next time you go to a preserved line.

 

Numbers are stamped on parts and in many cases rods have been swapped between class members or class members from other classes that used the same components. Its what caused some confusion about which loco was which at Barry - not least 4965 vs 4983

 

Cheers

 

Phil

[The Johnster]

When a steam loco goes into workshops for a full overhaul, it is generally ready to go back into service after about 14 or 15 working days (give or take), but the work on the boiler is more complex and may take a week or two longer.  In order to return the loco to service without having to wait for the boiler to be ready, railways carried a stock of spare boilers and swapped them between members of that class, and indeed any other classes that that particular design of boiler would fit on to.  For example, the Swindon Number 2 boiler could be fitted to all the large prairie classes except the 31xx and 3150, as well as 56xx, several rebuilt absorbed and constituent company's locos in GW rebuilt form, and some 4-4-0s.  Boilers were swapped indiscriminately between all of these classes, and this was the normal practice everywhere throughout the British railway industry until the end of steam.  

 

As for tolerances, they are irrelevant as long as they match within stated allowed limits across a range of standardised locos.  The Swindon claim that they withdrew locos from service with finer tolerances than Derby built them may or may not be true, but Derby were using building methods and a philosophy of design that worked very well for the Midland Railway for many years, and I would be reluctant to state that theirs were not as good as Swindon's.  The compounds were among the most efficient locos in the country when they were first built, though continuing to build them for the LMS a quarter century later was pushing things a bit!

 

As well as boilers, other components were swapped where they were standardised as well.  On the GW, AFAIK, only the frames and coupling rods were stamped with the loco's number and kept together for the life of the loco.  As Phil says, the availability of spares and less concern with the free running of the locos at high speed on preserved lines has led to some swapping of these as well.  

 

It leads to legitimate questions as to what the actual identity of a steam locomotive is, beyond name, number. and works plates; sometimes the answer is debatable!

[bike2steam]

This is completely wrong.  Component exchange was practicsed from the 1840s at least.  Daniel Gooch supplied patterns to some loco manufacturers to ensure that members of the same class had interchangeable parts.  It doesn't matter whether the tolerances are small or large - as long as the parts are within the tolerances they can be swapped between locos.

Indeed, by the 1930's most had grasped the concept of standardisation - except Gresley. Steam locos had general running/ clearance- fit tolerances because of the greater temperature variations than diesel, to use precision tolerances on steam locos would be fatal.

[LMS2968]

I'm sure I've said this before, but the different works varied in their attitudes to retaining parts with an individual loco; Crewe and Horwich stencilled the loco number on all parts prior to strip down so that they would indeed return to their parent. This is the left hand union link from 13268, currently under overhaul. You will notice the various stampings: L = Left Hand Side; C136 was her build number at Crewe Works; 13268 her original LMS number and 2968 her LMS number from 1934.

 

 

All but two motion items are stamped either 13268, 2968 or 42968; there are no numbers from other class members. The two items mentioned are not stamped. They are the left hand return crank and rod, both remade by the Fund during the initial restoration as these parts were missing when the engine went into Barry.

 

Take a look at 6201's boiler handrails. She and 6200 had the feed delivery pipes to the top feed external to the cladding as built. All others had them recessed, and these two were modified to conform later. But today her original handrails have a joddle beneath the top feed to clear those original external delivery pipes.

 

Take their big sisters, built as either streamlined or non-streamlined. The boiler handrail stanchions for the non-streamlined locos were long and slender, but the streamlined casing of the others required an extension piece fitted to the mountings on the boiler, into which much shorter stanchions then fitted. On destreamlining this system was retained, and all non-streamlined locos retained the long slender stanchions, all destreamlined ones had a noticeably thicker short stem with the short, slender stanchion fitting into it. With two exceptions: 6225 got non-streamlined handrails and 6255 got the destreamlined type. The similarity in numbers - 6225 and 6255 - is obviously the key.

 

With the same class, destreamlined locos appeared with a wedge-shaped smokebox top which had fitted under the streamlined casing. It took fourteen years to bring them all to standard, and that involved a lot of boiler exchanges in that time, but have you ever seen a photo of an originally non-streamlined loco with the wedge-topped smokebox? No, because the smokeboxes did not stay with the boiler but with the loco, so each type retained the correct sort.

 

Having said all that, there is a photo of 6232 under overhaul in Crewe works, but with the front numberplate from a Jubilee. All Stanier smokebox doors for his tender engines were the same size, so this is one case when an original part was not retained.

[pH]

Take their big sisters, built as either streamlined or non-streamlined. The boiler handrail stanchions for the non-streamlined locos were long and slender, but the streamlined casing of the others required an extension piece fitted to the mountings on the boiler, into which much shorter stanchions then fitted. On destreamlining this system was retained, and all non-streamlined locos retained the long slender stanchions, all destreamlined ones had a noticeably thicker short stem with the short, slender stanchion fitting into it. With two exceptions: 6225 got non-streamlined handrails and 6255 got the destreamlined type. The similarity in numbers - 6225 and 6255 - is obviously the key.

 

Just out of interest, what system did 46242 have when 'rebuilt' after Harrow? It was a de-streamlined engine, but had some "non de-streamlined" features after that e.g. footplate drop-ends.

[roythebus]

I could always mention the London Transport overhaul system at Aldenham, but that's another argument for another forum! 

 

I'd suggest you read the recent RAIB report on the West Country/BB that dropped the side rod on the South Western a year or so ago, that'll tell you a lot about engineering tolerances and standards.

[Jeff Smith]

Indeed, by the 1930's most had grasped the concept of standardisation - except Gresley. Steam locos had general running/ clearance- fit tolerances because of the greater temperature variations than diesel, to use precision tolerances on steam locos would be fatal.

I would strongly contest the last statement. The tolerances would have been designed to take account of the operating temperature range. This may have allowed greater clearances when cold but unless the appropriate (design) tolerances were applied parts would either be too loose at operating temperature or too tight and seize!

[LMS2968]

Just out of interest, what system did 46242 have when 'rebuilt' after Harrow? It was a de-streamlined engine, but had some "non de-streamlined" features after that e.g. footplate drop-ends.

The front drop plates were replaced by those from the non-streamlined types, but as far as I know, everything else followed the destreamlined types.

 

The running plate drop plates are often quoted as evidence that the front frames from 6202 were used complete to repair 6242, but a look at photos show that these were damaged on the left hand side, and destroyed on the right hand side.

[LMS2968]

Tolerances on steam loco0s were tighter than might be obvious; Horwich was using micrometers measuring in thous in the early 20th Century.

 

Alignment is vital. The cylinders had to be completely aligned with the driving wheelset nominal centres, especially in the transverse plane. Likewise, the wheelbase had to be accurately set up. A wheelbase of, say 7' 11 15/16" on one side and 8' 0 1/16" on the other will cause riding problems. Similarly, the distance between centres of two adjacent wheelsets and associated coupling rods has to match exactly. Assuming the above dimensions, there is a mismatch of 1/8", enough to rapidly wear out the coupling rod bushes.

 

It might not be appreciated that the greater the play between a bush and its journal, the greater the stresses, and the greater the wear rate. This is because the surface areas actually in contact are reduced, and stress is, by definition, force divided by area. Building in large tolerances would result in rapid run down and frequent works visits. This is also why an item will run happily for many miles, but once serious wear does start, it deteriorates very quickly.

 

[OFFTHE RAILS]

..........................Alignment is vital. The cylinders had to be completely aligned with the driving wheelset nominal centres....................

 

 

Not on Churchward's larger GWR locos. The 29xx Saints, the 28xx, 38xx Counties, 2221 County Tanks, 31xx/3150 prairies and 43xx classes were built with the cylinder centres 2.5" above the driving wheel centre 

Ian

[LMS2968]

Hence the comment, 'especially in the transverse plane'. The vertical alignment would alter due to movement of the loco on its springs.

[10000]

I can honestly say I did not see a single occurrence of that. Now, that may have been because I didn't think of checking they were the same - seeing either one of them was enough for me to note the number. Any examples, preferably with pictures?

 

Sorry no photos. The occasions when we noticed this was on shed walking between the rows of locos. From what I recall we only ever noticed this on Black 5's and Std 5's and it was 1966 the first time we saw this in Crewe south.

The reason I remember the year is that it was a friend who spotted it. I only met him in 1965 and we did not start going train spotting together until early 1966.