Our new ticketing site is now live! Using either this or the original site (both powered by TrainSplit) helps support the running of the forum with every ticket purchase! Find out more and ask any questions/give us feedback in this thread!
As per title, had BR adopted a TOPS/Austrian/German style numbering system, formed class+subclass+individual loco numbers system in 1948, what might it have looked like? What thoughts do people have?
Could BR have got away with two digit class numbers to cover all types? Or would three have been needed? My thought is maybe the later, with tender types two digits, then 1xx for tank locos, and higher ~xx series for diesel, electric, etc. types
As per title, had BR adopted a TOPS/Austrian/German style numbering system, formed class+subclass+individual loco numbers system in 1948, what might it have looked like? What thoughts do people have?
Could BR have got away with two digit class numbers to cover all types? Or would three have been needed? My thought is maybe the later, with tender types two digits, then 1xx for tank locos, and higher ~xx series for diesel, electric, etc. types
I would expect that - at least to some extent - not too dissimilar classes (same power class, wheel arrangement, origin,...) would have been grouped as sub-classes to one main class. To some extent dependant on operators and engineers agreeing on 'similar'. And whether the individual loco parts of the number would have been limited to three digits.
The LNER was sort of partly there in 1946, dividing tender and tank locomotives and then further subdivisions, so a scheme could have followed on. The GWR had a sort of scheme but they were obsessed by using the second figure of a four figure number so they got into a bit of a pickle with the 57xx series pannier tanks and the Castle and Hall classes. The LMS tried numbering the Black 5s in the 5xxx series and then clashed with the Jubilee (5XP) 55xx series so had to fiddle about when building further locomotives. The 8Fs went into the 8xxx series but then the Princess/Coronation Pacifics, although 8P, ended up in the (4)62xx series reflecting their wheel arrangement. This and other schemes and new builds required all sorts of spot renumbering.
Once you start trying to reconcile all the numbers of all the locomotives, you soon start to see why the post-nationalisation renumbering was done the way it was. There would be so much messing about and time-wasting with trying to cram six-figure serial numbers on little tank engines with no real benefit to man nor beast. And then the GWR had all those number plates. And don't forget that Grouping had only occurred 25 years earlier, so there would be plenty of stuff and staff left over from that event.
As per title, had BR adopted a TOPS/Austrian/German style numbering system, formed class+subclass+individual loco numbers system in 1948, what might it have looked like? What thoughts do people have?
Could BR have got away with two digit class numbers to cover all types? Or would three have been needed? My thought is maybe the later, with tender types two digits, then 1xx for tank locos, and higher ~xx series for diesel, electric, etc. types
Why would they have done this? What would the value have been in a pre-computer era, and what would the other costs have been to recategorise all of the records?
As it happens, BR did adopt a very simple classification along LMS lines - power and purpose (e.g. 4MT) for sort, and then wheel arrangements for differentiation.
I did once, as a paper exercise during Lockdown, look through "Longworth" to see how such a system might work.
BR inherited 20,076 steam locomotives, and built or acquired another 3,104 between 1948 and 1960. The definition of a "class" varied from one company to another, but Longworth lists
112 distinct classes from the GWR
91 from the SR
121 from the LMS
169 from the LNER
14 BR standard and WD types
Total 492 different classes
I can't find my working now, but I think it involved a three character prefix
A number indicating the power classification, following LMS practice
A letter indicating the wheel arrangement, following LNER practice
Another number distinguishing one class from another
Followed by a three digit individuial identifier
So Flying Scotsman might become 7A3.069, and Caerophilly Castle 6B1.001
I soon gave up, as there were far too many different 0-6-0 types in the 3F and 4F power ranges to give each a unique 3Jx or 4Jx classification
You might be able to have a 3-digit class where the first digit was the power classification, or possibly by wheel arrangment - the commonest was 0-6-0, of which BR inherited nearly 7,500, but again there were too many distinct 0-6-0 classes there to give each a separate xx (2-digit) code - the LNER alone had 43 different 0-6-0 classes.
== Doublepost prevention - post automatically merged: ==
In a similar vein, what would have happened if BR had not adoted the D/E system inn 1957 but continued to number diesel locomotives in the 1xxxx range. The new scheme relased very few extra numbers (and then only because pre-nationalisation designs retained their 1xxxx numbers), so everything should fit!
With the 10xxx, 11xxx, 12xxx 13xxx 15xxx and 18xxx ranges already partuially occupied, perhaps the pilot scheme designs would have been
Type 1 14xxx
Type 2 16xxx and 17xxx
Type 4 19xxx
SR electrics from 20100 and 21000 (electro diesels)
AC electrics from 23000
Later developments
The standard shunters, on reaching 13999, continuing down from 12999
The later Type 2s (D7500-7677 in our reality) numbered downwards from 15999 (to 15822)
The "Teddy Bears" (D9500) from 15300 (or possibly 14800)
Type 3s from 18000 (the gas Turbine locos both having been withdrawn by 1960)
Deltics in the 10xxx range (maybe from 10101)
Big Type 4s also in the 10xxx range (by 1964 only 10001 was the only number still in use in that range)
Why would they have done this? What would the value have been in a pre-computer era, and what would the other costs have been to recategorise all of the records?
As it happens, BR did adopt a very simple classification along LMS lines - power and purpose (e.g. 4MT) for sort, and then wheel arrangements for differentiation.
The advantage would have been a (relatively) simple system of knowing what was what. And renumberings (of differing scopes) were not uncommon - the effort was manageble.
While BR's power classification was fine for its purpose, when you wanted finer distinction, distinguishing one class from another - that might be necessary, even if they were similar - could be complicated.
BR inherited 20,076 steam locomotives, and built or acquired another 3,104 between 1948 and 1960. The definition of a "class" varied from one company to another, but Longworth lists
112 distinct classes from the GWR
91 from the SR
121 from the LMS
169 from the LNER
14 BR standard and WD types
One of the biggest issues would have been 'what is a class (or sub-class)', and deciding how to group or divide them. Longworth's 492 would be a starting point, but I suspect only a starting point...
The advantage would have been a (relatively) simple system of knowing what was what. And renumberings (of differing scopes) were not uncommon - the effort was manageble.
For small scale renumberings, I can see. But for large scale wholesale renumberings - what would the economic value have been for commercial enterprises?
While BR's power classification was fine for its purpose, when you wanted finer distinction, distinguishing one class from another - that might be necessary, even if they were similar - could be complicated.
The GWR had a simple scheme, locos were classified by colour, red the heaviest, generally the express locos, then down through blue, yellow, and no colour. Just four groups and a coloured spot was on the loco cabsides. Then routes were classified similarly, there were coloured maps all around the system, in every shed foreman's office, etc, showing the colour of each section of line. You couldn't run a loco of greater colour on a line. Simple, worked well, even lasted to the end of WR steam and onto early diesels. You can see this in any (colour) picture of a GWR/WR loco.
The GWR had a simple scheme, locos were classified by colour, red the heaviest, generally the express locos, then down through blue, yellow, and no colour.
They also used letters in the same coloured circles as the weight restriction to denote power class arranged by notional tractive effort in lbf - from 'no letter' through A, B, C, D, E to 'special', which was over 38,000lbf and thus only applied to Kings.
By the mid 1960s the UIC (International Union of Railways) had already developed a numbering system applied first to wagons and coaches, and then locomotives. The arrival of TOPS in the UK should (in my view) have coincided with BR adopting UIC numbering conventions. Instead we ploughed our own lonely furrow.
By the mid 1960s the UIC (International Union of Railways) had already developed a numbering system applied first to wagons and coaches, and then locomotives. The arrival of TOPS in the UK should (in my view) have coincided with BR adopting UIC numbering conventions. Instead we ploughed our own lonely furrow.
Surely even now the format and meaning of the “actual locomotive number” (positions 5-11) of the UIC number is determined by the national authority, not the UIC.
The GWR had a simple scheme, locos were classified by colour, red the heaviest, generally the express locos, then down through blue, yellow, and no colour. Just four groups and a coloured spot was on the loco cabsides. Then routes were classified similarly, there were coloured maps all around the system, in every shed foreman's office, etc, showing the colour of each section of line. You couldn't run a loco of greater colour on a line. Simple, worked well, even lasted to the end of WR steam and onto early diesels. You can see this in any (colour) picture of a GWR/WR loco.
Erm, that classification was about weight and what was permitted to work where - there was more to (G)WR classifications than that. It's just one facet of the many features that a railway needs to consider
There were copious route restriction books.
I have an old Souther Region one, and it lists every section of line on the SR and which loco classes are permitted. It considers every class then on the Souther, so some of the availabilities are a bit academic. Presumably done theoretically in an office somewhere, rather than sending a hapless R1 (for instance) on a jolly around the system.
As to how someone would know what an unfamiliar loco was - locos rarely strayed from their home depot, and the shedmaster would know what each of his stable was capable of (which depended on overhaul history and other individual features as well as actual class)
Classes seem to have been defined differently by different companies.
The LNER adapted the GNR system of allocating a letter to each wheel arrangement in use (A for 4-6-2, B for 4-6-0 etc to Z for 0-4-2) and giving each individual class a number, eg A4, J94. There may have been a system to the allocation of letters, but over time the introduction of new classes with novel wheel arrangments had to be fitted in wherever there was a spare letter.
The GWR used the number of the first example (not always the lowest number) eg 5700 (commonly seen as 57xx) although many classes also had names, eg the 49xx "Hall" class.
The SR seems to have simply perpetuated the classification systems of its constituents, each of which used an alphanumeric system; as I understand it the LSWR went A1, B1, C1 etc Z1, A2, B2.......... in order of introduction. There was some duplication - for example there were ex-LSWR and ex-LBSCR B4 classes (0-4-0T and 4-4-0 respectively). Conversely there was no physical difference between the Southern's BB and WC classes other than the themes of the names applied to them (Battle of Britain and West Country respectively)
The LMS, following Midland practice, classified anything within a given tractive effort range as the same class - although because of the Midland's "small engine" policy the number of classes had to be exapnded tfrom 4 to 7 to accomodate certain LNWR and LYR classes, and eventually to 8 (and even 9 in BR days)
Indeed, the Midland's "small engine" policy led to many nonsenses. The idea was that, for a given weight of train on a given route, a particular class (ie power rating) of locomotive was required. The class specified was very conservatively defined, as the policy was that any loco of that class could do the job, whetgher it was due for an overhaul or fresh from one. If the train was just a few tons over the limit for a Class 3, a Class 4 had to be found, or the train had to be double headed. No matter that the Class 3 was fresh from overhaul and more than capable of keeping to time.
(It qalso was an indirect cause of the Hawes Junction disaster - had the Midland had more powerful locomotives, fewer trains would have needed pilots over the "Long Drag" and the signalman at Hawes Junction, near the summit, would not have been so overworked with the extra movements caused by detaching pilots and sending tham back "light engine" down the hill, that he wouldn't have overlooked the light locomotives involved in the collision (if only because they wouldn't have been needed in the first place)
Surely even now the format and meaning of the “actual locomotive number” (positions 5-11) of the UIC number is determined by the national authority, not the UIC.
By the mid 1960s the UIC (International Union of Railways) had already developed a numbering system applied first to wagons and coaches, and then locomotives. The arrival of TOPS in the UK should (in my view) have coincided with BR adopting UIC numbering conventions. Instead we ploughed our own lonely furrow.
The UIC numbering system for wagons and coaches was based around a relatively standardised set of types, widely interchanged across continental Europe (which is where it was used) - beyond that area (into Britain, or the Soviet Union), where very little was interchanged with, and pretty well everything was very non standard relative to, continental European systems, it didn't make much sense - and in those days BR was far more separate from the European network than it is today, so BR doing its own thing was understandable.
True enough. I just think it would have made more sense to move directly to the UIC system in full, rather than the semi-compliant system we have now.
As I see it, the basic reason for our semi-compliant numbering system is that the railway system it applies to is semi-compliant, relative to the European norm.
== Doublepost prevention - post automatically merged: ==
The RA system was introduced on the LNER (initially) during the war to (at least partially) simplify this, with more 'unfamiliar' locos turning up. But while this may have resolved weight questions, there were also ones over loading gauge.
As to how someone would know what an unfamiliar loco was - locos rarely strayed from their home depot, and the shedmaster would know what each of his stable was capable of (which depended on overhaul history and other individual features as well as actual class)
Every so often locos would stray - and there were periodic reports of locos turning up where they shouldn't have, and a lot of head scratching to work out how they could be sent 'home'.
Another thing to remember is that in those days, loco crews were expected to work whatever turned up, even if they'd never seen anything like it before.
Classes seem to have been defined differently by different companies.
The LNER adapted the GNR system of allocating a letter to each wheel arrangement in use (A for 4-6-2, B for 4-6-0 etc to Z for 0-4-2) and giving each individual class a number, eg A4, J94. There may have been a system to the allocation of letters, but over time the introduction of new classes with novel wheel arrangments had to be fitted in wherever there was a spare letter.
When Ivatt introduced the system on the GNR, there was a system (see in, e.g., the relevant part of the RCTS history of GNR locos), but developments overwhelmed this (but I do wonder why the LNER changed some codes that the GNR used.
The GWR used the number of the first example (not always the lowest number) eg 5700 (commonly seen as 57xx) although many classes also had names, eg the 49xx "Hall" class.
The SR seems to have simply perpetuated the classification systems of its constituents, each of which used an alphanumeric system; as I understand it the LSWR went A1, B1, C1 etc Z1, A2, B2.......... in order of introduction. There was some duplication - for example there were ex-LSWR and ex-LBSCR B4 classes (0-4-0T and 4-4-0 respectively). Conversely there was no physical difference between the Southern's BB and WC classes other than the themes of the names applied to them (Battle of Britain and West Country respectively)
The LMS, following Midland practice, classified anything within a given tractive effort range as the same class - although because of the Midland's "small engine" policy the number of classes had to be exapnded tfrom 4 to 7 to accomodate certain LNWR and LYR classes, and eventually to 8 (and even 9 in BR days)
Any 'TOPS' style class + number system would have necessitated a completely new set of class ids - and possibly careful consideration of what 'class' meant, and how they should be differentiated.
When Ivatt introduced the system on the GNR, there was a system (see in, e.g., the relevant part of the RCTS history of GNR locos), but developments overwhelmed this (but I do wonder why the LNER changed some codes that the GNR used.
The GNR 1900 system originally allocated the letters A to J (except I) in
ascending order of number of driving axles, then descending order of leading axles, then ascending order of trailing axles. Other wheel arrangements were added as required, although K and L happened to fit the original pattern Thus
A 4-2-2
B 2-2-2
C 4-4-2
D 4-4-0
E 2-4-0
F 0-4-2
G 0-4-4
H 2-6-0
j 0-6-0
K 0-8-0
L 0-8-2
M railmotor
N 0-6-2
O 2-8-0
In 1922, the GNR built its first 4-6-2, and allocated classification A1, which by then was spare
The numbers were initially allocated in descending order of driving wheel size, with new classes added to the end of the list.
By 1922, the A, B, F, and M ranges were obsolete
The LNER adopted a similar system, but had to accomodate several wheel arrangements used by other constituents. A distinction was made between passenger engines (A to H) and goods engines (J to T) and miscellaneous (X to Z), and in the former arranged them in descending oder of driving axles, allowing the A1s to remain where they were, and using the now-vacant B range for the large number of 4-6-0s inherited from the NER, GCR and GER.
Changes were
B 4-6-0
F 2-4-2
H 4-4-4
K 2-6-0 (was H under the GNR system)
L 2-6-4
M 0-6-4
P 2-8-2
Q 0-8-0 (was K)
R 0-8-2 (was L)
S 0-8-4
T 4-8-0
X 4-2-2 (A having been re-used) and 2-2-4
Y 0-4-0
Z 0-4-2
letters U, V and W were later used for Garret, 2-6-2 and 4-6-4 arrangements respectively.
Class numbers were allocated
- Tender engines first, then tank engines
- ex GNR, then GCR, GER, NER, NBR, GNSR
- descending order of driving wheel size
- oldest first
New classes filled gaps left by withdrawn classes if any were available. Hence Thompson's A2 class of 1944 - a classification previously allocated to the NER Raven-designed Pacifics
New classes filled gaps left by withdrawn classes if any were available. Hence Thompson's A2 class of 1944 - a classification previously allocated to the NER Raven-designed Pacifics
Did A1 and B1 also fill gaps? Weren't A3s A10s before rebuilding.?
What were A1 and B1 in the 1920s and 1930s?
On the GWR did the second number have significance? So the express 4-6-0s were mostly x0xx, the 2-6-2 tanks were x1xx whilst many Pannier tanks were x4xx etc.
But what “UIC system” do you mean. As far as I am aware, only Austria had 4digit class number with three digit serial number, which is what is effectively needed to fit the long UIC system.
But what “UIC system” do you mean. As far as I am aware, only Austria had 4digit class number with three digit serial number, which is what is effectively needed to fit the long UIC system.
UIC 438-3.
Unfortunately to access the full leaflet you need to pay to download it.
The system is sufficiently flexible to accommodate railways with 2-digit class numbers (e.g. BR), or 3-digit (e.g. DB), or 4-digit (e.g. OBB).
Rewinding 50 years, renumbering the BR fleet and the introduction of a new stock control system (TOPS) could have been smoother and simpler if co-ordinated with fellow European railways going through the same process at the same time, rather than re-inventing an imported American system.
Did A1 and B1 also fill gaps? Weren't A3s A10s before rebuilding.?
What were A1 and B1 in the 1920s and 1930s?
On the GWR did the second number have significance? So the express 4-6-0s were mostly x0xx, the 2-6-2 tanks were x1xx whilst many Pannier tanks were x4xx etc.
No, A1 and B1 did not fill gaps - the existing classes were renumbered A10 and B18 respectively. The same thing happened with 01 and L1 - the existing classes became 03 and L3 respectively.
The GWR eventually used the second number for blocks of similar types. There weren’t exclusively reserved so a new type could start to “take over” a set of blocks used by another type or types such as the use of vacant x6xx by pannier tanks that also used many blocks in the x7xx series. The x6xx series was also used by GWR 0-6-2T locos.
Withe the discussion about steam & diesel loco number allocation perhaps I can air a theory that might just be true? Perhaps someone has inside knowledge.
The GWR, on the eve of Nationalisation, ordered 200 pannier tanks, the 9400 series. These were big & powerful for a tank engine, and intended not just for shunting but also trip workings in the Valleys etc.
When dieselisation came the WR ordered locos that performed exactly the same purpose as the 9400s, and they were numbered in the (D)9500s, obviously 100 up. In the diesel series this had no logic (the other D9000s were the Type 5 Deltics of course) but just possibly it was to reflect their iheritage.
Another irony is that the last steam 9400s weren't delivered until 1956 and some had very short lives - 4 years. The life of the D9500s, at least on BR, was very similar...
The LMS, following Midland practice, classified anything within a given tractive effort range as the same class - although because of the Midland's "small engine" policy the number of classes had to be exapnded tfrom 4 to 7 to accomodate certain LNWR and LYR classes, and eventually to 8 (and even 9 in BR days)
The LMS did allow for differentiation between tractive effort ranges - I have a scan of a 1953 Eastern Region route availability book which lists the route availability of a large number of ex-LMS classes, and it shows how locomotives of the same "Load Class", which may have differed in their route availability, were differentiated. In addition to the load class, also listed are the wheel arrangement, whether it is a tender or tank locomotive, original company ("Standard" is used for locos of LMS design), number of cylinders, and whether it has a tapered or parallel boiler (not every loco has all of these listed). So for example you have "2-6-4, Tank (Standard), 2 cyls., TB, 4MT".
But what “UIC system” do you mean. As far as I am aware, only Austria had 4digit class number with three digit serial number, which is what is effectively needed to fit the long UIC system.
The wikipedia page has a reasonably comprehensive summary. Within the 'National Block' there is space reserved for a 4-digit class number and 3-digit serial number, so TOPS numbers can easily be integrated into the UIC system by adding 1 or 2 leading zeros.
Rewinding 50 years, renumbering the BR fleet and the introduction of a new stock control system (TOPS) could have been smoother and simpler if co-ordinated with fellow European railways going through the same process at the same time, rather than re-inventing an imported American system.
Is there any reason TOPS could not have coped with simply allocating UIC-compliant numbers? Would it have coped with UIC-compliant wagon and coach markings as well?
The wikipedia page has a reasonably comprehensive summary. Within the 'National Block' there is space reserved for a 4-digit class number and 3-digit serial number, so TOPS numbers can easily be integrated into the UIC system by adding 1 or 2 leading zeros.
Indeed, Eurostar seem to manage perfectly fine between the two systems.
Is there any reason TOPS could not have coped with simply allocating UIC-compliant numbers? Would it have coped with UIC-compliant wagon and coach markings as well?
TOPS copes perfectly well with 10-digit UIC wagon numbers, so presumably (with suitable programming) could be made compatible with all UIC rolling stock numbers.
The point I was making was that BR seemed intent on inventing it's own independent system, rather than co-operate with other European railways to produce a common system.
They filled gaps, in some cases created for them by reclassifying the original classes.
The original A1s were the first Gresley Pacifics, most of which were eventually rebuilt as A3s. A handful were still in original condition in 1945 and were reclassified A10 to make way for a new class A1 (apart from the first of the Class, which was rebuilt as the first of the new A1 class - the rest of the new class were new build)
A2 did fill a gap, left by the short lived ex-NER Pacifics withdrawn in 1937
The original B1s and B2s were Great Central designs. recalssified B18 and B19 in 1943 to make way for the Thompson B1s and B2s, and all withdrawn in 1947
The orignal b2s
We are aware of an issue with emails from the Forum to Microsoft-based email accounts (hotmail/outlook/live.com email addresses). This is being looked into currently, thanks for your patience meanwhile.
