JimC

There's more inside at least the GW tanks than one might expect. This page is mainly about promoting the tech the company used to repair the tank, but there are excellent photos of the interior. The MR ones might have done without all the internal bracing, I don't know. https://www.corrolesseastern.co.uk/tank-lining-and-protective-coating-didcot/

Didcot is certainly covered.

That's a very good point, it hadn't occurred to me, but I think you're right, it was the classes with drum head smoke boxes that had supports to the footplate at the front end. Hence 94 and 15 . The usual GWR exception seems to be the 1366, which had brackets to the running plate but not a drum head box. I can't think of any drum head boxes with the support brackets though.

According to Ken Cook, GWR Works manager in the 40s, his predecessor, Hannington, was known to swim in a water tank from time to time. Apparently he also took a cold bath each morning in a pond in his garden, breaking the ice if necessary! In the end his love of water was his undoing as one day he dived head first into a 5 foot deep swimming pool at his daughter's school and fatally hit his head on the bottom.

I suppose the trouble was that whilst nearly everyone in the drawing office would have worked on inside motion during their apprenticeship, they would have done so as skinny teenagers...

I don't believe so. They are drawn on none of the saddle tank diagrams I just looked at and all of the pannier tank ones. Having the brackets resting on the frames rather than riveted to the firebox is probably a better solution bearing in mind heat expansion of the boiler. 57xx, 8750 and 1366 show a smokebox bracket, 54xx, 94xx, 15xx etc show both ends supported from the frames.

GWR locomotive committee minutes are full of agreements being made about sourcing water. Canals, streams, reservoirs, bore holes, landowners, water boards, other companies, they are there in enormous variety. It was clearly a significant preoccupation at board level.

Thanks for that (I think) Id forgotten that Vulcan list. Lot of work to do presenting some of them. I'd love to know the economics of these reconstructions. Going back to mid 19thC the tools and facilities were so different from how we'd do stuff today I have to wonder whether they found it economic to alter stuff we'd think crazy to reuse. I recall being gobsmacked by Holcroft's description of cutting the holes in a saddle tank for handrail pillars *by hand with a chisel* as late as the turn of the century because they had no powered tools that would function on a curved surface.

962 was an 1874 built 727 class, but by this stage 1076, 727 and 1134 classes had merged. According to RCTS it was withdrawn Mar 1931, having only received pannier tanks Dec 1927, quite late. 100 wagon coal trains had been run, so unless there were a load of long flats, macaws or something further back longest train seems unlikely.

Coming back to these, I've just started working up a sketch of the 1866 incarnation of numbers 34 and 35. These numbers seem to have been reserved for oddities! The original 34 & 35 were a pair of locomotives built by the Vulcan Foundry which the Shrewsbury and Chester Railway bought off the shelf in 1853, and one may suspect at a bargain price. They could be described as long boiler 0-4-0 tender engines, but the drive was not to either wheel axle, but to an intermediate crank axle, somewhat in the position that the middle driving axle of a long boiler 0-6-0 would be. In 1866 Joseph Armstrong took these weird contraptions in hand and reconstructed them. They reappeared as long boiler 0-6-0s, the only ones of this configuration to be built by the GWR, although a fair number of others were taken over in the early days. I'll probably do a separate blog page for them when the sketch is done.

Three small side tank locomotives from Sharp Stewart, delivered in 1866. They received a significant reworking in Cambrian days which altered their appearance considerably. All survived to the GWR, who proposed to scrap two of them immediately, but they were reprieved and numbered 1192, 1196 and 1197. They soon received the full GWR treatment above the footplate. The boilers were thoroughly overhauled with top feed added and they were given new GWR smokeboxes, tanks, cab and bunkers. Thus utterly transformed they resumed work, normally on the very weight-restricted Tanat Valley line. One was scrapped in 1929 but the other two soldiered on until 1948, becoming British Railways locomotives and running over a million miles each. A difficult one to draw, because they all had minor variations and the GWR changed the bunkers twice. Lets not be too precious as to whether every feature in the sketch was on any one locomotive at the same time, but this is intended to show the second bunker enlargement. This drawing is very much influenced by the 7mm drawings in Welsh Railway Records Vol 4 - Cambrian Railways Drawings Part 1 : 1853-1892, which has just been published and is much recommended. I've also grabbed photos and so on. I find the safety valve cover rather unconvincing in some drawings and models. Maybe this one is better! This little class has a somewhat higher profile than one might expect due to being the subject of an old school white metal kit in 4mm, and an etched brass one in 7mm. Also, I suspect because they were notably cute in GWR form and more photographed than one might expect.

In Cook's 'Swindon Steam' he states that maximum allowable flange depth was 1 1/2 inches. Also that under Collett the GWR increased the minimum depth to 1 5/16 from 1 1/8 which was standard across the big 4.

Yes indeed, but flange wear was a major problem too. Especially on tightly curved lines leading and trailing flanges would wear thin and a horrendous amount of tread depth needs to be removed to get flanges back to required thickness. But for your purpose you can assume flange depth to be pretty much constant.

It would be interesting to know more about how the diagram books were used operationally. Even more interesting to know more about how locomotives were allocated to sheds. The shed master has to work the diagrams he has with the locomotives he has, and juggle with his allocation when necessary, accept non ideal choices when he has to etc. The key decisions must surely have been in allocating locomotives to sheds. Was it imposed on the running shed, or did the shed master have influence in what he got? The running side certainly fed back to the factory what they felt they needed.

The GWR issued different engine diagrams for different tender types.

Sadly we don't have the Churchward diaries to read and find out his thinking, but my guess is that firstly it was a truer comparison, and secondly if atlantic had been preferred they wouldn't have to make changes.

Cox strikes me as being much less than a dispassionate observer. It's true of anyone who was personally involved in something of course, we all have our own viewpoint of what went on around us, but it seems to me especially striking with Cox' writings.

Probably nothing. Aluminium was very expensive in those days. The GWR did use nickel steel for boiler barrels on some 2-6-2Ts (in the 1930s) where there was a desire to increase working pressure without increasing weight. I believe it was also used on the 280psi boilers for the Counties. It doesn't seem to have been used on anything else, only where there was an especial need to reduce weight. Presumably the Manors, for instance, could have had a slightly larger boiler with the use of nickel steel, but to the best of my knowledge they used standard steel like other classes.

The King boiler was 250psi. What you are suggesting is so left field that it's hard for me to comment without going way beyond any credibility I might have.

On the GWR at least 2-6-0s and 4-4-0s were pretty much the same size and 4-6-0s and 2-8-0s also much the same, but a size bigger boiler. Cylinder/motion/wheel clearance was definitely a design issue.

The early GWR system certainly worked like that. Renewals, which were replacements for existing locomotives containing at least 0% (!) of the parts of their predecessors, were given the numbers of the locomotives they replaced, but additions to the fleet were added on at the end of the list.

But more often without an article at all, which is even wierder.

Very messy as Dean had to introduce a second block and small wheels (eg 850/1901/2021) were separate from larger wheels. So there are 0-6-0Ts under 1300 in the sequential series, 1501-1900, 1901-2160 (small wheel) and 2701-2800 in Dean's blocks, and then the later classes under Collett. And note how Hawksworth reused 15xx and 16xx, but starting at 0 not 1. I worked this out in some detail in my book (see sig - shameless plug!), which has over 2,800 words on GWR numbering! Its all in RCTS in one place or another, but it was quite a bit of work to try and bring it all together in a (hopefully) logical form.

More complicated than that! At one stage post grouping the GWR had at least four different numbering schemes in use concurrently! In Gooch and Armstrong days locomotives were pretty much numbered sequentially, with complications for renewals. Most of these were never renumbered. Dean introduced blocks for wheel arrangement, but with no attempt to separate classes, instead each lot followed on sequentially. These blocks started at one eg 3201 - 3500 was passenger 4 coupled, and started like this: Numbers Class 3201-3205 Stella (2-4-0) 3206-3225 Barnum (2-4-0) 3226-3231 3226 (2-4-0) (built 6 of these which complicated things!) 3232-3241 3232 (2-4-0) 3242-3251 3232 (2-4-0) 3252-3253 Duke (4-4-0) 3254-3261 Duke (4-4-0) 3262-3272 Duke (4-4-0) 3272-3291 Duke (4-4-0) 3292-3311 Badminton (large wheel 4-4-0) Some of these (mostly 4-4-0s) were renumbered to bring classes together in 1912 Churchward we know about (blocks starting 0, second digit indicating wheel arrangement, classes kept together where possible ) And then the absorbed classes at the grouping were ordered in wheel arrangement and power reusing available numbers below 2000.The arrangement by power was particularly odd because it split up locomotives of the same class which had derated boilers, and that got even odder when they were reboilered with GW boilers and the numbers not changed.

Or just twenty eights. There don't seem to have been hard and fast rules. A variation I've seen in some WR documentation at least is to refer to a class by the most recent number range, eg Castles as 70xx, 57s as 36xx, 28s as 38xx.