JimC

Running numbers and Lot numbers... How important to list these do you think. Lot numbers seem rather arcane, but are really rather important because changes were normally introduced with new lots. Running numbers are really rather less unimportant in development, but rather important if you are trying to work out what you are looking at. At the moment I still don't know if I like the format. Its rather cool to be able to page down, look at the sketches and see how the general style changed from domeless Gooch, early small dome Armstrong and when I get that far various dome positions under Dean, plus inside frames outside frames, sandwich etc. But on the other hand it feels as if the average entry could be something like: XXX Class 2-4-0 1866 [image] Lot ZZ, numbers 456-499 and 1024-1037. 14*24 cylinders and 5ft driving wheels. This was the first XXX Lot to have bushes rather than cottered bearings on the connecting rods. See p13 for earlier class members, page 22 for the next batch built and p45 for the 1878 renewals. or alternatively (see discussion below) For the previous Lot ZQ see P13, for the next Lot AAB see P22, and for the 1878 renewals see P45. Well maybe not quite as bare as that, but you get the point. But as Mike said with RCTS available again there's little point in going for RCTS levels of detail, but I think the consistent series of sketches has something to offer... I've also been playing with indices, having figured out a peculiar bug/feature in MS Word which was stopping me creating multiple indices. At the moment I'm playing with Illustrations, Classes, Year/Class, Lot, and CME/Class. It occurs to me, as a thought, whether rather than having all the class numbers in the text it would be good to have an index of class number. Ferocious bit of work, and you'd end up with multiple entries for some numbers, but it would be a useful thing to have even if some entries looked like 149-163 - page 75 157-164 - page 13 157-164 - page 35 159 & 159 - page 82 Easy to end up with a volume that had ten pages of inexes at the end though! What do folks think would be useful? Also what about cross references like the "see P 13" in the example. It seems to me vital that one can readily follow a class through this annal format which does mean lots of Xrefs. Having that based around Lot numbers might be helpful, as in the variant above.

Foks, its worth considering the two main 060T/pannier tank routes separately. The large ones had 4'6 (19thC) or 4'7.5 wheels and 7ft3+8ft3 wheelbase, and had this same wheelbase from the 633 class of 1871 right through to the 94s, although the overall length steadily increased from 29'10 of the 633s to 33'2 of the 94s. These included 645s, 655s/1741s and 2701s from Wolverhampton and 1813, 1851, 1701 and 2721 from Swindon, followed of course by 57s, 8750s, 9700s and 94s. All these may well be manageable with a single basic chassis design, although the upperworks varies considerably - side tanks, saddle tanks and pannier tanks over the years. The small ones were more disparate. Like the large ones they got longer over the years, but there were also different wheel sizes and wheelbases. 850s and 1901s had a short (7''4+6'4) wheelbase and 4' ->4'1.5 in wheels. 2021s were similar, but had a longer wheelbase of 7'4+7'4, as did all their successors, the 16s which were basically modernised 2021s, and the 54s 64s and 74s, which had larger wheels.

2-2-2 tank engines are very rare birds anyway. I think the Pembroke and Tenby ones were the only ones that came near GWR hands. Add a requirement for outside cylinders and you're getting close to rocking horse crap territory I fear. Funnily enough I knocked up a sketch of the P&MT2-2-2T a few weeks ago, and there are photos of some others in the blog post discussion here .

Yes indeed, I'm looking at developments as much as I can for the 19thC developments, injectors, cabs, all that sort of thing. If the idea comes off it should be possible to page through and see the design progress in successive illustrations. In an ideal world I'd combine my profile sketches with a good selection of photographs, but the trouble is I find the cost of library photographs quite prohibitive, and all the older stuff is with Locomotive Publishing Co collection at the NRM. I wonder at the economics of books full of photographs.

So at the moment I'm thinking that maybe I should mark boiler capacity in the "too difficult" column and omit it. After all it would need something quite extraordinary to be able to distinguish poor steam generators from good ones by dimensions.

Yes, I too have bitten nails in frustration over boiler diameters... I think realistically the amount of data you need to do a proper job for modellers is beyond my capabilities. Typically one of my profile sketches takes two or three days, depending on the level of detail I have in sources and the number of components I already have available. At one extreme I can do a fictional locomotive from the GWR parts kit in a couple of hours. At the other extreme if I have a hazy GA drawing with lots of detail but nearly impossible to read, and want a better level of detail than something weight diagram sourced it can be days. But if I was trying to do something for a modeller... (None of this will be news to Mike) Of course the first problem is sourcing data, You need elevations and probably sections. If the drawings are available they're probably in the NRM collection, and at £7.50 a scan it soon adds up - I have done something in the region of 200 sketches for the book, and if you were to budget 20 or 30 quid for NRM files for each then it starts to hurt! And then the amount of work in getting end elevations. Profile photos are relatively common and useable (and I can count the number of spokes!). When really desperate I've tried de-perspectiving (there'll be a word for that) 3/4 or at least 7/8 photos, but its hard work. Getting elevations from photos, well I find the thought quite intimidating. I would think its got to be weeks not days to produce a three view drawing. Anyway, yes, thanks for that, it all helps crystallise my thinking.

Boiler dimensions... The last book had a big table of leading dimensions for each class, Builder, Dates Built, Number Built, Route Colour, Power Class, Tractive Effort, Driving Wheel Size, Cylinder Dimensions, Wheelbase ,Front Overhang , Rear Overhang, Boiler Class, Dates Withdrawn (overhangs mainly/only 0-6-0 classes where it differentiates between classes and was often altered). Now all well and good, but the big tables distract from readability and I'm trying to give this a bit more flow, and also in general make it a smaller volume*. So I was thinking that just driving wheel size, cylinder size and some measure of boiler size in the text would be enough to get the picture. So I thought it would be a good idea to quote heating surface of the boilers as a measurement of real size and to illustrate capacity increase. I'm realising, though, that in the mid 19thC they were really cramming tubes into some boilers, to the extent that a new built 645 had a deal more heating surface than a 57xx, and very nearly as much as an unsuperheated 94xx, which was a far larger boiler. I've read, too, that these had the tubes crammed in too much to be optimal. So I guess I have to abandon that measure as being misleading. The question is what to substitute. The best measure of boiler capacity of course is to list the whole damn lot, physical size, heating surface, the lot but that's getting into the dense lump of numbers I was trying to avoid. I suppose barrel diameter and length and firebox heating surface might be the closest to a simplified measure, but that seems to be a nasty mix of measurements. Would barrel L & D plus firebox length be too simplistic? Or am I still getting over buried in numbers, and it would be better to skip the boiler data completely? But if I say OK, let's have a table at the end, effectively that's repeating the RCTS Boiler Appendices. and even if one simplifies its still a massive body of data and pretty damn indigestible. With RCTS now readily available on line at moderate cost it seems somewhat pointless to replicate a sub set of it. Any thoughts, suggestions? *which as I seem to be doing a fair number of extra sketches may be doomed to failure

The Birkenhead Railway was 'vested jointly' (which I presume means taken over and split up) in the GWR and the LNWR on 1st January 1860, and the GW and LNW got 21 of their locomotives each. Their main routes ran from Chester to Birkenhead and Chester to the LNWR near Warrington and became a joint line of the two companies.

The choice of inside valve gear on the GWR four cylinder locomotives had some advantages too. I don't think there is actually room for rockers behind the cylinders. A significant advantage of having the gear inside was that it gave much more flexibility in layout without the wheels to consider. If measurements and sketches I made a few years ago were correct then it would actually be impossible to have outside valve gear driving the inside cylinders of the Star and Castle without considerable changes to the locomotive layout and/or valve gear setup.

Here's a bit of fun, and probably some procrastination as well. This is the current contents of my 2-4-0 sketches directory. There are a good few more to come I reckon, which is rather intimidating. The main takeaway, of course is how all GWR locomotives look the same... I think at the moment I like the appearance of the 111 early best.

Yes AIUI the straight footplate on the 824 photo above was a Swindon rebuild. I seem to have given myself a project and a half to sort all these 2-4-0 variations out.

And now I've got to the 717s, and hey ho, what is there? That's a super clear profile photo, and will be a big help getting the frames because it doesn't seem as if they were much altered, but without the big Giffard injector on the side of the firebox, a polished dome, enclosed splashers and a cab that has got to be a good few years on. There's another 717 photo in RCTS, which must be a bit earlier as it has the Giffard injector, but its still cab and polished dome. It has both splashers open and a narrow cab, but RCTS is suggesting that originally they had a large side panel covering the trailing wheel splasher and only the leading one was open. I don't think I'm going to do much better with the 806 series either. There's a slightly earlier photo of one in RCTS, but we are now getting past the days of completely open locomotives, and I'm not clear whether the vestigial sides to the weatherboard as per photo D68 in RCTS and the photo of 442 above were from new or added just a bit later with these classes. At the moment its feeling as if each successive class is more hassle to sketch, and gosh I'm glad that I skipped over them in the previous book. Trouble is I don't think I can skip here in this format, its a significant era.

Well, I'm progressing through an initial draft of this format. I'm basically up to Swindon 1870. I think realistically I have to get well into the 1870s, maybe 1880s before I can make a judgement of whether I think the format works. It's going more slowly than I might have hoped, because I keep getting stopped because I don't have a sketch of a new class in original form. I'm hoping the drawing list will be more complete a few years on. At the moment I'm listing Swindon and Wolverhampton separately for 1864-1877 and 1877-1892. That may need to be rethought, because I've just got to the 1076s, and have to note that they appear strongly influenced by the 1016s at Wolverhampton, but at the moment Wolverhampton 1864 is after Swindon 1877 in page order. Another thought is that this might work better as an electronic document than paper. If one has hotlinks on each section to the previous and next entries for each class it would be a lot easier to read the whole history of a class link by link electronically than it would be on a paper equivalent. 'm also wondering if I should use lot numbers extensively in headings and indices. It's a very convenient way of separating successive builds of classes, and its also the way the GWR worked. I tended to stay clear of them in the last book as an unnecessary complication, but with this format it might well help readability and referenceing. Running by date certainly works in some respects. I'm picking up things that I hadn't really noticed before, introduction of injectors for instance, and the replacement of cottered bearings with plain bushes on the coupling rods. Also spotting things that seem to be missing. RCTS seems very quiet about the changeover from coke to coal, and that was a very big deal indeed at the time.

Not really, they are facsimiles of the originals - in other words they've scanned good copies of the books. They haven't gone back to the original photographs. The pdfs are searchable though which is quite handy.

After a little diversion into working out an as built Armstrong Goods, I've now got to the 481 class. I'll be glad when I get nearer the end of the 19thC and there are a few more images available. The "As built" 481 seems to be another with one half reasonable photograph in RCTS. I've just bought the electronic version of the RCTS books, and I don't reckon I can scan their images any better than they have, so I can use those and save a bit of time. Need to make another trip to Leatherhead and see if there's anything better in "The Locomotive". Also figure out a way of getting the pages flatter when photographing without any risk to the books. I guess I can experiment here on any old book!

Fairbairn originally, according to RCTS, but that must be post 1889, by which time it had not only had near enough everything changed above footplate but most of what was below too!

Reckon this is about as good as I can do on the Bicycle for now without going mad(der). Been a bit of a struggle this one. Think I'm going to give myself a break before tackling the next.

Yes indeed, but consider how many tenders arrived with absorbed lines, and whilst the locomotives might have been problematic, one supposes tenders are easier to get right. Also I suspect that once you've got standard wheel sets, brakes and axle boxes on there was probably less value in standardising the rest. More photos are definitely a good thing, keep it up please. I can't see us getting buried in 19thC 2-4-0 photos!

Yes indeed, those are the Swindon renewals of 1887-9, and nice useful photos they are too. But the wheelbase was quite different so there aren't too many cues. Trivia: 17 481s went into the Swindon factory to be renewed, and 18 rebuilt 481s came out. The other 3 481s went into Wolverhampton, one came out with new frames of a different design, one was relatively little changed, and one came out as a renewed 439, which was a rather larger locomotive than the 481!

When I wrote the first book I was rather guilty of somewhat glossing over the 2-4-0s in the Armstrong and Dean eras. There were so many of them, they were rebuilt so much and I just found them confusing and, dare I say it, not that interesting. I'm paying for it now! Working up my experimental chronologically based GWR locomotive history I'm in into the late 1860s, early 1870s, and they are becoming impossible to avoid! I have to wonder, incidentally, why, with standard goods engines and standard tank engines in numbers there were so many different ones. The old Gooch era 149s built by Englands and the Joseph Armstrong's 111 class from Wolverhampton were adequately documented, but then... Next was the 439 class. Intriguing beats, because they were an early Joseph Armstrong product at Swindon, and looked almost exactly like broad gauge engines. When they were rebuilt/renewed later just about everything was changed, so there are no clues there. So what do we have? Russell has nothing. No drawings at GWS or NRM. Ahrons in British steam has nothing I can see. RCTS has one rather unclear photograph and a bare minimum of dimensions. There was a thread here some years ago, but even @MikeOxon doesn't seem to have found much other than a slightly better version of the same photograph. So I wondered about Ahrons original article in "The Locomotive". You may be aware that Ahrons wrote a whole series of articles on early GWR locomotives for the Locomotive, typically illustrated with his simple line drawings, which have been widely reproduced, notably in Holcroft's books and his own "British Steam Locomotive", and which I have made wide use of for my drawings. I discovered, to my surprise and delight, that the RCTS archives are at Leatherhead station, just a few miles from where I live, and they have a complete run of"The Locomotive". So I joined up and yesterday spent a useful but very chilly couple of hours perusing the bound issues. And yes, Ahrons does cover the 439s in the magazine issue. But the article was written a few years after the previous one, and he is eschewing his line drawings for photographs, which for the 1870s are presumably increasingly available. And I turned to it and: Yes, its the same photograph, although the reproduction is better so it's a lot clearer. It's a nice profile at least, I could do a hell of a lot worse. Now this morning I've come to the 481s, which were the next batch at Swindon. Very much the same dimensions, but visually quite unlike. And another I happily glossed over in the book. And what do I find in RCTS? A similarly tiny profile photo. In Russell - only the renewals, again rather different, and in Holcroft little enough too. So, slightly discouraged, I'm writing this blog post to let off steam! Really I don't think there's much of a way round it, I need to produce something. Perhaps I should make them plain line outlines and much more diagramattic than my usual ones to make it plain they are, well, rather sketchy sketches!

I'm trying to find photos or drawings of the earlier 388s in as built condition, and not doing very well. A couple of rather indistinct photos in RCTS and not much more. I'm thinking particularly of the first three lots which started life with small domes, large safety valve covers and no cabs, but also the 1876 lot with side sheets on the spectacle plate. Anything with a cab is definitely too late. Amy suggestions?

There wasn't a 2-4-0 with as big a boiler as the Dean singles I don't think. Only the Armstrong 4-4-0s. Mind you, bearing in mind my limited technical knowledge, my understanding is coupling rods aren't the problem with balance. They go round and round in nice circles (well not really circles, but you know what I mean) and you just hang a weight opposite. I think its connecting rods that get all complicated. But correct me if wrong.

And I quite agree with you about the 3000s looking lumpish. Here's my sketches of an original Dean and an original Sir Daniel side by side.

Mike, you are welcome to have a vector version of any of my sketches if it would help you.

RCTS states "Up to 1865 the general practice was to use crosshead driven pumps... At that date the Giffard injector (invented in 1859) was introduced on the GWR.". Which begs the question, what classes, were they rapidly retrofitted etc etc. Does anyone know any more? There's a photo in RCTS (Part 4 D113) of a 322 (Beyer) no 334 "as built by Beyer Peacock in 1864" which would appear to have an injector fitted. Similarly D119 shows a 360 class with injector, but the caption makes it clear the photo is not quite in as built condition. An Ahrons drawing in Holcroft's "Armstrongs" (p66) shows 361 in very early condition with the injector, but drawings of the slightly earlier 110 and 111 do not. So from this very sketchy information one might surmise that injectors came to the GWR with the Beyer, and started to be fitted to Swindon products with the 360 class. Anyone know if that's accurate? The trouble is the RCTS photo could just as easily be an experimental fitting, or, if they were rapidly introduced, simply very early installations to existing locomotives. Thoughts?