JimC

Quite agree. But then even Webb just tried it as an experiment and never built another. It started life as a 2-4-0, so presumably the [grief, cant think of the word, not crank] place where the rod was connected is under the valance.

There's friction and friction. As a general rule friction is purely parasitic and best avoided. But what's going on here is effectively a clutch using friction in much the same principle as a clutch on an internal combustion engine. So provided the clutch is working fully effectively with no slippage there isn't any more loss than on a motor car clutch. That's the good news. The bad news? I don't suppose the tiny little contact area of that friction drive has any more chance of working efficiently without slipping than I have of winning a tug of war with a GWR 2-8-0 tank engine...

"an attempt to meet Brunel’s virtually impossible specifications." Its interesting isn't it. There's nothing particularly unusual, even these days, about a prospective client coming up with utterly unrealistic specifications, but a competent supplier should work out a way to agree a deliverable product. In those days, I suppose, with the whole business of heavy engineering being so new, perhaps it was more difficult. I doubt Brunel was an easy client too, but oh to be a fly on the wall for the discussions of the actual locomotive designers. Did they have grave doubts, or did they really believe - or hope - their interpretations of the specifications would make for useful locomotives? I well recall a situation in my IT career. I worked for a central IT department, but the divisions of the organisation had complete power to purchase their own systems. The IT manager of department X put together a concept and specification for a new system that on paper looked as if it solved all sorts of problems, but in practice was clearly going to be impossibly complex and nightmarish to run. Somewhat Brunellian really! Anyway the in house team put together a proposal with their normal external partner that discarded all this horrendous complexity and proposed something that would actually work. Meanwhile ICL (you can tell this was a long time ago) put together a proposal that met the specification closely and won the contract. At the centre we were awaiting with great interest to see how on earth they were going to make this confection work. At this point ICLs clearly highly competent technical sales people stepped in. They 'worked with' the customer to make some 'minor changes' to the design which actually totally discarded Dept X's IT manager's castles in the air, and replaced it with something that was very close to the proposal the in house team had worked up, with a few fancy features that were of little benefit or disadvantage, but made it look as if it had some relation to the original concept. I've always admired their superb customer management, and tried to emulate it...

I don't know about an actual map, but there is certainly a text list - 6 pages in the 1936 General appendix to the rulebook.

I'm sure I've heard of parts stamped 111, but these are surely more likely to date to her rather longer career as a Castle.

I have absolutely no connection with the running of this site, but a fair bit of experience with recovering broken IT systems. It's in the nature of these things that we (as in IT people) don't really have any idea how long a complex one off process is going to take until it finishes. This is especially the case when it's an intensive background process that will take processing power away from the main business of running the site, because if it causes too much of a performance hit then it has to be slowed up. There's also an age old dilemma that doing the work required to find out how long a recovery will take diverts resources from doing the recovery, so the engineer has to decide whether its more important to get the system back as quickly as possible, or whether to accept a slower recovery in order to give the users a better sense of when full service will be resumed. So pragmatically, at this stage my uninformed opinion is that it's sensible to assume it will be a few weeks, and consider whether there's any material that is so much referred to and so important that it's worth you taking some time to reload - and maybe review the content at the same time.

Its interesting too that the word is being used of an actual existing mechanism that is nothing at all like Capek's artificial humanoids, and in no sense mimics limbs or other human function.

Also begs the question of whether it would be advisable to recreate something that was demonstrably not that great in the first place.

Must be quite an early use of the word robot, The film must be mid 30s I suppose with a GWR roundel, and the word was only coined in 1920.

As numbers around 100 were used for pretty much all prototype locomotives in the Dean/Churchward era I think its pretty safe to say yes, coincidence. For example:

You know I wouldn't be surprised if runaway pushchairs and wheelchairs on the platform were as great a concern as runaway trains these days.

I don't dispute a word of your comments. But the problem is that society in general and press and pundits in particular have a very poor appreciation of relative risk, and some industries are held up to much higher standards than others. And has often been said, "fools are so very ingenious".

Trouble is it's hard to make a rational evaluation of risk and cost/benefit look good in the newspapers or at a public enquiry if something bad happens.

These were saddle tanks with double frames. Five were built in 1891 by the Vulcan Foundry. They had started life with a polished brass safety valve cover as well as the dome, not to mention elaborate lining out. As shown here the lining is omitted completely and its intended to be a post WW1 configuration. The frames are rather different to the general run of 19thC RR saddle tanks. By GWR days three had been converted to L1 Class 0-6-2ST, one of which had already been withdrawn. The two remaining 2-4-2s were numbered 1324/5 and acquired GWR safety valves (and cover) and other minor features, and the air brakes were removed. They were scrapped in 1928. So there's really not much difference between the RR and GWR configuration. From the few photos I've seen they didn't have any GWR lettering on the tanks, which wasn't unusual for saddle tanks, not that these sketches are intended to portray accurate liveries. The sketch looks very bare between the leading and driving wheels to me. I think in real life the connecting rods and valve gear would be filling that gap. I noted with some amusement - if I interpret the drawings I have correctly - that the leading wheels appear to have bearings in the outside frames only, the driving wheels bearings in both sets of frames, the second set of drivers just the outside, and the radial trailing truck only frames inside the wheels.

This is intriguing me more... I've tried manipulating the late photos of 64, and as well as a much higher pitch I am getting a significantly larger smokebox diameter - its a very imprecise exercise, but I get around 4'6". According to the dimensions I found in RCTS and WRR1 (at least if I'm not having a senior moment) the K boiler, although longer than the original L boilers, was still only 4ft diameter. So I wondered if the note claiming it to be a repaired K boiler was wrong, since there seems to be no reason for a K boiler to be pitched higher. That got me wondering about A or P class original boilers, of which there must have been some knocking about after A1/P1 upgrades. However they're really big at 4'9 diameter, and although it definitely looks bigger than a standard L/K boiler, I can't convince myself it can possibly be as large as the later ones. The smokebox looks more the size of a Std 3/Std 10 to me, which is about 4'5. Strange! So *if* my guesstimate is correct it appears to be about the size of the earliest and very unsatisfactory M class boilers, which had belpaire fireboxes, so even cut down in length would hardly be a great choice under a saddle tank. What's more there would be a considerable reduction in water capacity. It doesn't feel like a viable hypothesis. Getting back to the drawing I think I'm going to have to take an educated guess at the K class boiler in the lower pitched configuration. I wonder if the extra inches went forward, or back into the cab?

The RR passenger locomotives all had air brakes, but most of them had the pump on the R/H side of the smokebox. Oddly the Ls were l/h drive, and the freight classes r/h drive not sure about the other passenger classes.

That's great:-) You can even see some of the motion, which is as prominent as I feared!

Here's the GWR weight diagram as reproduced in Russell. It demonstrates the issues with weight diagrams, especially grouping era ones, in spades. I agree with Nick, the boiler pitch must have been raised - at least on 64 - so that's one inconsistency. The curved footplate over the leading wheels is wrong, as discussed, and the shape of the cab step is wrong too FWIW. The safety valve cover doesn't look like a GWR one, but the original RR ones were years gone by this stage, and it must be extremely doubtful whether the L1s, which never received their GWR numbers, were given GWR safety valves. The Ls didn't receive them until well after the L1s were cut up. I don't think there's anything about this weight diagram I would trust without independant confirmation! I think it must have been a very quick tracing of a 2-4-2T diagram.

I hadn't found the small RCTS photos helpful in the past, so hadn't looked... Thanks for all the helpful info. The frames aren't truncated in length, it's the front axle box area that I cut back. Hopefully I can have a good go at the boiler changes with the dimensions. [later] Isn't that a good reproduction by the standards of the RCTS volumes. Scan and blow up I think. See what can be done with it. The motion I was expecting to see would be under the frames rather than round the splashers. The change in the reversing rod suggests the cylinders and motion may have been moved upwards, which would explain a change in boiler pitch. Altogether it seems a much more comprehensive rebuilding than I had assumed...

The L1 class were built as L class 2-4-2 saddle tanks with double frames. By GWR days two of them had been converted to 0-6-2T, given new design boilers based on those of the K class, and called class L1. A third conversion had been scrapped in 1921. The rebuilds presented an odd appearance, since the 2-4-2s had a small rise in the footplate over each crank on the drivers, but this was not repeated over the new leading driving wheel. The 0-6-2s, allocated diagram J, were scrapped in 1922 and 1923, and never carried their allocated GWR numbers, 1324 and 1325. This is an updated version of the sketch, and hopefully further updates will follow. The sketch attempts to represent RR No 64 between 1911, when it was converted to 0-6-2ST, and 1917. At this stage it was still carrying a modified L class boiler. Some time after 1917 it received a new boiler, documented as being a K class boiler, and photographs show it to have been considerably higher pitched. At the moment I am uncertain as to whether 62 and 63 had the higher pitched boiler as I have yet to find any photographs. Published information states all three had low pitched boilers as per the sketch, but the K class boilers were a few inches longer. This is one of the least well founded drawings. The material I have found for the L1s so far consists of drawings and photographs of the L class 2-4-2ST, a single rather indistinct photograph in WRR Vol 1No 64 in this form, two photos of 64 in post 1917 configuration and a particularly sketchy - in more than one sense - GWR weight diagram which includes a rise in the footplate over the leading wheel which didn't actually exist. So this was created by taking my drawing of the L and truncating the frames in what seemed to be a reasonable manner and hoping for the best. If anyone can urn up some good photos of the L1s I'd be grateful. More than most drawings this one suffers from not having found any kind of source for the inside rods and motion, which I think should be particularly prominent between the leading and driving wheels. I would like to thank contributors to this blog for comments that predate this drawing, which have been extremely useful in making improvements.

The A class could be considered an enlarged S class 0-6-0T, and the S class a smaller R class 0-6-2T. The As had a smaller boiler than the Ms and Rs.They had 4ft 4½in driving wheels like the S class, the Rs having 4ft 6in. Stephenson’s delivered 10 As in 1910, and Hudswell Clarke 6 more in 1911. The A1s were essentially the same as the As except that the boiler was pitched higher, had a Belpaire firebox and was fitted with top feed. There were also small differences to the frames. Three came from Hudswell Clarke in 1914, two more in 1916, and three from Stephenson’s in 1918. Several of the A class were given A1 boilers, which put them in the A1 class. By the time of the grouping the A class was reduced to ten and the A1 class increased to fourteen. They were allocated numbers 52-75 by the GWR. The sub class complications continued under the GWR and some A1s received A boilers and vice versa as boilers were exchanged at heavy repairs. From 1929 standard 10 boilers were introduced and these were fitted to about half the class. GWR bunkers appeared on most, and GWR cabs on some. Withdrawals started in 1948 with some of the earlier locomotives, and the majority went between 1953 and 1955. The GWR created the same boiler fitting drawing for both variations. The Std 10 boiler sketch above shows 1914 (A1) frames, but the differences are small. I also idly wonder whether the frames from the three different builders of A1s were identical. There seems little point in drawing both. As usual with the RR classes I owe a substantial debt to the Welsh Railway Research Circle's excellent publication on the Rhymney, Welsh Railway Records Volume 1, but a GWR drawing reproduced in Russell was also very important, especially in sorting out the differences between the frame types. Sandboxes... You'll see the A has the front sandbox above the frame, and the A1 below. Some As definitely had the sandboxes changed to under the footplate. Another in life change was that the front suspension springs were moved to under the frames on a couple of the locomotives in the early 1950s.

They certainly got GWR whistles - both of them. They also lost the air brakes. Plus you need the sound of the crew grumbling about being given one instead of a 28.

I'm not sure the frames are the limiting factor. Doesn't the piston rod have to be on the centre line of the cylinder? In which case won't the need for the connecting rod and crosshead assy to clear the wheels be a defining factor, as would the width of the bearings on the coupling rods? I have a table of GW bearing sizes and if I read it correctly on the 47xx the coupling rod bearing on the driving wheel is 4 15/16 long (=wide) and the connecting rod 5 31/32, whilst on the 30xx (ROD) they are 3" and 4 15/16", which it seems to me, if I'm interpreting the table correctly, suggests the ROD cylinder diamter is at the expense of bearing size.

I didn't feel like spending a lot of time on this, so I grabbed the first dimensioned large US locomotive I found ( https://en.wikipedia.org/wiki/USRA_Heavy_Santa_Fe ) and did a quick hack of a drawing. It scaled to 80% to fit in the UK loading gauge as shown. I haven't attempted to correct for wheel gauge. Look up what 80% of standard gauge is in 4mm model terms, it might amuse:-) Anyway that leaves the cylinders at 24" diameter which is surely still too big for the UK gauge, especially so low slung, so this is a very crude approximation. Driving wheels scaled to about 4'6. I hadn't a readily scaleable drawing of a 9F, so I added a GW 47. As you can see its bigger then any UK locomotive - big surprise as a 2-10-2, but not outrageously so.

I wonder how big the US Mallets would be if you scaled one down to fit the BR standard loading gauge... I imagine the cylinders might be the main limit since the track gauge couldn't be changed of course. I might go hunting for a scale drawing to try...