RLBH

The only common locomotives between the two studies were the 2-Do-2 and the 1,200hp Bo-Bo - the LMS plan called for just eight of the latter, compared to 91 2,100hp Co-Cos! Oddly enough, the LMS didn't see a need for any shunters either, sticking to purely main line electrification and letting steam handle the yards. The two mixed traffic locomotives, and the two heavy goods locomotives, were mutually exclusive - the LMS study called for the 1,800hp 2-Co-2 and the 2,100hp Co-Co, whilst the LNER reckoned the 2-Do-2 would do all the regular passenger work needed and standardised on the Co-Co for everything else. Even the Bo-Bo was proposed to have high commonality with the 1,800hp Co-Co, although steam thinking does seem to have led them to try to finely matched locomotive power to train size. Speed improvements were expected, but through better acceleration and maintaining speed on gradients rather than higher maximum speeds. The authors of the reports were aware that higher speeds were possible, but considered them inadvisable with unbraked wagons and traditional axleboxes. Their remit was, after all, electrification - not the general improvement of rolling stock and operating practices. For the same reason they recommended electric boilers for train heat. One gets the impression that the LNER had a reasonably good handle on how to electrify and what it was capable of, and wanted to use government money to try and sell it to the board, while the LMS were trying to make the case that it was too expensive to bother with. The LNER proposal to electrify everything on the former Great Northern section, except for the most marginal of branches, showed much better return on investment than the LMS study for heavily trafficked, steep main lines and nothing else.

One of the findings of the studies done for the 1931 Weir Report was that if you totally electrified a railway system, about half of the mileage was in goods yards and sidings. But if you only electrified those parts of the goods yards and sidings where a locomotive might run, you still did virtually all of it. Which points to the importance of diesel shunting locomotives, that way you only need to electrify the exchange sidings. Interestingly enough, both studies (one LMS, one LNER) done for the Weir Report were done by Merz & McLellan, and they found that electrifying the branch lines was a better proposition than doing main lines only, because of inefficiencies brought in by changing engines. The LNER study proposed that a lot of branch line working would be done by three-coach multiple unit trains, which may have been replicated in the GWR study - this would explain a surprisingly small number of locomotives. The 1931 locomotives were: A 2,400hp, 110 ton, 2-Do-2 for heavy express passenger work A 1,800hp 2-Co-2 for light express passenger and fitted freight work A 2,100hp, 108 ton, Co-Co for 40mph goods trains A 1,800hp, 108 ton, Co-Co for 35mph goods trains up to 900 tons A regeared version of this with a train heat boiler was proposed for fitted freight and excursion traffic A 1,200hp, 72 ton, Bo-Bo for 35mph goods trains up to 600 tons, or 1,200 tons working in multiple A 720hp, 60 ton, Bo-Bo shunting locomotive Some of these sound similar to the proposals for the GWR scheme, with some evolution.

An obvious missing route occurred to me recently: seemingly the only valley in the whole of South Wales that doesn't have a railway up it is the Usk Valley above Abergavenny. Such a route might well have been promoted to give the GWR access to Brecon and to the Mid-Wales Railway, competing with the Brecon & Merthyr. It would also serve Crickhowell en route, and no doubt a number of smaller places.

American railroads and the British government take the same capital-limited approach to electrification - while it saves money later, they'd rather not spend the money now to get the benefit. Diesel traction may be more expensive in the long run, but you can buy more or fewer locomotives depending on the budget, rather than having to commit to an expensive programme of electrification up front then stick with it to see the benefits.

I seem to recall that the LMS pair actually did less damage to the track than the later BR 1Co-Co1 arrangement, despite the higher axle load, because the suspension was superior.

I've imagined the hills around Shap being much higher to force the main line to Scotland across Stainmore, so all those extra hills must have come from somewhere!

Which is indeed why my 'Domhnall Beag' scaled-down Big Boy is intended to work 60-wagon trains over a steeply graded route with lengthy sections of single track... unfortunately, I believe there's only one such line in the UK, and it only saw that level of traffic when there was a war on!

It would look decidedly less odd as a Co-Co, I think. Not much use as a way of using spare 91 bits, though.

More Deltics are always welcome, but surely a single 18-cylinder engine would require less maintenance?

A bit more, in fact. As a unique, diesel-hydraulic, equivalent to a Deltic, it would probably have been immensely popular with enthusiasts, regardless of how well it performed!

Unless of course you have eyewatering vertical curvature. But if you're trying to avoid buffer locking in the vertical plane, it's probably a better solution in the long run to find the Civil Engineer and beat them repeatedly with a copy of the design standards until they fix the alignment.

Indeed, and my terminology was deliberately vague because of that. When a horsepower figure is quoted for a steam locomotive, it's usually 'at rail' or drawbar power - which of course aren't quite the same thing. When it's quoted for a diesel locomotive, it's almost invariably the brake power of the prime mover, which is entirely different again and much smaller than either.

It's a bit of a stretch, but you could probably equate the firebox to the diesel engine itself, the boiler to the generator, and the cylinders to the traction motors. Electrical machinery typically has about 90% efficiency, so that transmitting 80% of the engine's output to the wheels is fairly straighforward. Early diesel-electrics achieved a bit less, so the 65 to 70% of a well designed firebox and boiler doesn't look too bad. Inconveniently, the thermodynamic cycle takes place at opposite ends of the system, but it's them that lock in the inefficiency of steam. The best steam locomotives built only used 15% to 18% of the energy in the steam - Porta's ideas about improving thermal efficiency notwithstanding. That limits them to about 12% of the energy in the fuel being turned into power at the rail. By comparison, a realistic diesel engine can turn 40% to 50% of the energy in the fuel into crankshaft motion, which with realistic transmission losses gives them 30% to 35% efficiency overall. On top of which you get the advantages in availability and labour saving. Steam is really a losing proposition unless you have cheap labour and cheap coal. Missing the transmission losses is one of those mistakes that seems obvious in hindsight, but was probably forgivable - those writing the specifications had likely never had to think about it, because transmission losses aren't really a concern for steam or early DC electric locomotives. In the latter case, they do exist, but are palmed off on the electricity generating company and the electrical engineering department, so the locomotive engineers don't really have to deal with them. And it's remarkable how well 'fixing' the transmission efficiency issue solves the issue of diesel Types mapping on to steam Power Classes. Just looking at the Riddles classes for BR and the War Department: Duke of Gloucester is right at the top of the Type 4 band; the other express Pacifics she worked with are firmly into Type 5. The 9F, Britannia and Clan all equate to Type 4s - the first two are pretty close to the Peaks, the Clan is at the bottom end of the range and very comparable to a Class 40. The 5MT and 4MTs all equate to Type 3s, as do the WD 2-8-0 and 2-10-0s. The 3MT, 2MT and Hunslet 0-6-0 are all Type 2s. That all seems broadly correct in terms of work undertaken. Steam was generally limited by tractive effort rather than by power. So it makes sense that in a classification system which is based on power, the smaller steam locomotives will be rated fairly highly. That's why almost any steam loco could put in a decent performance on a branch line freight working, but a Class 08 diesel shunter isn't much use for anything outside a goods yard. The rough rule is, a steam locomotive can pull at speed anything it's capable of starting, whilst a diesel can start anything it's capable of pulling at speed. Even the smallest, least powerful diesel locomotives had tractive efforts comparable to the largest diesel locomotives - the sole exception being the LNER U1, which cheated! I've tried coming back the other way to equate diesel & electric traction to steam locomotive power classes; that's easy for freight (all the main line locos are 9F!) but very difficult for passenger rating. Partly because the BR power factor formula is a bit of a fiddle to make things come out 'right', and partly because diesel locomotives have a serious lack of grate area and boiler tubes.

What I suspect happened was a two-pronged failure. I think they did accurately determine the power that could be sustained by steam locomotives, without fully appreciating the benefits that being able to mortgage the boiler brought in terms of temporarily exceeding what would be called a 'continuous rating' on a diesel locomotive. Perhaps more significantly, I don't think they accounted for drivetrain losses. A steam engine rated at 2,000hp delivers that straight to the wheels. A diesel-electric locomotive loses about one-quarter of its' rated power between the engine and the wheels, and a diesel-hydraulic tends to lose about one-third. So, equalling an 8P rated at 2,000hp needs a 2,650hp diesel-electric, or a 3,000hp diesel-hydraulic.

You'd expect any engine replacement to be with the standard Paxman engine - unless parts for the Rolls Royce engine were readily available at the relevant depots, their improved reliability might well be offset by increased costs resulting from being oddballs.

AIUI, not only were the two Rolls-Royce engined locomotives more satisfactory, but the Beyer-Peacock built examples (with Crompton-Parkinson traction motors, rather than Clayton's GEC motors) also did better. The ideal Class 17 would probably be built by Beyer-Peacock with Rolls-Royce engines and Crompton-Parkinson electrics. And still be mostly useless because the traffic disappeared, but you can't win 'em all.

Since unreasonably large locomotives have come up, the North American loading gauge is approximately a seven-eighths scale version of the British loading gauge. This is demonstrated by the ability of American prototypes in H0 scale to run on British 00 layouts. So why not scale down American locomotives in the same proportion? Taking the Union Pacific Big Boy for starters. The resulting unreasonably large locomotive has four cylinders, of 21" bore and 28" stroke. We know this fits in the loading gauge, a fair few locomotives have done it. Driving wheels are down to 60", which we know is perfectly fine for fast freight turns. Working pressure drops to 225psi, thanks to the smaller boiler. This is all very reasonable, and gives us a tractive effort of 78,720 pounds. Grate area would be about 88 square feet, with a power stoker and an unreasonably large tender being mandatory to keep it fed. The locomotive ought to weigh in at about 230 tons, with an axle load of 20.5 tons. The boiler is going to be immense - 19'3" long with a 7'9" diameter - which will cause major loading gauge issues. Directly scaling from the American locomotive winds up at 14'3" tall. By reducing the clearance between the boiler and running gear, it should be possible to get down to 13'8", which is at least within the realms of possibility on some British lines. Making it a Garratt would probably remove the objection altogether, but we'll quietly ignore that point... As far as tender capacity goes, scaling down from the Big Boy gets us 16.75 tons of coal, which seems reasonable for the size of the grate. It would also give us 15,000 gallons of water, which is disproportionate by UK standards. Scaling off the 9F's coal capacity, anything between 9,300 and 13,700 gallons seems reasonable. Go for 11,500 gallons. Using the Big Boy tender's load-to-weight ratio, that's a 118.7 ton tender; scaling from a BR.1, could be anywhere from 120 to 130 tons. Call it 120 tons on 6 axles, probably a bogie tender as well, though if it's at the higher end of the range a 7-axle 'centipede' tender might be needed. Might even be worth a Vanderbilt tender to save on weight. Obviously, this is a big machine. Realistically, the Big Boy is designed to haul long freights up big hills at reasonable speeds, and this will be similar. Shap and Beattock don't cut it, in my book, to give the 4-8-8-4 a chance to stretch its legs. They're reasonably steep, true - but they're not long enough. They can be tackled by building up a good head of steam, charging the hill, and mortgaging the boiler to get over the top. Ais Gill is long, but not steep enough - trains will hit a length limit before the tonnage limit (1). There's only one place in Britain that I can think of with a big enough hill to truly compare to the duties the Big Boy was meant for. And fortunately, the company that built it was a 'big locomotive' railway with a history of doing new things - it built, for instance, the first British 4-6-0s. I am talking, of course, of the Highland Railway's main line from Perth to Inverness. Operating conditions on the Highland Main Line are generally more similar to American railroads than they are on other British lines, with long, steep hills where the boiler will need to work at equilibrium and long sections of single line which dictate running longer trains in preference to more frequent trains. If you're going to try and fit an American-style locomotive, it's the place to do it. Southbound from Inverness, it's 22 miles of climbing, mostly at 1 in 60, from a dead stop at sea level to 1,315 feet. Northbound, Drumochter isn't quite as steep but is longer. This is comparable to the grades that Union Pacific had to deal with in the Rockies. On a first approximation, it should be able to handle a 1,145 ton freight southbound from Inverness over Slochd - ruling gradient 1 in 60 - and 1,415 tons northbound from Perth over Drumochter at 1 in 70. That's actually relatively sensible, working out at about a 55 wagon train - which could and did run further south. So there shouldn't be too much of a problem with coupling strength. A British-sized 4-8-8-4 actually makes some sense in the Highlands. 'Big Boy' is an unlikely name for such a machine. I call it 'Domhnall Beag' - or, translated from Gaelic, 'Little Donald'. It seemed fitting. We'll need to bore out or skylight the tunnels at Killiecrankie and Dunkeld, but more challenging civil engineering has been done. The only problem is that there isn't enough traffic to justify it. The modern Oxwellmains to Inverness cement working today is about right - it's a 1,400 ton train, usually entrusted to a Type 5 diesel-electric and running as a Class 6 fully-fitted freight, permitted up to 60mph. The ideal size and class of train for Domhnall Beag to haul. Unfortunately it only runs a couple of times a week, and that's with today's traffic requirements. Justifying it would need a lot more freight running to (or from) Inverness during the steam age, and a railway willing to spend the money on a specialised locomotive to handle it. That might make Ais Gill a better bet. Just lengthen the loops to handle 80 to 100 wagons. Can't be that difficult, right? Alternatively, it's not too far off of being a pair of 9F engines with a common boiler, so would be appropriate for any turn of duty that routinely needed double-headed 9Fs. The Ebbw Vale and Consett ore trains, for example. Axle load is higher than the 9F, but any line seeing that kind of regular heavy freight traffic ought to be able to handle it. A similarly scaled Challenger 4-6-6-4 is well within the realms of sanity - it's pretty comparable to Powells proposed 4-6-2+2-6-4 for Preston to Glasgow and Perth. The passenger equivalent, a 7/8ths scale Union Pacific FEF-3, is actually slightly smaller and appreciably less powerful than the proposed LMS fast mixed traffic 4-8-4. (1) Tonnage limit on Ais Gill would be about 2,150 tons, requiring 85 16-ton coal wagons or an equivalent.

The immediate thought would be whether the LMS had any thought of fitting coal wagons with vacuum brakes at the time the first batch of Garratts were ordered. This would explain why dedicated freight locomotives were fitted with vacuum equipment. Such a scheme would surely have been abandoned by 1930, between simple cost considerations and the impact of the Great Depression, so there'd have been no requirement for the second batch to be so fitted.

Although it should be remembered that the original thoughts for a Standard Class 8P locomotive in 1948 were for a Coronation given BR Standard features, in the vein of the Class 4 2-6-0 or Class 5 4-6-0.

The only place modern coal-fired steam would have a hope of competing economically would be hauling massive coal trains. Very much along the ACE 6000 line of thinking, which means not in the UK. Maybe in China, Australia or some other massive coal exporter. But almost certainly not – diesel-electric and straight electric have all the cards.

Yes, the main driver for the Austerities was overseas service with the War Department rather than domestic service. What domestic service they did see during the war was mostly for running-in and training. It's hard to see why the War Department would want a 4-6-0 version of the 2-8-0. What might be justifiable is a War Department Austerity 2-8-2. The US Army Transportation Corps had the S100, S160 and S200 for military use - the S100 is pretty close to the Hunslet 4F, and the S160 to the Austerity 2-8-0. The S200 was designed to a British requirement for use in the Middle East, and I believe was compatible with the British loading gauge, but there was no direct British equivalent. It's not too much of a stretch to imagine Riddles having a crack at designing one. Actually modelling one might be a challenge, though.

In principle, the 2,000hp rating on a 12CSVT is equivalent to the 2,700hp rating on a 16CSVT as used by the Class 50 (and the 1,330hp rating of the 8CSVT on the CP Class 1400 - basically the ultimate Class 20) so should be reasonably achievable. I've heard that EE offered the higher rating to BR for one of the later batches of Class 37s but it wasn't thought necessary.

Very roughly, an ETH index of 1 corresponds to 5 kW, but it's not necessarily a 1:1 relationship.

Coal being heavy, and satellites being light, you can only get 29 tons of coal in before it's overloaded. And because the International Space Station is in an awkward orbit, you can only get 15 tons 13 cwt there. Less if you want some sort of container to keep it in. Steam powered rockets are definitely a thing that people use, though. And there was a serious NASA study for building a railway on the moon at one point - I may have to see if I can find it. Electrified, of course, and I believe aluminium rails.

I would imagine that in the case of multiple factories, one might be assigned numbers 1 to 1000, another numbers 1001 to 3000, and so forth. Any given factory would start making them in consecutive order, simply for administrative convenience, but you'd probably get the odd case where number 273 had to be sent back for rework so was completed out of order. But you'd quite probably wind up with number 1001, being the first produced by Factory B, being completed well before 1000, the last from Factory A. Unless of course you didn't, because Factory A had other priorities, or was on strike, or the contract was signed late, or any number of other things.