Discussion in 'Steam Traction' started by Eightpot, Mar 23, 2020.
Possibly because Beyer Peacock held the patent?
There's much in this (and @Cosmo Bonsor's post), though probably worth noting that the principal strongholds of Garrats were on mountainous lines. Much of South Africa is a plateau, over 5000ft up and when compared, on the same diagram, the EAR mainline (Mombasa - Kampala) makes UK summits at Slochd and Beattock look like mere pimples! The other main reason for adopting Garrats was heavy traffic on mainly single track routes, the majority of these being narrow gauge (the same considerations behind the FfR's adoption of articulated power.
With the marginal performance of the LMS Garrats (thanks a heap, Mr.Anderson) doing little to help perceptions, plus the additional maintenance burden, including on number of unfamiliar features, such as high pressure flexible steam joints, bogie bearings, compatatively comlex reversing and braking arrangements, payment of patent fees (were construction to have been undertaken in-house) and operational considerations, such as driving outlook and increased grate sizes .... both likely to have exercised ASLEF .... and we're probably more than halfway to understanding why Garrats never really gained a significant foothold here. Recall also, Gresley's reported pre-grouping comment to the effect he was grateful that conditions on British lines rendered the question of articulated locos irrelevant (he must've loved the U1 proposal, inherited from the GCR!).
I note too, that in recent years, much work has been undertaken at Blodge, to improve flexible high pressure steam joints on the FfR's bendy fleet, suggesting that even late designs, such as the SAR (Oz) class 400, EAR classes 59/60, RR classes 20/20A, SAR/SAS classes GMA & GMA/M, NGG16 hadn't completely perfected this tech.
I think another downside of Garratts in a UK context is probably length (and I speak here as somebody who likes Garratts). If we go by the African designs, where trailing pony trucks came in pretty quickly after experience with vicious flange wear on the trailing coupled wheels, I'd suggest that a 2-6-2+2-6-2 is probably a good wheel arrangement; certainly if you basically put LMS/BR Standard trucks on it I don't think there'd be any trouble with ride. But with the boiler in the middle like that, a Garratt tends to be intrinsically quite long, which takes up lots of space in sheds, at platform starters, on turntables, in headshunts, etc. Most Garratts were designed to replace two tender engines (on the SAR the GM family are two 19 classes, a GL is two 14Cs, etc), so this isn't a problem, except for turning facilities.
Leader was supposed to be a tank engine replacement; a design that would probably end up considerably longer than an N class was going to be an interesting sell. Instead, Leader was a Kitson Meyer in drag, where by positioning the boiler such that only the firebox is between the two power bogies, you take out quite a lot of length. Kitson Meyer's weren't unsuccessful, particularly in South America, but the Garratt eclipsed them quite quickly in the rest of the world. A more conventional Kitson Meyer design might have been an interesting concept, but whether they were ever successfully used at anything like 60mph, I've no idea. The nearest I know to them being used in Europe would be the Du Bousquet locomotives, which were intended (wait for it) to eliminate double heading on heavy freight work. From the description, I'm not sure they're as different from Kitson Meyers as the author implies.
*Note that Beyer Peacock basically cracked the high pressure steam connection problem, using a similar design to the early Fairlies but with cast iron seats and mechanical lubrication. Phil Girdlestone wrote at some length on this, being apparently slightly exasperated by an FR tendency to reinvent the wheel because they hadn't investigated what anyone else had done to solve the problem.
My hunch with Garratts is that two locos are more flexible than one for the traffic on offer. There is also the issue that to exploit the power of the loco, you have knock-on consequences for infrastructure, for example length of loops, sidings, platforms etc. As an example, the longest possible train out of Waterloo is about 13 carriages which is within the capability of a Merchant Navy, at which point why do you need anything more powerful?
In the early 1930s, Maunsell drew up a scheme for a 4-6-2+2-6-4 locomotive, with 6'3" wheels, six 16" * 26" cylinders,
that would have been about 100 feet long. It would have had a tractive effort of about 50,000lbf, 3,400 sq ft of heating surface, a grate area of 52 sq ft and weighed 209 tons. There was evidently enough interest round about 1933 - 34 to get detailed costed proposals worked up with Beyer-Peacock.
The Civil Engineer agreed to the use of the loco on about 130 miles of the mainline outside London, with a maximum speed of 75mph. The CME and the Traffic Manager worked out a scheme in which ten such locomotives could replace 64 conventional locomotives (54 * class 5/6 passenger and mixed traffic 4-6-0s; 8 * class 4 mixed traffic 2-6-0s; 2 * class 3 goods 0-6-0s) provided the manufacturer could guarantee mileages of 140k - 160k between heavy overhauls - to which the manufacturer agreed, subject to the provision of a certain quantity of spares in addition to the order for ten locos.
Needless to say, the scheme never proceeded. To have been feasible, it would have meant running a small number of very large trains. But the platforms at the major stations weren't long enough; the goods loops weren't long enough; the sidings weren't long enough; the draw gear on wagons may not have been strong enough. In the end, a greater number of smaller locos gave a more flexible service - and a better passenger service.
Could be the case. Again there were a large number of Beyer Garratts at work in the world by the mid 1940s and the type was well proven but not in the UK. Those built for UK railways were not the best examples that were built, they did not have the best of records. The Garratt might have been seen as a heavy freight design but this was not the whole story passenger types were built and the Algerian 4-6-2 + 2-6-4 achieved 82mph, the highest speed achieved with articulated types according to some sources.
Bulleid would have wanted to create his own solution to meet the requirements of the 1944 traction review. I suppose it is worth reminding ourselves of the railway that he came from to join the Southern, the LNER. Here he had been involved in some remarkable work, the more remarkable considering the financial constraints this company operated under. But he also operated as part of a team and worked within the constraints of the same. This was not purely disadvantageous, in this situation his innovative nature could find an expression but it could also be kept in check.
When he took on the position of CME on the Southern the situation changed. He was in charge of the team and there was no Gresley to, how shall we say, maintain the enthusiasm but limit the excesses. And the financial restraints were not the same.
Yes, Beyer Peacock could have produced a locomotive that would meet the requirements but this would not be the brainchild of Oliver Vaughan Snell Bulleid. That is the difference and I don't believe that the Garratt option was ever considered at the time and even if it were it would not be seriously so. As for the patent costs, it would in hindsight have been worth biting the bullet.
And that is absolutely the key point. If you can get enough power from a rigid-frame loco, then why go to an articulated? On the freight side, remember that the LNER had difficulty in making use of the full power of Gresley's P1 2-8-2s.
Great Britain had an advantage from higher permitted axle-loadings (during the steam era) than almost anywhere outside the USA, so we could get by with fewer wheels. But right across Europe, the "big engines" were mostly 4-6-2s, 2-8-2s and 2-10-0s, with only a relatively small number having more wheels. The Soviet Union was slightly different, but even there the JS 2-8-4s and FD 2-10-2s (at around 135 tons engine weight with 20 ton axle-loadings) were deemed adequate.
Russia's sole excursion into Garrat territory (the largest such ever built) was less than successful.
Orenstein& Koppel sales literature before 1914 explain it quite well.
For short- heavy use railways like brickworks,Steel works etc it is best economy to buy heavy rails and locomotives with 4 drivers.
For railways in woods,sugar plantations etc ligther rails and 6 0r more drivers.
Giesel Gieselingen(He made satisfactory locomotives ) states that ten drivers with for example 100 tons adhesion pulls better than twelve with 120 tons in two uncoupled groups.
Bulleid could have made a Heissler but then it was not a Bulleid.
Interesting answers to the Garratt question, thanks.
A Q1 boilered tank engine would have been fun. Pig to oil up though.
On a personal note I found the Q1 somewhat unpleasant to fire, a bit draughty and dirty.
It was very good at turning burning coal into an accelerating train though.
Interesting observation about Garratts and leading/trailing wheels. I wonder whether the Leader design would have been similarly hard on the wheel flanges?
I'm sure Tom's right, a 262-262 Garratt would have been comically oversize on the Southern, but say a 242-242 would have been viciously expensive but not that much advantage over a non articulated locomotive. With overhaul costs dominating cost of ownership, presumably a Garratt would be a lot more expensive to run than an unarticulated locomotive,which is fine if its replacing two similarly expensive to run antiquated 0-6-0s, but otherwise dubious.
Deleted, as too inaccurate (sorry ,,, brain in one place, books in another!). See post #79 below by @andrewshimmin for correct summary of what isn't here.
Leader might have been less harsh on its flanges than a Garratt of the same wheel arrangement. One of the defining features of the Garratt is carrying the fuel and water on the power bogies, which I think was generally considered a positive; it reduced the weight on the pivots and avoided having a big cantilevered frame above the bogies to carry the bunker and tanks*. It may well also have made them slightly more stable, but all that extra mass makes the bogies less enthusiastic about changing direction. Whereas on the Kitson Meyer/Leader configuration, the fuel and water are on the main frames, so the polar moment of inertia of the bogies should be a fair bit lower. It may reflect the intended duties, but a lot of Kitson-Meyers were 0-6-0+0-6-0, or even 0-8-0+0-6-0, whereas there were a few 0-4-0+0-4-0 Garratts and everything else had at least leading pony trucks, as far as I can recall. An advantage specific to Leader is that with all cylinders inside they didn't have to overhang the wheelbase as far as on an outside-cylindered design, again keeping the moment of inertia lower.
*A couple of builders marketed pseudo-Garratts with coal and water on extended main frames, so as to get round the Garratt patents. NBL sold them to SAR as 'Modified Fairlies'. You could think of them as really long Kitson Meyers with tanks in Garratt positions. These were 2-6-2+2-6-2 locomotives, but by all accounts were not a success.
Was there scope for Bulleid to take a more evolutionary approach though?
Rather than testing one new feature per design, he seems to have introduced masses of new ideas on every loco he designed. Was there a reason for this approach?
My guess would be the time imperative, what with impending nationalisation. Attlee's manifesto comittment was clear enough, so it was pretty much a 'given' after the '45 election, which neither the LMS nor LNER were in any fit state to mount an industry operation to oppose. Without those two, the Southern and GW would've been on a hiding to nothing.
How long did the leader take from drawing board to building? Likewise, although the railways were nationalised in 1948 there is a period before that where they have to be drawing up plans as to how it was to be organised so it wasn't necessarily obvious from the outset that Bulleid would be out on his ear.
I am not sure given the major 'disinvestment' in Railways in the inter war period that the finances of the GWR & The Southern were significant better
Beyond reiterating my understanding of the "broadest brush" postwar situation regarding the Big 4's attitudes to nationalisation, there's not much I could usefully add to what I've said without authoritative references, which I don't currently have to hand.
A slight correction here: the 5'3'' Garratts of the Sao Paulo Railway were of two classes: a 2-4-0+0-4-2 type (Q class) for the short coastal run to the bottom of the inclines up the Serra do Mar mountains, and the much larger R class for the plateau trains from the top of the incline, through Sao Paulo itself, to the end of the line at Jundiai. The R class were originally 2-6-2+2-6-2 but after a couple of years were rebuilt as 4-6-2+2-6-4, mainly for additional stability at speed. Very handsome they were too, with a fair turn of speed and both classes served the railway well until electrification (on the plateau) and dieselisation (on the coast) in the 1950s.
Meanwhile the Sierra Leone Government Railways (2'6'') had some 2-6-2+2-6-2 Garratts built before the War, and then some magnificent 4-8-2+2-8-4s built in the 1950s: a masterpiece of design to get such powerful locos with such low axle loading on such a small gauge. One of the latter survives in the Freetown museum.
But your general point, that leading bogies and trailing pony trucks became gradually more common on Garratts is valid.
It's also worth noting, re the Bulleid comments, that the Garratts where BP were allowed to used their experience and build to their own standards tended to be more successful than the ones where they were obliged to use standards and parts from the purchasing railway (the LMS ones are just one example).
Sent from my Pixel 3a using Tapatalk
While we're on the subject (albeit not the subject of the thread!) it's worth noting that the LMS considered various passenger Garratt proposals in the early days when the motive power situation was a bit unsettled, shall we say?
According to Robin Barnes, in his excellent "Locomotives that Never Were" the LMS approached BP to produce some compound 4-4-2+2-4-4 based on the Midland compounds.
A bit later BP offered their own concept, a 4-6-2+2-6-4 of a fairly standard Garratt type.
Both would have been very interesting, the latter probably magnificent, but neither got anywhere beyond a concept.
Barnes has illustrations of both types in the book.
Sent from my Pixel 3a using Tapatalk
In about 1936 the LMS tried one of its Garratts, No. 4999 (only the first three, 4997-99 were vacuum fitted) on an experimental passenger train between Derby and St Pancras. Alas, the loco got no further than Leicester were it was removed from the train, having succumbed to the LMS Garratt malady of a hot box. Frank Carrier photographed it at Borrowash, and his collection is held by the Kidderminster Railway Museum. The photo depicts the vey clean engine with the ex-L&YR dynamometer car at the head of the stock. The test was not repeated.
Separate names with a comma.