Imaginary Locomotives

[JimC]

3 hours ago, Compound2632 said:

 

(Though if economy of maintenance really mattered to Churchward, why didn't the GWR adopt outside valve gear?)

Pros and cons. Connecting and coupling rod bearings could be attended to without disturbing the valve gear.  Loco preparation was a piece work task, so little saving by taking a few minutes off the job. 

[JimC]

12 hours ago, The Johnster said:

Atlantics performed well on other UK railways, free running and free steaming, 

Probably fair to say, though, that a larger boiler and cylinders can be fitted on a ten wheel locomotive (even within UK loading gauge) than four driving wheels can provide adhesion for. 

[Hroth]

5 hours ago, Compound2632 said:

 

Is there any data on this? Did Churchward or one of his staff present a paper to I Mech E or ARLE? Economy means more than just fuel consumption of course, there's also other running cost such as maintenance to take into account. (Though if economy of maintenance really mattered to Churchward, why didn't the GWR adopt outside valve gear?)

 

Anecdotal, I suppose, plus the GWR didn't proceed with compounding. Was TCO a consideration a hundred years ago, or would the cost of coal per mile be the main consideration?  Maintenance of inside valve gear didn't seem to be a concern for many British railway companies, labour was relatively cheap.

 

As for outside valve gear, it ain't pretty!

 

[The Johnster]

All of which illustrates that designing locomotives that are successful on the road and economical in terms of fuel and maintenance to satisfy the bean counters who actually run the show is a bit like juggling on a tightrope.  Especially during the Edwardian era, when train loads were increasing (except on the Midland where they double-headed and put on more, but comparitively short, trains, enabling them to stay with 4-4-0s for the top link jobs) and Traffic were demanding faster timings and longer non-stop runs because bogie gangwayed stock meant that you did not have to break journeys for refreshments. 

 

CME's were under pressure, and some did not make a great success of the new, bigger, engines (Drummond on the LSW comes to mind, but he wasn't the only one, a sure hand with an inside cylinder late Victorian style 4-4-0, but less so with the Paddleboxes).  The culture was small c conservative, hardly surprising given the career-threatening result of failure, and it was only a decade or so since Dean, Johnson, and Stirling had taken the availability of steam sanding to revert to singles for fast work.  Innovation was needed, but fraught with risk, so the CMEs were between a rock and a hard place and very much in the eye of the public and their comany's shareholders.

 

The new century brought new problems to be solved, and Churchward coped with this with a combination of workshop practice such as optical alignment of frames and study of what was working well eleswhere, notably France and the US.  From France came the de Glehn 4-cylinder layout and from the US the cyldiner-and-half-saddle front end, both of which became a staple of GW loco practice.  His locos were considered shockingly ugly and utiltiarian at the time, and the response was to soften the brutal lines with curved drops to the running plates. 

 

Inside motion wasn't much of a maintenance problem on Churchward locomotives, as there was plenty of room to access it between the boiler and the inside edge of the running plate without much difficulty, and men had been used to doing this for years anyway.  It was only when the larger boilers of the Castles and Kings made access more problematic and required scrambling about under the loco that it became an issue, perhaps an indication that Collett was pushing the Churchward concept too far and might have done better designing from scratch than developing the Stars.  Now, there is fertile ground for imaginary locomotives!

Edited May 19, 2022 by The Johnster

[MikeOxon]

John Gibson, who always loved a good conspiracy theory, claimed in his book 'GW Locomotive Design' that the inside Walschaerts gear on the 4-cylinder engines was the result of a 'cover up'. 

 

The first 4-cyl engine, No.40 'North Star', was fitted with a crossover arrangement of Stephenson gear, known as 'scissors' gear.  While there has been a suggestion that this infringed a Patent by Deeley, Gibson claims that the real reason was that it was very difficult and time-consuming to set-up the scissors gear correctly.  What was acceptable on a single engine would not do for a large class.  Churchward realised that the existing design could be converted to inside Walschaerts gear very discreetly, so hid his 'mistake' in this way.  It did mean that this arrangement of Walschaerts had higher friction from the eccentrics, each driving two valves, than the usual arrangement of a return crank outside the wheels.  Gibson claimed that this friction inhibited the performance of the 'Kings' at high speed.

[rodent279]

Inside valve gear on a 2 outside cylinder engine I can understand-there's plenty of space between the frames for the gear,  and for access. 

On a 4 cylinder engine,  with 2 inside cylinders,  inside gear seems a little like making life difficult for oneself. 

Edited May 19, 2022 by rodent279

[tythatguy1312]

23 minutes ago, rodent279 said:

On a 4 cylinder engine,  with 2 inside cylinders,  inside gear seems a little like making life difficult for oneself. 

honestly, with all the issues the Kings had for their power, I'm beginning to think that Gresley's A1's truly were superior for their smaller power output. Don't get me wrong the A1's aren't fun to maintain but at least there wasn't 4 sets of valve gear jammed between the frames

[JimC]

50 minutes ago, MikeOxon said:

John Gibson, who always loved a good conspiracy theory, claimed in his book 'GW Locomotive Design' that the inside Walschaerts gear on the 4-cylinder engines was the result of a 'cover up'. 

Yep, and Gibson's conspiracy theories often don't stand up to detailed analysis.

I know of four issues with the scissors gear in various sources, and of course they could all be true! Limp home on two cylinders after a failure, accuracy of valve events, Deeley's patent and valve setting as per Gibson.  Personally I don't find any of them very convincing on their own, but add them all together...

As for the inside Walschaerts, I've tried drawing outside Walschaerts for the Stars and Castles I can't make it fit, hence the 4 sets of gear on the first Stanier Pacifics. To have outside gear with rockers to drive the inside cylinders it either needs rockers in front of the cylinders, with all the valve timing issues from expansion, or else a complete reworking of the cylinder/wheel relationship as with the Duchesses. 

 

9 minutes ago, tythatguy1312 said:

 at least there wasn't 4 sets of valve gear jammed between the frames

Only two sets, and being Walschaerts there were only two eccentrics plus the big ends, as opposed to four eccentrics and the big ends on all the thousands of inside cylinder locomotives with Stephenson's gear.

Edited May 19, 2022 by JimC

[Compound2632]

3 hours ago, Hroth said:

As for outside valve gear, it ain't pretty!

 

Churchward wasn't much interested in the prettiness of his engines. We're inured to the distinctive Churchward look of GW engines  and forget just how shocking what he was doing was to the aesthetic expectations of the day. It's like how we hear Stravinsky's Rite of Spring.

Edited May 19, 2022 by Compound2632

[MikeOxon]

1 hour ago, JimC said:

Limp home on two cylinders after a failure, accuracy of valve events, Deeley's patent and valve setting as per Gibson. 

The fact that No.40 retained its scissors gear suggests that the Patent argument was spurious and that the valve events were satisfactory.

[The Johnster]

3 hours ago, tythatguy1312 said:

honestly, with all the issues the Kings had for their power, I'm beginning to think that Gresley's A1's truly were superior for their smaller power output. Don't get me wrong the A1's aren't fun to maintain but at least there wasn't 4 sets of valve gear jammed between the frames

 

Well, there's different measures of 'superior'.  The Gresley A1s were 'inferior' to Castles in that the Castles could do the same work to the same timings with the same loads for less coal, and Gresley responded by upgrading them to the A3 specification.  It was never determined if that made them equal to or better than Castles in any respect, coal consumption for work done or otherwise, but FS as an A3 managed the first 'official' 100mph and the longest ever non-stop run for a steam loco, so it was a pretty good engine in it's day!  How an A3 compared in any way to a King I have no idea, and the differences may be too great for any meaningful comparison to be made.  The nearest equivalent to a King, an 4-cylinder 4-6-0 that BR would class as 8P for power, was a Lord Nelson, and I'd say that there probably wasn't much in it until the Kings were given new superheaters and double chimneys in the 50s, which probably gave them the edge, but at the cost of shortening their useful lives by breaking the frames, a problem with Ivatt's upgraded Royal Scots as well.

[JimC]

2 hours ago, MikeOxon said:

The fact that No.40 retained its scissors gear suggests that the Patent argument was spurious and that the valve events were satisfactory.

Don't forget 40 only retained the scissors gear until 1929: it wasn't retained when she got Castle cylinders. There's an argument to say that there's a region between "not good enough to repeat" and "good enough to live with". And if I understand what Don Ashton had to say correctly the vast majority of locomotives had worse valve events than the GWR ones. Cook says in his book "GW locomotives had extreme regularity in their exhaust beats" .   

But I must admit I have my doubts about the patent argument too, since 40s gear was rather different (and superior in detail) to Deeley's concept and there had been others in the same vein before. But patents are funny things, so much is in the fine detail of the wording as to whether something is covered or not. Holcroft says that the timing of the patent application and grant and 40s construction was such that 40 was in the clear, but any more would have had to pay a license fee.

[Compound2632]

36 minutes ago, The Johnster said:

Well, there's different measures of 'superior'.  The Gresley A1s were 'inferior' to Castles in that the Castles could do the same work to the same timings with the same loads for less coal, 

 

Well, yes, but they were designed to do different work on different coal. 

 

38 minutes ago, The Johnster said:

FS as an A3 managed the first 'official' 100mph and the longest ever non-stop run for a steam loco, so it was a pretty good engine in it's day!  

 

Wasn't Flying Scotsman an A1 at the time? As to the longest ever non-stop run, in 1928 the LMS staged demonstration Euston - Glasgow and Euston - Edinburgh runs with Royal Scot No. 6113 and Compound No. 1054, both longer distances and over a harder route than Kings Cross - Edinburgh.

[tythatguy1312]

1 hour ago, The Johnster said:

the Castles could do the same work to the same timings with the same loads for less coal

realistically speaking, that was due to the coal itself. Welsh coal, which the GWR extensively used, was almost competitive with fuel oil for locomotives, and actively proved superior for the day-to-day operation of Great Western Trains. However, crucially, the A1's were designed with Newcastle Coal in mind, which was lower quality and needed a larger ashpan and firebox to be used effectively. Hence, the Castles could get away with a smaller firebox for the same performance as the A1's.
It's also of note that the Castle has 4 cylinders, whilst the A1 has 3. This meant the Castle was more powerful, yet the A1's were easier going on maintenance for its work, even with Gresley's conjugated valve gear.

[kingmender]

4 hours ago, tythatguy1312 said:

honestly, with all the issues the Kings had for their power, I'm beginning to think that Gresley's A1's truly were superior for their smaller power output. Don't get me wrong the A1's aren't fun to maintain but at least there wasn't 4 sets of valve gear jammed between the frames

As Jim C pointed out there are only 2 sets of valve gear, the rocker arms drive the outside pistons. Plenty of room inside the frames too, even with the airpump. Not too sure about smaller GWR locos though!

[The Johnster]

The GW was lucky to have the South Wales coalfield within it's territory, which meant that the cost difference between using Aberdare Steam Coal and coal from other areas was lessened.  That said, the LMS as successor to the LNWR had access to South Wales coal as well, with lines in the Sirhowy (Tredegar) valley and the pits on the Central Wales Craven Arms-Swansea Victoria route, yet they neve exploited this for loco fuel.  And, IIRC, Pendennis Castle in the 1925 exchange trials used Yorkshire coal while working on the LNER, while Victor Wild used Welsh coal while working on the GW, the GW engine turning in better coal consumption figures in both cases.  I am not completely convinced that Welsh coal was as important to GW loco practice as is sometimes claimed; it happened to be conveniently available and was therefore used, even in the southwest of England where it was imported by sea.

 

If you compare a Hall to a Black 5, or B1, or BR standard 5MT, they are broadly similar engines with similar boilers and narrow bottomed fireboxes, the main difference being the Hall's valve gear.  So, if superior performance was predicated on Welsh coal, the Halls should have been noticeably better than the others, but were in fact on a par by and large.  The other classes were allocated to sheds from Bournmouth to the far north of Scotland, and used many different types of coal, but I have not heard that the different coals made any significant difference to performance; a Black 5 performed more or less the same everywhere, even when they visited South Wales sheds in late steam days and were fed Welsh coal.  Certainly different coal required different firing techniques, and a less experienced fireman having to deal with unfamiliar coal might come unstuck because of this, but this is not really a matter catered for in the design of the locomotives.

[JimC]

10 hours ago, tythatguy1312 said:

realistically speaking, that was due to the coal itself. 

I believe this is overstated. On both exchanges, LNER and LMS, the Castle used the local coal. 

[Mike 84C]

We are all concentrating on fuel and valve gear. One link in the chain is the regulator, the GW type was not just streets in front it was light years in front. Sensitive, light to use and pretty much instant in response. Why it was not adopted for use on BR stds beats me. 

  Is my memory playing tricks but I seem to remember that the Stevenson valve gear was set to admit to lead steam with a longer port opening which made the loco "sit down" on the drivers, hence the lack of slipping and the GW bark.

  I do have experience of all the locos mentioned in Johnsters post

[The Johnster]

My childhood memories of up trains starting from Cardiff General supports the ‘sitting down on the drivers’ impression.  The exit from p1 and p2 is sharply curved left, and slightly uphill for a short distance, then level over the canal bridge before dropping steeply to take the SWML under the CEJ bridge.  A Britannia would make a proper fuss over pulling out; slipping, stalling, blowing a hole in the sky, sparks and hot coals all over the place, not really getting to grips (literally) until it had managed to persuade the first few coaches on to the downhill part, even on a dry rail.  This is of course par for the pacific course, but a better regulator might have helped.  It would take several minutes for the last coach to clear the end of the platform
 

A Castle, or for that matter a Hall or a Hawksworth County, seemed to simply sit back and plod steadily off in the up direction, accelerating gently but steadily and displaying the sharp barking exhaust you mention.  No fuss, the fore stayed in the firebox, and much less sense of an ‘event’!  The Kings simply laughed at the task.  
 

 Watched ‘Britannia’ itself take 12 bogies out of p2 in 2019, and was looking forward to some photogenic fireworks, but in the event there was very little bother compared to the good old days.  ‘Tornado’, 12 and a dead 47 from p3 on a wet rail that November, proved every bit as capable as a King; 8P makes a difference!

[AlfaZagato]

Squatting is a known aid in traction & acceleration.   That's why drag racers first did the elevated front suspensions (Gassers,) then the 70's trend for high, but relatively weak rear springs.

 

I'm curious how throttle operation causes a steam locomotive to 'squat.'  Can't imagine it was good for hammer blow.   

 

I'm also not suggesting a 'gasser' locomotive, but I'm curious how the T9's fared in comparison to other similarly dimensioned 'Americans' with more conventional, lower-mounted boilers.

[Miss Prism]

4 hours ago, AlfaZagato said:

I'm curious how throttle operation causes a steam locomotive to 'squat.' 

 

http://www.clag.org.uk/41-0rev.html#annex4

 

  

[tythatguy1312]

4 hours ago, AlfaZagato said:

I'm also not suggesting a 'gasser' locomotive

this isn't a remotely serious proposal but a squatting 4-2-0 might be both a speed demon and hilarious to observe

[RedGemAlchemist]

4 minutes ago, tythatguy1312 said:

this isn't a remotely serious proposal but a squatting 4-2-0 might be both a speed demon and hilarious to observe

Agreed. A Crampton locomotive with drag suspension would be absolutely hysterical and also kind of terrifying.

Edited May 20, 2022 by RedGemAlchemist

[MikeOxon]

Good to see 'imagination' returning to this thread - a Crampton Dragster - now that would be something :)

[PhilJ W]

How would it stay on the rails?