Sir Nigel Gresley - The L.N.E.R.’s First C.M.E.

It makes me wonder though, supposing Gresley's junior draughtsmen had been better at designing reliable and maintainable bearings for both big ends and conjugated gear, would we be having this dialogue?

 

Given that conjugated gear in general was never widely adopted outside the LNER I suggest that bearings alone might not have been the issue

 

I saw "Heavy Harry" a couple of times when I was working in Melbourne a few years ago as I had some time to visit the museum - a very impressive machine. I still can't get over the humble little Class 08 making its way there though!

 

Slightly tangential perhaps but given that Harrison adopted Caprotti Valve Gear for his Pacific design does this have any bearing on Gresley and his attitude to that form of valve gear ? IIRC some of Robinson's GCR 4-6-0s (LNER B3 = 6166-68; had this form of valve gear and I believe that Gresley adopted Lentz rotary valve gear on some of his locomotives (Holden B12 rebuilds by Gresley; Gresley D49 62736 - 62775); is there any papers on the results of this experience ?

 

Other railways had designs of inside big ends that gave better performance so perhaps Gresley should have told his design team to find out why. Is it possible to compare the failure statistics for the Gresley engines with the Royal Scots, Patriots and Jubilees?

Regarding the conjugated gear, wear in all valve gear bearings (and sliding components) has an effect on the middle cylinder valve events. So making improvements is much more difficult than for single St's of valve gear. (I have not expressed this very well)

Sent from my SM-A105FN using Tapatalk

 

That statement was indeed perfectly correct but it was your making the contrast with divided drive (rather than with the third set of valve gear) which suggested that divided drive was significant. As someone has said up thread, the numbers of locos with divided drive and the numbers with unified drive suggest that that design choice in itself is not very significant, except insofar as it can influence other design choices.

 

Some Continental engines had a third inside valve gear driven from an external crank. Most obvious on Czechoslovakian 3-cylinder engines, where the extra crank was placed on the axle behind the one with the cranks for the outside cylinders. This option might have overcome a possible lack of space for an inside eccentric arrangement, but obviously introduces other issues including that of British loading-gauge width constraints.

http://www.railfaneurope.net/pix/cz/steam/387/CSD_387043_Luzna.jpg

I think the Victorian Railways were unique in trying two different conjugated valve-gear solutions - Gresley gear on their S-class Pacifics and Henschel on the "Heavy Harry". Neighbouring New South Wales used Gresley gear on the D57 4-8-2 and then a "rack and pinion" arrangement on the subsequent D58. But the Australians reverted to just 2 cylinders for their final generation of steam express engines, the NSW C38-class 4-6-2 and Victorian R-class 4-6-4.

We've seen an interesting discussion in the last few pages about unified versus divided drive, where it seems to have come down to engineers' judgements and personal preferences as to which way they wanted to go - unified drive on either first or second coupled axle, or divided drive with equal (or nearly equal) or unequal connecting rods. Britain managed to use just about all possible arrangements for both 3-cyl and 4-cyl types, although there appears to have been only a single example of a 4-cyl unified drive onto second axle - Gresley's 1915 rebuild of GN Atlantic No 279.

https://railway-photography.smugmug...-Atlantics-/Ivatt-C1-C2-Atlantics/i-vhrswRf/A

 

I hope no-one here would disagree with that (except perhaps on the aesthetics and slight benefit of the A4s' streamlining). But was there anything to choose between them in the 50s and 60s after the A4s had been fitted with Kylchap exhausts and improved middle big-ends and the conjugated gear was being properly maintained?

 

Mr Cox mentions that the Royal Scots big ends were trouble free and Stanier/ Gresleys were not.

 

An interesting statement by Mr Cox. The inside big ends on 11 Royal Scots were fitted with roller bearings in the late 1950s.

 

After I posted my comments it occurred to me that the answers might have been in the Cox report. It would be interesting to see a comparison of big end failure figures year by year. Perhaps the reduced maintenance in WW2 had a greater effect on the Gresley engines than the 3 cylinder LMS ones.



Sent from my SM-A105FN using Tapatalk

 

I'll be interested to see Simon's response, but I really recommend Townend's books Top Shed and East Coast Pacifics at work. Not at all pricey and worth it for the photos alone.

You are talking about the 1957-61 period. Ideally you would want a whole series of performance indicators --- (a) direct operating costs, mainly coal ; (b) maintenance costs over the cycle ; (c) availability ratios, stopped on shed, failed in service ; (d) crew comfort ; (e) type of work done.

My conjecture is that the A1 would win on (a), (b) and (c). On (d), probably the A3s and A4s would win, the A1s being a bit susceptible to high speed oscillation though not as hard a ride as the Britannias.

On (e) clearly there was a high degree of overlap between the classes in work done, but perhaps not 100% overlap. For the heavy trains eg KX to Leeds, Bradford and Hull portion trains, the A1s would be preferred. For the light trains on special timings (Talisman) the A4s would be preferred. My impression is that the Newcastle to London lodging turns tended to be A3s and A4s rather than A1s in that period, but that might be memory playing tricks.

 

Yes, and it was the main comparison given in ES Coxs report, and the stats are interesting. I’ll copy them in and post them here again when I can.

 

The relevant paragraphs are here:

 

Yes, mileages and availability were still significantly higher. The A4s were averaging around 40,000 miles a year and the Peppercorn A1s around 70,000 miles a year, on similar work (sometimes, arguably, more strenuous by way of the tare load being higher).

The issues of the conjugated valve gear never entirely disappeared. Most of the LNER locomotives fitted with it were still handicapped to 65% cut off (under Thompson , six A4s were modified to have 75% cut-off and they were the only ones so fitted), the conjugated gear would wear and cause steam distribution issues, over-travel was never entirely eliminated.

When maintained well and with the kylchap they were of course good engines, but it was only bringing them into line with the original four so fitted (Mallard, Capercaillie, Seagull and Peregrine).

The point that is being made is not that they could not be made to run well, but that as an asset to the railway, by the 1950s and 60s the Peppercorn A1 was performing better in mileages and availability and, arguably, was fully capable of doing the same work as the A4s. This is the thing we need to be more clear on: nobody is denigrating the Gresley A4s by acknowledging the excellence of the Peppercorn A1s. If the A1s weren't better than the A4s, something would have had to have gone badly wrong in-between (which we now know of course, it didn't, despite secondary evidence giving us some views to the contrary).

 

Perhaps a point worth considering is the time scales involved: The A4s lasted for over 30 years (i.e. 1935 - 1966); the A1s lasted for 20 years (i.e. 1946 - 1966) but the life of steam was influenced by the introduction of "modern traction". That timescale does not denigrate either design as, given the evolution of steam traction, a 20-year working life would normally expect a replacement design to be in vogue hence the A1s replacing the A4s is surely a part of the natural cycle of progress.
An interesting comparison methinks is that of the evolution of traction on the LNER routes; the Class 55 lasted for 20 years (1961 - 1981), the replacement HSTs lasted for 40 years (i.e. 1977 - 2017); the Class 91s lasted 35 years (i.e. 1988 - 2023) and the IEPs introduced in 2018 are expected to operate for at least 20 years.

Looking at these timescales it would appear that the Gresley Pacific designs were deemed capable for a lengthy period of time despite the considered "weakness" of the 3-cylinder conjugated valve gear arrangement. I venture to suggest that staff were so familiar with this foible that it was never identified as a problem at shed level but simply as "a job to be done" hence it was only at management level where comparisons of locomotive availability identified the foible as a serious problem.

As an aside perhaps one wonders about the contribution of the A3s to the discussion given that they were considered part of the express Pacific stock and given that the A4s were nominally improved A3s with streamlining.

 

Humans are prone to a certain complacency or acceptance of mundane failings, so things that damn well ought to be fixed are lived with and effort put on more exciting projects. I was about to label it as a steam era problem until I thought of plenty of examples from my IT career. I recall proposing we did an analysis of help desk calls, pick out the top 5 issues that were generating calls and kick off a project to improve them. 'Great idea' says management. Did it happen? Nope. There was always something more exciting, or some executive's pet project.

 

I forget the book but the author wrote that the A1 was a fine loco and was the ideal choice for the heaviest trains but as loads decreased, the 50 sq. ft. grate was less than ideal as the fireman would be shovelling as much to keep the grate covered as to generate steam. A4/A3 being the better choice for these duties as the smaller grate could still produce all the power needed for lighter trains.

 

Something that is fortunately beginning to change, as organisations start to realise how much they need to ensure that they eke out improvements using what they have.

Much of this discussion has revolved around data. It’s worth noting, as with much of railway history, that the availability of data has been limited and usability worse - the IT of the last 60 years has transformed the ability of managers to identify and drill into phenomena. Taking the questions around conjugation and divided drive, I’m sure that had modern telemetry been available, tangible data might have been available to support interpretation of which factors had what influence on performance.

Looking at history, we need to remember that those we’re studying lacked those advantages.


Sent from my iPad using Tapatalk