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

Does anyone here have a Grace's guide subscription? I think the correspondence following Gresley's original paper on 461 might help illuminate this discussion.

 

I did a bit of Googling yesterday afternoon. One thing that is interesting is the very large variability in the lifetime mileages of the A3s, with the highest (60059 mostly 34A) being more than twice the lowest (60095 of Carlisle Canal). It would be interesting to look at the annual mileages of the NER A3s because I would probably concede that the work profile of the fleet at Neville Hill in particular is more like the profile of the MNs.

I realise that Simon has individual record card data for the LNER fleet. Does equivalent data exist for the other companies/regions and what are the best sources to read up on it? I presume it must exist because it is widely stated that 46233 was the second highest mileage of the Duchesses at 1.64 million.

 

Probably stating the obvious but a 34A loco would have seen much more work than one from Carlisle Canal.

 

B17s and the V2s are both conjugated and more or less same age and crankshaft between frames do more or less same work.
Can a comparison of repairs indicate if unified drive was best?

 

Yes indeed, the context of my comment was how to control for variations in work diagrams between different classes of Pacific in comparing annual mileages. But reading it up I realised there are quite big variances within classes too. As well as location, there are differences in what happened towards the end of life, especially when the diesels came along. Maybe there should be a cut off date such as 1961 after which annual mileages are too heavily constrained by opportunity to be meaningful.

 

Regarding annual mileages, IIRC there came a point where these were either not recorded at all or at least not with any great detail.

 

Yes, I don't recall the numbers but 60009, the highest mileage A4, is quoted as ' x plus about y after it went on the Aberdeen duties'.

 

I would argue that the V2s did quite different work to the B17s much of the time. Plus the derived motion on the B17 was behind the cylinders as opposed to in front of the cylinders on the V2. Not a "like for like" comparison IMO.

 

That's interesting. The Henschel form of conjugated valve gear was used on the more than 1400 G12 3-cylinder 2-10-0s, but on very few other classes. I don't think any of the later Reichsbahn standard types ever used conjugated gear.

The only German express engines to have the feature appear to be ten 3-cylinder Pacifics built in 1917/8 for the Saxon Railways, subsequently designated by the DR as Class 18.0. These were very Gresley-like, with unified drive on the middle coupled-wheels as well as conjugated gear. In photographs, one can see the extra link in the outside valve gear, connecting back from the piston valve rod to toggle one of the cross-shafts. Unfortunately, available photos do not show the arrangement of levers between the frames.

https://en.wikipedia.org/wiki/Saxon_XVIII_H

These engines may not have had any influence on later German designs but must have been considered satisfactory, as most worked in East Germany until the early 1960s, without having undergone any major changes. They seem to have always kept to their home territory, operating from Dresden to Leipzig and Berlin. None were preserved.

 

Been pondering this for a few days and I was wondering if someone could explain to me why we need absolute like for like comparisons between locomotives?

Speaking with my asset engineer's hat on, absolute like for likes are not always possible, and in modern railway engineering we'd look at the different factors between rolling stock before making a decision on retention/disposal/renewal/overhaul/etc.

One thing which strikes me about the Gresley story that doesn't seem to happen with other engineers is there is this sort of innate desire to prove the worth of the conjugated valve gear, sometimes at the detriment of the gear's reputation.

Case in point: I know I can produce pretty good comparisons between the LNER Pacifics and make a case that those with conjugated valve gear during the war were doing poorer on the whole than those later engines with divided drive and three sets of valve gear. (These comparisons are made between classes doing the same work in the same areas, for instance).

Telling me that we can't make an informed judgement because one drives onto one axle and one drives onto two and therefore they're not comparable, seems inherently wrong. Of course we can compare them. We cannot definitively say the conjugated gear is the definite defining factor in one locomotive class doing less work for the company, but we can look at the overall primary evidence base and see it is a strong factor (case in point, ES Cox report).

The individual engine record cards are an eye opener too and allow even more close analysis between the classes. It's why I value Great Northern more now than I did when I was uninformed and much younger, and why I have a greater respect for the work Peter Townend did on the Gresley locomotives post-war, where changes to their maintenance structure, pooling and some upgrades resulted in locomotives performing much better.

But it doesn't lead us to the conclusion that the A4 Pacific was the best Pacific the LNER produced: that accolade goes to the Peppercorn A1 which was the better all rounder, borne out by doing the same work and having better mileages/availability.

Is this denigrating the A4 Pacific? Absolutely not, we can give credit to that design by saying (rightly) it was Gresley's crowning achievement, but a product of the 1930s and a locomotive designed in 1947/48 and in production by 1949 should rightly be better all round than the earlier locomotives. If it's not, something is badly wrong on the LNER - and it's also why the (often misinformed) criticisms of the Thompson Pacifics are more damaging to having an informed debate on LNER matters.

There is the fair point of certain sheds achieving substantially less mileages than others - that's where we look at the data for the lifetime of locomotives at those sheds and compare their datalines along their life to see if the lower mileages were a trend in work, or something else.

The point I am trying to make, in a roundabout way, is just because we can't have an absolutely exact comparison between locomotives, doesn't mean we can't comment on, and make informed comparisons between them whilst looking at the primary data.

 

To eliminate as many variables as possible?

 

If it's the case that we are not able to make exact comparisons of locomotive types, perhaps all those who have denigrated Thompson's work on the other thread over the last decade would like to retract their statements, as they have not provided like for like comparisons of the locomotives in question?

No? Didn't think so!

 

This is an image of the gear on what looks like an accurate model of Victorian Railways H220 "Heavy Harry" (not my image). It is difficult to get such a comprehensive picture from the actual loco.

 

What can be seen in that photo looks like a rocking shaft giving 180º out of phase for one inside cylinder; which would imply four cylinders in antiphase pairs as on a Castle or Stanier Pacific. If that's not the arrangement, please can we have some more explanation?

 

There were 124 Prussian S10.2 4-6-0 three cylinder (1914-16) with conjugated gear that was not repeated but not rebuilt either.
They were more powerfull than the fourcylinder simples,but much weaker than the fourcylindered Compound S10.1.
Most of the three cylindered worked to postWW2 and were the cheapest in maintenance and repair

WP_20230111_004[1] by Hermod posted Jan 11, 2023 at 4:25 PM

 

There are two cross-shafts whose motions are combined, the one driven from the combination lever on the other side is obscured underneath the nearside shorter shaft in the image of the model.

 

If you look up Victorian Railways H Class on Wikipedia there are some more photos there of Heavy Harry's Henschel conjugated valve gear.

Peter

 

Well, I think I may have provoked this question a little, so I'll do my best to explain what I was getting at. The statement of yours I picked up on was this:

On the contrary, we know divided drive was better in terms of maintenance and availability, and we can evidence that by the Use of Engine Power document, and the engine record cards. The Peppercorn A1s records alone compared to the best Gresley conjugated class (A4) proves that beyond reasonable doubt.

You can say the Peppercorn A1 was a better asset for the period in question. The data supports it quite convincingly for the relevant period and I'm willing to accept for the moment an assumption of sufficiently similar duties. Age of the asset comes into it as well, mind you, the Peppercorn A1s were also newer than everything else. But you've then said that therefore divided drive is better than unified drive and I don't think you can. Not with such a massive second variable as conjugated gear or not conjugated gear also in the equation, when you have emphasised previously the issues that could emerge with conjugated gear. We know that the conjugated gear degraded in a way that led to imbalanced work between cylinders, particularly favouring the middle cylinder and changing the valve events in a way that increased load on the inside big end and RH driving axlebox, and that for much of the time period the LNER Pacifics had a suspect inside big-end, so here we have a hypothesis for the cause of trouble and data that also points back to the source, to say nothing of contemporary reports on the problem.

So given that, if you did want to take a closer look at the unified/divided drive question, then you really want to be looking at designs where that's the only significant variable, and the LNER don't have them. Not 4470, not the Peppercorn A1s, not any of the A2s, because they all have both changes. All of this also ignores any questions of detail implementation; successful 3 and 4-cylinder classes were built in the UK and across Europe both with unified and with divided drive. My personal opinion from that is that neither was inherently inferior or superior enough to make the choice between them a significant factor, only how well either was designed.

I apologise if this is just re-framing the exact same point with the words in a different order. But you seem from the quote to think that I'm saying we can't compare conjugated and non-conjugated because of the unified/divided drive question. I'm actually saying the opposite; that it's much harder to compare unified/divided drive because of the conjugated gear muddying the waters.

 

But I'm not arguing against unified drive versus divided drive. That's never been my contention that one is better than the other.

The issue specifically for the LNER is whether conjugated valve gear was superseded by the divided drive setup, and we can show that it was, absolutely.

Besides which, comparisons between valve gear setups which are not like for like has been made throughout railway history. Not least the 1925 and 1948 exchange trials. We've seen debates between conjugated onto one axle, divided drive, three and four cylinders, two and three cylinders, chain driven gear, and more. It's a fallacy to say we can't make reasonable comparisons of assets and make an informed view on what we think the issues are.

The 1948 exchange trials could not have happened without being able to make observations on the different setups (and - point of order - the conjugated gear and middle big end was specifically criticised in the official report due to failures during the trials, in comparison to the four and three cylinder setups with four and three sets of valve gear being used, irrespective of whether they drove onto one or two axles. That did not factor into their thinking or comparisons at all).

If it was the case that we can't make reasonable comparisons of conjugated valve gear and divided drive because one is unified drive, then there's an awful lot of criticism out there for the Thompson Pacifics that needs complete retraction, as it's not a "like for like" comparison. As I said previously, that's unlikely to happen.

If you want to have the debate about unified/divided drive, go and have that. I'm not singling that out for criticism. I am commenting on the fact that the failure rate, mileages and availability achieved of the conjugated gear locomotives versus the later divided drive locomotives setup can be compared, with the later setup undoubtedly proving the better asset when all primary evidence is taken into consideration.

So we're effectively talking at cross purposes.

If I may say so though, it's a bit suspect to bring unified/divided drive into the discussion. Nobody as far as I can see has singled out unified drive for criticism, only in the specific instance where conjugated valve gear is concerned.

And - just in case people are thinking I am criticising Gresley for the choice of conjugated gear - I am not - it is reasonable for locomotive engineering to go through phases of change and improve over a thirty to forty year period. If it didn't, we would be in trouble!

 

Genuine non-loaded question. Would it have been possible to design and build a Peppercorn A1 in 1936? Or were there any technical, materials or knowledge advances in the following decade? For example, were the roller bearings, which seem to be associated with particularly low maintenance cost and high availability both for 60156 etc and for 46256/7, known technology pre-war?

I suppose another proposition for testing is that the A4 was the best in class on pre-war technology and the A1 was the best in class on post-war technology.