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

The question came up in the thread earlier as to why we haven’t discussed the more everyday locomotives. It’s a good question.

I am doing a talk for the Gresley Society over zoom this Saturday, which will like the MRC lecture, showcase the research I have undertaken for which I will be covering all of Gresley’s designs. Please feel free to join and observe/ask questions.

 

based on the table from earlier, entering it into my spreadsheet that calculates aerodynamic drag, (assuming 11 meters squared frontal area.)

bhp for 100mph - second number is bhp divided by Meters

HST - 1543 - 5.14
91 - 1487 - 7.55 *(is that length of the wind tunnel model!?)
390 - 1358 - 6.23
47+ mk1s - 1318 - 9.62
150 - 602 - 10.03
165 - 385 - 8.37
319(4) - 828 - 10.48
319(8) - 1479 - 9.36
357 - 723 - 8.82

And for the HST speed (mph) and bhp required for drag just to show how the numbers are fairly significant even before 100mph.

30 - 41
40 - 99
50 - 193
60 - 333
70 - 529
80 - 790
90 - 1125

pretty significant values and that's for a streamlined train.

The loco + 10 Eurofirma saves 100bhp compared to loco + 10 SJB7 at 70mph, i suppose not a huge saving, not sure what SJB7 carriages look like though, and that's with the 11 meter squared frontal area, so might be more given European loading gauge.

 

Out of interest, and just because it floated through my head, do you consider there to be a difference between (among the everyday classes) clearly Gresley designs and what I might call evolutionary designs? So as an example, the V1/V3 family are Gresley in every obvious way: 3-cylinders*, Doncaster aesthetics. Whereas the N2, which is a superheated, large-boilered N1, to me feels like a class built as an expedient (in considerable numbers). The same might go for the J38/J39 family, which have strong hints of NER 0-6-0s. Of course, we know Gresley made more direct use of pre-grouping designs in the form of the Directors and the N7s, to name a couple. Although any distinction between 'Gresley' and 'evolutionary' would be a bit arbitrary, it would be interesting to consider their relative merits as assets.

As an aside, I hadn't realised until fairly recently what a strong advocate of Sentinel products Gresley was. I knew about the Y1 and Y3 but had no knowledge at all of the large number of Sentinel railcars the LNER bought in his time (a non-trivial proportion of the total production), or of the Doble compound design (similar to the Colombian locos) that was ordered but never delivered, after Sentinel's MD strongly advised them to drop the order, on the basis that he did not think it would satisfy the LNER in performance or reliability.

*If the V4 design gets flak for being over-complicated for the task (as opposed to a good design, not necessarily at the right time), I don't know what the critics make of the V3.

 

The N2 is a Gresley upgrade of an Ivatt design, but it’s still under his direction. It’s a Gresley product.

The J38 and J39 are Darlington takes on the J6. I go into great detail on this in the book. And that final form of the J6 is still a Gresley product. It is in fact his first G.N.R. Design.

Yes, the sentinels were an interesting set of orders. I haven’t covered them in great detail, as they were ordered by Gresley rather than designed by Gresley.

So I actually commend the original design - the V1 - as one of the classes that “won the war” - in my book. Make of that what you will!

 

Supermarine were very proud of their flush riveting.

It was a production nightmare however and there was work done to work out where flush could be replaced with ordinary rivets without detriment to the Spitfires performance.

 

The first J6 locos were being built in 1911 just as Gresley took over from Ivatt on the GNR, so I think the design work must have been done while Ivatt was still in office. It was in any case a development with superheater and piston valves of the previous Ivatt design (LNER J5). However, Gresley's first 2-6-0 emerged the following year and was certainly a completely new design.

The J38/J39 was a much heavier loco than the J6, or any of the other 0-6-0 types inherited by the LNER. That would have restricted its route availability, but there would have been more than enough of the older 0-6-0s to cover work on weight-restricted routes.

The V1/V3 was not Britain's only class of 3-cylinder tank engine, but it was the most numerous (with the runner-up being the ex-NER H1 4-4-4T, rebuilt by Gresley to A8). One might have expected the V1/V3 to be on Edward Thompson's list for rebuilding to 2-cylinder form, but I don't recall seeing that in accounts of Thompson's plans. Perhaps he judged that the V1/V3s were working satisfactorily so could be left alone.

 

If I remember what I read correctly, they built a entirely flush riveted prototype and over several test flights glued on as many dried split peas as they could before the performance dropped below the required threshold? Quite a practical example of lots of small differences adding up to something greater.

 

There are actually a multitude of differences between the Ivatt design of 0-6-0 and that Gresley put out, including a re-design of the horn blocks, boiler placement and piston valves fitted. The Gresley variant is enough of a design difference to justify Gresley's name over Ivatt's.

That's true, but the J39 still ended up as Gresley's most built class.

Actually there were some investigations as to a 2-cylinder V1/V3 but were later dropped - there are drawings of this . I have a feeling, though I can't prove it, that Thompson must have examined the maintenance records and Use of Engine Power document statistics in some way as the Gresley V1/V3s were performing satisfactorily during the war. The V3 was later chosen as a non-standard type to be maintained, so that indicates a change of thinking from Thompson.

Drawing below is in the RCTS Green Book, Volume 9B, Page 44.

 

Really ?? I thought that Stanier built 37 Class 4 2-6-4T (LMS 2500 - 2536) in the 1930s for service on the London Tilbury and Southend route between Fenchurch St and Shoeburyness but then those were never used elsewhere hence remained hidden to many; 2500, however, IIRC is available to view in the NRM

 

I think he may be referring to it being the only three cylinder 2-6-2T- I am unsure if there are any others extant in the UK?

 

He said it “was not Britain’s only class of three cylinder tank engine,” as opposed to “Britain’s only class of three cylinder tank engine”. I think you’ve misinterpreted the post.

 

Its an interesting point though. If one considers a more obvious evolutionary design, the GWR pannier tanks you can see a straightforward evolutionary development through the careers of 5 CMEs. Doubtless each CME would have made strategic decisions:-
lets try superheating; lets abandon superheating;
and even tactical ones :-
too many firebox tube plate cracks in the corners; what can we do about it?
But they weren't clean sheet of paper designs in the way that other designs were. The CME was of course responsible for success and failure either way and would have made the decision to produce an evolutionary design rather than something all new, but instinctively one feels there is a qualitative difference.

 

Thank-you. I think the word "not" had been overlooked in my previous post.

For the record, the story of British 3-cylinder tanks began in 1903 with James Holden's experimental 0-10-0T.

https://www.gersociety.org.uk/index.php/locomotives/j-holden/a55

 

Agree with you Jim. It makes for interesting reading though. In my book, I've listed 25 locomotives design as directly by Gresley. There's a further 11 that had modifications or were built new, with modifications. It means Gresley effectively had a hand in a further 445 locomotives of 11 classes.

So nearly 3000 steam locomotives from Gresley overall for the L.N.E.R. in his role as G.N.R. and L.N.E.R. C.M.E.

 

Always strikes me that there is a lot of inconsistency in such things. For example, a 1936 built Southern Railway S15 differs from a 1920 example primarily in slightly smaller cylinders and slightly higher boiler pressure; revised valve events and cab profile / chimney height to fit the composite loading gauge. Yet the 1920 example is a "Urie S15" and the 1936 one a "Maunsell S15".

Compare with a 1923 and 1950 Castle, and I'd suggest the differences are at least as great between the two extremes, but they are all "Collett Castles" even though Collett had been retired 10 years before the last ones were turned out.

There are other examples I'm sure, but I don't think there is any consistency in how such things are named; and I think a lot goes to latter-day enthusiast (and writer) perception. Collett was a "big name" and the BR(W) CME of 1950 wasn't, so Collett gets the credit for any Castle even those with modifications that post-date his tenure. Whereas in Southern circles, Maunsell is a "bigger name" than Urie, so Urie doesn't get the same post-office credit for development of his designs; they get ascribed to the more famous name. (By the same token, Urie took Drummond 700 class 0-6-0s and fitted new superheated boilers, but they remained "Drummond" 700s in any book. It just seems Urie isn't a well revered name even though he did an awful lot to improve the rather poor legacy left by Drummond. The T9 is justly celebrated - but largely for what it was after Urie sorted out the boiler ...).

Not something to get too hung up on until you start wanting to look at statistics. "How many locos did Maunsell design?" Hmm, that's not easy to answer. Most books credit the S15 (and N15 / King Arthur) to him, but realistically he did precisely what a good CME should do: made recommendations (which were worked up via the drawing office) to areas where the original could be improved (valve events; route availability) but left well alone where there was no gain to be made (the basic frame layout and fundamental design of the structure of the loco). I'm sure the same with Gresley: the important thing isn't whether the N2 is a Gresley or Ivatt loco, but whether in the design changes Gresley made, he concentrated on what needed improving and left unchanged what was perfectly serviceable from the original.

Tom

 

Ah yes, Urie is rather poorly served by losing credit in both directions, entirely without consistency.

To go back to my original point, I wasn't trying to suggest that the N2 (for instance) was an Ivatt design for which Gresley is now given credit. But, as you say, the demands of evolving an existing design are subtly different to those in starting with a clean sheet. Some engineers are superb at evolutionary work, identifying weakness in an existing design and sympathetically updating it, or identifying potential and significantly modifying it to meet a new requirement. But maybe when given a completely clean sheet design the lack of an existing concept to work from throws them completely. Others are much happier with clean sheets and their attempts at updating existing designs have a habit of losing the virtues and introducing new vices in the process of trying to address problems that didn't exist. Gresley was responsible for both evolutionary designs and clean sheets, both with success. His big engine classes are a mix of clean sheet and evolution of his own work, but the medium power engines include some entirely of his own school and some that evolved from an existing lineage. I was wondering idly if there was any obvious trend in which approach resulted in more successful classes, in availability and mileage terms.

And yes, I am aware that such idle speculation conflicts with some of my previous comments about the difficulty in drawing robust and statistically valid inferences about different classes, on different duties, in different parts of the system.

 

And this takes us back, again, to the exam question of "what makes a good CME?". An engineer might be good at either or both of these things, and a "good engineer"; the role of CME goes beyond those facts to something more about what they did for the railway that they worked for.

So, taking Stanier as the example, we have someone who was not especially innovative in his engineering practice, but achieved a high level of standardisation in design and practice within an organisation that was previously far from standardised, while doing so with designs that effectively delivered the requirements of the day. That, to me, makes him a "good CME" even if not a great engineer (a point on which opinions will vary).

 

Ref the J6, I see from the RCTS green book that the first 15 of the class, built in 1911 and called the "521 series", differed from the remainder. Compared with the previous Ivatt J5, they introduced superheater and piston valves but did not have the additional modifications that you describe above, introduced by Gresley in 1912 for the later members of the class. It seems that the GNR regarded J5 and J6 together as a single class (J22 in the GNR classification).

It will be interesting to read your thoughts on the J39 as they develop. It is evident that Gresley produced a machine substantially larger and more powerful than any of the pre-Grouping 0-6-0s inherited by the LNER. It was also substantially larger and more powerful than the Midland 4F "Big Goods" 0-6-0 that the LMS proceeded to build in large numbers. I am not sure which was the better design choice here.

Thanks for the info and thoughts on the V1/V3. I'm unsure why the V1/V3 might have been apparently less troublesome than other conjugated gear classes. Perhaps being smaller, components were less stressed?

 

Noted; my apologies.

 

One quibble, I believe that the N2 boiler is the same diameter (albeit with a superheater) as the N1, but is pitched higher because the N2s have piston valves above the cylinders, rather than slide valves between them as on the N1.

Re Sentinel, the LNER also had a couple of cranes, the chassis was essentially a stretched loco type. Both were based in Scotland. I'll see if I can find a photo of one.