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

Smooth sides and boxy structure are better aerodynamically than typical steam locomotives, even if purely a result of function rather than form.

I admit i was sceptical given the claim intially about the W1, but i will admit i was wrong. Although it's implementation is a bit like fitting a streamlined top (windscreen upwards) to an old landrover with it's bluff front and exposed wheels maintained.

 

I

I’m inclined to the same opinion as maddog. You may have shown that Gresley may have set out to produce a streamlined loco but in terms of this I still think it was about as successful as the LNS Austin Seven was as a steam loco.

 

Hi Steve,

I’m inclined to point out to you that the locomotive is the earliest example of Gresley’s streamlining, and was required to be multi-functional at the front end.

I have now seen something of the primary evidence where the W1 is concerned, and in terms of its aerodynamics it’s actually quite good for a first attempt. It actually does some things better than either the A4s or the Coronation Scots, surprisingly.

It’s certainly by far better than the GWR’s “efforts” - !

As a locomotive I think the W1 has been highly misunderstood by us all, and actually from my research, I am inclined to stick up for the W1. I think Gresley was onto something rather cool and inspirational with his approach. It’s also pretty clear that the W1 was used as a basis in some respects for aspects of the A4, something that hasn’t been acknowledged before in print I think.

Gresley was very ahead of his time with the W1 in the late 1920s. It wasn’t far off working as intended, as William Brown has ably demonstrated with his work. If only Gresley had kept with it, but the A4 was more pressing in terms of development. And the work on the W1 didn’t entirely go to waste.

It is a pity that it retained such a poor reputation amongst LNER writers but I think that is down to lack of understanding rather than anything heinous.

 

I’ve no problems with the W1 as a locomotive and agree it was a fine attempt at improving on the basic Stephenson concept, which largely carried on unchanged except for superheating until the end of steam.

 

One thing that pretty well always amuses me, is that the original "water tube" boiler outlived the rest of 10000. As when it was decided to end the project and rebuild the engine with a "conventional" boiler, the original boiler ended up at North Road [Darlington] Works supplying steam/hot water etc to the works. The rest of the engine being withdrawn in IIRC 1960+/- and the "water tube" boiler being withdrawn/condemned in 1965-6 when North Road closed, so living on another 5-6 years beyond the rest of the 10000.

 

Oh, come now Simon, please give credit where credit is due. The GWR’s efforts were multifunctional too - they made both engines look fugly…

 

I'd love to see the different stream-liners properly assessed in CFD. It's complicated, just a bit, by the interaction with the ground and lots of bits spinning around and exhausts etc, also the fact that the locomotive is not in isolation, being at the front of a set of carriages which are probably equally, if not more important than what is on the front.

 

Has there been any significant push back other than to the streamlining claim? That boils down mostly to semantics, of exactly what the term "streamlined" implies for the designer's original intentions and for the eventual outcome.
Edit: As pointed out in the next post, there has been push back on several aspects of Gresley's work, but I had (perhaps wrongly) read Simon's comment as relating specifically to the streamlining of the W1.

 

From my perspective, there has been push back around conjugated valve gear, three cylinder layouts, the P2s, and whether Gresley was right to back steam v electric (and whether Raven would have done differently).

I don't think we've heard very much at all one way or the other about the works organisation, the running shed organisation, the carriage stock, and relatively little about the 'non-glamourous' 80% of the loco stock.

I think the first exam question Simon has to answer to his own satisfaction is ---- how do you judge the quality/greatness of a CME? It is clearly a very multi-talent type of job. In a way that was at the heart of the Gresley/Chapelon discussion.

 

I have found this thread fascinating as it shows how advanced much of the work being done on steam locomotives was in the early 20th century. But then it would have attracted the brightest and the best, as it was the fastest way of travelling at that time, when those engineers were starting their careers.
I am also thinking of OVS Bulleid and how he benefitted from being part of the Gresley design team. The work on sleeve valves was also part of the engineering challenges particularly current in the 1930's and 40's.
I believe OVSB may be the subject of Simon's next book?

 

Hope not. There’s enough been written about OVSB - how about a combined volume on the Riddles/Harrison/Bond era? Or just Harrison who, of course, might well’ve been a future LNER CME if it’d survived into the ‘50s but maybe the 1950s & 60s steam CME’s didn’t do enough each to justify a book each compared to their predecessors since steam development was winding down by then.

Of course Simon, if he ever wants to write any more books, will no doubt follow his own inspiration.

 

This table is from Johansen's paper to the IMechE in 1936, which Gresley anticipated in his presidential address. Apparently the LMS and LNER jointly hired the NPL to run the tests. The paper runs to 124 pages. The -ve resistance of the Royal Scot tender is because it had a smaller profile than the loco.

 

So what conclusion can be drawn from those figures?

 

While on the subject of 10000 here is my photo of it as 60700 taken around 1958 at Brookmans Park, between Potters Bar and Hatfield. Bearing in mind that it started off from Doncaster there is still a surprising amount of coal left in the tender.

 

The trite observation is that the apparently streamlined W1 seems no better than the entirely traditional 4472!

The caution I'd add would be that wind tunnel testing can be quite sensitive to scale; and certainly requires careful calibration, but what is presented is data at different scale and (presumably) from multiple different wind tunnels and experimental set ups. So I'd want to see the whole paper, but would be cautious about concluding too much from the table in isolation.

The comparison between 10000 and 4472 is a bit surprising but on the assumption that those two at least were under similar experimental conditions, does rather suggest that if the intention in the design of 10000 was to reduce air resistance, the LNER didn't do a very good job of it!

Tom

 

But the W1 is designed specifically to scoop air INSIDE the casing from the front end, AND also push it up and over the casing, both points on which it succeeded on most definitely - we know this because in the former, that’s how the locomotive’s draughting worked for the water tube boiler, in the latter we know this from photographs and reports on the locomotive’s smoke deflection properties.

I do not know the context of these figures but they seem to be suggesting that the tender of no.10,000 had more wind resistance than Scotsman’s, which given they are virtually identical designs is more than a trite puzzling. 10,000 does come out better than Scotsman on the locomotive, but not by a clear enough margin.

Personally, I’d like to see the context of Johanssen’s address before commenting further on the figures. They don’t add up to me and they are at odds with the wind resistance figures I have seen elsewhere.

I was also not aware that 4472 had been wind tunnel tested (!) - this is a new one on me.

 

Maybe this is something that could be done for a future book. Would be interesting to see how the various streamlined locomotives do!

 

I’m not sure the tender figures mean much. My hunch is that a measurement was done on the loco alone; and on the loco plus tender; and the tender figure was then just calculated by subtraction. But if so (and that is a guess) I don’t think the methodology is exactly sound - otherwise you’d just couple a bunch of Royal Scot tenders together and watch them magically pull a train unaided! The critical figure is the whole vehicle, not individual components.

The ground also has an impact: modern practice for testing cars is to have a moving surface under the car (on a continuous hand) to mimic the interaction between the vehicle and ground.

I think the best that can be said is that there is no evidence from those figures that any meaningful level of drag reduction was demonstrated for the W1. If there are other figures, feel free to share.

Tom

 

I am compelled to agree with you Tom on the basis of these figures, but Scotsman even being tested is something that hasn’t come up before in secondary evidence, and I have done a pretty deep dive, as has Tim Hillier Graves, on the primary research and neither of us knew this either (I think - I will go back to Tim and ask if he knew!)

It’s really interesting to see that the Royal Scot tender by virtue of being thinner than the locomotive ended up with negative figures. I have a hunch that the W1 tender came out worse than Scotsman’s because the W1 is, at the cab end, narrower than Scotsman and the tender, thus the flow of air from the W1 loco to its tender probably creates drag. An unintentional design flaw by way of utilising a standard tender design rather than one to match the locomotive’s design.

It’s fascinating stuff. I may well pay out for the full paper towards the book’s research today.

 

I now have the full paper and I will be reading it this week. Initial impression however is that there is a suspicion that the W1 is actually better streamlined at the locomotive end but the tender relationship creates greater drag on the whole, more than 4472. Really interesting stuff.