Did this improve performance? Streamlined King + Castle

They would if they could.
The Motor Industry Research Association has a wind tunnel where they can do full size testing of cars. The industry considers it sufficiently valuable to have spent the money.
Part of the cost restrictions in Formula One are on wind tunnels: https://www.formula1.com/content/fom-website/en/championship/inside-f1/rules-regs/Testing.html

This NASA fact sheet gives a little introduction to the subject.

http://www.nasa.gov/centers/langley/news/factsheets/windtunnels.html

 

Well frankly I did not feel the urge to go to the largest scale when I had the Hamburg windtunnel perform my tests of a drilling rig on a ship back in the seventies. One does not really need to know all the finer details of a locomotive test, it is comparative, one design against another. Above all, only a few questions need answering in a locomotive a) how much hp is saved and b) will it lift the exhaust gases sufficiently in most cases. This is not car/airplane/rocket science!
Kind regards
Jos Koopmans

 

Two other things to remember about the intended style of streamlining for the proposed 5AT:

This aspect of the design, and indeed the whole aesthetic treatment of the locomotive (the dreaded l***** word) probably had not been given any detailed consideration at the time the project was put on ice, though concepts of how the 5AT might have looked were published; and

The role of the 5AT would have been to haul steam excursions, and it would have had to LOOK like a traditional steam loco to be successful. Hence, I think, the rather minimal streamlining in the published designs. Adopting the most effective physically streamlined outer casing would not have worked if the result had looked like a Pendolino (or Leader?) with all the working parts hidden away. The punters would not have seen the point of paying extra for steam haulage if that had been the case.

"case" huh ? I'll get my coat ...

 

It should be remembered that the Gresley A4 design was wind tunnel tested with large scale models. OS Nock has stated in a few books that the thumbprint behind the chimney was accidental, but this seems incredibly unlikely given the amount of testing that was done with Humorist and a number of other A3s prior to the A4 design being finalised. The thumbprint behind the chimney is only part of the overall aerodynamic package and the complex curves on the smoke box itself also add to the streamlining and wind deflection significantly. When you look at how the chimney and smokebox arrangements were changed for Humorist in particular you can understand the level of understanding and development Gresley and his team went into to produce a design which could deflect smoke the most efficiently.

The idea that Gresley's work on the A3/A4s/W1 and P2s for smoke deflection and streamlining is in any way experimental hit and miss is absurd when you look at the amount of design work went into those classes and particularly the experimental front ends such as on Humorist, the Hush Hush and Cock O' The North.

On a side note: sorry Jim, but I don't think that you do know how I feel when I've been writing about Edward Thompson.

In the last six years when I've been researching and talking to people about Edward Thompson, I've received electronic and hand written hate mail, a large number of abusive messages via Facebook and on Twitter and found that anyone who does have new information on the man is unwilling, or wary of sharing it for fear of a backlash from a wider following elsewhere. Several of my sources have confided in me and there's a number of things I can't print because of that fear that they too would suffer a backlash for giving it. It is the only situation I have encountered like this in railway preservation and I fear that even if I were to publish my book, anything that I print would be largely discredited, not on its own merit or by rebuttal but purely through an almost malicious and cultish manner by the extreme viewpoint on the other side of the debate.

Best example I have of this is that because I had spoken with the author of an upcoming book on the P2s - he had asked for my views so he could gauge his own against his own knowledge - a gentlemen then stated he wouldn't buy said book because of my views and my "influence!" This despite my assurances that my involvement with and contributions were simply of an editorial nature and I pointed out that on certain points I disagreed with the author (but then nobody is going to agree on everything).

 

Why? He was very much around at the time and well connected. Why on earth would he write about it if the 'story' wasn't true?

 

The reason I question it is that if you look at a number of the weight diagrams for the A4s and some of the LNER's early drawings for them, the thumb print is present quite early on - and if you look at Humorist's design evolution up to the double chimney and including the sloped front with small deflectors, the gap behind the stovepipe chimney is the first thing to be changed on the design for the A4 after the early Cock O' the North style casing applied.

The other thing is I sometimes question OS Nock's reliability. He may have been well connected but if he's the only one saying it and it's not something recorded officially elsewhere, can we really rely on it to be absolutely true? This isn't the only OS Nock anecdote that gets questioned and I do sometimes wonder reading a few of his LNER themed books whether he was working from source. One particular thing which sticks out for me is his writing on Thompson which if you look at it chronologically is very praising, very supporting during the Thompson reign (he wrote several publications for the LNER whilst Thompson was CME) but after nationalisation and into the 70s the story completely flips on its head. Just doesn't seem consistent.

But hey, to be fair to the chimney thumbprint story stranger things have happened in railway history so who knows? I personally just question it as much of the design evolution in LNER history seems quite logical and the streamlined casing plus the stovepipe chimney all seems to be rational engineering development based on testing of prototypes.

 

I can't really comment on his LNER books, but I get a definite sense that as he got older, the quality of his written work went down, with an increasing sense to me that he was simply cranking a handle to produce books by the yard rather than from genuine interest or unique knowledge of the specific subject.

He was also clearly better on some subjects than others. For example, as a professional signal engineer, his descriptions and insight into major resignalling schemes are often excellent. But from a loco point of view, too often he ends up just giving a series of logs, with comments such as "an excellent run" or "a disappointing run" but without considering the conditions that may have lead to such. Too often I have felt that he can be unduly critical of the crew when reporting a poor run, without any feeling for the conditions (coal, state of the engine etc) under which they were working.

Tom

 

There are any number of contradictory stories from back in the day. Sometimes people mishear things, sometimes memory fails, sometimes stories grow in the telling, sometimes people have a weakness for a good story, sometimes people state as fact what is really only supposition, and sometimes there are chinese whispers... Sometimes perhaps people are tempted to say what they think the boss wants them to say, and even, if you believe some stories, sometimes things are covered up!

 

OK

 

Yes, I know all about scaling effects, I work with scale hydraulic models for my day job and have done aerodynamic wind tunnel testing in the past. The scaling effects can be taken into account, with suitable setup. Of course larger models are more accurate, but I think if the effects are not significant at small scale, I doubt they would be significant in operation.

 

By the way, we should not denigrate "experimental hit-and-miss", that is the scientific process! Albeit hopefully somewhat guided by theory and collaboration rather than just random variations.

 

If you want a wind tunnel I have access to one. The only problem is the 18'000 Euros per day for full use and analysis. If you want something bigger try my friends in France.
http://windtunnel.onera.fr/capabilities
They did the A380 studies. There is a picture of us during a visit on my son's blog(in Ialian)
http://jamesaviationblog.blogspot.it/2015/05/visita-alle-gallerie-del-vento-di-onera.html?m=1

 

This is very sad. Anyone would think you've been writing a book saying what wonderful chaps Hitler and Pol Pot were and how badly they've been misunderstood. The reaction to your work on Thompson is a classic case of "My mind is made up: please don't cloud the issue with facts".

 

I find that contemporary witnesses tend to mishear things the more their evidence disagrees with the preconceptions of those who have a contrary opinion.

 

On the subject of wind tunnels: A friend of mine is an aerodynamics team leader for a championship-winning F1 team. They ran models at 50% scale, and also needed a continuously moving belt underneath to simulate the interaction between the car and road, which are moving relative to each other. From what little I understood of the field, scaling up results is non trivial, not least because while you can scale some parameters (such as size) and control others (such as wind speed), some properties remain at resolutely 304.8mm:ft scale, such as air density. These days an awful lot is done by CFD in a computer rather than in a wind tunnel, but that would not really have been available to Gresley, Bulleid et al, even though the underlying equations were known.

One significant difference in modelling between the aerodynamics of aeroplanes and of ground-based vehicles (whether cars or trains) is that the inter-relationship between the moving vehicle and the static ground has a big impact which you don't get on aeroplanes.

Tom

 

Apart from seaplanes of course where their behaviour on water has to be considered but that involves a different type of testing.

 

With fluid dynamics, scaling has to be based on the governing non-dimensional number: for shallow-water flows Froude number, for low speed viscous effects (e.g. drag) Reynolds number, for supersonic Mach number, etc. So long as that number is kept constant (or in the same range) the scaling will work.
So e.g. you keep the same Reynolds number (assuming your viscosity is the same because you use the same fluid) by keeping the product of velocity and critical dimension (e.g. length) the same.
The issue comes when more than one phenomena is at work - e.g. viscous/inertia and depth - when it is difficult to keep all the relevant non-dimensional numbers in the right ranges.
This is why there are limits on scaling down - e.g. if you go too small, surface tension or adhesion or buoyancy or evaporation/condensation or other smaller effects start to go out of scale and behave differently.
However, for simple drag comparisons of the sort we are talking about, fairly small scale should be absolutely fine. If we're talking a few % differences in drag, then it won't happen in operation, whatever the laboratory results say.

 

A few slightly related points:

The importance of the moving floor when simulating an F1 car is that the relation of the car with the moving ground is critical to how it generates its downforce; in fact the amount they can extract from ground effect is very much restricted by current regulations, so modelling it correctly becomes rather significant. Tom, that description must really narrow down who your friend is to anyone familiar with the field ... I'm not enough of a rabid fan and never have been. For exhaust deflection on a steam locomotive, it's really the top half that becomes significant, so a moving floor is probably less important, especially when you consider how aerodynamically dirty the running gear is.

The Reynolds number is the important scaling factor for most work at the scale of a locomotive; if two objects of identical form but different scales are subjected to flow with the same Reynolds number the drag coefficient will be the same. However, it scales linearly with size and velocity, so if we were to use a model at say 1:8 scale (think 7.25" gauge, or fairly hefty by anyone's standards) we therefore need to multiply the velocity by 8. At 50mph this may be ok, for 90mph this has us firmly in the transonic regime so the scaling relation breaks down and the model is not representative at all. If we're at 1:4 scale, then compressibility effects become less significant but we still need a ~400mph wind tunnel.

However, if we stay subsonic then most of the unpredictable effects in aerodynamics occur at relatively low speeds, or when you're trying to extract a lot of lift (or downforce) from the flow. With a locomotive we are not asking the air flow to work for us in the same way as with a plane or racing car, we just want the train to slip through with minimum disturbance. Going back to our (still rather large) 1:8 model, if we have several of them then, as @JJG Koopmans and @andrewshimmin have said, relative performance will be fairly realistic, especially if all the rough and knobbly bits have been cleaned up with a streamlined casing. I reckon that models of an A4 will be rather more accurate than of Manorbier Castle...

 

Did not steam loco streamlining originate from an attempt to appear as up to date as the new diesel streamliners of the 1930s, in the USA and Germany for instance. As regards it's effectiveness at speeds below, say, 100 mph it is perhaps interesting to look how the easily streamlined diesel loco evolved. By the late 40s into the 1950s a more subtle, though related, rounded off at the edges design, such as 10000, Deltic and the Warship was adopted, by the 60s we see designs such as the 47s, Westerns, Hymeks etc which can in no way be described as streamlined. The one exception being the Blue Pullman which, interestingly was one of the 'ancestors' of the streamlined 125 mph HSTs of the 1980s.

But regarding 90 mph trains the 2nd generation express Sprinters again made no attempt at streamlining whatsoever.

With modern materials it is not hard to design a streamlined 'nose' but designers clearly don't think it worthwhile at the sort of speeds relevant to the steam age.

 

Streamlining became a "thing" in the 30s for everything. Possibly to try to emulate the new all metal passenger aircraft coming into service, or possibly just to appear "modern" - or both.

Some decided to try to find the best design, others just slapped on what they thought looked cool.