Edward Thompson: Wartime C.M.E. Discussion

I think the D49 must have had a pretty close approximation to the original Holcroft concept, as with the B17, although that would have needed an extension to the middle valve spindle. Certainly both of those classes did have the gear behind the cylinders. I'd guess the B17 arrangement was also theoretically possible for the Peppercorn Pacifics, but when they were being designed the LNER big end problems hadn't been fully solved, so reviving the over-travel of the middle valve issue wouldn't have been very desirable, and it does have the disadvantage of reducing the accessibility of the conjugated gear pivots for lubrication.

The only actual Holcroft implementation of the gear, on the prototype N1 and K1, was also pretty inelegant, with the gear ahead of the cylinders but the drive taken directly from the combination lever with separate rods alongside the valve chests to drive it. Both of them also had the heavily offset centre valve chest, in the Gresley style. C.S. Cocks wrote that it was also prone to over-travel of the middle valve, but the route by which it was replaced was pretty circuitous.

The N1 and K1 had been built and fairly thoroughly tested, but before the production N1 and U1 locomotives were built the Z had been introduced. The Civil Engineer objected to the conjugated gear on the Z at the design stage, because it made the front overhang (already pretty large) completely unacceptable, so a centre valve gear had to be devised (Walschaerts with two eccentrics). This was evidently successful enough to be adopted for the production three-cylinder Moguls, and eventually the prototypes were converted to match the rest.

Edit to add: Holcroft still wanted conjugated gear on the Schools class, but as with several other features he desired, the Eastleigh drawing office managed to lose it in the detail design stage.

 

I suppose one could conclude that no-one in the UK really solved all the design problems of conventional valve gear on large 3 cylinder locomotives. AIUI even the Peppercorn A1s had different events on the middle cylinder to the others, even though they got pretty close, and they were arguably (as they should have been of course) the best of the breed.

If anyone hasn't seen it, this is a crude sketch of the original Holcroft layout, rear of locomotive to the top. The N1 and K1 were quite different to this.

 

It is interesting that you mention valve events Jim.

The often decried equal connecting rods layout of a Thompson A2/2, A2/1 and A2/3 is set up to give precisely that - equal valve events.

This being achieved through identical geometry on all three pistons, albeit with the centre cylinder further forward and driving the front axle (divided drive).

The A2/2s suffered with frame flexing and smokebox saddle cracks as their original P2 frames were split between the front set of driving wheels and second set of driving wheels on the mikado layout.

When rebuilt, the second set of driving wheels became the front set, with a new set of front frames quite literally bolted on to the rear three quarters with new cylinders and the new front bogie.



The reality of this is that the conversion used as much of what was already there.

Wartime restrictions from the war office and LNER emergency board precluded building entirely new - Thompson’s team in the short time they had produced a Pacific design relatively expeditiously.

The testing conducted on Thane of Fife was rigorous and took a year.

You are correct in saying the valve events on the Peppercorn A1 and A2 were not equal - but then the middle cylinder has a different length connecting rod and different geometry.

Compare Thompson, Peppercorns and Jarvis’ approaches on their 6ft 2in three cylinder Pacifics. If we were being entirely theoretical, there is nothing wrong with the logic behind Thompson’s teams choice for equal valve events.

However the divided drive setup, though elevating some stresses of driving all onto one cylinder, creates a less compact locomotive.

Who was right? Is there such a thing in locomotive design as “right?” Are all designs not compromises in some way?

All I know is that my time with a certain Bulleid Pacific of late convinces me that Jarvis was probably closest to an optimum Pacific design in this country, even if my preference remains for those in apple green liveries.

 

Thank-you. That is interesting. I did not know that the Germans changed to three-cylinder pacifics to enable higher speed.

The German Class 03.10 three-cylinder locomotive seems to be equivalent to the British Southern Railway "West Country" Class.

I think that Germany was the only country with more three-cylinder locomotives than Britain. Preferences in both countries seemed to swing to and fro between two-cylinder and three-cylinder options. But the most numerous three-cylinder engines in Germany were 2-10-0 heavy freight, not passenger engines as in Britain.

 

Thank-you for your thoughts here. The digression onto wheel diameters does actually lead back to a wider issue that is very pertinent to discussion of LNER locomotive policy, namely the debate between standardization and the making of design adjustments to suit local conditions. I referred earlier to the Midland Railway, whose engine fleet appears at first sight to have been well standardized. But look a little more closely, and you realise that Samuel Johnson was not a practitioner of strict standardization in the manner of, for example, James Stirling on the South Eastern.

 

8126,

What are your sources for the above statements? I think I have read everything Holcroft ever wrote, and I do not recall anything of the sort.

Cheers,

Julian

 

Prussian P10/Baureihe39/baureihe22 was made three -cylindered due to connecting rod length.
It was planned to make it two-cylindered with short conrods to the second driver set.
Calculations showed that piston pull , when going backward, could lift wheel from track.
It was then proposed to drive unto third driver set but nobody dared make so long conrods at that time.
For the three cylindered freigth engines it was mostly concern for the strength,size and fatique of the big outside driving crankpins of the corresponding two cylinder design.
The P10 is very close to LNER P2 dimesionwise and did 135 kmh on level track on acceptance test.
That had not been nice with two cylinders and 1750mm drivers.

 

I'm of the opinion we can misunderstand standardisation. To my mind the biggest advantages are in consumable components that will be regularly changed, and wheels are arguably not that. On the other hand we need to think it through: the wheels may not need to be swapped very often, but the brake shoes do, and they are dependant on wheel size!

 

Don't forget too that although the wheels may not change size or need changing, the tyres do as they wear/are turned during reprofiling.

 

Yes, I wondered about that. I couldn't imagine that different profile brake shoes were stocked for every possible diameter after turning, I kinda assumed that the sheds probably ground them to as close a fit as could be managed. What happens now?

 

Brake blocks are a standard size for whatever class of Loco or item of rolling stock they are for. New brake blocks quickly bed in to slight variations in wheel diameter, if you grind them to fit perfectly you are wasting a good deal of labour and material in the brake block.

 

I suspect it makes almost no difference in profile of the block, which will rapidly wear to fit the actual profile.

Imagine a driver diameter of, say, 6’8”. That’s a circumference of about 251.5”, of which the brake block itself occupies an arc that is perhaps only 4% of the circumference (for a block about 10” long). By time the wheel has worn to say 6’6”, the wheel circumference is still 245”. If you had a block that was a perfect fit across about ten inches of arc on a wheel of 245” circumference and then placed it against a wheel of 251.5” circumference, there would me minuscule local high spot at the centre, but it would wear away very quickly until block and wheel matched.

Tom

 

Although equal valve events on all three cylinders are ideal, in reality there will be usually be some differences for various reasons, which will increase as components wear. As long as the differences are not too great, they shouldn't matter. Having equal-length connecting rods avoided building in one cause of differences, but at the cost of a very odd front end layout. Did any other designer go for equal-length rods in combination with divided drive?

 

Brake blocks are usually made new so that the tread radius is slightly greater than the wheel radius so that they touch in the middle and not at the ends when first installed. This avoids putting a bending stress in the block a cast iron doesn't like such things. They soon bed to the wheel radius. In theory, though, the block could be of infinite radius (i.e. straight) and it would still work because F= µN is independent of contact area, It's not quite as simple as this in reality, which is why new brake shoes need to bed in.

 

From some comments my Father made I believe the blocks were put on and then wore in on their own.

 

GWR four-cylinders?

 

A question if I may. Wouldn't the conjugated fleet also have had the flawed big end design?

Dave

 

It's not that odd though - look at the Chapelon Pacific, the Stanier Princess, the GWR Castle or King - all short connecting rods, cylinders astride or near the rear bogie wheel.

I don't know - I will have a think.

Hi Dave. Do you mean "wouldn't the non conjugated fleet also have had the flawed big end design?" - as it was only specifically on the conjugated fleet, and in particular, the big engines, that the marine type of big end was an issue.

Better maintenance regime together with workshop setup post war under Cook in the 50s almost eliminated middle big end failures compared to those suffered throughout WW2 and in its immediate aftermath.

 

No, not really. Poor availability has been attributed to the conjugated gear, but I was wondering if big end problems might also have had an impact on the availability of these engines?

Dave

 

It is probably worth reading the Cox Report and making your own mind up on that. There is an argument that a better big end design would have resolved the issues; equally, neither Gresley, nor Thompson, nor Peppercorn's teams managed to design a better version of the big end, despite a number of trials (on a few of the A4s in particular).

Short answer is - yes, the big end did affect availability.