Chapelon and related Matters

The wording that Simon highlights in his quote from the patent in post #158 is typical in patent specifications, preventing someone from coming along and evading the patent by changing something without changing the essential principle (in this case deriving the motion for the middle valve by a system of levers from the outside valve gear). Any particular change to the details, such as that reportedly suggested by Holcroft allowing the levers to remain in one plane despite the cylinders being differently inclined, might or might not be held to infringe the patent if that came into question, but in this case no such issue seems to have arisen, unless perhaps any of the overseas engineers who adopted the Gresley arrangement might have sought to avoid paying royalties.

Nothing said so far in this thread establishes for certain whether Gresley adopted that arrangement at Holcroft's suggestion or whether he had already decided on it before Holcroft suggested it. When Holcroft said "How about doing (so-and-so)" (assuming that he did), Gresley may have said "Just what I was already intending" or he may have said "Thank, you, good idea".

 

Interesting that that letter is addressed to what (I presume) was Holcroft’s private address, rather than the SECR offices at London Bridge or Ashford. Which does tend towards the supposition that Gresley had a bit more in mind than a cosy fireside chat about valve gears!

(Which, as per “Locomotive Adventure page 93, Holcroft batted back saying that if he wanted to go through with that route, he should approach Maunsell formally).

Tom

 

But..........

Presumably as 3 cylinder loco's became more common the idea of a conjugated valve gear must have been considered by several locomotive engineers, and potential workable options would be limited?

 

(Slightly off topic ...)

"The President:

I have not had much experience with outside cylinders, but there are two locomotives running at the present time on the South Eastern & Chatham Railway - a 2-6-4 tank engine and a 2-6-0 tender engine and I can say from personal experience, and also from the experience of others who have ridden a great deal on these engines, that there is no swaying. The tank engine is remarkably steady running, and at high speeds especially so."
​

That aged well :-(

Tom

(p.s. - and I realise that the track was sub-standard as well).

 

I note in your link @Jimc that the discussion piece includes Gresley talking about inclining the cylinder, his original issue being not having the valve spindle ends on the outside offset, something that had changed by 1919 with the K3 development.

In all, I just don’t think there’s enough evidence to turn Gresley’s conjugated valve gear into “Gresley - Holcroft” - and my view hasn’t really changed based on what we’ve read here.

I do think it is pretty telling that Gresley himself never deigned to change the name of his own gear, nor did Holcroft seek to have his name attributed to Gresley’s.

The conjugation (if you’ll pardon the pun) of their names onto the one distinct type feels like an enthusiast wanting to provide Holcroft for more credit than is due.

 

Not sure we can quite say that. Locomotive Adventure Volume 1 page 93...

 

Interesting Jim! Thanks for posting. But who specifically was referring to it as Gresley-Holcroft?

I will of course revise my view on Holcroft’s beliefs.

 

So I have a suspicion in mind, so I will voice it on the grounds that this is only what I currently think, I am open to having my mind changed.

I do not think Gresley was willing to name his valve gear variants as anything other than “Gresley conjugated valve gear”.

I think the main difference between Holcroft and Gresley is the approach to the use of the spindle ends in relation to the 2:1 gear. Yes, Gresley’s next locomotives after his talk with Holcroft altered the angles of the cylinders, in particular the centre cylinder became steeply inclined.

I think it is fair to say Holcroft suggested this - however I think given the timing of the various designs, Gresley already had this in mind and Holcroft confirmed his thoughts.

Gresley offered him a job and reached out to Maunsell directly, who rejected the offer on Holcroft’s behalf.

It does make you wonder what would have happened had they actually collaborated in an official capacity.

 

According to Holcroft's autobiography he was still working for the Government at the time of the meeting with Gresley. He reports that Gresley said he should get back to locomotive work to which he replied there is nothing I would like more!

 

Does the cranking of the King/Castle rocker make it out of phase or modify the travel? I find it very difficult to visualise what the effect on the valve movement is of cranking the rocking lever. If you take a piece of wire, say a straightened paper clip, as you start to crank it from the straight and push it around a fulcrum in the middle, part of the resultant action is transferred to a push to the side (i.e. it becomes akin to a bell crank) rather than in the opposite direction. The rocker on a King (based on the Swindon June 1927 drawing 83262 - I do not have a Castle drawing) had arms of unequal lengths: 1' 5 3/4" and 1' 7 3/8" (the longer arm activating the outside valves), so admittedly it seems that the cranking would not be needed to modify the relative outside travel. I know the cranking seems to be effective in the Dockstader model but is there anything in Swindon-issued material that speaks of the thinking of the design behind cranking the rocker?

The Duchesses (and 46205) had straight rockers as you suggest, with arms of equal length. As the LMS valve tables put it (the LMS helpfully put valve tables on its drawings of motion), if you take the average of the valve events of adjacent cylinders, they are very good events! Having said that, as we are on a Gresley-related thread, the valve events and indicator diagrams of his conjugated gear were anything but even in practice, but saved by boilers which produced lots of steam (even though Ell managed to signifcantly increase steam production when he got his hands on the V2), so presumably even if we are converted by Don Ashton's zeal for geometrically perfect events, we perhaps should not worry too much about the geometric precision?

 

Way above my pay grade here. I may have maintained Don Ashton's web site for a few years, but that doesn't mean I understood all of it!

I note though, in my own field of networking infrastructure, we had a saying "broken networks work". What these means in practice is that there's a big area between a completely functional error free network and something that doesn't work at all, and in that region you can have huge inefficiency without it being particularly obvious to the end user. It seems to me that the steam engine is a device that wants to function, and will accept all sorts of levels of inefficiency and even partial failure while still seeming to work adequately. What mechanical knowledge I have is from working in motorcycle shops with what are basically finicky high stressed engines, and what seems to have been tolerated as acceptable in steam days I find just mind boggling!

 

Isn't the effect similar to that obtained with the backset of an expansion link with Walschaerts valve gear? That is necessary because the return crank rod is not parallel to the radius rod. I don't know as I haven't studied swindon practice but I ca nsee some logic in this.

 

I have followed the Gresley/Holcroft discussion, although as a Chemical Engineer I learnt
early on that the minutiae ( small but essential ) of valve gear design was not for me.
However the mention of the late Don Ashton made me wonder if he had anything to
say on the subject.

QUOTE “ Perhaps the most celebrated example of conjugated valve gear for three
cylinders is the Gresley LNER arrangement in which Holcroft was consulted “ END QUOTE

That IMHO appears to encapsulate the position neatly ?

Michael Rowe

 

That seems fair to me.

 

Going a bit OT (maybe worth a thread on its own?)

I came across these two videos on Youtube and (allowing for US terminology, i.e throttle vs regulator, main rod vs connecting rod etc) they are the clearest explanation about valves and valve gear design I have ever seen, with really informative animations.

The first one covers valves (primarily piston valves), how they move relative to the piston; the design of cylinder castings to allow steam passages, and the concepts of lap, lead and exhaust clearance measurements.



The second one then builds up Walschaerts valve gear from first principles, showing what each rod does, how it moves through a cycle and geometrically how it forms part of imparting the correct motion to the valve head to allow reversal and variable cut off.



You can probably watch the second in isolation to get a view of the geometric complexities of Walschaerts valve gear - obviously the derived gear is then a further abstraction (not covered) that combines two motions to impart the necessary valve movements to a third valve.

Tom

 

The cranking of the King/Castle rocker obviously does something, but the movements of the outside and inside valves must remain in anti-phase. I think it does something to the relative amounts of valve motion at the beginning and end of the stroke, which indeed may be similar to the effect of the backset of the expansion link with Walschaerts valve gear. Both of those exercises in geometry are too subtle for my old brain.

BTW, splitting the Chapelon-related discussions from the general Gresley thread did seem like a good idea, but it hasn't worked very well: here we are on the Chapelon thread discussing Gresley's apparent attempt to poach Holcroft, conjugated gears, and even the geometry of GWR valve gear.
Edit: corrected a typo and inserted a missing word.

 

Yes, some tidying up undoubtedly needed, my apologies to the mods for my part in the back and forth across threads...!

 

It's all become a bit of a chapeleon really.

 

All explained (hopefully) on P.25 of O.S. Nock's 'The Stars, Castles and Kings' Part 1.

 

Further to my last post, for those that don't understand what I was waffling on about this may help better explain things. If the cylinder is on back dead centre the expansion link needs to be centralised so that, when moving the radius rod from forward to reverse, there is no movement of the valve. With no backset, the swing of the expansion link is not equal and thus, the valve travel is not equal. Adding the correct backset corrects that error.

Edit: drawing slightly altered as I realised there was an error in the original.