Chapelon and related Matters

Strictly speaking, the expansion link needs to be perpendicular to the centre line of the motion at front and back dead centre. As you say at front and back dead centre, if the gear is properly set up, in theory the die block can be moved from top to bottom and vice versa without moving the valve.

Angularity applies to the eccentric rod as well as the connecting rod. As the NYMR diagrams basically show, the angles of swing of the expansion link fore and aft have to be equal which is usually achieved through the so-called backset by locating the drive pin to the rear of the expansion link fulcrum. As an aside, it would be interesting to know what the valve diagrams of the Southern 2-6-0s look like as they appear to have no backset.

The NYMR diagrams are copied from Yoder and Wharen (fig .170 p.158). This whole chapter of Yoder and Wharen (Locomotive Valves and Valve Gears, dating back to 1917) is on distortions of Walschaerts and measures that can be taken to mitigate them including the method of die block suspension. I suspect that the problem the GWR sought to address is derived from angularity, but the purpose of the cranking of the King rocker remains a mystery to me.

 

What the King rocker does, I think, is play games with angularity effects. I've tried and given up trying to understand the detail, but it seems clear that if one end of the crank moves from say -15 to +15 degrees, with angularity effects at the extremes, the other end of a 15 degree crank would be moving say 0 degrees to +30 degrees, with a greater angularity effect at one extreme and no angularity effect at the other.

 

The Maunsell 2-6-0s actually have negative backset. The valve events tabulated on the drawing are pretty good for 30% and 40% fore gear but the back gear events are quite poor. Whoever did the detail design really knew what they were doing

Sent from my SM-A105FN using Tapatalk

 

Die block suspension has a significant effect with Walschaert's gear. Locos that have a pendulum lifting link in front of the expansion link to support the radius rod (i.e. S15's) have little die block slip in forward gear compared with those having a slide block arrangement on the lifting arm to control the position of the radius rod (i.e. Black 5's). However, having a pendulum link means there is substantial die block slip in back gear. I am of the opinion that this was a significant factor in the unfortunate fatality on the NYMR when the S15 changed direction and crushed a person. In fore gear there is little kick at the reverser but, in back gear, because of the slip that kick becomes substantial and it needs the reverser power lock engaged to prevent it moving into fore gear. It has happened with me in the past. I'm fairly certain that the accident wouldn't have happened with a Black 5 or similarly arranged Walschaert's gear loco.

Coming back to the King's and the question about the rocking lever, I've no idea but will have to find time to study it

 

There must be something wrong here. The prototype N1 was not completed until 1923 and the K1 in 1925, so could not have been discussed at a meeting in 1918. ????

 

That paragraph is about the two cylinder types, N and K, not the 3 cylinders. Holcroft's paper had suggested that 3 cylinder locomotives should intrinsically ride better than those with two outside cylinders.

 

He's talking about the lack of swaying of the two cylinder locos, i.e. class N / class K - in contradiction of Holcroft who had asserted that an advantage of three cylinder designs was that they reduced the swaying of two cylinder types. The president is saying that he had not detected such swaying, and indeed the two cylinder SE&CR 2-6-4Tanks rode very well.

Tom

 

Nock is a bit of an enigma. One assumes that he had some engineering aptitude but this passage reads like a press release. It generally explains that if you set the gear up for the inside, then you cannot simply turn it through 180 degrees for the outside due to conflicting angularities, which is fine as far as it goes, but it does not explain how cranking the rocker achieves the correction, nor does it note that the rocker arms were of different length, which is quite a major point. His statement that "In all well designed valve gears there is a geometrical device for giving the valve the correct non-symmetrical motion." is of the same ilk - come on Ossie, what was that device? He may be referring simply to backset.

 

If the LMS drawing office didn't understand the subtleties, then its hard to blame Nock for not doing so either... The different lengths was one of the few bits I *think* I did understand. Lets say that at 50% travel the "straight" arm is at 90 degrees to its valve rod. If the valve rods are parallel, and equidistant from the pivot, as one might expect, then the "cranked" arm must be somewhat longer to reach the plane of its valve rod.

 

It also has to be remembered that Nock was a signal, not locomotive, engineer, a senior manager at Westinghouse.

 

This is an interesting discussion which I for one would like to pursue, but preferably in a thread of its own. How best do we organise that?

Churchward and his successors must have considered that the de Glehn layout of four cylinders had significant advantage(s) to justify the considerable extra complexity over two cylinders. The most obvious apparent advantage is a great reduction in net reciprocating masses, which should allow easier balancing -- except that (as I understand from some earlier discussion in another thread on here) the hammer blow of the Stars was actually worse until the balancing was modified.

If you're going to have four cylinders you can eliminate some of the complexity by using only two sets of valve gear. If you then just ignore the difference in angularity, the power outputs of the inner and outer cylinders will be a bit different, but how much would that really matter? Anyway Churchward or someone in the Swindon Drawing Office presumably felt that they might as well deal with the difference in angularity, and they came up with some fiendishly clever geometry.

Without rigorous analysis, I can understand intuitively that the cranked levers cause the outside valves to more a bit further on half the stroke and a bit less on the other half, thus (if the details are worked out correctly) compensating for the different angularity. But unequal lengths would surely result in different total valve travel, which makes no sense to my limited understanding.

Edit: corrected a typo

 

Well, one may request it of the mods. But I do wonder how many people there are alive who actually fully understand the topic! I freely admit that I can feel frying neurons when I try!

I don't think it does if you can tune the angularity effects sufficiently. I did try sketching something, but it needs a CAD package and skills I don't have to work it out properly.
But the way it goes in my head is this:
Imagine a rocker that at each end comprises an equal arc. So the arc has a lateral component and a lengthwise component.
Now imagine a connecting rod on each end that connect to parallel valve rods.
If both connecting rods are the same length then the valve rods have identical travel, because the angularity balances out.
Now imagine one connecting rod is very long, and the other very short.
The angularity effect is very large for the short rod, and very small for the long rod.
The end result is that travel of the valve rod with the short connecting rod is less than that for the long one.
Now consider the cranked rod. Instead traversing an equal arc each side of the centre, its traversing an arc where one end is parallel to the valve rod, and the other has much more than double the angularity. Now there are still more angularity effects.
At this point it then occurs to me that the pivot of the rocker need not be equidistant between the valve rods, and the plane of the valve rod may intersect the arc the end of the rocker describes at different places, and I hastily retreat before the smell of frying neurons becomes quite unacceptable, and my admiration for a man who could sort all this out with no more assistance than a notebook and a book of logarithmic tables increases further...

 

It would have given better valve events ,less mass, wear and costs if each cylinder had had its own valve gear.

Where inside cylinders were a must for gauge issues and tracking it will be a better solutions to have valve gear outside;

https://upload.wikimedia.org/wikipedia/commons/2/27/Gr625.017.jpg

the reverse of GWR that used inside Stephensons for outside cylinders.

The layout of valvery and cylinders on UK locomotives was not rational but dictated by fashion and morale and permissible skirt rim altitude.

 

That Italian Loco is a bit - interesting

 

It is based on a compound 2-6-0 from 1904 type 600 that became type FS 625 in 1910 superheated and two simple cylinders.
It is a Walscharts as was the Bulleid chain thing.
Italians made about 300 and they still have 25 and some that run.
My last holliday with wife in spring 2013 was behind one from Bergamo to a lake somewhere.

https://anno.onb.ac.at/cgi-content/anno-plus?aid=lok&datum=1909&pos=5

on page 169 and forward and page 242 and forward some drawing and photos.
A system from before 1900 have being used in earnest in a coal importing country until 1970 and then quite some being heritaged must have been usefull.

I find it interesting because the german baureihe 24 069 two cylinder compound (and 25 bar)
used 4.9 kg of steam per ihph and a Britania 6kg.
The next generation all round UK lokomotive is in my view a K1 with one outside high presure cylinder as is but lifted 50mm and put forward 75mm and a big inside lowpresure between frames and a boiler made of 9f plates
The small movement of the remaining out side cylinder allows a Zara truck and this allows 75 mph on 5 feet drivers and best british steam economy.
Where can I find a K1 drawing without needing to sell a grandchild?(Or two)
I have a very good LMS Crab book but K1 cab and footplate looks nicer and that is very important in all steam enginering.

 

The drawing shows a 7" offset, using Pythagoras the length of the inclined side would be the square root of 1' 5 3/4" squared plus 7 squared giving 19,08 " instead of the real 19,375". As a consequence the outside valve is shifted a little more
than the inside valve. However if this was made equal any wear and tear in the pins would result in a smaller valve movement, that is not really desirable.
Kind regards
Jos

 

From Don Ashton's web site (https://www.donashton.co.uk/html/more_cylinders.html) here's his sketch of the King valve gear movement.
For convenience I've added in my 'handwriting' the total movement. The effect of the cranked lever is quite marked.

 

I feel like we should have a proper conference on some of this stuff, with people submitting papers. I have sat back because I profess, I am struggling to understand, but it is entirely fascinating.

 

The trouble is how many people are there who have a sufficient understanding of the topic. I tiptoe round some of the edges, aware, I hope, of just how little I do understand, but I certainly couldn't hope to present a paper.

For instance its clear that the GWR draughtsmen spent an immense amount of time in getting as close as they possibly could to an ideal of identical valve events on inside and outside cylinders, whereas their opposite numbers at the LMS seem to have metaphorically said "**** it, that's close enough for hand grenades." But I haven't a clue how much difference it makes, or even what difference it makes, and whether pragmatically speaking the LMS decision was reasonable.

Don Ashton tells us that Peppercorn's staff did a superb job on making the events of the inside gear of the A1 match the outside ones (apparently a very difficult task with the connecting rods rather different in length) only to mess up at the very last second by making the operating arm for the outside cylinders longer than that for the inside when it should have been the other way round. But how much difference does that little error make, bearing in mind the A1 is still arguably the best British express locomotive?

I've tried to understand Don Ashton's work, but it appears to me that although he puts an awful lot of effort into instructing his audience in how to get it right, I can find very little about the consequences of getting it wrong. Maybe that's understood in the model engineering fraternity that are his primary audience?

 

I cannot comment on the details of valve gear design, which is beyond my level of technical comprehension. But perhaps I can make a general observation.

When Churchward and his team were designing the Star Class valve gear around 1906, they did not have a lot of precedents from which to learn. The de Glehn 4-cylinder compounds had been around for 20 years, but they had four separate valve gears. Some other 4-cylinder compounds made do with two valve gears plus rocking levers, including the Webb 4-cylinder compounds in Britain (with inside Joy gear) and the Maffei compounds of the South German railways (with outside Walschaerts gear).

4-cylinder (non-articulated) simples were a rarity in 1906. There had been a few isolated examples in the 19th Century, and Dugald Drummond had built his F13 4-6-0s in 1905, with four separate valve gears. Churchward opted for a different approach for his Star-class, but he seems to have been very much designing the valve gear from scratch rather than following any previous model (leaving aside the briefly-used "scissors gear" option).

PS: Apologies for adding to a discussion that is completely off-topic for this thread!